Read Microsoft Word - BK_ZWE 02 2011 final.docx text version

Zimbabwe's Foreign Trade Performance During the Decade of Economic Turmoil

Will Exports Recover?

Bartlomiej Kaminski and Francis Ng 2/2/2011

Key words: Zimbabwe's economy, macroeconomic stability and policy, economic governance, foreign trade performance, export competitiveness, natural resources, factor intensity, revealed comparative advantage, export specialization

JEL classification: F10, F13, F14, O11, O13, O24

A revised version of a background paper prepared for Zimbabwe's Diagnostic Trade Integration Study in Africa Region (AFTP1). Without in any way asking them to share responsibility for opinions and interpretations expressed in this paper or for errors that might remain, the authors are grateful to Angelica Katuruza, Praveen Kumar, Sydney Mabika, Gift Mugano, Simon Nyarota, for their useful comments and suggestions on the draft of this report. In particular, we would like to thank Ian Gillson for numerous discussions and involvement in all stages of our work on this project. Emails of correspondence: "Bart Kaminski"<[email protected]>; "Francis Ng"<[email protected]> Political instability and economic mismanagement have extracted a heavy toll on Zimbabwe's economy. Over the last decade or so, traditional surpluses in agricultural products and industrial raw materials have either diminished or disappeared turning Zimbabwe into a net importer of agricultural products. The volume of exports has been in decline each year since 1998 while both contracting exports and imports have shifted away from the EU to the South African market. Yet, considering the extremely poor quality of Zimbabwe's economic governance and the antiforeign trade bias of its institutions and economic policies, it is rather surprising that all exports--except for food products--have not contracted more. Following the restoration of macroeconomic stability in 2009 after more than a decade of inflation and hyperinflation, the major barrier to economic activity has been removed. Yet, macroeconomic stability alone will not put the economy on the path of sustained growth unless weaknesses in the country's investment climate are addressed. The damage inflicted by a decade of illconceived economic policies has been huge but it can be reversed once policies conducive to the revival of specializations in line with Zimbabwe's endowments in natural resources, relatively high quality of human capital and attractive natural environment are in place.

1 | P a g e

Contents

Executive summary ................................................................................................................................... 4 Introduction .............................................................................................................................................. 8 A. Determinants of foreign trade performance: governance and macroinstability ............................ 8 B. Observations derived from empirical analysis of Zimbabwe's foreign trade ................................. 10 C. Structure of the report ................................................................................................................... 12 . 1. Zimbabwe's lost decade: economic governance, macroeconomic disequilibria and other self inflicted wounds ...................................................................................................................................... 14 1.1. Quality of economic governance in 19962009 in a comparative perspective ........................... 14 1.2. Economy in a frenzied free fall in 19992008: implications for foreign trade ............................. 18 1.3. Zimbabwe's foreign trade performance in a regional perspective.............................................. 25 1.4. Concluding observation ............................................................................................................... 27 2. . Highlights of Zimbabwe's trade performance: openness, dynamics, and direction of trade ........ 29 2.1. Deceptive increase in openness .................................................................................................. 29 . 2.2. Trade in services: transportation and tourism ............................................................................ 31 2.3. Dynamics of foreign trade in goods in 19972008 ....................................................................... 34 2.4. Zimbabwe's export rebound in 200307: was it priceled or quantitydriven? ........................... 36 2.5. Changes in Zimbabwe's direction of trade: a shift in exports from the EU towards Southern African markets ................................................................................................................................... 41 2.6. Zimbabwe's export performance in regional markets................................................................. 44 2.7. Concluding comment ................................................................................................................... 45 3. Drivers of Zimbabwe's trade performance: evolving patterns of specialization ............................ 47 3.1. Diversity in exports has declined ................................................................................................. 47 3.2. The degree of processing embodied in exports and imports ...................................................... 49 3.3. Agriculture and agroprocessing: contraction of net exports ...................................................... 53 3.4. Factor and technology content of foreign trade and revealed comparative advantage ............ 55 . 3.5. Victims and survivors: new and vanishing specializations ........................................................... 59 3.6. Revealed comparative advantage: predictable shifts from agriculture to extractive sectors ..... 62 3.7. Conclusion .................................................................................................................................... 69 4. Moving forward: how to revive exports? ....................................................................................... 71

2 | P a g e 4.1. Restoration of macroeconomic stability: its initial impact and the challenge ahead ................. 73 . 4.2. A new setting for foreign trade policy: implications of the multicurrency regime .................... 75 4.3. Expanding the export base calls for new investments but barriers persist ................................. 79 4.3. Conclusion .................................................................................................................................... 82 References .............................................................................................................................................. 84 Appendix 1: Note on foreign trade reporting in Zimbabwe ................................................................... 87 Appendix 2: Note on geography and policyinduced costs of trading across borders .......................... 99 Appendix 3: Taxation as the most binding formal constraint to do business in Zimbabwe ................. 103 Statistical Annex .................................................................................................................................... 106

Tables

Table 1: The values of aggregate quality of governance (SAG) indices for Zimbabwe and selected SADC countries in 1996, 1998, and 200209 ............................................................................................................................................. 5 1 Table 2: Dimensions of economic governance quality in Zimbabwe in1996, 1998, 2000, 200209 ........................... 6 1 Table 3: Exports of selected minerals in terms of volume and value in percent of their peak performance and share in total exports in 2010 (in percent) ............................................................................................................................ 0 2 Table 4: Imports of capital equipment and FDI flows to Zimbabwe and Zambia in 19952008 (in millions of US dollars) ......................................................................................................................................................................... 3 2 Table 5: Foreign trade growth performance of selected SubSaharan African countries in 200108 (in millions of US dollars, dollars and percent) ........................................................................................................................................ 6 2 Table 6: GDP and its growth in real and value terms according to different sources over 19982008 (in millions of US dollars and percent) ............................................................................................................................................... 0 3 3 Table 7: Zimbabwe's trade in services in 2000 and 20082010 (in millions of US dollars) .......................................... 1 Table 8: Receipt from international tourism according to Zimbabwe's balance of payments statistics and the World Bank' world development indicators database in 200008 ......................................................................................... 3 3 Table 9: Exports and imports of goods in terms of value and exports in percent of imports in 19972008 (in millions of US dollars) ............................................................................................................................................................... 5 3 Table 10: Three phases of export growth in 19902008 (in percent) .......................................................................... 6 3 Table 11: Exports of selected commodities and total exports as reported by Zimbabwe's authorities and importers 3 of Zimbabwe's products in 200008 (in current prices)............................................................................................... 6 Table 12: Major Zimbabwe's exports in terms of natural units in 200008 ................................................................ 7 3 Table 13: Prices of major Zimbabwe's exports in 200008 (in US dollars per unit) ..................................................... 8 3 Table 14: Cumulative annual change in prices and quantities exported in 200108 and 200307 (in percent) .......... 9 3 Table 15: Total values of exports of selected commodities expressed in constant quantities and prices in 200008 (in millions of US dollars) ............................................................................................................................................. 0 4 Table 16: Direction of trade in 200008 (in millions of US dollars and percent) ......................................................... 2 4 Table 17: Exports to neighboring countries, EU27 and ROW in 19972008 (in millions of current US dollars and percent) ....................................................................................................................................................................... 4 4 Table 18: Zimbabwe's export performance in neighboring countries in 200008 (in percent) ................................... 5 4

3 | P a g e

Table 19: Various indicators of concentration in 1994, 1997, 2002, 2007 and 2008 (in percent and numbers) ........ 8 4 Table 20: Shares of top ten in total exports and values of HerfindahlHirschman index of total exports and of 4 exports to South Africa and European Union: fourdigit SITC items ........................................................................... 9 Table 21: Developments in Zimbabwe's exports in terms of enduse product categories ......................................... 0 5 Table 22: Developments in Zimbabwe's trade by enduse product categories in selected years over 19972008 (in millions of US dollars and percent) .............................................................................................................................. 1 5 Table 23: Direction of Zimbabwe's exports by enduse to South Africa, European Union and rest of the world in 1997 and 2008 (in percent) ......................................................................................................................................... 2 5 Table 24: Net exports of raw food and agricultural products in 19972008 (in current millions of US dollars) ......... 4 5 5 Table 25: Change in factor content of Zimbabwe's trade in goods in 199497, 19972003 and 200207 .................. 6 Table 26: Exports in terms of technology intensities in 1997 and 200208 (in millions of US dollars and percent) ... 8 5 Table 27: Dynamics of Zimbabwe's trade in terms of technology and factor intensities in 199497, 19982002, 200307, and 2008 (in percent and millions of dollars) ............................................................................................... 9 5 Table 28: Victims and survivors in 2008 against the average in 199497, 19982002 and 200307 (in millions of US dollars and fourdigit SITC product lines) .................................................................................................................... 1 6 6 Table 29: Low technology labor intensive products with the values of RCA exceeding unity in 1997 and 2008 ........ 2 Table 30: Resource intensive products with values of RCA exceeding unity in 1997 and 2008 ................................. 3 6 Table 31: Export specialization in South African and EU markets in 1997, 2002 and 200708: number of SITC three digit products with values exceeding unity in terms of technology, natural resource and labor content .................. 4 6 Table 32: Agricultural `losers:' values of RCA and exports in 1997, 20022008 (in millions of US dollars) ................. 5 6 Table 33: Exports and values of RCA of threedigit SITC manufactured products that lost comparative advantage in world markets in 199708 and their technology and factor content .......................................................................... 6 6 Table 34: Exports of threedigit SITC sectors that acquired revealed comparative advantage in 200708 (in millions of US dollars in 1997 and 200208) ............................................................................................................................. 8 6 Table 35: Projected exports of selected agricultural and nonagricultural commodities in current prices and volumes in 2010 (in percent) ....................................................................................................................................... 4 7 Table 36: Zimbabwe's exports and imports of sensitive agricultural and nonagricultural products in 200209 (in percent) ....................................................................................................................................................................... 7 7

Figures

Figure 1: Moving in different directions: Zimbabwe' quality of economic governance against that of Mozambique and Rwanda in 1996,1998, 2000, 200209 .................................................................................................................. 6 1 Figure 2: FDI in Zimbabwe and Zambia in 19942009 (in millions of US dollars) ........................................................ 2 2 Figure 3: Two different tales: Imports of machinery and equipment (index 1996=100) and FDI into Zimbabwe and Zambia in 19942008 (index 1997=100) ...................................................................................................................... 3 2 Figure 4: GDP and foreign trade in US dollars per capita and percent in 19972000 and 200308 ............................ 0 3 Figure 5: Zimbabwe's exports and imports of goods in 19972008 (in millions of current US dollars) ....................... 4 3 Figure 6: Exports to neighboring countries, EU27 and ROW in 19972008 (in millions of current US dollars) .......... 3 4 Figure 7: Developments in trade balances of traditional inputs and Zimbabwe's total trade balance in 199408 (n millions of US dollars) .................................................................................................................................................. 1 5 Figure 8: Exports of lowtechnology labor intensive products, resource intensive and medium to hightech products in 19942008 (in millions of US dollars) ....................................................................................................................... 8 5

4 | P a g e

Executive summary

The objective of this report is to provide preliminary answers to three intertwined questions that are essential to the development of a strategy that would boost Zimbabwe's foreign trade performance: How a decade of economic chaos has affected Zimbabwe's export base? Whether lost export capacity can be restored? And what policy actions are now necessary to set Zimbabwe on a path of sustained, exportled growth? The short answer to these questions can be summarized as follows: As a result of economic policies pursued in 200009, exports of goods in terms of quantities are well below the levels achieved in the 1990s, albeit with some notable exceptions discussed below, and the balance of trade has been in deficit since 2006. For services, revenue from tourism has increased steeply over 200910 to reach the levels previously recorded in the 1990s, which has helped somewhat in offsetting significant expenditures associated with the costs of transporting goods into Zimbabwe. But it also suggests that the potential for doubledigit growth in this sector, at least in the shortterm, has probably already been exhausted; Despite significant contractions in the real output of exporting sectors, not all capacities have been lost and some can be restored relatively quickly. The most affected sectors were labor intensive agriculture and miscellaneous industrial sectors. Less affected were capital intensive extractive industries. Output of tobacco can be probably increased in a relatively short period, as can agro processing. Restoration and the development of new export capacities requires an overhaul of existing economic policy. In particular, without the removal of policy barriers to domestic and foreign investments, the development of a diversified exports base will not be realised. However given the extent of current bottlenecks, their removal may (a) be relatively straightforward assuming political commitment; and (b) release powerful forces stimulating economic activitity. Put differently, the payoff from economic reforms can be quick and significantthanks to an existing dissonance between the country's endowments and its subpar performance.

These observations derive from an empirical analysis. The diagnostics underlying them may be usefully divided into two groups: foreign trade performance and economic governance.

Foreign trade performance

While there is a commonly shared view in Zimbabwe that the main problem facing the economy is underutilized production capacity in most sectors, an examination of Zimbabwe's export performance does not provide empirical support to this view. The exports base has shrunk, although probably less than data on the contraction of GDP might suggest. Exports and imports have undergone a huge transformation during the period of economic turmoil in 200008. These changes can be summarized as follows: The contraction in exports and imports was quite substantial although not as large as the fall in real GDP per capita: o Had Zimbabwe exported the same quantities in 2008 as it did in 2000, the value of its exports would have been around 50 percent higher than they were in 2008. In real terms,

5 | P a g e exports in 2008 were significantly lower than a decade earlier. In 2000 prices they were at least 30 percent lower in 2008 than in 2000; o Had Zimbabwe's exports grown at the region's average, their value in 2008 would have been 2.3 times higher than they actually were. Zimbabwe's share in the region's exports fell from 5 percent in 200002 to 3 percent in 2005 and, further, to 2.2 percent in 2008. In order to retain its export share in 2002, the value of exports in 2008 would have had to be US$5 billion or around 2.3 times higher than it actually was that year; o Had imports grown in line with the region's average, the value of imports would be 1.8 times higher or US$4.7 billion rather than US$2.7 billion in 2008. The contraction of imports has been less pronounced than that for exports. Relative to other countries in the region, the share of Zimbabwe in the region's imports fell from 3.8 percent in 2000 to 2.2 percent. o Consequently: Zimbabwe's trade openness as traditionally defined has paradoxically increased. The fall in GDP per capita there was accompanied by an increase in foreign trade as a percent of GDP solely because both exports and imports per capita did not fall as steeply as GDP. Deficits in goods trade replaced earlier surpluses. Zimbabwe ran a significant trade surplus in 19992005 followed by growing deficits in goods trade. Export coverage of imports fell from 130 percent in 2000 to 83 percent in 2008 and 51 percent in 2009. The structure of Zimbabwe's exports has changed rather substantially mainly as a result of differences in the pace of contraction of different product groups as well as both exit and entry of exporters; o The contraction in agricultural exports was much more pronounced than in exports of minerals. Industrial raw materials replaced agricultural foods and feeds as major exports in 2005: the value of exports of agricultural foods and feeds fell from around US$1 billion in 19972001 to US$400 million in 2005 while that of industrial raw materials rose from around US$150 million to US$400 million in 2005 and US$800 million in 2008; o The contraction in exports has been accompanied by a shift from lowtechnology unskilled laborintensive to nonprocessed natural resourceintensive products with the share of medium to high technology exports remaining unchanged; o The degree of processing embodied in Zimbabwean exports has remained unchanged as one set of products (industrial raw materials) has replaced another (food and feeds), both being low value added traditional inputs; o ... while the concentration of Zimbabwe's exports has increased dramatically especially to Zimbabwe's two major markets (South Africa and the European Union) e.g. The number of SITC fourdigit exports exceeding the value of US$1 million has steadily fallen from 140 in 199497 to 133 in 200307 and 114 in 2008; The values of the HerfindahlHirschman Index of export concentration for: South Africadestined exports surged from an average of 0.018 in 199497 to 0.19 in 200307 and 0.22 in 2008; EUdestined exports increased from 0.11 in 200307 to 0.14 in 2008.

6 | P a g e Zimbabwe's foreign trade has also undergone a fundamental change in terms of its geographic composition triggered by a redirection of trade to its neighbors, mainly South Africa, at the expense of falling exports to the EU and the rest of the world. Zimbabwe is the second most trade dependant economy in Southern Africa. The turning point was in 2003, when the value exports to its neighbors increased 53 percent in that year alone overtaking exports to the EU, which contracted 9 percent: o The shift in the geographical pattern of Zimbabwe's trade was therefore not triggered by expanding exports. Redirection of Zimbabwe's exports has been due to a redirection of exports from the EU to South Africa. The change in the geographical pattern of Zimbabwe's imports has been less pronounced, yet displays striking similarities with those for exports. The share of South Africa in Zimbabwe's total imports was 63 percent in 2008 and much higher than its exports (34 percent).

Quality of economic governance

Zimbabwe's economy was in the state of a free fall in 19992008: its GDP was falling every year while inflation was increasing to reach an astronomical 500 billion percent in 2008. The introduction of the multicurrency regime in April 2009 marked a turning point not only in restoring macroeconomic stability but also because it ended trade suppressing foreign currency regulations and other confiscatory measures. The major findings concerning the evolution of the quality of economic governance over the last decade can be summarized as follows: The quality of economic governance as measured by regulatory quality, political stability, and government effectiveness dramatically deteriorated. Zimbabwe moved from an average performer in the regional context in the 1990s to a poor one already in 2000. Subsequently, the quality of economic governance continued falling, elevating Zimbabwe to one of the worst run and least competitive economies in the world. Despite the restoration of macroeconomic stability in 2009, the quality of economic governance has not significantly improved according to various international rankings. Reasons include poor quality of economic regulations and disregard of the government for private property rights as exemplified by its indigenization program prompting capital flight and erecting new barriers to foreign direct investment. While other countries in the region have introduced economic reforms and seen significant improvements in the quality of their business environments, Zimbabwe was moved in the opposite direction. Ease of doing business has deteriorated in all dimensions and lags behind most other countries. Not surprisingly, the incidence of corruption, as captured by Transparency International's corruption perception indices, has moved in step with the deterioration in the quality of economic governance. Despite notable improvement in 2009, Zimbabwe remains perceived as the most corrupt country in the region.

Macroeconomic stability alone will not revive exports: policy reforms and new investments will

Most constraints in Zimbabwe are policyinduced weaknesses whose removal does not require significant amounts of time or resources, and so could be relatively quickly addressed provided there is political commitment to liberal economic reform which is indispensable to revive growth and reduce

7 | P a g e unemployment. Complementing macroeconomic stability with measures friendly to private sector development would go a long way to boosting economic growth in the country. This calls for the development of a comprehensive program of economic reforms that would address all aspects of Zimbabwe's economic regime. Without significantly reducing regulatory hassle and the fiscal burden of conducting business in Zimbabwe, the revival of laborintensive exports will be impossible. Reform is not only necessary for job creation but also to alleviate the very real risk that the country will become even more dependent on foreign aid and continue underperforming relative to its significant potential. However, the reversal in economic fortunes in any significant way is not likely to be achieved without inflows of FDI (foreign direct investment). FDI has stayed away from Zimbabwe for the last ten years going to its neighbors instead. Without a dramatic increase in FDI inflows, modernization of Zimbabwe's extractive and manufacturing sectors will simply not occur.

8 | P a g e

Introduction

The Global Political Agreement (GPA), signed on September 15, 2008, and the introduction of a multi currency regime in April 2009 have provided a modicum of political certainty and restored macroeconomic stability. The latter put an end to hyperinflation and removed some of the most damaging measures introduced during the economic chaos in 200009. A key question is whether these measures will suffice to restore economic growth and boost exports? A. Determinants of foreign trade performance: governance and macroinstability This study provides an indirect answer to this question through examining both economic governance and foreign trade performance. At the most general level, the answer is that macroeconomic instabilities were not the sole cause of Zimbabwe's disastrous economic performance. Instead, it was also the product of low quality of economic governance and the government's disregard for private property rights. Government policies have failed to strike a balance between redistributive functions and growth generation. Rentseeking behavior, aimed at transferring or keeping wealth thanks to political connections, was encouraged while activity that increases the economic pie was suppressed. Restoration of macroeconomic stability without setting institutions that "reward socially useful entrepreneurial activity once started (...) as otherwise individuals cannot be expected to take the risks of losing their money and their time in illfated ventures" (Baumol et al. 2007, p. 7) will fail to produce sustainable economic growth. At a more pedestrian level, another cause of the country's poor growth record is that Zimbabwe has not had institutions and structures supporting protection of property and contract rights combined with a low cost of doing business, i.e., easiness to enter and exit business activities, low taxation and regulatory compliance cost, high flexibility of labor markets, and a reasonably wellfunctioning financial system. Only under these conditions, could firms have had incentives to expand and cut costs in order to stay competitive. Indeed, all success stories of economic development over the last two decades have been based on exportoriented policies combined with businessfriendly environment, political and macroeconomic stability, and low or declining barriers to imports of goods and services, and openness to FDI inflows. Successful countries have also developed infrastructure commensurate with growth in internal demand for inputs and for expanding transportation of goods, storage facilities and ports. Zimbabwe appears to score satisfactorily in terms of physical access to external markets. Its major urban and industrial centers are linked with paved roads and railroads both tied into an extensive central African transportation network with all its neighbors. But the government has so far failed to follow up on opportunities stemming from restoring macro economic stability. It has been acting as though macroeconomic stability by default would end under utilization of capacities and restore employment. The problem is that some human capital might have been lost as a result of longterm unemployment. More importantly, as our analysis of Zimbabwe's exports performance suggests, these production capacities may no longer exist. The scope of firms engaged in export activities significantly declined in the 2000s. Some survived by selling in domestic markets, but even those that did did nowbadly need new capital investment to stay competitive. High real interest rates combined with an unfriendly investment environment do not make these investments

9 | P a g e possible. Except for the removal of some restraints on current business activities (e.g., the foreign currency surrender requirement, foreign exchange controls), there has been no improvement in the investment climate. Put differently, the quality of Zimbabwe's business environment has improved but only marginally as other barriers that put Zimbabwe at the bottom of several international rankings have remained in place. The size of Zimbabwe's economic reversal was unprecedented during peacetime, so the pace of rebound should be without precedent had the issue been of only activating existing capacities. GDP per capita contracted by around half and so did Zimbabwean exports and imports. Agricultural exports collapsed. Minerals and low processed industrial raw materials have replaced them as Zimbabwe's largest exports, despite the contraction in terms of volumes exported. Exports of manufactured goods stagnated in terms of value and fell in terms of quantities. And the unemployment rate hovered at around 7080 percent. Under these circumstances, one would expect strong economic growth had the only barrier been the absence of macroeconomic stability. But this is yet to take place. Zimbabwe's GDP growth of 5 percent in 2009--albeit the first recorded expansion since 1998--and the projected GDP growth of 5.4 percent in 2010 strikes one as rather modest considering the unprecedented size of the country's economic contraction. It is tempting to compare Zimbabwe with the economies that transitioned from central planning following the collapse of the Soviet bloc in 19892001. There are some things common to them, but there are many other features that put them aside from Zimbabwe. The most common theme is that both crises were humanmade, i.e., they were the result of wrong institutions and bad economic policies. By the same token, nothing short of a complete overhaul of the economic regime could revive the economy. But the similarities end there as reformers in the Soviet bloc recognized that the whole framework of economic policy making had to be overhauled while Zimbabwean policy makers appear to be unable to recognize the contribution of their economic policies and institutions to the collapse of national economy. While radical reformers in Central Europe quickly overhauled excessive regulations and liberalized foreign trade and foreign investment regimes, Zimbabwe is yet to recognize that economic revival calls for similarly deep and bold measures. Policies, rather than external and infrastructural constraints, remain responsible for Zimbabwe's performance which remains well below its potential. Like Central European transition economies, Zimbabwe borders a much higher developed economy, South Africa. Although landlocked, Zimbabwe does not suffer from--to use Paul Collier's phrase (2007, p. 5)--the "trap of being landlocked with bad neighbors." Both Botswana and South Africa have superb quality of economic governance by world standards and rate highly in trade logistics. While Mozambique and Zambia rank lower than them, neither of these countries has a dysfunctional government.1 Not unlike Central European transition economies, Zimbabwe also has high quality human capital relative to other countries in the region. Its

According to the World Bank's survey of the cost of doing business in 2010, Zimbabwe ranked third in the world in terms of the highest cost of importing and exporting after Chad and Central African Republic. But traders from Zambia, Zimbabwe's northeastern neighbor, paid 35 percent less for imports and 19 percent less for exports per container. This clearly cannot be explained by geography alone.

1

10 | P a g e workers are praised in neighboring countries for their work ethics and the country does not seem to suffer from a shortage of entrepreneurial talent. Thus, broadly conceived quality of economic governance remains Zimbabwe's biggest problem choking off foreign trade and economic growth. It rather dramatically deteriorated in 19972008 and the incidence of corruption has been on the rise since 199798. Zimbabwe's fall to the bottom of various international rankings In the 2000s points to a very unfriendly business and investment climate. Indeed, in nominal terms, FDI inflows peaked in 1996 at US$52 million, fell to US$28 million in 1997 and subsequently never exceeded US$10 million averaging barely 0.06 percent of the GDP.2 Domestic investments were also very low. Zimbabwe suffers from the trap of bad governance. While in the presence of geographical barriers, poor infrastructure and a shortage of entrepreneurial skills, improvements in institutions and policies can do little to put a country on a path of sustainable economic growth. But these barriers are not particularly binding for Zimbabwe. With literacy rates at over 90 percent Zimbabwe has a population better educated than the average in Africa. Moreover, climate, fertile soil and ample endowments in natural resources, all create huge opportunities for export activity. Yet, despite its many assets, Zimbabwe has failed in generating economic growth throughout most of the period covered by this study, i.e., 19942008.3 The last decade or so was largely a lost decade in terms of economic development. Zimbabwe got stuck in a lowincome trap suffering from pervasive corruption, gross economic mismanagement and political instabilities. Financing the budget deficits through printing money led to hyperinflation that would reach 231,000,000 percent in July 2008 until the local currency was abandoned as "...the printing press was not able keep pace with the expansion of local currency" (IMF 2009: p. 7). Cash shortages and a significant divergence between the parallel cash and electronic transaction exchange rates led to the official adoption of hard currencies for transactions in early 2009. Faced with shortages of foreign currency triggered by poorly designed economic policies (e.g., allocation of foreign exchange at subsidized exchange rates to parastatals and poor farmers), the authorities introduced measures (surrender requirements on exports proceeds, the retention of foreign exchange earnings above mandatory surrender levels, and the confiscation of foreign currency deposits) further suppressing foreign trade activities. Furthermore, the government's program of seizure of white owned farms by squatters and forced landacquisition program expelling around 3,000 white farmers in 200002 dealt a blow to the agricultural sector, traditionally, a backbone of the Zimbabwean economy.4 B. Observations derived from empirical analysis of Zimbabwe's foreign trade Zimbabwe provides a textbook illustration of the negative impact of economic mismanagement and political instabilities on foreign trade performance. An empirical investigation of developments in Zimbabwe's foreign trade performance leads to the following major observations. First, paradoxically foreign trade has turned out to be more resilient to the folly of economic policies than overall economic

2 3

Derived from data in UNCTAD `s FDI database. The availability of foreign trade data from external sources has determined the time coverage of this analysis, as the reliability of Zimbabwe's statistics leaves much to be desired. 4 For a detailed timeline of critical events in the history of Zimbabwe, see British Broadcasting Corporation's website at http://news.bbc.co.uk/2/hi/africa/country_profiles/1831470.stm (accessed on May 31, 2010).

11 | P a g e activities as captured by the GDP. The openness of Zimbabwe's economy significantly increased in 1998 2008 as GDP per capita was falling faster than foreign trade. Both exports and imports per capita in current prices recorded a modest positive growth during this period. These developments produced an intersection of two linear trends of falling GDP and rising share of foreign trade in the GDP crossing at around 200405. In all, Zimbabwe's export performance over the last decade has been lackluster, albeit quite impressive when cast against its dismal economic growth performance. Second, viewed in a longer perspective of the last two decades, one may distinguish three phases in Zimbabwe's export performance: phase of growth in 199097; phase of contraction in 19982002; and the phase of moderate recovery in 200307, although only in terms of the value of exports, which came to a halt in 2008 when exports fell 8 percent. External factors were decisive: 200203 marked the beginning of the fiveyear commodity boom notable for its exceptional duration, the number of commodities involved, and the heights that prices reached reflected the resilience of developing country growth during this period (WB 2009). Zimbabwe has failed to take advantage of opportunities offered by the boom. Its exports in 2008 would have been around 30 percent higher had it exported the same amounts as in 2000. It also underperformed visàvis its neighbors, as its share in total exports of neighboring SADC countries fell from 5 percent in 200002 to 3 percent in 2005 and further to 2.2 percent in 2008. Third, the differences in exports growth performance between the 199097 and 200307 phases raise concerns over the depletion of levers of growth in the future. Zimbabwe's exports have become less diversified and increasingly concentrated on products with low level of processing and high natural resource content. This does not augur well for the future export growth performance unless some measures encouraging the development of the internationally competitive private sector are taken. Fourth, there are strong indications that the industrial base supporting exports has significantly shrunk. Consider the following: the share of capital equipment in total imports fell from around one third in 199098 to around onefifth in the 2000s; the degree of processing embodied in Zimbabwe's exports fell in the 2000s as exports shifted towards industrial raw materials and resource intensive products; so did the technological content marked by a dramatic shift from low technology labor intensive products to natural resourcebased products; export offer has become less diversified with the fall of the number of products with exports exceeding US$500 thousand; and net exports, as measured by the difference between exports of a product and its imports, have been rapidly falling. Fifth, Zimbabwe's surplus in trade in natural resource intensive products has fallen rather dramatically mainly due to the combination of falling exports of foods and feeds and their rising imports. While exports coverage of imports was on average 510 percent in 199497, it fell to 224 percent in 200207 and 164 percent in 2008. Put differently, Zimbabwe exported around five times more natural resource intensive products in 199497 and more than twice as it imported in 2008. Surplus in trade of agricultural foods and feeds fell from US$1.1 billion in 1997 to US$ 99 million in 2008. Excluding feeds, Zimbabwe has moved from a net exporter to net importer of food. But its agroprocessing business appears to have survived the economic chaos decade, although its base has been diminished. Sixth, the rise to prominence of natural resources in Zimbabwe's exports has not been the result of different rates of expansion of products with different endowments in factors of production. Except for

12 | P a g e natural resource intensive products, exports of human capital intensive products have been stagnant and those of unskilledlabor intensive products have been falling. Both developments are particularly worrisome considering that Zimbabwe has a large pool of cheap skilled and unskilled labor. Paradoxically, exports of unskilled labor intensive products experienced the second largest decline: its negative LSG (least square growth) rate of 4.3 percent was larger than the pace of decline of capital intensive exports of 4.0 percent but lower than that of skilled labor intensive products. However, exports of skilled labor intensive recorded the largest decline in 2008. Their share in total exports fell from an average of 6.4 percent in 200207 to 3.2 percent in 2008. Seventh, a great shift in geography of Zimbabwe's exports from the EU to South Africa has taken place since the beginning of this century. The shift in geographical patterns of Zimbabwe was not triggered by expanding exports to its major external markets. Neither was it triggered by the surge in exports to all of its neighbors as exports increased only to South Africa. The EU was taking almost half of Zimbabwe's total exports in 19942000 and accounted on average for 24 percent of Zimbabwe's imports during this period. By 2008, these shares fell to 21 percent for exports and 7 percent for imports. Hence, the change in direction amounted to redirection of exports from the EU to other countries, mainly to neighboring South Africa. This was not the result of the emergence of new export capacity, as exports in real terms were stagnant, if not falling, but a simple redirection as the presence of Zimbabwean exporters in EU markets fell rather precipitously. So did it in other than South Africa regional markets, albeit to a smaller extent. Last but not least, contrary to what one might expect considering proximity of South African markets, Zimbabwe's South Africanoriented export basket has converged towards that of the EUoriented one. In consequence, Zimbabwean total exports have become less diversified and increasingly concentrated on a few minerals. While the traditional measures of concentration do not necessarily reveal it, as Zimbabwean exports have always been dominated by a few large items--tobacco, nickel, the difference is that the number of exported items above US$500,000 has significantly declined indicating the disappearance of small and medium enterprises in the export sector. C. Structure of the report The remainder of this report is organized as follows. The first chapter asks about links between the quality of economic governance, overall economic performance and foreign trade performance in a regional perspective. Chapter 2 provides a bird'seyeview of Zimbabwe's foreign trade performance: it ponders over the paradox of expanding economic openness, as traditionally defined by the percentage of foreign trade in GDP, in spite of a disintegrating economy; it discusses developments in trade in services; identifies three phases in Zimbabwe's export development between 19942008; and, assesses the extent to which the last phase of moderate growth was driven by priceincreases; it compares Zimbabwe's trade performance with that of its neighbors; and, last but not least, it seeks to identify changes in Zimbabwean geographical patterns of trade. Chapter 3 turns to an examination of change in Zimbabwe's exports in terms of patterns of specialization, factor intensities, degree of processing and technology content. It also identifies losers and survivors of hyperinflation in 200208 and seeks to assess changes in revealed comparative advantage of Zimbabwean exports. The last chapter tries to sketch prospects for a revival in exports: it argues that some traditional exports are still viable and those

13 | P a g e that did survive the vagaries of economic policies in the 2000s could be expanded. But this will not come by default, as serious policyinduced barriers to reviving exports are still in place. These barriers go well beyond trade policy: they stem from bad policies and a weak institutional framework. Without addressing these key issues, prospects for exportled growth are slim.

14 | P a g e

1. Zimbabwe's lost decade: economic governance, macroeconomic disequilibria and other selfinflicted wounds

Neither geography nor shortages of entrepreneurial talent and, more generally, high quality human capital put up insurmountable barriers to Zimbabwean economic growth based on integration into regional and global markets. Zimbabwe is amply endowed in natural resources and skilled labor, its moderate climate is friendly towards agriculture, and, although it is landlocked, it has easy access to seaports and cheap maritime transport. The question addressed in this section therefore concerns the quality of economic governance. Leaving aside macroeconomic stability as a necessary condition for staying on a path of economic growth, the quality of economic governance defined here broadly as a set of institutions, regulations and policies assuring political stability, protection of property rights and accountability of those in power. Good governance is critical to employment and economic expansion because it is private enterprise that creates wealth and funds government, not the other way around. 1.1. Quality of economic governance in 19962009 in a comparative perspective In order to assess Zimbabwe's quality of governance, an equivalent of government directive capacity, we use the results of the World Bank's "Governance" surveys systematically conducted since 1996, albeit not on an annual basis.5 The international survey covering more than two hundred countries across the world gives numerical estimates ranging between () 2.5 (worst case) and (+) 2.5 (best case) together with ranking in percentile6 for the following six dimensions of governance: voice and accountability, political stability, government effectiveness, regulatory quality, rule of law and control of corruption. We use the average of estimates of political stability, government effectiveness and regulatory quality as an approximation of the quality of economic governance--hereafter referred to as a single aggregate index of economic governance (SAG)--for two reasons. First, these indicators are critical dimensions of business climate. Political stability impacts investment decisions and in extreme situations of its absence, i.e., violence, it may completely disrupt economic activities. The inclusion of political stability as a component of economic governance can be justified on the grounds of its influence on investment decisions: political instabilities suppress private investment activities even more than macroeconomic instabilities. More generally, democratization can be viewed as an investment in social capital improving governance and economic growth performance. The quality of regulation is of little relevance unless supported by a government having the capacity to enforce regulations, i.e., government effectiveness. Second, three other indicators pertinent to such dimensions of governance as the rule of law, control of corruption, and voice and accountability are not taken into account as they do not yield extra information. They are strongly correlated with the selected three indicators, with the values of correlation coefficients equal or above 0.97. Table 1 presents values of the SAG index for selected SADC countries normalized for each dimension of governance so that the best performer is 100 and the worst performer is zero. Thus, the scores for other countries will show the distance separating them from the best and the worst performer in terms of the quality of governance. In general, three groups of countries can be identified: decent performers with

For an explanation of methodology, see Kaufmann D., A. Kraay, and M. Mastruzzi (2007). Percentile rank gives the percentage of countries worldwide (above 200 but their number has varied) that rate below the selected country (subject to margin of error) with higher values indicating better governance ratings.

6 5

15 | P a g e the value of SAG above 60 percent; average performers with the values of SAG between 40 percent and 60 percent; and dismal performers with the value of SAG below 40 percent. South Africa (since 2000), Botswana, Mauritius, and Namibia (since 2004), have decent levels of economic governance. The remaining countries were within world median except for Zimbabwe, which slipped from this group beginning in 2000. Zimbabwe's quality of governance was already below the region's average in 1996, but it did not stand out as a pariah.7 Subsequently, there was a strikingly deep regress with Zimbabwe replacing Rwanda as the worst performer among comparator countries (Table 1). The good news for Zimbabwe is that its immediate neighbors--listed in Table 1, above the row with the data for Zimbabwe--have rather good quality economic regimes: put differently, Zimbabwe--in contrast to a number of other African landlocked countries--is fortunate in not being surrounded by `bad neighbors.'

Table 1: The values of aggregate quality of governance (SAG) indices for Zimbabwe and selected SADC countries in 1996, 1998, and 200209

Botswana Mozambique South Africa Zambia Zimbabwe Malawi Mauritius Namibia Tanzania Memorandum: Rwanda

1996 1998 2000 2002 2003 2004 2005 2006 2007 2008 2009 71.0 45.3 49.9 50.7 44.5 46.8 65.7 66.4 50.5 12.0 73.4 54.1 57.8 49.3 42.2 53.2 71.1 64.7 49.8 24.6 73.7 53.4 62.0 46.3 29.0 49.2 70.4 58.6 48.7 30.3 71.4 50.0 61.8 41.5 19.7 45.4 68.9 59.7 46.4 28.6 72.8 47.6 62.7 43.3 20.0 46.1 71.3 59.9 44.2 34.5 70.8 45.8 64.6 44.8 20.4 46.0 69.4 60.4 45.1 38.0 71.8 47.4 64.8 42.5 15.1 45.8 69.2 59.6 45.5 34.5 69.0 50.8 65.5 46.9 21.5 43.9 69.7 62.1 47.9 43.0 69.5 50.3 65.1 48.2 18.3 47.5 71.0 62.8 48.4 45.7 70.4 51.2 64.3 49.7 15.7 45.7 73.7 66.1 48.2 47.9 70.2 53.1 61.5 49.9 14.4 46.2 70.6 62.4 49.2 48.2

Source: derived from data discussed in D. Kaufmann, A. Kraay, and M. Mastruzzi, 2009. Governance Matters VIII: Governance Indicators for 1996-2008 and available at www.worldbank.org/wbi/governance

The ground lost against other comparator countries has been outstanding. While Rwanda moved from the value of SAG of seven percent in 1996 to a very respectable 48 percent in 200809, Zimbabwe moved in the opposite direction with the value of SAG falling from 45 percent to 14 percent over the same time span. In order to appreciate how low this score is for Zimbabwe, bear in mind that Rwanda was in 1996 only two years after the civil war ended and three years before its first local elections.8 In order to appreciate the distance lost, compare Zimbabwe with Its Eastern neighbor, Mozambique, which scored on a par with Zimbabwe in 1996 and stayed in an average group in the 2000s. As it can be seen in Figure 1, it stayed the course making significant strides in building policy and institutional foundations of competitive markets.

Zimbabwe scored also well below the countries listed in Table in annual rankings compiled under the "Index of African Governance" (earlier, in 200708 named the Ibrahim Index of African Governance). For 2009 ranking based on data for 2008, consult the website http://www.worldpeacefoundation.org/africangovernance.html. For methodology and theory, see Rotberg and Gisselquist (2009). 8 The Tutsi rebels defeated the Hutu regime and ended the killing in July 1994. Domestic political stability has been solid, although under threat of several thousand Hutu insurgents operating out of the neighboring Democratic Republic of the Congo. Rwanda held its first local elections in 1999 and its first postgenocide presidential and legislative elections in 2003.

7

16 | P a g e

Figure 1: Moving in different directions: Zimbabwe' quality of economic governance against that of Mozambique and Rwanda in 1996, 1998, 2000, 200209 60 50 40 Mozambique 30 20 10 0 1996 1998 2000 2002 2003 2004 2005 2006 2007 2008 2009 Rwanda Zimbabwe

Source: As in Table 1.

A closer look at the constituent dimensions of the SAG for Zimbabwe indicates that regulatory quality has consistently negatively contributed to the value of the SAG as it has been lagging behind other dimensions of governance--political stability and government effectiveness (Table 2). By 2006, the government has become dysfunctional: its ability to implement measures has all but disappeared while the quality of regulation has even further deteriorated. The latter, however, no longer mattered as the government lacked the capacity to implement regulations. Excluding the value of regulatory quality dimension, the average for other two dimensions would be 8 percent in 2008. Hence, although political stability was maintained, the state became largely dysfunctional. And so did the economy with hyperinflation and steeply falling aggregate output (see section 1.2).

Table 2: Dimensions of economic governance quality in Zimbabwe in1996, 1998, 2000, 200209

Political Stability Government Effectiveness Regulatory Quality Source: As in Table 1. 1996 47.5 43.8 42.1 1998 43.3 40.3 42.9 2000 2002 29.9 31.5 25.5 23.0 31.1 5.1 2003 28.2 29.5 2.2 2004 27.4 30.7 3.3 2005 25.4 22.0 2.0 2006 38.4 22.7 3.3 2007 32.6 22.5 1.9 2008 30.5 14.8 1.8 2009 29.3 13.9 2.0

Deterioration in governance quality has moved in tandem with creating new opportunities for rent seeking. The incidence of corruption, as measured by Transparency International's CPI (corruption perception index) has also increased.9 By 2007 Zimbabwe became the most corrupt country amongst comparators with the value of CPI falling from quite a respectable score of 4.2 in 1998 to 1.8 in 2008 as compared with the average of four for the region (Annex Table 1). As the IMF (2009) report noted: "A further deterioration in the business climate exacerbated the economic decline in 2008. A tightening of price controls and exchange restrictions, a pickup in land invasions, the confiscation of foreign currency deposits, and frequent changes in business regulations made it more difficult to conduct business in Zimbabwe." While other countries made progress in making conditions of doing business more

The corruption perception index (CPI) assumes values from one (maximum incidence of corruption) to ten (no corruption)

9

17 | P a g e transparent and kept the incidence of corruption at unchanged levels, Zimbabwe moved in the opposite direction. But in 2009, the value of CPI moved up to a still very low score of 2.2: yet, this is the first time that an improvement was noted and Zimbabwe's score in terms of the average for comparator countries raised from 45 percent in 2008 to 59 percent. A weakened state and collapsing regulatory quality, also captured in the World Bank's Doing Business annual surveys,10 have raised transaction costs of conducting business activity and undercut the competitiveness of Zimbabwe's economy in global markets. According to the 2009 Africa Competitiveness Report (ADB 2009) and Global Competitiveness reports (WEFA 2009), Zimbabwe was ranked as one of the countries with the least competitive economy: it ranked 133rd on the overall Global Competitiveness Index out of the 134 countries' surveyed, scoring 2.9 points out of a possible 10. It was ranked higher than only one country worldwide, Chad, also in earlier surveys 200709 (WEF 2009). The major reason was that since around 2000, the growing gap between government expenditures and revenue was covered increasingly by printing money and implementing other businessunfriendly measures including expropriation of hard currency profits and administrative rationing of access to foreign currencies. Although Zimbabwe has a very long way to go in order to establish policies and institutions capable of reversing the trend of economic decline, political and economic changes that took place in 200809 have already improved the quality of economic governance, which is yet to be captured in most international assessments. Except for the Transparency International ranking, which duly registered improvement in its measure of Corruption Perception Index, the verdict of other organizations on Zimbabwe has remained unchanged. Thus, it is rather surprising to note that two major events, i.e., signing on September 15, 2008 of the so called Global Political Agreement (GPA), between the Zimbabwe African National UnionPatriotic Front (ZANUPF) and the two formations of the Movement for Democratic Change (MDC), on resolving the challenges facing Zimbabwe, and an agreement adopted in April 2009, to abandon printing of the Zimbabwean dollar, and replace it with the US dollar and the South African Rand did not have any perceptible, positive impact on Zimbabwe's rankings in 2009,11 although they, taken together, have significantly altered the political and economic landscape in Zimbabwe. While the GPA has not lived up to the task and despite complaints about a political stalemate, the current arrangement negotiated under the GPA is clearly superior to the one existing prior to it. It is also not clear why dumping of the Zimbabwean dollar was not enough to boost international assessments of the quality of economic governance. After all, this has ended hyperinflation and has helped remove some restrictive regulations directly linked with administrative attempts to contain hyperinflation (e.g., price controls, surrender requirements). But, without some degree of macroeconomic and financial stability, the economy cannot grow: it usually contracts. Hence, the introduction of a stable currency has been a huge improvement. Both events have clearly improved Zimbabwe's economic regime and contributed to political stability. Yet, there are still reasons for concern. In particular, except for the removal of measures that became redundant once local debased currency was dropped, not a single measure has been taken to address

10 11

For information, see the World Bank's site at http://www.doingbusiness.org/. To the contrary, the value of SAG declined in 2009.

18 | P a g e other weaknesses of Zimbabwe's investment and business climate. Macroeconomic and financial stability are a necessary condition but not a sufficient one to ensure that current and potential investors can retain not only the returns to their investments but also investments themselves, as the rule of law remains capricious. Together with incentives, these are critical conditions to attract investments, both domestic and foreign, that remain unmet so high levels of static and dynamic inefficiency in the allocation of resources persist in the economy. 1.2. Economy in a frenzied free fall in 19992008: implications for foreign trade Zimbabwe's economy, which was in a tail spin for almost a decade, has two dubious world records of the early 21st century: highest inflation, or hyperinflation, and deepest contraction in aggregate output experienced during peacetime. Some observers note, however, that the income loss in 19942004 alone exceeded even the income losses incurred by Côte d'Ivoire, Democratic Republic of Congo, and Sierra Leone during their respective conflicts (Clemens and Moss, 2005). In 199698, Zimbabwe was the fastest growing economy in Africa; in 19992008, it became the fastest shrinking economy in the world with an average contraction of the real GDP of 6 percent per year. The trough appears to have been reached in 2008 with 15 percent fall in real GDP as compared to 2007, on top of 48 percent cumulative decline in 19992007. The GDP in constant 2000 prices fell from US$8.3 billion in 1998 to US$4.3 billion in 2008. In 2009, real GDP increased 6 percent for the first time since 1998 and is projected to be 5.4 percent in 2010 (Figure 2).

Figure 2: Real GDP growth and contraction (in constant 2000 prices) and annual rates of inflation in 19952008 (in millions of US dollars and percent) 9,000 8,500 8,000 7,500 7,000 6,500 6,000 5,500 5,000 4,500 4,000 432 282 302 140 23 21 19 32 59 56 77 500 400 300 200 100 0 7.7

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 (p)

Inflation, consumer prices (annual %) GDP (constant 2000 US$)

100

Notes: Right axis--the values of CPI in percent; Left axis--GDP in millions of 2000 US dollars Sources: World Bank WDI database, IMF Article IV Consultation Staff Reports, 2009 & 2011, and The 2010 MidYear Fiscal Policy Review, Ministry of Finance, Harare, 14 July 2010.

There were two intertwined channels through which Zimbabwe's exports performance was negatively affected: escalating macroeconomic instabilities combined with the overvalued official exchange rate; and the dramatic contraction in output triggered the government's program of "indigenization of

the economy" including the illdesigned land reform. While other countries suppressed inflation to

single digit levels in the early 2000s, Zimbabwe was moving in opposite direction: after Zimbabwe's

19 | P a g e confiscation of whiteowned farmland and its repudiation of debts to the International Monetary Fund, hyperinflation set in and lasted to 2009. Inflation exceeded 50 percent already by 1999 and continued its march upwards to reach 432 percent in 2003 (see Figure 2). Inflation, as measured by CPI (Consumer Price Index), slightly fell in 200405, albeit it was still very high with threedigit annual growth rates. Subsequently, however, it rebounded exceeding one thousand percent in 2006 and swiftly moving to stratospheric levels in 200708 to peak in September 2008 "... at about 500 billion percent" (IMF 2009, p. 5). The decision, adopted in April 2009, to abandon printing of the Zimbabwean dollar, and instead to use hard currencies (especially the US dollar and the South African Rand), with the Rand as the reference currency, has put an end to hyperinflation and CPI, with the value of minus 7.7 percent, moved to a negative territory in 2009 for the first time since independence in 1980 (see Figure 2). While hyperinflation alone would have undercut both current economic activity and investments, as the experience of a number of South American economies in the 1980s has demonstrated, Zimbabwe's misery was aggravated by two government's interventions: one typical for a number of countries undergoing hyperinflation and the second unique to Zimbabwe, i.e., the government's program of "indigenization of the economy." The first intervention was the usual government temptation to `solve' hyperinflation by outlawing it and instead dealing with its symptoms, rather than with the root of the problem, i.e. deficits financed by printing money. The government introduced price and interest rate controls, exchange controls, state monopsonies (e.g., Grain Marketing Board) and foreign currency surrender requirements on exporters. The latter amounted to the confiscation of a portion of revenue from exports. These measures contributed to further distortions in the economy; increasing shortages of daily necessities; and, predictably, the emergence of a black market and profiteering. Furthermore, since controlled prices were set at a fraction of market prices, there was a shift to produce goods not subject to central price controls further distorting the allocation of resources. Another consequence of growing macroeconomic instability was that convertibility of the Zimbabwean dollar for current account transactions became increasingly limited. The domestic currency became increasingly overvalued in 200003, which prompted the government to establish special regimes for tobacco and gold exporters, while the parallel market premiums exploded (Muñoz, 2006). Black markets emerged where the US dollar would be traded at exchange rates several times higher than the official rate.12 Already in 2002 (the end of third quarter), US dollar was exchanged at ZWE$740 in parallel markets while its official exchange rate was ZWE$55 (Sandawana 2007). Subsequently, the gap between two rates expanded rapidly. Exporters were of course negatively affected. The overvaluation of the official exchange combined with extremely lax fiscal and monetary policies undercut competitiveness of Zimbabwean exports. The growing of the parallel market premium led to smuggling at the expense of legal exports. Muñoz (2006) empirically shows a strong negative relationship found between the parallel market rate depreciation and the value of legal exports in Zimbabwe in 200004. Furthermore, with the price of imports rising and growing difficulties in accessing foreign currency, producers dependent on imported inputs would often

Various ad hoc measures, such as the introduction of a managed foreign exchange tender system early in 2004 or the gradual relaxation of the surrender requirements have failed to prevent the apreciation of the official exchange rate, simply because "... the demand for foreign exchange continued to pick up" (Muñoz, 2006 p. 4).

12

20 | P a g e go out of business if no domestically produced substitutes were available. Those who did not increasingly turned to customers abroad than sell their products domestically. While strict foreign currencysurrender requirements, often amounting to confiscation of up to half of foreign currency earnings by the government, was clearly a disincentive to export,13 the rewards of having access to hard currencies and escaping domestic inflation offset these losses for some products. On balance, one might have expected some diversion of sales away from the domestic to foreign markets during this time and, therefore, a lower contraction of exports than that of aggregate output. The disastrous macroeconomic environment combined with political instabilities and the government's program of "indigenization of the economy" was bound to erode one, if not all of three pillars that historically shaped Zimbabwe's external performance, i.e., tourism, agriculture including agro processing and other simple manufactures, and mining. Tourism is always the first victim of political instability and violence and the first to recover once the situation warrants the removal of travel warnings by foreign governments. In fact, by 2006, the current value of revenue from tourism exceeded the peak value reached in 1996. In mining, given the size of `sunk capital' this sector was less affected by macroinstabilities than agricultural production. Nevertheless, extraction sectors have also not flourished: to the contrary, exports of most minerals fell in terms of volume in the 2000s. Except for diamonds and platinum group metals, which significantly expanded both production and exports in the 2000s, exports of other minerals in terms of volume stood well beyond their peak levels in 200001. Thanks to sharp increases in prices of gold and copper, the value of their respective exports exceeded previous record levels in the 2000s set for gold in 2006 and for copper in 2000, despite lower volumes exported (Table 3).

Table 3: Exports of selected minerals in terms of volume and value in percent of their peak performance and share in total exports in 2010 (in percent) Volume, Index Values, Index Share in total exports peak peak Minerals 2010 year=100 2010 year=100 In 2010 Platinum Group Metals (`000 ounces) Gold (`000 ounces) Diamonds (`000 carats) Nickel (tons) Copper (tons) Agricultural products 63.6 30.7 115.4 33.3 95.5 2008 2001 2009 2000 2000 76.8 106.3 89.3 16.5 268.8 67.3 126.8 5.2 24.2 2008 2004 2005 2007 2000 2001 2000 2000 2002 18.9 14.5 2.0 1.9 1.1 20.4 3.4 0.1 1.1

Fluecured tobacco (tons) 38.5 2001 Sugar raw (tons) 80.0 2000 Sugar refined (tons) 4.7 2002 Flowers (tons) 23.9 2002 Source: Derived from data provided by the Reserve Bank of Zimbabwe

The agricultural sector--the backbone of Zimbabwe's economy--witnessed huge drops in output as the prime victim of Zimbabwe's chaotic land reform program, albeit not the only one. In 2004, the maize production stood at around onethird of its level in 1999; wheat production at around 8 percent; and

The surrender requirement is a tax on exports, as exporters are obliged to convert a portion (50% in 2003) of foreign currency earnings into domestic currency at official exchange rates several times below black market exchange rates. Since every dollar converted loses almost all of its value, exporters have incentive keep money abroad further aggravating the severe shortage of foreign currency.

13

21 | P a g e tobacco production at around onefourth (Clemens and Moss 2005). Agricultural output fell below the levels needed to feed the local population and Zimbabwe ceased to be one of the largest exporters of foods in Africa and now relies on food aid. So did the volumes of agricultural exports fall and to a much larger extent than those of minerals: except for raw sugar whose exports stood at 80 percent of their volume in 2000 but with refined sugar exports at 5 percent of their level in 2002, exports of tobacco was at 38 percent of their volume in its peak year 2001 and cut flowers at 24 percent of its peak in 2002. The values of their exports were also well below their respective peaks in 200002 (Table 3). A decade long mismanagement of the economy combined with violations of private property rights and ad hoc confiscations of private property by the government was also bound to extract a heavy toll on manufacturing. New production capacities did not emerge in the 2000s and some existing industrial capacities, deprived of investment, disappeared. Furthermore, since the farm sector supplied about 60 percent of the inputs to the manufacturing base (Richardson, 2005), its collapse led also to the breakdown of manufacturing output. In contrast to most other African countries, Zimbabwe had a sophisticated manufacturing base. In the 1990s, Zimbabwe was one of four African countries with the share of manufactures (mainly textiles, cement, chemicals, wood products, and steel) in total exports exceeding 20 percent: the other countries were Kenya, Mauritius, and South Africa (Wood and Jordan, 2000). Although exports of manufactured goods continued contributing more than 20 percent to the total in 200008, the value of exports of manufactured goods of US$556 million in 2008 was US$10 million lower or 2 percent less than in 1997.14 Moreover, this relatively high share was mainly the result of a slight increase in the value of exports of steel and iron products. Consumer goods including textiles and clothing were particularly negatively affected: their exports in current prices stood in 2008 at 58 percent of their peak level in 1996. Hence, hyperinflation combined with a heavy dose of state micromanagement of most economic activities including foreign trade was bound to bring about the retrenchment in exports and, consequently, imports. Yet, the foreign trade sector had displayed quite remarkable resilience and its contraction, although rather dramatic, strikes one as less steep than one might have anticipated given the conditions of doing business in Zimbabwe in 200009. The second action that negatively impacted investment in Zimbabwe was the government's program of "indigenization of the economy," launched in 2000. Although initially its main target was confiscation of around 4,000 largescale commercial farms--accounting for 70 percent of Zimbabwe's arable land and owned exclusively by white Zimbabweans--under the socalled "fasttrack land reform" announced in February of 2000,15 the program was subsequently expanded to foreign investment with the adoption of the Indigenization and Economic Empowerment Bill of 2007. But already in 2000, the investors in other

Exports of manufactured goods, defined as SITC 5 through 9 minus 68, are based on imports from Zimbabwe reported by its trading partners to the UN COMTRADE database. 15 As Clemens and Moss (2005) note: "Most Zimbabweans (including white farmers) say that land reform was both necessary and inevitable. The tragedy of Mugabe's approach is that it has harmed those whom a wellordered, selective redistribution program could and should have helped. Generally the farms have not been given to black farm managers or farm workers." Moreover, although the official goal was to divide the farms into hundreds of thousands of small plots for traditional black farmers, most plots ended up in the hands of Mugabe's political supporters and government officials (Richardson, 2005).

14

22 | P a g e sectors of the economy felt also threatened, as many whiteowned urban businesses and offices of foreign corporations were plundered (Power, 2003). The government's program of "indigenization" that requires that 51 percent equity in firms worth over US$500,000 be transferred to black Zimbabweans has remained a huge barrier to FDI inflows.16 Despite a dramatic fall in FDI inflows in 19992008 they failed to rebound with the improvements in domestic politics following the signing of the Global Political Agreement in September 2009 (Figure 2). The increase in FDI inflows from US$52 million in 2008 to US$105 million in 2009 not only falls well short of the investment needed to jumpstart the economy but FDI inflows are projected to fall in 2010 to US$85 million.17

Figure 2: FDI in Zimbabwe and Zambia in 19942009 (in millions of US dollars)

1,400 1,200 1,000 800 600 400 200 0 Zimbabwe Zambia

Sources: UNCTAD WIR database.

Two measures of Zimbabwe's investment `deficit' illustrate the depth of undercapitalization of the Zimbabwean economy as a result of economic chaos in 200008. The first derives from benchmarking FDI flows to Zimbabwe against those to Zambia. The second relates to imports of machinery in a historical perspective and also in comparison to Zambia. For the first, Zambia has actively sought FDI to supply capital and know how, whereas Zimbabwe's policies have actively discouraged foreign investment. While in 199499 flows to both countries were of similar magnitude, except in 1998, when they peaked in Zimbabwe at US$444 million against Zambia's US$238 million, in 200009 FDI into Zambia continued its upward movement while those into Zimbabwe almost completely dried up. Total FDI inflows of US$878 million over 199499 to Zimbabwe were larger than those into Zambia of US$709 million. But this was reversed, subsequently, as total FDI invested in Zambia in 200408 amounted to US$3.6 billion, whereas FDI in Zimbabwe was just US$323 million over the same period.

The indigenization program is still in place encouraging transfers of profits abroad from Zimbabwean firms not owned by "disadvantaged" ZWE citizens and discouraging foreign investments. It also continues in the agricultural sector: the Harare newspaper The Standard reported taking over by the government of the Tuli cattle breeding farm ("Worldfamous Tuli cattle breeders thrown off farm," The Standard, August 15 to 21, p. 89). 17 Based on balanceofpayments statistics of the Reserve Bank of Zambia, Harare, August 2010.

16

23 | P a g e While very low inflows of FDI deprived domestic industries of access to modern technologies and, by the same token, kept labor productivity from growing, there are indications that the manufacturing base was not completely erased. First, while the economic climate in the 2000s was hostile to investment activities, they did not appear to come to a complete halt. In the absence of domestic data, the best information shedding light on investment activities can be derived from statistics on imports of capital equipment.18 The data shows a dramatic decline in the value of these imports, which peaked in 1996 at US$667 million and fell 70 percent to US$217 million in 2002. Yet, despite inflation followed by hyperinflation, these imports began recovering in 200307 and rose in 2007 to 80 percent of the 1996 peak level (Figure 3A).

Figure 3: Two different tales: Imports of machinery and equipment (index 1996=100) and FDI into Zimbabwe and Zambia in 19942008 (index 1997=100)

A. Zimbabwe

100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

B. Zambia

700% 600% 500% 400% 300% 200% 100% 0% 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

FDI inflows (1997=100) Imports of machinery (1996=100)

FDI inflows (1997=100) Imports of machinery (1996=100)

Source: Own calculations based on partners' trade data reported to the UN COMTRADE database and FDI data from UNCTAD WIR database.

Zimbabwe's imports of capital equipment have therefore fallen much less than FDI inflows and stood in 2008 at 75 percent of their level in 1996. However, Figures 3A and 3B present another stark contrast by benchmarking Zimbabwe's imports of capital against Zambia's. They show that while machinery imports by Zambia were almost six times higher than their value in 1996, in Zimbabwe, over 19952008, they were almost identical (Table 4). Neither the current imports of machinery nor FDI inflows are therefore likely to elevate Zimbabwe's exports to bring them in line with the country's potential. The comparison of machinery imports and FDI into Zimbabwe with the flows into its northern neighbor, Zambia, gives some idea as to the magnitude of flows that might be necessary to turn around Zimbabwe's economy. Consider the following: the value of machinery imported into Zimbabwe in 199598 was almost three times higher than into Zambia and

As elsewhere, these are based on exports to Zimbabwe reported by other countries to the UN COMTRADE database.

18

24 | P a g e total FDI flows over this period were 8 percent higher (Table 4). While total imports of machinery in 199903 exceeded those into Zambia by 13 percent, total FDI inflows into Zimbabwe amounted to seven percent of foreign capital invested in Zambia. The discrepancy further increased in 200408: Zambia's total machinery imports were more than twice as large as Zimbabwe's and total FDI into Zambia over this period was more than ten times larger (Table 4).

Table 4: Imports of capital equipment and FDI flows to Zimbabwe and Zambia in 19952008 (in millions of US dollars) Total 2007 2008 199598 Machinery, excluding automobiles and parts (US$ millions) Zimbabwe Zambia Ratio of Zimbabwe to Zambia FDI inflows (in US$ millions) Zimbabwe Zambia Ratio of Zimbabwe to Zambia 533 1,333 0.40 52 1,324 0.04 510 1,166 0.44 60 939 0.06 2,592 925 2.80 Total 199903 1,463 1,292 1.13 Total 200408 2,140 4,537 0.47 323 3,599 0.09 Average 200408 22 27 0.82 25 25 0.99 6,195 6,754 0.92 1,108 5,198 0.21 Average 199508 Total 199508

719 65 669 930 1.08 0.07 Average Average 2007 2008 199598 199903 Machinery, excluding automobiles and parts (share in total imports in %), Zimbabwe Zambia Ratio of Zimbabwe to Zambia 23 33 0.70 22 23 0.96 32 26 1.22 23 23 0.98

Sources: UNCTAD WIR database (FDI data) and UN COMTRADE database (trade data).

Indeed, the evolution of machinery imports and their share in total imports and developments in FDI inflows shed light on the difference between the economic paths that the two countries embarked upon. Zambia's industrial basis was expanding in the 2000s, whereas that of Zimbabwe was either flat or shrinking. The share of machinery imports in total imports was on average significantly larger in Zimbabwe's imports in 199598; roughly equal in 19992003; and five percentage points lower in 2004 08. Over the last decade or since 1999, Zambian economy imported 62 percent more machinery in terms of value than Zimbabwe did. It is quite likely that a significant portion of these imports into Zambia was related to foreign investment. The decade of economic mismanagement has extracted a heavy toll on the export capacity of Zimbabwe's economy. Although mining was somewhat insulated, it has not survived without wounds revealed in lower outputs. Despite suppressed investment activity, imports of machinery and parts continued indicating that some industrial capacities have survived, albeit in desperate need of modernization. Modernization can only occur insofar as the climate for private sector development significantly improves. Macroeconomic stability is a necessary condition but without a genuine reduction in the cost of conducting business there the current recovery cannot be sustained. While large mining and farming (tobacco) operations tend to be more immune to unfriendly business climates, small scale agricultural and industrial activities are highly sensitive to illdesigned economic policies and weak economic governance. Indeed, as we shall show in Section 3, exports that have been most negatively affected were agricultural and simple industrial products. Exports of these low

25 | P a g e technology labor intensive agricultural and miscellaneous industrial products have been consistently declining since 1997. They have been largely responsible for Zimbabwe's disappointing overall export performance. 1.3. Zimbabwe's foreign trade performance in a regional perspective Although the links between institutions, policies and foreign trade performance, are complex, one would expect that Zimbabwe would be a laggard. Its quality of economic governance deteriorated and was the worst among its neighbors while hyperinflation largely disorganized economic activities across most sectors of the economy. Indeed, statistical data corroborate this expectation. Due to weaknesses in foreign trade statistics in Zimbabwe (WB 2010), we use foreign trade data derived from its foreign trading partners reporting to the United Nations. Zimbabwe's exports include CIF (cost of insurance and freight) and imports are FOB (free on board). By the same token, the data exaggerate the value of exports and lower the value of imports: the latter should be borne in mind when comparing foreign trade performance against that of other countries. Barring some major changes in the composition of exports, this has no significant impact on the export performance over time as freight and insurance usually do not undergo volatile changes over short periods of time. For reasons of geography and policy, members of SADC (Southern African Development Community) would be the first candidates against whom Zimbabwe's foreign trade performance record can be benchmarked. In terms of geography, they are located in the same region. As for policy, Zimbabwe together with most other SADC members is also a participant in the SADC FTA (free trade area) which came into force in 2008 but has been trading on preferential terms for much longer. The comparator countries chosen, however, neither include all SADC member states nor are they limited to them, as Rwanda has been also included (Table 5). The lack of data has eliminated such countries as Angola, DR Congo, Lesotho and Swaziland. While the first two do not report trade data to the UN, two SACU (Southern African Customs Union) members have gaps in foreign trade reporting: for Lesotho the data are available only for 200004 and Swaziland for 200007. The inclusion of Rwanda derives from shared geography (both are landlocked) and the legacy of major manmade disasters. Moreover, it may provide some indications of the dynamics that one may expect when a country takes strong measures to recover from an economic crisis. SADC is a highly diversified regional grouping in terms of the size and level of economic development of its members, quality of economic governance as well as geography. The GDP per capita of the richest, Seychelles, is almost fortytimes higher than that of Malawi--the poorest member of SADC. South Africa towers over all other countries with its GDP amounting to around threefourths of the region's and population to less than onethird of the total. The quality of economic governance is significantly higher in SACU countries and Mauritius than in other SADC countries with Zimbabwe scoring the lowest amongst them all in the 2000s in the World Bank's surveys of the quality of governance covering 213 states across the world (see Section 1.1). Last but not least, geography and tourist attractiveness profoundly impact foreign trade performance. Seychelles and Mauritius are islands as well as attractive destinations for tourists. Exports of goods are not their main earners of foreign currency. On the other hand, Botswana, Malawi, Rwanda, Zambia and Zimbabwe are landlocked depending on infrastructure

26 | P a g e and quality of economic governance in other countries to get their products to seaports providing the cheapest means of transportation.

Table 5: Foreign trade growth performance of selected SubSaharan African countries in 200108 (in millions of US dollars, dollars and percent) Annual growth rates 2006 2007 2008 1.7 34.5 8.8 36.4 34.8 6.2 11.8 11.9 11.5 108.3 12.8 12.6 30.4 4.5 1.3 19.7 33.2 5.2 21.7 14.7 22.5 8.4 4.6 1.2 7.8 10.0 17.1 117.1 31.7 15.5 45.9 10.4 7.6 Average 200108 8.3 12.6 6.3 30.6 20.0 35.0 4.4 14.7 22.0 28.4 2.5 14.2 Average 200108 15.7 20.5 10.9 17.7 18.3 22.3 14.2 16.7 23.3 24.9 9.0 16.7 LSG rate 200008 9.1 10.8 6.3 23.9 18.1 24.5 6.2 14.8 18.6 24.9 3.6 14.5 LSG rate 200008 8.7 17.3 12.3 17.1 18.3 21.1 12.2 19.5 23.2 23.1 13.1 18.7 Value of exports per capita (d) in % of GDP 2,431 59 1,876 122 2,252 37 2,826 1,510 76 430 196 39.8 21.1 25.2 27.8 54.4 8.1 25.6 27.1 15.3 37.6 56.9

EXPORTS

Botswana Malawi Mauritius Mozambique Namibia Rwanda (a) Seychelles South Africa Tanzania Zambia Zimbabwe (b) Total

2008 (c) 4,838 879 2,401 2,653 4,729 398 246 73,966 3,121 5,099 2,230 100,561

IMPORTS

Botswana Malawi Mauritius Mozambique Namibia Rwanda (a) Seychelles South Africa Tanzania Zambia Zimbabwe (b) Total

15.6 19.0 13.8 Annual growth rates 2006 2007 2008 3.4 3.6 15.3 19.1 11.2 34.2 12.2 24.4 39.4 20.2 20.1 22.4 30.6 14.2 7.1 6.3 43.9 25.1 13.4 16.7 30.8 30.3 7.6 18.4 27.9 59.9 19.7 31.4 16.5 64.4 6.1 9.7 36.6 26.3 6.1 14.6

605 27.9 Value of imports 2008 (c) per capita (d) in % of GDP 5,099 2,204 4,670 4,008 4,689 1,146 912 87,593 8,088 5,060 2,680 2,562 147 3,648 185 2,233 107 10,481 1,788 197 427 235 759 41.9 52.8 49.0 42.0 53.9 23.3 94.9 32.1 39.8 37.3 68.4 40.1

123,467

Notes: (a) Rwanda is not a member of SADC; (b) trade data derived from partners' foreign trade statistics; (c) in millions of US dollars; (d) in US dollars. GDP data are in current US dollars, and LSG rate stands for Least Square Growth rate. Sources: derived trade data as reported to UN COMTRADE database; and population and GDP from World Bank Development Indicators and UN social indicators databases.

Performance in exports of goods reflects these differences. Landlocked Zimbabwe with the worst quality of economic governance in 20022008 had by far the worst performance amongst comparator economies: the average growth rates were between four five times lower than the weighted average growth rates for countries displayed in Table 5. A simple eyeballing of data leads to the following general observation: Zimbabwe was the worst performer amongst comparators in terms of growth of exports. The growth in 200008 was by far the lowest: it was more than two percentage points below the second worst case of Seychelles. However, exports per capita remain respectably decent: in fact, they are larger on a per capita basis than in Rwanda, Tanzania, Mozambique and Malawi: however, if the differential in export growth rates between them, on the one hand, and Zimbabwe, on the other, is

27 | P a g e sustained over the next ten years, all will overtake Zimbabwe. Zimbabwe stands out in another respect: the value of its exports in terms of the GDP was the highest amongst these countries. Zimbabwe, a laggard among exporters, is also at the bottom in terms of import dynamics. Imports into Zimbabwe grew faster in terms of LSG (least square growth19) rate over 200008 than into frugal and richer Botswana, but slower when measured as an average of annual growth rates. Zimbabwe's double digit LSG rate was mainly the result of a rebound following the sharp contraction in the value of imports of more than one third in 200001. In 200206 imports grew at an average rate of 16 percent but still 10 percentage points below the growth experienced by neighboring Zambia. Other countries have experienced strong doubledigit rates of imports growth in 200008. Thanks to either surpluses in services trade or economic aid, or both, other countries, with the notable exception of Zambia, had imports per capita significantly higher than exports per capita. Thus it comes as no surprise that Zimbabwe's share in total region's foreign trade has been on the decline. Its share in exports fell from 5 percent in 200002 to 3 percent in 2005 and further to 2.2 percent in 2008. In order to retain its share in 2002, the value of exports in 2008 would have to be US$5 billion or 2.3 times higher than it actually was in 2008. This shows the scope of decline in exports over 200008. But the decline was smaller for imports. The contraction in the share of Zimbabwe in total imports was less pronounced falling from 3.8 percent in 2000 to 2.6 percent in 200406 and 2.2 percent in 2008. Zimbabwe's imports would have to be 1.8 times higher or US$4.7 billion rather than US$2.7 billion in order to retain its share in region's imports in 2000. In the meantime, however, a traditional surplus in trade in goods moved to deficits in the balance of merchandise trade. 1.4. Concluding observation The weakened state and collapsing regulatory quality led to a rapidly shrinking economy, as the cost of conducting business operations skyrocketed and the competitiveness of Zimbabwean firms in global markets was greatly eroded despite constant devaluations. According to most international assessments comparing the quality of economic governance and competitiveness, Zimbabwe was consistently ranked at the bottom with the worst investment climate and highest incidence of corruption. While the restoration of macroeconomic and financial stability has been achieved thanks to replacing the domestic currency with the US dollar and South African rand, other barriers--such as low government effectiveness and poor quality of regulations--are still in place. The discussion of the quality of economic governance points to a significant challenge facing the government in overhauling the existing economic regime. This calls for deep structural reforms. Without them, economic performance will remain wanting, albeit likely superior to that experienced in 200008. Given the low quality of governance, i.e., the absence of institutions and policies supporting competitive markets, combined with restrictive foreign exchange controls, confiscatory taxes imposed on exporters through surrender requirement, inflation and hyperinflation that were all trademarks of the economic landscape during 200008, it comes as no surprise that, cast against other countries of the region,

According to the standard definition, the leastsquares growth rate is estimated by fitting a linear regression trend line to the logarithmic annual values of the variable in the relevant period (for a discussion, see OECD Glossary of Statistics, available at http://stats.oecd.org/glossary/detail.asp?ID=6687).

19

28 | P a g e Zimbabwe's relative position as an exporter and importer visàvis other comparator countries has drastically deteriorated. But it is rather surprising that foreign trade had displayed such strong resilience in the face of devastating macroeconomic policies and an unfriendly business environment. Note that exports in terms of value still recorded growth in the 2000s, as did imports.

29 | P a g e

2. Highlights of Zimbabwe's trade performance: openness, dynamics, and direction of trade

Zimbabwe is unique. Except for the collapse of central planning across the former Soviet bloc and the aftermath of civil or interstate wars, it would be difficult to find a country whose economy has been declining for more than a decade at such a quick pace. The shrinking economy has been bound to affect foreign trade as fewer goods were produced and fewer would be available for export. The question addressed in this chapter concerns the long term perspective for development in trade and its linkages with the country's overall economy as captured by GDP.20 Both call for a careful statistical assessment as hyperinflation made estimates of output and trade subject to wide margins of error. Yet, all signs point to a more pronounced decline in Zimbabwe's GDP than in its foreign trade. Three other issues addressed in this chapter concern development in trade in services; a comparison of price and volume changes in exports; and, shifts in geographical patterns of trade. For services, tourism and transportation overshadow other areas in Zimbabwe's services sector: tourism has been an important source of foreign exchange earnings while Zimbabwe, as a landlocked country, has historically run deficits in transportation services. In terms of prices, these have been favorable for Zimbabwean exporters, although they did not take full advantage of them as their exports either stagnated or declined in terms of volume. Although the period 200007 witnessed a dramatic reorientation in Zimbabwean foreign trade, Zimbabwe has not been successful in taking advantage of rapidly growing import demand in other countries in the region. The competitiveness of Zimbabwe's exports in regional markets--as measured by their share in partners' imports--has been on the decline with the exception of South Africa. 2.1. Deceptive increase in openness Zimbabwe's shrinking economy and stagnant foreign trade in goods in current prices led to an unexpected increase in the openness of Zimbabwe's economy as measured by its share of foreign trade in GDP. Since foreign trade in goods has been stagnant and the economy has been contracting, Zimbabwe's openness has increased over the last decade or more.21 The stagnation in foreign trade was due to a combination of falling exports and expanding imports in terms of value. Zimbabwe's economy has dramatically shrunk since 1998--the last year when GDP experienced growth of 3 percent in real terms.22 While real GDP fell every year over 19992008, Zimbabwe's GDP in current US dollars displayed a highly volatile picture of dramatic increases and declines with GDP increasing almost 400% in 200102. The period raising the most serious suspicions of statistical inexactness covers the early 2000s. Recorded rates of GDP per capita growth swung from minus 25 percent in 1998 and

As for foreign trade statistics, we shall rely on foreign trade statistics reported by Zimbabwe's trading partners to the UN COMTRADE database for the reasons discussed at some length in a background note prepared for this study (see Appendix 1). While the quality of Zimbabwe's imports reporting is remarkably good, its exports statistics raise concerns. 21 It is important to note that ratio of trade to the GDP must not be interpreted as a trade policy indicator. The former depends on the size of an economy whereas the latter refer to the extent of policyinduced barriers to trade. 22 In real terms (in 2000 prices), its GDP was falling every year at an average pace of minus 7 percent over 1999 2008: its total GDP in real terms stood in 2008 at 35 percent.

20

30 | P a g e minus 15 percent in 1999 to annual rates of growth of 40 percent, 58 percent, and 140 percent each year between 2000 and 2002 (see Table 6). Since the population remained at roughly the same level of 12.5 million, this spectacular growth episode had nothing to do with a sudden dramatic increase in population.

Table 6: GDP and its growth in real and value terms according to different sources over 19982008 (in millions of US dollars and percent)

WB GDP * IMF GDP (1) IMF GDP (2) 1998 6,066 5,983 6,851 25 2.9 100 1999 5,964 6,011 5,801 2000 2001 10,256 12,891 12,883 58 2.7 86 2002 21,897 30,873 30,856 140 4.4 81 2003 7,397 7,918 10,515 66 10.4 71 2004 4,712 4,696 4,713 55 3.8 67 2005 3,418 5,792 4,627 2 5.3 62 2006 3,711 3,711 5,596 21 6.3 56 2007 3,553 3,553 3,583 36 6.9 49 2008 3,180 3,180 3,929 . 10 14.1 35 Index 2008, 1998=100 52 53 57 Cumulative 70 63

7,399 7,470 8,136 Memorandum: annual change in percent Current IMF GDP(2) Real GDP Real GDP (1998=100) 15 3.6 96 40 7.9 89

Notes: * There is no data available from WB from 2006 to 2009. (1) Based on IMF IFS database through official exchange rate and IMF Consultation Article IV paper 2009; and (2) Based on IMF World Economic Outlook database 2003/4, 2009, 2010; data for 2003-09 are estimates.

Neither did it have anything to do with positive developments in the real economy. Only external intervention could explain the recorded increase in GDP per capita from US$486 in 1999 to US$2,466 in 2002 and subsequent contraction to US$376 in 2004 but this seems unlikely. One should also note a significant divergence in IMF and World Bank's estimates of Zimbabwe's GDP in 2002 with the former being 41 percent larger. Yet, in 2008, three estimates indicated a similar level of overall contraction in relation to the 1998 level. Since the GDP statistics in current prices display rather unusual developments in 200102 with GDP increasing almost four times, which can be only attributable to a large statistical inaccuracy, we exclude these years from our assessment of the change in the significance of foreign trade to the Zimbabwean economy over time. Hence, years 2001 and 2002 do not show in Figure 4.

Figure 4: GDP and foreign trade in US dollars per capita and percent in 19972000 and 200308

800 700 600 500 400 300 200 100 0 1997 1998 1999 2000 2003 2004 2005 2006 2007 2008

180% 160% 140% 120% 100% 80% 60% 40% 20% 0%

Trade in GDP (in percent) GDP (IMF) per capita Exports per capita (in US$) Imports per capita (US$) Linear (Trade in GDP (in percent)) Linear (GDP (IMF) per capita)

Right Axis: in percent; Left Axis: US dollars Sources: GDP based on IMF IFS database through official exchange rate and IMF Consultation Article IV paper 2009; population from the World Bank's World Development Indicators database; and trade from partner's statistics as reported to the UN COMTRADE database.

31 | P a g e These comments notwithstanding, the contraction in GDP was significantly larger than in foreign trade. Imports per capita in value terms in 2008 were 21 percent higher than in 1997 and exports 1 percent higher down from 9 percent a year earlier, whereas GDP per capita in 2008 was 64 percent lower than in 1997. The LSG rate of the contraction in GDP per capita over 19972008 was 8.8 percent, whereas imports grew at 3.5 percent and exports at 3.5 percent. As a result, foreign trade in terms of GDP grew from 50 percent in 1997 to 154 percent in 2008. This increased openness as traditionally defined is not, however, the product of a successful outward oriented strategy but, rather, the ability of the foreign trade sector to better absorb the devastating effects of bad economic policies pursued during this period. While the value of exports and, to a much larger extent, that of imports continued to rise, Zimbabwe's GDP was contracting. As Figure 4 above graphically illustrates, the fall in GDP per capita was accompanied by an increase in foreign trade as a percentage of the GDP solely because both exports and imports per capita fell much less. 2.2. Trade in services: transportation and tourism Detailed data on trade in services are always problematic not only in developing countries but also in highly developed ones. Zimbabwe is no exception. The quality of data is particularly low through most of the 2000s but has improved recently. Consequently, the analysis is limited to three years--2008 and 2009 together with projections for 2010. The balance of payments data made available by the Reserve Bank of Zimbabwe are not only relatively aggregated but also, except for foreign trade in goods, cannot be independently assessed in terms of their reliability. In the area where this is possible, i.e. foreign trade in goods, Zimbabwe's balance of payments statistics fail the `foreign trade mirror' test in 200006, especially so on the exports side (see Appendix 1). They converge, however, in 200708. Another `quality' test can be derived from data on international tourism. Except in 200304, there is huge divergence between the Zimbabwean data and those from the World Bank WDI (World Development Indicators) database. While no information for 2008 is available in the WDI database, Zimbabwe's estimates for 2008 are in line in terms of magnitude with the WDI data for 2007 indicating improvement in the quality of balanceofpayments reporting observed in other verifiable areas of these statistics. The developments in 2009 point to significant changes in Zimbabwe's trade in services.23 The most notable change has been the surge in exports of services. The value of these exports increased in 2009 by around 2.5 times or from US$219 million in 2008 to US$467 million in 2009 (Table 7). Although imports of services also increased, the deficit in services trade contracted rather dramatically from US$269 million in 2008 to US$112 million in 2009 and US$146 million in 2010. The deficit also declined considerably relative to 2000, when it stood at US$322 million. Tourism, the largest net exporter of services, was responsible for this improvement. Had net revenue from "travel" stayed at its

Nonfactor services include both commercial and government services. Unfortunately, Zimbabwe's balanceof payments statistics do not allow for an easy differentiation between the two. But this is not a significant problem as receipts from commercial services tower over government ones. Total trade in services includes `nonfactor services,' i.e., receipts for shipment, passenger and other transport services, and travel, as well as current account transactions not separately reported, and `other services.' The major item in the latter are fees for professional services and operational expenses of foreignowned companies in Zimbabwe.

23

32 | P a g e 2008 level, the deficit in trade in services would have reached US360 million rather than US$112 million. The largest contributor to deficit has been transport, i.e., shipment and other transport taken together, followed by professional, technical, and administrative fees.

Table 7: Zimbabwe's trade in services in 2000 and 20082010 (in millions of US dollars) Total services Shipment, of which freight and insurance transit freight Other transport, of which passenger fares port services Travel Other services, of which fees (professional, technical, administrative) local operational expenses Memorandum: Nonfactor services */ Diaspora remittances Receipts (exports) Payments (imports) Balance in services 2000 2008 2009 2010p 2000 2008 2009 2010p 2000 2008 2009 2010p 390 41 22 19 165 70 95 89 95 29 21 294 0.4 219 33 19 14 33 10 23 102 51 11 15 169 75 467 30 17 13 35 11 24 350 52 13 15 415 198 500 35 20 15 36 11 25 375 54 13 16 446 229 712 226 187 39 190 47 148 79 217 97 0 495 n/a 488 281 234 48 77 9 68 17 113 47 0 376 n/a 579 345 287 58 79 10 70 22 133 63 0 447 n/a 646 399 331 68 84 11 73 25 138 64 0 507 n/a 322 185 166 20 25 23 52 10 122 68 21 201 n/a 269 248 215 34 44 1.2 45 85 62 36 15 207 n/a 112 315 270 45 44 0.8 46 328 81 50 15 32 n/a 146 364 311 53 48 0.7 48 350 84 51 16 61 n/a

*/ total services excluding "other services;" (p) projected Source: derived from the Reserve Bank of Zimbabwe balanceofpayments statistics.

Not surprisingly, given that Zimbabwe is landlocked and has large deficits for trade in goods, the major expenditure on the services side is transport, i.e. freight and insurance, as well as transit fees. The deficit in goods trade raises the cost of shipments as trucks delivering imports return empty to their origin. In relation to the value of imports (CIF), the cost of freight at 11 percent in 200809 appears to be in line with other estimates for African economies.24 Studies, using freight payments as a percentage of total imports, show roughly similar values of this ratio for African landlocked countries. For instance, Stone (2001) estimates that out of 15 landlocked African countries, 13 of them had a ratio higher than 10 percent, and for seven the ratio was even higher at 20 percent, compared with 4.7 percent for industrial countries and 2.2 percent for the United States.25 But the caveat to be borne in mind is that, as Arvis, Raballand and Marteau (2010) convincingly show, freight payments as a percentage of total imports are complex to assess and prone to errors as the actual calculation of freight payments is often difficult to conduct. These comments notwithstanding, there is clearly a need to carefully examine the full logistics costs of foreign trade in Zimbabwe and the reliability of supply chains servicing this trade in order to identify sources of excessive costs. In particular, while the transportation costs for Zimbabwe's imports appear acceptable compared to other African landlocked countries, they nevertheless seem to be rather high considering that Zimbabwe is in a much better situation than most landlocked African countries. This is

The cost includes fees paid for transit. The ratio slightly fell from 13 percent in 2000 to 12 percent in 200204, and 11 percent in 200510. 25 Quoted in Arvis, Raballand and Marteau (2010, p. 2).

24

33 | P a g e because Zimbabwe not only has relatively good infrastructure but is also surrounded by `good neighbors' with very good infrastructure and respectable qualities of economic governance. For example, around threequarters of Zimbabwe's imports come from neighboring countries, mainly South Africa. High costs therefore point to other reasons such as low levels of competition between shipping services providers, concentration in the trucking industry market structure and restrictive transport regulations in Zimbabwe.26 While transportation remains the biggest import item in services, tourism has been traditionally the largest net foreign currency earner. The increase in revenue from international tourism in 2009 to US$305 million from US$102 million in 2008 was spectacular and solely responsible for the growth in services exports during that period. But it seems that this estimate may be below the actual revenues. In marked contrast to official estimates of trade in goods grossly exaggerating their values, especially in the early 2000s, the differences between official data on travel and World Bank's estimates were overall smaller in 200004, but huge in 200607. The World Bank's estimates of receipts from international tourism exceeded reported receipts from travel by the Reserve Bank by the factor of more than two in 2006 and three in 2007 (Table 8).

Table 8: Receipt from international tourism according to Zimbabwe's balance of payments statistics and the World Bank' world development indicators database in 200008 WB: World Development Indicators RBZ: Reserve Bank of Zimbabwe (travel receipts) Ratio of WB to RBZ (in %) 2000 2001 2002 2003 2004 2005 2006 2007 2008 125 89 140 81 58 140 76 53 143 61 61 100 194 194 100 99 147 67 338 150 226 365 117 312 n/a 102 n/a

Sources: WB; World Bank WDI database and RBZ; Reserve Bank of Zimbabwe.

Which estimate is closer to the real world? The answer to this question would call for an analysis going beyond the scope of this study. Instead, we offer two observations pointing to the fact that revenues were larger than estimated in official statistics. First, the potential for tourism in Zimbabwe is huge. From internationally reconciled data, we know that in the 1990s there was substantial growth in Zimbabwe's exports of services as a result of increases in tourism. By 1994, travel (tourism) became the most important item in Zimbabwe's services exports accounting for almost half of its services exports in that year. However, tourism receipts have fallen back sharply since then in the wake of the crisis. Tourism receipts fell from over US$232 million per year at their peak in the 1990s in 1996 to US$61 million by 2003 and overseas arrivals fell from 331,000 in 2004 to 223,000 in 2008. Average room occupancy rates have been about 40 percent since 2004 (see ZTA 2008). Second, as in the case of goods trade data, which largely converged with the mirror statistics beginning in 2006, the official estimate of revenues from travel of US$350 million for 2009 and projected receipts of US$375 million in 2010 would fit more the World Bank's trend than that of the RBZ. Since there is no other information that would support a sudden increase in numbers of tourists coming to Zimbabwe in

Considering high expenditures on shipment and freight, their reduction would have significant positive impact on prices of imported products and would increase competitiveness of Zimbabwean exports. To our knowledge, there has been no recent study of the transport sector, regulations and cost structure in Zimbabwe.

26

34 | P a g e 2009, it seems that the Reserve Bank of Zimbabwe has simply improved its methodology of estimating receipts from international tourism. Furthermore, both the World Bank and the Reserve Bank give the same estimate of receipts from tourism in 2003 and 2004. But in 2006 and 2007, the official estimates were at 44 percent and 32 percent of the WDI data in spite of a significant increase in the number of people coming to Zimbabwe: the increase was 35 percent in 2007 over 2004. This is rather implausible that this increase be accompanied by a 40 percent fall in receipts from travel as reported in official statistics. Hence, it seems that revenues from tourism in 200910 may be even larger than reported in Zimbabwe's balance of payments statistics. Whatever the exact level might be, it seems that Zimbabwe's performance in the sector has significantly improved since 2003. Receipts per arrival significantly improved from an average of around US$50 in 200005 to almost US$150 in 200607. However, they were still slightly below average amounts spent by tourists in Zambia, but not by as much as in 200005 when they were on average four times as large (Annex Table 2). Tourism has again become a very important source of foreign currency earnings for the country creating new jobs and offsetting expenditures on transportation of goods into Zimbabwe. Promotion of tourism through the implementation of measures that would facilitate expansion of services attracting tourists appears to be an area where government involvement might produce tangible effects if negative perceptions about the country can be overcome. In contrast, the fall in current account receipts from passenger fares for air and railroad transport services, both of them monopolized by the state, suggests that these are areas where government involvement should be carefully assessed. 2.3. Dynamics of foreign trade in goods in 19972008 Zimbabwe's foreign trade has displayed rather an unusual dynamic: its imports plummeted in 1997 2001 and then quickly recovered reaching the 1997 level in terms of value around 200506, while exports after a decline in 1998 slightly increased and then fell only to rebound at rather flat growth rates. Because of different time profiles, Zimbabwe ran a significant trade surplus in 19992005 followed by growing deficits in trade in goods. Exports coverage of imports fell from 130 percent in 2000 to 83 percent in 2008 (Figure 5).

Figure 5: Zimbabwe's exports and imports of goods in 19972008 (in millions of current US dollars)

2,800 2,600 2,400 2,200 2,000 1,800 1,600 1,400 1,200 1,000 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Exports Imports

Source: derived from partners' data available in the UN COMTRADE database.

35 | P a g e High growth in the world economy and trade in 200007 helped Zimbabwe recover from a contraction in its exports in 20002002 but it did little to significantly improve its exports performance. Over the whole decade of 19972008, the value of exports reached a bottom in 1998 and the value of imports bottomed out in 2002 (Figure 5 and Table 9). In contrast to exports, imports have a distinct `U' shape: they fell rather precipitously between 1997 and 2000 contracting each year by between 1114 percent. Subsequently they have been growing at an average Least Square Growth rate of 11 percent. As a result, the value of imports was 88 percent higher than in 2000 while that of exports was 20 percent over its value ten years earlier in 1998 but merely six percent over the value of Zimbabwe's exports in 1997.

Table 9: Exports and imports of goods in terms of value and exports in percent of imports in 19972008 (in millions of US dollars)

Exports Imports Exports in percent of Imports Exports (1998=100 Imports (2000=100) 1997 2,095 2,146 98 118 151 1998 1,778 1,860 96 100 131 1999 1,919 1,651 116 108 116 2000 1,855 1,424 130 104 100 2003 1,867 1,532 122 105 108 2004 1,924 1,702 113 108 120 2005 1,974 1,954 101 111 137 2006 2,227 2,348 95 125 165 2007 2,414 2,526 96 136 177 2008 2,230 2,680 83 125 188 LSG rates 2.6 (a) 10.7 (b)

Note: (a) LSG (least square growth) rate over 1998-2008; (b) LSG rate over 2000-08. Source: derived from partners' data available in the UN COMTRADE database.

When set against imports, the time profile of export performance differs in three respects: first, the bottom was reached two years earlier; second, the growth was weaker; and third, the value of exports contracted in 2008 by 8 percent, while imports increased 6 percent. The fall of exports in terms of value is partly attributable to the Global Financial Crisis which began in 2008 and spilled over from the financial sectors in highly developed countries to the world's real economy. World merchandise output fell 0.5 percent, while the rate of growth of goods significantly slowed down from 9 percent in 2006 to 6 percent in 2007 and 1.5 percent in 2008.27 Viewed in a longer perspective of the last two decades, one may distinguish three phases in Zimbabwe's export performance: two phases of moderate growth in 199097 and 200307 and a phase of contraction in 19982002. During the first growth phase an average annual growth rate was almost 10 percent, followed by contraction at an annual growth rate of 3.5 percent, and a rebound at an average annual growth rate of almost 8 percent (Table 10). The third phase came to a halt in 2008 when exports fell 8 percent. Leaving aside the contraction phase, the differences in export growth performance between the 1990 97 and 200307 phases point to a significant contraction in the country's export base. The first phase of growth was significantly more robust than the second one five years later: For starters, it was much longer extending over 7 years rather than 5 years. Second, the value of exports during the peak year in 1997 stood almost 80 percent above the level in 1990, whereas the peak in 2007 was only 40 percent above the level in 2002. Last but not least, the value of exports in 2002 stood at 82 percent of its peak level in 1997 and that in 2007 only 15 percent. The value of exports in 2008 was only six percent above its level in 1997.

27

Downloaded from http://www.wto.org/english/res_e/statis_e/its2009_e/its09_world_trade_dev_e.htm on May 1, 2010.

36 | P a g e

Table 10: Three phases of export growth in 19902008 (in percent)

Average annual growth rate Growth phase, 199097 Contraction phase 19982002 Moderate rebound, 200307 Contraction in 2008 9.8 3.5 7.0 7.6 Least Square Growth rate 8.6 2.7 6.4 n/a Index, peak low=100 179 82 140 n/a Share of EU first year 56.9 46.8 30.8 21.3 last year 48.6 36.5 22.0 n/a Share of neighbors first year n/a 16.4 31.7 42.9 last year 13.3 22.4 43.6 n/a

Source: Partners' trade data reported to the UN COMTRADE database.

An interesting question is whether the fall in exports in 2008 sets the stage for a new contraction phase or was it a temporary blip? Official data suggest that the contraction was limited to 200809. Since partners' foreign trade statistics for 2009 were not available as of this writing (September 2010), it is impossible to give a precise figure. According to the balance of payments statistics, compiled by the Reserve Bank of Zimbabwe, exports contracted 4 percent in 2009. But they are projected to grow 21 percent in 2010. Production statistics seem to suggest that indeed exports were rebounding in 2010. The critical question addressed in the next section concerns the extent to which increases in exports have been driven by increases in prices of exported commodities or by increases in the volumes exported. 2.4. Zimbabwe's export rebound in 200307: was it priceled or quantitydriven? The value of exports during the moderate rebound phase was around 40 percent above its value in 2003. But was this growth merely the result of increasing world prices for commodities during the global economic boom which took place during 200207 or did real exports increase during this time? Unfortunately, detailed information on Zimbabwe's exports in natural units is not available except for a selected group of products traced in the balanceofpayments statistics compiled by the Reserve Bank of Zimbabwe, which raises questions about the reliability of official data. As discussed in Appendix 1 and also shown in Table 11, the quality of official export reporting appears to be wanting. Leaving aside absolute differences in estimates of the values of exports, the trend, which was similar in 200102, becomes different in 200307: while mirror statistics point to the growth, official statistics point to the collapse of exports in terms of value beginning in 2005. Beginning in 2006,28 the value of total official exports is below the value of total mirror: while this is the way it should be considering that exports are FOB (free on board) and imports from Zimbabwe are CIF, i.e., include cost of insurance and freight, the difference cannot be explained by CIF. It is rather unlikely that the cost of insurance and freight was so high at 29 percent in 2006 and increasing to 35 percent in 2008 as the ratio of CIF to the value of Zimbabwean imports was `only' above 10 percent.29 Yet, none of these should discredit the official data on selected commodities exports (for their list see Table 12 below) and suggest that they are not worth a serious analysis. To the contrary, there are reasons to believe that the use of official disaggregated statistics will not produce unreliable results as

One wonders whether a sudden change in 2006 with officially reported exports suddenly falling well below mirror exports had anything to do with firms' attempts not to reveal exports in order to avoid confiscation of a growing portion of foreign earnings through the surrender requirements. 29 See the earlier discussion of shipping costs in Section 2.2 on trade in services.

28

37 | P a g e distortions in official reporting seem to have been related to other products than those for which data in unit values are available. Consider the following: First, data on selected commodities were not the source of grossly overvalued exports: note that as the quality of export reporting was improving, the share of these commodities in total exports was increasing from 21 percent in 2000 to 7071 percent in 200608 (Table 11). This indicates that exports of other products were overvalued to a much larger extent. Put differently, other exports `were downsized' rather than those of selected commodities.

Table 11: Exports of selected commodities and total exports as reported by Zimbabwe's authorities and importers of Zimbabwe's products in 200008 (in current prices)

Total exports of selected commodities*/ Share in total exports Total exports Total mirror exports Ratio of exports of selected commodities to total mirror exports 2000 1,305 21% 6,354 1,855 70% 2001 1,158 24% 4,809 1,809 64% 2002 945 26% 3,641 1,728 55% 2003 946 27% 3,461 1,867 51% 2004 1,173 34% 3,478 1,924 61% 2005 1,065 37% 2,860 1,974 54% 2006 1,222 71% 1,721 2,227 55% 2007 1,297 71% 1,819 2,414 54% 2008 1,161 70% 1,657 2,230 52%

*/ Table 2 lists the commodities covered by this analysis. Source: Reserve Bank of Zimbabwe and UN COMTRADE database.

Second, the ratio of exports of selected commodities to total mirror exports was relatively stable in 200108 indicating again that official reporting of exports of selected commodities was decent. Note also that the surge in the share of selected commodities exports in total exports from 37 percent in 2005 to 71 percent in 2006 coincided with the first year that official exports were not overreported: yet, the ratio of selected commodities exports to mirror exports remained unchanged at around 55 percent. This indicates an improvement in the quality of reporting. Last but not least, another piece of evidence suggests there are no substantial differences in exports of some of them, i.e., those for which finding an equivalent in mirror statistics is relatively straightforward (e.g., cut flowers, raw sugar, tobacco), as reported in Reserve Bank's statistics and foreign trade data. The official data point to a significant contraction in volumes of exports of both agricultural and mining products, albeit with two exceptions. One exception stands out: exports of PMG (platinum group metals) skyrocketed in 200308 increasing more than tenfold in 2003 alone (Table 12). This was the only bright spot in an otherwise grim picture: the result of Impala Platinum, South Africa, taking over Zimplat mines from BGP Delta, an Australian mining company, and turning them into profitability. Cotton Lint was another exception, albeit with a caveat: exports in 200508 were below the levels of 114 million tons recorded in 2000 and 2004 (Table 12). Otherwise, there were significant decreases in quantities exported across the board. The largest decline was in registered exports of gold: volumes shipped abroad in 2008 stood at just 13 percent of exports in 2000. The contraction was particularly acute in 200508 with entry into force of the law granting the Reserve Bank of Zimbabwe exclusive rights to export gold which remained in effect until the introduction of the multicurrency system in early 2009. Since this amounted to 100 percent surrender of hard currency earnings in return for (what became largely worthless) domestic currency, gold output sharply declined. Hence, except for PGM exports, mining was not spared from the significant contraction in exported quantities exacerbating the effects of a collapse in domestic agricultural production:

38 | P a g e volumes of agricultural exports in 2008 were between 34 percent (flowers) and 49 percent (raw sugar) of those in 2000. Exports of more processed refined sugar also contracted more strongly than exports of raw sugar in 2008: they stood at 35 percent of their level in 2000 as compared to 49 percent (see Table 12).

Table 12: Major Zimbabwe's exports in terms of natural units in 200008

Fluecured tobacco (mlns tons) Sugar raw (000 tons) Sugar refined (000 tons) Flowers (in 000 tons) Gold ('000 ounces) Platinum Group Metals (PGM) Nickel (000 tonnes) High Carbon Ferro Chrome ('000 tons) Cotton Lint ('000 tons) 2000 177 150 101 18 778 21 9.0 245 114 2001 195 86 95 22 827 36 6.3 219 80 2002 141 59 107 23 513 12 4.9 293 52 2003 102 69 50 20 417 128 7.3 264 48 2004 69 79 61 15 672 369 7.4 248 114 2005 64 62 46 13 429 436 6.6 245 95 2006 67 82 61 10 335 536 6.9 191 98 2007 73 41 48 11 226 574 6.5 152 94 2008 71 74 35 6 105 574 4.2 82 95 Index 2007, 2003=100 72 60 96 55 54 448 89 58 196 Index 2008 2000=100 40 49 35 34 13 2720 47 33 83

Source: Reserve Bank of Zimbabwe.

Fortunately for Zimbabwean exporters, the prices of most commodities were moving up in 200008, albeit at different rates depending on the product. Except for flowers and cotton lint, the prices of other exports tended to grow over the period. Prices of refined sugar were lower in 200205 than in 200001: they rebounded in 200607. So did the prices of chrome, albeit with a caveat: the prices rebounded earlier in 2004. Prices of tobacco were flat throughout the whole period in 200008. So were the prices of cut flowers in 200108. The strongest upward movement in prices was for gold and high carbon ferro chrome. Volumes of exports for these commodities, however, declined rather dramatically: exports of gold stood in 2008 at 13 percent of their level in 2000 and ferro chrome at 33 percent (see Table 13). The most volatile prices were for nickel, which fell in 200102 below their level in 2000, but subsequently increased spectacularly in 200307 before contracting by almost onehalf in 2008 thereby significantly contributing to the contraction in total exports that year, which, in turn, was also exacerbated by the fall in the volume of nickel exported (Table 13).

Table 13: Prices of major Zimbabwe's exports in 200008 (in US dollars per unit) Index 2008, 2000=100 106 177 155 77 323 153 214 198 86

2000 2001 Fluecured tobacco ($/kg) 3.050 3.000 Sugar raw ($/kg) 0.347 0.350 Sugar refined ($/kg) 0.431 0.420 Flowers ($/kg) 4.540 3.600 Gold ($000/1000 ounces) 0.278 0.273 PGM ($000/1000 ounces) 0.541 0.492 Nickel ($ per kg)) 8.700 5.587 High Carbon Ferro Chrome ($ '000/ton) 0.500 0.300 Cotton Lint ($/kg) 1.400 1.000 Source: Reserve Bank of Zimbabwe.

2002 3.030 0.456 0.346 3.780 0.311 0.540 6.480 0.400 1.000

2003 3.120 0.544 0.348 3.780 0.365

2004 3.250 0.417 0.341 3.600 0.391

2005 3.170 0.427 0.367 3.780 0.445

2006 3.070 0.508 0.644 3.780 0.601

2007 3.180 0.503 0.650 3.780 0.680

2008 3.240 0.614 0.670 3.500 0.897

0.603 0.473 0.532 0.581 9.386 12.927 14.500 23.144 0.500 1.400 0.748 1.070 0.641 1.010 0.614 1.100

0.600 0.828 34.974 18.575 0.933 1.100 0.990 1.200

39 | P a g e To what extent did increases in world prices compensate for the contraction in exported values of selected commodities? In order to answer this question, we calculate cumulative changes in prices and quantities exported separately over 200108 and the phase of "mirror" export rebound in 200307. The last column of Table 14 is the sum of cumulative changes of prices and quantities: zero indicates that falling/increasing prices equal to increasing/falling quantities, whereas negative values indicate that either prices or quantities drove exports down.

Table 14: Cumulative annual change in prices and quantities exported in 200108 and 200307 (in percent) Cumulative change in prices 200108 200307 Cumulative change in quantities 200108 200307 53% 8% 44% 64% 48% 1240% 38% 60% 112% Total change 200108 200307 68% 68% 1% 105% 3% 1295% 89% 25% 49% 48% 8% 39% 64% 38% 1255% 244% 48% 132%

Fluecured tobacco 6% 5% 74% Sugar raw 70% 16% 1% Sugar refined 66% 83% 64% Flowers 23% 0% 82% Gold 131% 86% 133% Platinum Group Metals 54% 15% 1241% Nickel 139% 205% 49% High Carbon Ferro Chrome 107% 108% 83% Cotton Lint 0% 20% 48% Source: Own calculations based on the data from Reserve Bank of Zimbabwe.

Two observations can be derived from examining data in Table 14. First, prices tended to grow while quantities of exports were contracting. As mentioned earlier, except for cut flowers, prices of all other commodities increased.30 In terms of quantities exported, except for PGM and cotton lint, all other exports had negative growth rates in 200108. Exports of nickel increased but only in 200307. Second, the improvement in prices neutralized the fall in exported quantities of most commodities and positively contributed to the increase in the value of exports in the case of all commodities except for fluecured tobacco and cut flowers. For the whole period 200108, one should also include gold whose exports collapsed in 2008 due to the state taking them over. The largest increase in 200307 was for two commodities--PGM and nickel--whose both prices and quantities exported displayed strong growth. Since the composition of exports has also been changing, the data in Table 14 do not allow us to estimate the aggregate impact of changes in prices and quantities on total exports of these commodities. As the first step to address this question, we calculate the values of exports of these commodities using (a) constant prices and (b) constant quantities.31 Table 15 presents the results. The values of exports in Table 15 at the diagonal are actual values of exports, i.e., quantities are expressed in prices for the same year. Values in a row record the values of an export basket in successive years in constant prices. For instance, the value in the row "in 2004" and in column "2007" is the product of 2007 quantities multiplied by respective export prices in 2004. Put differently, as one moves across the rows, successive changes in values are the result of changes in the volumes of exports

Cut flowers experienced a cumulative decline in prices of 23 percent, almost exclusively due to a 21 percent contraction in 2001. The price of cotton lint displayed huge volatility with annual contractions of 29 and 24 percent and increase of 40 percent in 2003. 31 In more formal terms, we calculated for each year the values of exports according to the formula (pIt*qit) where pi and qi stand for price and quantity of a good i, and t for years.

30

40 | P a g e in successive years expressed in terms of prices in the base year. In other words, the first row contains information about exports in 200008 in 2000 prices. Symmetrically, the entries in a column represent values for quantities exported in the base year using prices in each year over the period 200008. For instance, the first column provides information about the values of exports in 2000 quantities valued in prices in successive years. Hence, were the 2000 export basket reproduced in 2007, the value of exports in this year would be US$1.982 billion rather than the actual US$1.297 billion and their value in 2007 prices would be 52 percent higher than in 2000 prices (see the last row). The data in the last row and column are similar to Lespeyres and Paasche indexes, i.e., they summarize the impact of prices and volumes on exports in 2007 expressed in terms of either volumes or values in a base year (the value on the diagonal).32 Corresponding indexes were calculated for 2007, because this was the last year of the moderate rebound phase of Zimbabwe's total exports.

Table 15: Total values of exports of selected commodities expressed in constant quantities and prices in 200008 (in millions of US dollars)

Column: Base volume in current prices in 2000 in 2001 in 2002 in 2003 in 2004 in 2005 in 2006 in 2007 in 2008 Index 2007 100=t i Rows: exports in successive years in base prices 2000 1,305 1,151 1,231 1,400 1,472 1,490 1,716 1,982 2,058 152 2001 1,291 1,158 1,231 1,375 1,452 1,472 1,673 1,905 2,026 164 2002 1,012 885 945 1,056 1,147 1,144 1,284 1,492 1,559 158 2003 895 768 833 946 1,016 1,024 1,165 1,382 1,411 146 2004 1,067 909 992 1,157 1,173 1,208 1,415 1,650 1,794 141 2005 962 815 890 1,034 1,039 1,065 1,224 1,429 1,545 134 2006 980 840 912 1,056 1,030 1,063 1,222 1,408 1,522 115 2007 946 814 876 1,007 970 1,002 1,135 1,297 1,396 100 2008 833 728 780 893 823 852 948 1,051 1,161 ... Index 2007 100=t i 73 70 93 106 83 94 93 100 ....

Notes: (1) For a list of products covered, see Table14; (2) index is a ratio of the value in 2007 to that in the diagonal of the table; (3) in bold are diagonal values of exports in current prices and quantities. Source: own calculations based on the data provided by the Reserve Bank of Zimbabwe.

The scope of the overall decline in export volumes was large and progressing every year. As can be seen from the last row `in 2008' showing the values of quantities exported in successive years over 200008 in 2008 prices, except in 2004, the values were decreasing. Had Zimbabwe exported the same basket in terms of quantities in 2008 as in 2000, its exports earnings would have been US$2.058 billion rather than US$1.161 billion or 77 percent higher. Had the exports prices been the same in 2008 as in 2001, the value of these exports would have been US$728 million or 37 percent less than in current prices. Indeed, except in 2003, the value of exports in 2008 prices expressed in quantities of a base year was significantly higher than in the preceding year: in terms of 2003, it was 21 percent less than for the 2004

The data in the last column can be thought of as the equivalents of the Laspeyres index: the ratios of the value of exports in 2007 expressed in prices for each year over the period 200008 to the value of exports in quantities and prices of the base year or (pIt0*qit1)/( p It0*qit0) where pIt0 is the price of good i in base period, qit1 is quantity of good i in period 1 or 2008, and pIt0 and qit0 are prices and quantities in base periods. The Laspeyres index captures the change in quantities. The data in the last row "Index 2007 100=ti" are equivalents of the Paasche index with the volume in a given (base) year expressed in prices in successive years or (p It1*q it0)/( p It0*q it0). It denotes the impact of change in exports prices.

32

41 | P a g e bundle. Yet, the value of 2003 exports in 2008 prices would be 1 percent higher than 2007 exports in 2008 prices. The 2000 exports bundle in 2008 prices would generate export earnings 47 percent above the 2007 export bundle. In relation to actual 2008 exports, these earnings would be 77 percent. In a nutshell, except for exports in 2008, any earlier export bundle would generate higher export earnings than in the `peak' year in 2007 (for more details, see Annex Table 3). Since products covered in this analysis account for the bulk of Zimbabwe's total exports, it is safe to conclude that higher world prices rather than increased exported quantities drove total exports during the modest rebound phase of 200307. The rebound in real terms was minimal, if at all. Quantities of these commodities exported in the trough years 200203, which witnessed the largest fall in the value of these exports, would yield higher export earnings than those accrued in the peak year of 2007. Considering that firms operating in natural resource extractive industries might have been somewhat immune to deficiencies in the overall quality of the business environment as they tend to be geographically concentrated and depend less on local supplies, this finding comes as somewhat of a surprise. It appears that extractive industries were not entirely spared from the reach of economic policies. 2.5. Changes in Zimbabwe's direction of trade: a shift in exports from the EU towards Southern African markets Zimbabwe's exports to neighboring countries--SACU, Mozambique and Zambia--weakened the size of contraction of export contraction between 1998 and 2002 and contributed to their rebound in 200307. Redirection of exports from the EU to neighboring countries, mainly to South Africa, has been significant and taken place at least since 1990, i.e., well before the turbulent 2000s. While we do not have information on neighbordestined exports until the mid1990s (Botswana until 2000), the share of exports going to regional markets increased during both the contraction and moderate rebound phases. It went up from the peak of the growth phase in 1997 from 13 percent to 16 percent in 1998, when total exports fell 15 percent, and further increased in the last year of the contraction phase to 22 percent indicating a cushioning effect of import demand in neighboring countries. Their export share further expanded during the moderate rebound phase in 200307. The redirection of Zimbabwe's exports has been unidirectional: the major lever of change was expanding exports to South Africa and shrinking exports to the EU. While the average annual growth rate of total exports in 200108 was an unimpressive 2.5 percent (LSG rate over 200008 of 3.6 percent was slightly higher), Zimbabwe's South Africadestined exports recorded an average growth rate of 22 percent. In contrast, exports to the EU were falling an average of 56 percent per year. The share of the EU in Zimbabwe's total exports contracted 18 percentage points and the share of South Africa increased 24 percentage points over the same period from an average of 14 percent to 37 percent indicating a significance increase in the reliance on a single export market (Table 16). Note also that the share of exports to the rest of the world in total exports also fell from 32 percent to 27 percent in this period. The change in the geographical pattern of Zimbabwe's imports has been less pronounced, yet displaying striking similarities with those of exports. For starters, although the increase in the share of South Africa in Zimbabwe's total imports in the 2000s was less pronounced, its share is much higher as almost three

42 | P a g e quarters of imports originate there (Table 16). Second, the increase in South Africa's share in imports was almost the same as the fall in the share from the EU. Third, amongst major exporters to Zimbabwe, only China has succeeded in increasing its presence in Zimbabwe's market. Table 16: Direction of trade in 200008 (in millions of US dollars and percent)

Shares (in percent) Avg. Avg. 200002 200308 2008 14.3 10.3 3.6 0.6 4.9 2.1 38.9 7.0 32.2 100 30.9 27.7 2.1 0.7 6.7 2.5 25.2 7.2 26.8 100 37.4 34.0 2.1 0.7 4.8 1.7 21.3 6.6 27.4 100 Index 2008 Avg.00 02=100 260.9 330.0 57.4 118.3 98.1 84.1 54.9 94.4 85.1 100 Value 2008 834 759 46 15 107 38 476 148 611 2,230 Growth rates LSG Average annual 200008 200108 18.7 17.1 22.2 21.8 3.5 4.0 9.2 18.4 4.6 14.9 3.1 9.2 5.6 5.2 3.0 5.8 0.7 0.1 3.6 2.5 15.3 Growth rates LSG Average annual 200008 200208 12.6 11.1 18.5 6.7 23.0 26.9 1.5 25.8 3.8 10.7 12.3 16.3 16.0 22.9 18.1 53.8 82.2 0.0 42.7 9.2 12.9 15.7 15.7

Exports

SACU, of which South Africa Botswana Mozambique Zambia Malawi EU27 China Rest of the world TOTAL Memorandum: Neighboring countries

19.4 37.2 41.6 214 928 Shares (in percent) Index 2008 Value Avg. 200002 60.5 53.0 6.4 4.1 4.1 0.6 16.1 2.6 12.1 100 67.6 68.9 76.5 Avg. 200308 67.9 54.3 9.5 2.6 2.6 1.4 9.1 5.6 10.9 100 2008 71.3 63.0 8.0 3.0 2.4 0.8 7.2 5.0 10.3 100 113 Avg.00 02=100 118 119 125 74 58 151 44 193 86 100 2,050 2008 1,911 1,689 215 81 64 23 192 133 277 2,680

Imports

SACU, of which South Africa Botswana Mozambique Zambia Malawi EU27 China Rest of the world TOTAL Memorandum: Neighboring countries

Notes: Neighboring countries include Botswana, Mozambique, South Africa and Zambia Source: Derived from trade statistics reported to the UN COMTRADE by Zimbabwe's trading partners.

The increase in overall geographical concentration of Zimbabwe's exports has been also accompanied by an increase in the concentration of its exports to neighboring countries. The share of South Africa in Zimbabwe's exports to its neighbors--Botswana, Mozambique, South Africa and Zambia--increased from an average of 53 percent in 200002 to 82 percent in 200708. Although a number of other SADC countries are highly exportdependent on South Africa, Zimbabwe stands out. While the share of South Africa in the total exports of Mozambique, Zambia and Malawi was between 10 percent and 15 percent, it amounts to around 30 percent for Zimbabwe's exports. Zimbabwe has become more dependent on

43 | P a g e the South African market than most SACU members are: its share is only lower than that of Swaziland which exports around threequarters of its goods to there.33 While large yearonyear variations in the annual values of its exports to Mozambique and Zambia characterize Zimbabwe's sales to these markets, commercial transactions with both South Africa and Botswana appear to have been more stable. For instance, a sudden fall in the value of exports by 36 percent in 2003 was preceded by a 74 percent increase and followed by annual growth rates of 0.9 percent, 63 percent, 55 percent and another contraction of 48 percent in 200407. Similar variability was in exports to Zambia with the rates ranging between 109 percent in 2003 and minus 39 percent in 2004. On the other hand, rates of South Africaoriented exports ranged between minus 11 percent in 2008 and 72 percent in 2003. The shift in the geographical pattern of Zimbabwe's exports was not triggered by expanding exports to its major external markets. Neither was it triggered by a surge in exports to all of its neighbors as exports increased only to South Africa. The shares of other neighboring countries in both exports and imports were falling so was the share of the EU: the EU was taking almost half of Zimbabwe's total exports in 19942000 and accounted on average for 24 percent of Zimbabwe's imports. By 2008, these shares fell to 21 percent for exports and 7 percent for imports (Table 16 above). EUoriented exports have been the main casualty in the shift of Zimbabwe's trade towards neighboring markets, which occurred largely in 200108. Exports to the EU fell rapidly from around US$1 billion in 1997 to US$476 million in 2008 while rest of the worldoriented exports were stagnant and ranged between US$650 million and US$800 million in 19972008, and those to neighboring countries increased from an average of US$313 million in 19941999 to US$355 million in 200002 and then began rapidly expanding (Figure 6). Except in 2007 when they increased 7 percent over 2006, EUdestined exports continued falling at an average annual rate of 8 percent per year, whereas exports to neighbors grew at an average rate of 16 percent per year over 200407.

Figure 6: Exports to neighboring countries, EU27 and ROW in 19972008 (in millions of current US dollars)

1,200 1,000 800 600 400 200 0 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Source: Partners' trade data reported to the UN COMTRADE database.

ROW Exports Exports to EU27 Exports to neighboring countries

South Africa takes about 10 percent of Botswana's total exports and 20 percent of Lesotho's exports (IMF Directions of Trade Statistics).

33

44 | P a g e The turning point in Zimbabwe's transition to a new geographic pattern of trade was in 2003, when the value of its exports to neighboring countries increased 53 percent overtaking its exports to the EU, which contracted 9 percent in that year (Figure 6). This was in marked contrast to developments in 19942002: exports to neighbors grew during this period at an LSG rate of 1 percent; total exports at 0.4 percent; and exports to the EU fell at 4 percent (Table 17).

Table 17: Exports to neighboring countries, EU27 and ROW in 19972008 (in millions of current US dollars and percent)

Exports to EU27 Exports to `neighbors' ROW Exports Total exports Source: As in Figure 6. 1994 1997 2002 631 388 709 2003 574 593 700 2004 2007 2008 476 956 798 LSG 1994 2002 1.6 1.0 2.2 0.4 LSG 2002 2008 4.0 15.3 2.7 5.1 Index 2008 2002=100 75 247 112 129

706 1,018 303 278 598 799

564 531 634 1,053 730 830

1,607 2,095 1,728 1,867 1,927 2,414 2,230

The largest change occurred in 20022008; the pace of contraction accelerated to 4 percent, while exports to neighbors expanded at a LSG rate of 15.3 percent. Although the value of `neighborsdestined' exports fell 9 percent in 2008, this contraction wiped out the annual increase of 9 percent in 2007. Consequently, the value of neighbororiented exports in 2008 was almost 2.5 times higher than in 2002, whereas EUdestined exports stood at 75 percent of their value in 2002. The shift towards South Africa and other neighboring countries cannot be explained solely in terms of the redirection of Zimbabwe's exports away from EU markets. Annual increases in neighbordestined exports over 200306 in terms of value were larger each year than the annual falls in EUdestined exports: in aggregate, the former increased US$578 million whereas the total contraction for the latter was US$136 million. In 2007 exports to both markets increased: to neighbors by US$87 million and US$35 million to the EU followed by contraction in 2008 by US$96 million and US$55 million respectively. In all, the value of total exports was 29 percent higher in 2008 than in 2002 (Table 17 above).

2.6. Zimbabwe's export performance in regional markets

Hence, regional exports were the only bright spot in an otherwise grim picture of Zimbabwe's foreign trade performance. An interesting question is about Zimbabwe's performance in these markets. Some general insights can be derived from comparing the changes in total imports of neighboring countries with the growth in imports from Zimbabwe. This is a very general measure of competitiveness that does not take into account developments in individual markets and changes in relative prices. Yet, it sheds some light on the responsiveness of export sectors to opportunities offered in adjacent markets within easy reach, at least in terms of logistics. Based on the data in Table 18, the answer is unequivocal: Zimbabwe's exporters increased their presence in the largest and most important market in the region South Africa whereas they were outperformed in other regional markets. South African imports from Zimbabwe expanded at an impressive LSG growth rate of 24.5 percent in 200008, albeit with a caveat. The caveat is that the

45 | P a g e contraction in imports in 2008 resulted in a fall of Zimbabwe's share in total South African imports from 1.1 percent in 2007 to 0.9 percent. This share is below its levels in 200407 and 199496.34

Table 18: Zimbabwe's export performance in neighboring countries in 200008 (in percent)

Average annual rates in 200108 (a) (b) Botswana Mozambique South Africa Zambia 15.7 17.7 16.7 24.9 4.0 18.4 21.8 14.9 LSG rates in 200008 (a) (b) 8.1 15.7 18.0 22.2 Share of Zimbabwe in total imports 2004 2005 2006 2007 2008 4.3 1.58 1.54 1.56 1.25 0.91 10.8 0.47 0.65 0.84 0.41 0.38 24.5 0.90 0.89 1.00 1.07 0.87 7.4 5.78 4.30 5.72 2.79 2.12 Index, (d)=100 (c)= 100 32.6 54.9 143.7 29.3 Memo: Share in total exports, 2008 2.1 0.7 34.0 4.8

Notes: (a) total imports; (b) imports from Zimbabwe; (c) average share in 2000-01; (d) average share of Zimbabwe in total imports in 2007-08. Source: Derived from trade statistics reported to the UN COMTRADE by Zimbabwe's trading partners.

Zimbabwe has lost market share in other neighboring countries, as well. The average share in 200708 stood well below its average in 200001 in Botswana's imports at 33 percent of an earlier average, Zambia's imports at 29 percent, and Mozambique's imports at 55 percent. Moreover, volatility in market shares suggests randomness in commercial transactions, which does not augur well for the future expansion or even sustainability of Zimbabwe's exports to these destinations. 2.7. Concluding comment Zimbabwe's foreign trade performance has been lackluster even against regional standards, albeit much better when cast against its poor overall economic growth performance record. Imports were much less affected by the collapsing economy than exports of goods. Thanks to increased tourism, exports of services dramatically expanded, although Zimbabwe remains a net services importer mainly because of high transportation costs associated with bringing goods in and out of the country. Zimbabwe clearly failed to take advantage of the boom in commodity trade in 200207. Total exports continued falling through 200002 and rebounded only in 200307 thanks to increasing exports prices. Growth performance was not impressive: their value reached the 1999 level only in 2004 and their value exceeded the earlier record level reached in 1997 only in 2006 with the value of total exports only 6 percent higher; US$2.2 billion as compared with US$2.1 billion. It appears that the increase was solely the result of improved terms of trade for commodity exporters during this period: except for cut flowers and cotton lint, the prices for other commodities exported significantly increased in the 2000s. The depth of contraction can be illustrated as follows: assuming Zimbabwe's exports of basic commodities were the same in terms of quantities as in 2000 and that their share in total exports was the same as in 2000, the value of its total exports would be 30 percent higher than they were in 2008, that is, US$2.9 billion rather than US$2.2 billion. In terms of volume, exports in 2008 were 36 percent below their level in 2000. The contraction in exports of major commodities in terms of quantities over 200008, raises concerns about prospects for the country's future export performance. While world trade contracted and

The average in this period was 1.12 percent. It subsequently fell to 0.63 percent in 1997: the average for 1997 02, i.e., before the surge in 2003, was 0.7 percent.

34

46 | P a g e commodity prices fell during this period, the fall in the share of Zimbabwe in South African imports may suggest a more serious problem in competitiveness of Zimbabwe's exports. The next section seeks to answer this question by taking a closer look at the evolution of Zimbabwe's export basket in terms of factor intensities and degree of processing.

47 | P a g e

3. Drivers of Zimbabwe's trade performance: evolving patterns of specialization

Zimbabwean businesses operated in a highly unfriendly environment in 200008. In addition to rampant inflation and damage inflicted on the agricultural sector by the land reform program, certainty about the government's disregard for private property rights and sanctity of contracts has created a highly unstable business environment hostile to domestic and foreign investments alike. An examination of foreign trade performance offers important insights on how the economy coped with policyinduced adversities. Its strength derives also from the fact that the quality of domestic statistics can be checked by `mirror' statistics, i.e., data collected by Zimbabwe's trading partners. Domestic developments viewed through the lens of foreign trade performance allow for the identification of sectors that survived and those that did not. More importantly, they also offer some glimpses at which sectors may recover once the economy recovers and the business environment becomes friendlier. With the restoration of macroeconomic and financial stability, the conditions for recovery have greatly improved, although other barriers have to be removed in order put the economy on a path of sustainable economic growth (see Chapter 4). Thus, questions addressed in this chapter boil down to the following: How diversified or concentrated has Zimbabwe's exports become? What do the changes in exports depending on the degree of processing and levels of technology embodied in exported products tell us about Zimbabwe's industrial capacities and its revealed comparative advantage? What characteristics do products, in which Zimbabwe specializes, embody in terms of technology and production factor intensities? Which sectors survived and which ones have disappeared from Zimbabwe's export basket?

3.1. Diversity in exports has declined

Zimbabwe's exports have always been relatively highly concentrated with a few commodities generating the bulk of export revenues. But the concentration of exports increased further in 19942008 by most measures, although the increase began to take place already in the late 1990s. One measure relates to the share of top exporters in total exports: Table 19 tabulates the shares of the top five, ten, and twenty fourdigit SITC sectors in total exports.35 These shares kept increasing for each turning point in the export cycle. In 1994 the remaining 80 percent of fourdigit SITC sectors contributed 28 percent to total exports; by 200708 their contribution fell to 1819 percent (Table 19). Other measures also point to the decline in diversification of Zimbabwe's exports in 199497, and acceleration of this trend in 200208. The number of SITC fourdigit products exports increased from 774 (out of 1018 items) in 1994 to 803 in 1997 and 859 in 2002. This points to an increase in diversification, albeit with a caveat. The caveat is that most of these exports were negligible, as simultaneously the number of products with the value of annual exports exceeding US$100,000 slightly fell from 377 in 1994 to 367 in 1997 and 351 in 2002. By the same token, the number of products with exports below US$100 thousand increased from 397 in 1994 to 508 in 2002: by itself, this was a positive development that, if sustained, could lead to a greater diversification. However, this trend was not sustained as both

For a list of top twenty fourdigit SITC products in Zimbabwe's total exports in these years, see Annex Table 5 and 6.

35

48 | P a g e the number of products exported at all and that with exports above US$100,000 fell in 200208. Hence, by these two measures, the 2000s witnessed a strong tendency towards concentration of exports.

Table 19: Various indicators of concentration in 1994, 1997, 2002, 2007 and 2008 (in percent and numbers) Index 2008 1994=100 134 120 113 67 96 72 111

Fourdigits SITC sectors TOP FIVE (in percent) in total exports TOP TEN (in percent) in total exports TOP TWENTY (in percent) in total exports REMAINING 998 SECTORS (in percent) in total exports Memorandum: Number of sectors involved in exports Number of sectors exporting more or equal to US$100,000 Values of HerfindahlHirschman index

1994 45 60 72 28 774 377 666

1997 55 69 80 20 803 367 950

2002 52 65 75 25 859 351 1,188

2007 61 72 82 18 791 296 927

2008 60 72 81 19 746 270 783

Source: Derived from trade statistics reported to the UN COMTRADE by Zimbabwe's trading partners.

Yet, another measure of concentration--the HerfindahlHirschman Index (HHI)36points in the other direction, i.e., towards diversification in 200208. During both export expansion in 199497 and contraction in 19982002, the values of the HHI kept increasing indicating an increase in the concentration of exports but decreased between 200208 suggesting diversification. Despite the decline in diversification in terms of shares of dominant products and the range of products exported, the values of the HHI suggest an increase in diversification. What is then the final verdict? In order to answer this question, Table 20 gives information on the average share of the top ten exports in total exports during each phase of export development and the values of HHI during these phases not only for total exports but also for Zimbabwe's exports to its two major export markets: South Africa and the European Union. The increase in average shares of the top ten exports is less pronounced as the averages appear to moderate an upward trend shown for individual years. Furthermore, the number of sectors with exports above US$1 million fell.37 Thus, both these measures point to the increase in the degree of concentration of Zimbabwe's exports. So do the values of HHI increase, although only for exports to South Africa and not to the European Union. Zimbabwean exports to South Africa have undergone a dramatic change towards greater concentration. The values of the HHI increased from a very low level of 18000, indicating reliance on a few crop exports, to 1,426 in 200307 and 2,208 in 2008. The latter indicates a high level of concentration. In contrast, EUoriented exports have remained moderately concentrated, although the trend has been towards higher levels of concentration. The average value of HHI slightly increased

The Herfindahl index (HI) is a measure of concentration equal to the sum of the squared shares of exports of products in total exports. It assumes values ranging from zero to 10,000: zero when shares are equal and 10,000 if only one product is exported. Values between 1,000 and 1,800 are considered as depicting moderate levels of concentration. 37 The number of these products was steadily falling from 140 in 199497 to 133 in 200307 and 114 in 2008: expressed in terms of the total number of fourdigit product lines of 1018, it fell from 14 percent to 13 percent and 11 percent. At a first glance, this might strike one as a small contraction. But this is not so. Consider that the US dollar significantly depreciated during this time: assuming, for illustrative purpose, that the 2008 US dollar equals $0.75, the number of product lines in 199497 would go from 140 to 165.

36

49 | P a g e during the export retrenchment phase in 199802, fell during the expansion phase in 200307 but--as was the case for South Africa--significantly increased in 2008. Since the values of HHI of both South Africa and EUdestined exports were significantly higher than their respective averages in 200307, this suggests that hyperinflation undercut the competitiveness of a large number of firms in 200608. It is also interesting to note that within sectors such as steel, the degree of concentration of exports significantly increased (Annex Table 7).

Table 20: Shares of top ten in total exports and values of HerfindahlHirschman index of total exports and of exports to South Africa and European Union: fourdigit SITC items Average 199802 63 567 1,426 140 138 133 Index 2008 199497=100 118 1229 128 81

Share of top 10 in total (in percent) Values of HHI for exports to South Africa Values of HHI for exports to European Union Memorandum: Number of products with exports above $1 million

199497 62 180 1,109

200307 65 1,937 1,051

2008 72 2,208 1,415 114

Source: Derived from trade statistics reported to the UN COMTRADE by Zimbabwe's trading partners.

In all, one might conclude that over the past decade: (a) Zimbabwe exported very similar products but often in lower quantities and (b) its export basket has shrunk and the level of concentration has increased. The dramatic increase in the concentration of Zimbabwe's exports to South Africa has coincided with lackluster performance in most markets. Considering the proximity and intimate knowledge of the South African market by most Zimbabwean firms, this is clearly a sign of a contracting export base because of deindustrialization and higher costs of doing business.

3.2. The degree of processing embodied in exports and imports

An examination of the composition of trade in terms of enduse product categories sheds light on the changes in domestic demand for various goods as well as in export composition depending on the extent of their processing. For instance, an increase in imports of machinery usually points to an ongoing industrial restructuring effort. In a similar vein, an increase in automobiles and parts may be related to the emergence of firms specializing in production of parts or automobiles. In contrast to machinery, this would call for an indepth investigation of imports. Furthermore, the increase in the combined shares of foods and feeds and industrial raw materials merely indicates that processed goods play a growing role in a country's trade. How have Zimbabwe's exports performed according to these criteria? As a first step in answering this question, it is interesting to compare developments in trade during the two phases of export growth in 199497 and 200307 (Table 21). Note, first, that except for automobiles and parts and textiles and clothing, exports of other product groups recorded relatively strong growth. As discussed earlier, the decline of the clothing and textile sectors did not start in the 2000s but was a continuation of a trend already observed throughout the 1990s. Leaving aside fuels, the product group that experienced the strongest growth was iron and steel followed by agricultural foods and feeds. In marked contrast, the phase of modest rebound in 200307 characterized a greater dispersion in export growth rates and a larger number of enduse products fell in terms of value.

50 | P a g e But there were other more telling contrasts between the two expansion phases. The most glaring contrast is a steep contraction in exports of foods and feeds and acceleration in the growth of industrial raw materials during the second phase. The average growth rate in exports of foods and feeds changed by 20 points: it moved from a positive growth rate of 11 percent to a negative growth rate of 9 percent that was already preceded by falling exports in 199802. Exports of industrial raw materials, which grew at an LSG rate of 3.2 percent in 199497 exploded to 31 percent in 200207 emerging as the single most important driver transforming Zimbabwean exports. With exports of foods and feeds falling and industrial raw materials expanding, the shares of other enduse product groups have remained relatively unchanged.

Table 21: Developments in Zimbabwe's exports in terms of enduse product categories Exports Agricultural Food & Feeds Industrial Raw Materials Iron and steel Machinery, excluding auto Automobiles & Parts Textiles & Clothing Consumer Goods Fuels Total Least Square Growth rate in Annual 199497 199702 200207 2008 11.4 1.8 9.4 0.6 3.2 1.4 31.1 26.8 17.7 8.9 11.4 42.1 9.8 4.2 2.4 23.4 2.6 6.7 18.4 52.5 15.2 4.1 5.0 37.1 14.4 4.1 3.9 9.0 32.2 17.6 5.1 86.9 9.9 2.9 6.3 11.0 Average share in 199497 199702 200207 57.6 61.0 40.5 12.8 12.2 32.5 10.4 9.2 10.1 2.1 1.7 2.2 0.4 0.3 0.6 5.2 3.5 2.0 10.3 11.1 10.9 1.1 1.0 1.3 100 100 100 Share in 2008 33.4 40.1 14.7 1.5 0.3 1.2 7.7 1.0 100

Note: (1) The end-use categories are defined as Agricultural Food & Feeds (SITC 0+1+2+4-27-28); Industrial Raw Materials (SITC 27+28+68);Iron and Steel (SITC 67); Machinery, excluding automobiles (SITC 7-78), Automobiles & Parts (SITC 78), Other Consumer Goods (SITC 5+6+8+9-67-68); Fuels (SITC 3); and All Goods (0 to 9): (2) The data does not include exports of aircraft etc parts (SITC 7929) valued at US$91 million in 2008 attributable to an error as there are no indications of any significant production of aircraft parts in earlier trade data and other available information. Source: Based on world's reporting to UN COMTRADE Statistics.

But the relative stability in the commodity composition of Zimbabwe's exports of other product groups (i.e. excluding industrial raw materials and agricultural food and feeds) falls well short of capturing the extent to which Zimbabwe's integration into global markets has changed since the peak of exports expansion in 1997. The most notable change is that the degree of import coverage by exports declined and imports of foods and feeds exploded. Although the level of processing embodied in exports, as captured by changes in the weight of foods and feeds together with industrial raw materials,38 has not changed over the last decade, the level of processing embodied in Zimbabwe's imports has fallen (Table 22). More disturbingly, however, the balance of trade, although still in surplus, in traditional inputs has dramatically deteriorated as a result of a threefold increase of these types of imports in terms of value between 2002 and 2008. Although Zimbabwe is a net exporter of traditional inputs, exports of agricultural foods and feeds as a proportion of the imports in these products fell from 530 percent in 2002 to 116 percent whereas exports of industrial raw materials in terms of their imports increased

It increased during the commodity boom in 200207 and slightly contracted in 2008. The aggregate share of foods and feeds and industrial raw materials, however, was even slightly lower than in 2008: 72 percent versus 74 percent.

38

51 | P a g e from 346 percent to 437 percent. However, this was only around half the value of this indicator in a peak year 1997 as compared with 2007.

Table 22: Developments in Zimbabwe's trade by enduse product categories in selected years over 19972008 (in millions of US dollars and percent) Agricultural Food & Feeds Industrial Raw Materials Iron and steel Machinery, excluding autos Automobiles & Parts Textiles & Clothing Consumer Goods Fuels Total Agricultural Food & Feeds Industrial Raw Materials Iron and steel Machinery, excluding autos Automobiles & Parts Textiles & Clothing Consumer Goods Fuels Total 1997 2002 2007 2008 Index 2007 in millions of US dollars 1997=100 1,248 239 224 44 6 71 202 25 2,059 693 803 321 7 2 57 29 132 99 1,044 233 128 37 9 52 175 13 1,691 530 346 470 17 6 90 46 28 148 707 1,152 217 42 15 40 177 12 2,361 203 323 377 8 7 85 27 13 104 703 843 308 32 7 25 161 22 2,101 116 437 569 6 3 51 29 25 95 57 482 97 95 264 57 88 47 115 17 54 177 95 141 88 101 19 96 1997 2002 2007 2008 Composition of exports (in %) 60.6 61.8 29.9 33.4 11.6 13.8 48.8 40.1 10.9 7.5 9.2 14.7 1.5 2.1 2.2 1.8 0.3 0.5 0.6 0.3 3.4 3.1 1.7 1.2 9.8 10.4 7.5 7.7 1.2 0.8 0.5 1.0 100 100.0 100.0 100 Composition of imports (in %) 8.7 1.4 3.4 32.1 13.9 5.9 33.6 0.9 100 17.2 5.9 2.4 19.0 13.1 5.0 33.3 4.1 100.0 15.3 15.7 2.5 23.4 8.9 2.1 28.3 3.8 100 26.2 8.4 2.3 22.2 11.1 2.2 23.8 3.8 100

(exports in percent of imports)

Note: (1) The enduse categories are defined as Agricultural Food & Feeds (SITC 0+1+2+42728); Industrial Raw Materials (SITC 27+28+68);Iron and Steel (SITC 67); Machinery, excluding automobiles (SITC 778), Automobiles & Parts (SITC 78), Other Consumer Goods (SITC 5+6+8+96768); Fuels (SITC 3); and All Goods (0 to 9): (2) The data does not include exports of aircraft etc parts (SITC 7929) valued at US$91 million in 2008 attributable to an error as there no indications of any significant production of aircraft parts in earlier trade data and other available information. Source: Based on world's reporting to UN COMTRADE Statistics. Figure 7: Developments in trade balances of traditional inputs and Zimbabwe's total trade balance in 199408 (n millions of US dollars) 1,200 1,000 800 600 400 200 0 200 400 1994 1997 2001 2002 2003 2004 2005 2006 2007 2008 Agricultural Food & Feeds Industrial Raw Materials Iron and steel Trade balance

Source: as in Table 22.

52 | P a g e The combination of falling exports and rising imports of foods and feeds has been responsible for the contraction in the surplus of traditional inputs in Zimbabwe's trade. Despite strong growth performance in net exports of industrial raw materials, the contraction in net exports of foods and feeds was steep. The surplus in trade of agricultural foods and feeds fell from US$1.1 billion in 1997 to US$ 99 million in 2008, whereas aggregate surpluses in industrial raw materials and iron and steel increased from US$364 million in 1997 to US$ 995 million in 2007 and fell in 2008 to US$905 million (Figure 7 above). Another indication of slow growth in the industrial base has been developments in imports of capital equipment and industrial raw materials. Although the latter grew at an impressive average rate of 38 percent per year in 200407, this does not appear to be the result of growing domestic production of more processed goods but rather high prices for primary commodities during this period. Consider that the average rate of exports during this period was 34 percent. Moreover, the average rates including the contraction of both exports and imports in 2008 were 22 percent for exports and 21 percent for imports. But there is no similar ambiguity if another indicator of industrial development (i.e. imports of capital equipment) is taken into account. The share of imports of machinery in total imports fell from an average of 32 percent in 199497 to 28 percent in 19982001 and 21 percent in 200208. The shift in the composition of imports towards agricultural products and away from consumer goods points to the rapidly deteriorating standards of living and growing pauperization of society over this period. The share of agricultural products in total imports has dramatically expanded since 1997. It increased from 9 percent in 1997 to 17 percent in 2002 and 26 percent in 2008. Simultaneously, the combined share of cars, clothing and other consumer goods in imports fell from 53 percent in 2006 to 37 percent in 2008. The average in 19942002 was 51 percent with a very small dispersion of annual data.

Table 23: Direction of Zimbabwe's exports by enduse to South Africa, European Union and rest of the world in 1997 and 2008 (in percent)

Agricultural Food & Feeds Industrial Raw Materials Iron and steel Machinery, excluding autos Automobiles & Parts Textiles & Clothing Consumer Goods Fuels Total 1997 2008 Index 2008 2002=100 182 247 76 333 52 994 71 84 317 1997 2008 Index 2008 2002=100 60 47 134 54 7 53 56 0 58 1997 2008 Index 2008 2002=100 126 37 80 59 122 88 130 118 83

Share of S. Africa 11 7 2 35 77 14 38 32 13 12 71 2 50 7 30 14 43 36

Share of EU 52 26 57 51 8 79 56 24 50 27 11 50 3 1 23 14 0 22

Share of ROW 37 67 41 14 15 7 6 44 37 60 18 48 46 92 46 72 57 42

Source: Based on world's reporting to UN COMTRADE Statistics.

Major markets for Zimbabwe's exports byend use product categories changed between 1997 and 2008: driven by declining exports to the EU. While around 80 percent of exports of textiles and clothing were shipped to the EU in 1997, only 23 percent were sold there in 2008 (Table 23 above). Since the value of these exports contracted around 70 percent, other markets absorbed only a small fraction of the lost sales to the EU. The EU, however, accounted for 50 percent of iron and steel exports in 2008, only 7 percentage points below the 1997 level. On the other hand, South Africa emerged as the major market

53 | P a g e for industrial raw materials: its share increased from seven percent in 1997 to 71 percent in 2008. Except for three product groups--food and feeds, consumer goods, and fuels, whose combined exports accounted for 43 percent of the total in 2008--South Africa and the EU take more than half of other exports. The preceding discussion leads to the following observations: First, the level of processing embodied in Zimbabwe's export basket has remained unchanged since 1994. The share of primary inputs has remained at slightly below threequarters of total exports independent of their levels. Second, exports of food and feeds, iron and steel and consumer goods were the main drivers of export growth during the growth phase in 199497. The picture changed during the period of recovery in exports in 20032007 in two important respects: first, food and feeds contracted, and second, consumer goods ceased to be a driver. Industrial raw materials topped them all. Third, except for iron and steel, positive net exports across all enduse product categories have been falling. Not surprisingly, the largest decline has been in food and feeds and industrial raw materials. Finally, the shift in imports from consumer goods to food products is an indication of declining standards of living and a declining surplus in agricultural output, as discussed below

3.3. Agriculture and agroprocessing: contraction of net exports

Since agricultural output significantly contracted in response to land reform, interesting questions worth exploring in more detail concern the impact this has had on foreign trade performance of raw foods and agricultural products in 19972008 and product groups in which Zimbabwe ceased to be a net exporter. In order to address these issues, we extract trade data, as reported by Zimbabwe's trading partners, grouped through the filter of agricultural products in terms of SITC items developed in a paper by Ng and Aksoy (2008). They distinguish between four broad categories: the first category is called "raw food" and it includes meats and dairy, grains, and fruits and vegetables; the second category, referred to as "cash crops," includes tropical foodstuffs, feeds, oil seeds and tobacco; other food including processed food and seafood; and agricultural raw materials or nonfood products. Table 24 presents net exports of these products in 19972008. Exports of all these products still exceeded their imports in 2008, although the surplus, (i.e. net exports) had kept falling since 2001. The contraction was particularly acute in 2006 08, when they fell from US$482 million to US$150 million. We do not have data for 2009, but it seems that net exports somewhat rebounded mainly thanks to an increase in exports of tobacco, Zimbabwe's largest exportable among agricultural products.39 While, with the exception of seafood, net exports in 200508 were lower than in 19972001, some fared better than others. The least affected were agricultural raw materials. The value of their total net exports was on average only around 20 percent lower than in 19972001 as compared to a 50 percent contraction in exports of cash crops (mainly tobacco). In terms of magnitude, by far the largest change was in net exports of raw food (mainly maize): the pendulum swung from an average of US$99 million of net exports in 19972001 to net imports of US$ 166 million on average in 200508. Net exports moved to negative territory in the third year into the land reform program in 2002.

According to the data from the Reserve Bank of Zimbabwe, the value of these exports increased from US$229 million in 2008 to projected US$301 million in 2009.

39

54 | P a g e

Table 24: Net exports of raw food and agricultural products in 19972008 (in current millions of US dollars)

Meats and Dairy Products, of which Beef, fresh, chilled or frozen Grains and Cereals, of which Wheat and meslin Maize (corn), unmilled Vegetables and Fruits, of which Vegetables, fresh, chilled or frozen Fruits, citrus etc. Tropical Products, of which Sugars, beet and cane, raw, solid Coffee, green, roasted or sub Tea and mate Spices Feeds, Oilseeds and Tobacco, of which Tobacco, unmanufactured; tobacco refuse Oil seeds and oleaginous fruits Processed Food, of which Meat & edible offals, prep. & preserved Milk & cream, preserved, concentrated Meals and flour of wheat, other cereal preps. nes Refined sugars and other products Bran, oil cake, meal fodder and other food wastes Misc. edible products and preparation Non alcoholic beverages nes Alcoholic beverages Tobacco manufactured Animal/vegetable oils and fats, processed Seafood, of which Fish, fresh (live or dead), chilled, frozen Fish, dried, salted or in brine ; smoked Fish, crustaceans and molluscs, prep. Agricultural Raw Materials, of which Hides, skins and furskins, raw Cork, wood, pulp and waste paper of which: softwood (2482) Textile fibres, silk, cotton, jute etc. of which: raw cotton, excl linters Crude animal and vegetable materials of which cut flowers Total above 1997 40.8 35.7 60 6 70 49 21 23 91 35 37 8 12 647 652 5 2 8 2 7 2 1 4 0 3 6 9 8 5 4 1 259 6 11 13 137 146 56 58 1,140 1998 38.2 28.0 18 1 15 52 23 24 121 57 34 10 20 487 485 0 10 8 0 4 0 3 3 0 2 2 6 4 1 0 3 246 5 5 10 119 126 62 62 920 1999 41.4 29.4 31 8 37 62 26 30 98 48 23 11 16 640 637 2 10 4 0 4 1 4 2 0 3 10 7 3 0 1 4 245 4 5 10 113 117 64 62 1,047 2000 46.0 30.5 2 5 3 55 22 28 88 42 15 20 12 540 531 8 12 3 5 6 2 1 0 2 1 2 6 2 1 1 6 295 7 9 11 141 146 73 68 1,041 2001 32.1 19.5 3 1 5 60 18 38 89 55 10 17 7 598 587 10 40 2 4 16 5 5 1 3 2 4 1 9 3 1 6 293 6 3 8 149 152 69 68 1,123 2002 11.6 1.0 88 4 82 55 18 34 77 40 10 15 12 613 615 2 12 1 3 9 5 2 1 0 2 10 3 10 1 1 8 234 7 7 8 89 92 67 67 926 2003 11.4 1.8 99 5 87 56 18 34 74 36 8 17 13 520 527 7 19 3 0 5 5 1 4 0 1 19 11 12 2 0 8 124 7 103 10 86 89 69 69 717 2004 9.4 0.0 102 9 85 27 3 29 101 66 10 16 9 368 369 1 18 5 0 8 14 0 8 1 1 25 13 12 4 2 10 313 8 17 17 185 190 53 55 748 2005 6.5 0.0 161 0 154 47 11 33 90 62 8 15 5 246 257 11 76 4 1 102 14 1 5 0 2 30 15 10 0 1 10 255 10 21 19 140 145 45 44 416 2006 11.2 0.0 92 6 77 42 13 26 91 67 7 14 3 189 190 1 31 4 4 5 15 5 1 0 2 28 14 11 0 1 12 199 12 23 20 110 113 30 30 482 2007 25.7 0.1 165 9 150 38 3 39 49 34 4 8 2 254 256 2 10 3 2 43 20 2 3 1 2 33 19 16 4 0 13 202 11 20 18 110 114 34 34 411 2008 4.9 0.1 385 21 354 21 7 24 68 55 3 8 3 292 272 20 79 1 1 62 15 1 11 4 13 21 22 18 5 0 13 210 16 18 15 116 117 32 30 150

Sources: Partners' statistics reported to the UN COMTRADE based on taxonomy of agricultural products in Ng, Francis and M. Ataman Aksoy. 2008. "Who are the net food importing countries?" Policy Research Working Paper 4457, World Bank, January.

While output has fallen across all raw food sectors, the largest contributor to the contraction in net exports of raw food was maize. In the 19942001 period imports of maize exceeded their exports twice

55 | P a g e in 199899. In most years until 2002, Zimbabwe was a traditional net exporter of this crop. The situation changed beginning in 2002: when Zimbabwe's imports started to exceed its exports each year with the gap reaching the record level of US$354 million in 2008. Excluding maize, Zimbabwe is a net exporter of raw food, although net exports have fallen from US$63 million in 2002 to US$55 million in 2007 and US$5 million in 2008. Furthermore, excluding maize total net exports of raw food and agricultural products in 2008 would amount to $448 million rather than US$150 million. The processed food sector, whose importance stems not only from its contribution to foreign currency earnings but also to employment creation, appears to have somehow withstood adversities of the economic environment. The value of exports of processed food was higher in 200507 than in the peak years in 199506. Although net exports were negative in 200708, the sector overall recorded a positive growth with an annual average LSG rate of 8.7 percent over 19972008. The strongest growth, LSG rates above 20 percent per year, was experienced in manufactured tobacco, refined sugars and other products, other sugar syrups, artificial fruits and vegetable juices. The overall position of this sector was negatively affected by the expansion of imports of cereal preparations, meals and flour of wheat. Excluding these products, the sector was a positive net exporter with an average of US$23 million per year in the 2000s except in 2008 (minus US$17 million). Although their contribution to foreign exchange earnings pale in comparison to revenue derived from tobacco and agricultural raw materials whose total net exports averaged around US$600 million a year in 200008, the importance of processed goods has to be measured in terms of their employment effect.40 In all, the smallest contraction was in net exports of "agricultural raw materials" followed by "cash crops" (mainly tobacco). Zimbabwe became a net importer of both "other food"--almost exclusively because of huge imports of meals and flour of wheat and other cereal preparations--and "raw food" due to the fall in output of maize. Relative to the average in 199701, the largest contraction in net exports was in "other food."

3.4. Factor and technology content of foreign trade and revealed comparative advantage

Changes in the pattern of trade in goods reflect differences in comparative advantage as determined by different factor endowments. A country tends to export those goods, which use its factors in relative abundance. Exploring a full causal chain linking factor endowments, comparative advantage and trade patterns is not relevant for this discussion. The general question here concerns an assessment of broad changes in relative factor intensities and technological content as revealed by Zimbabwe's total exports. Natural resourcebased products have driven Zimbabwe's export dynamics: two peak points of expansion phases represented the peak share of resourceintensive exports in total exports in each cycle except in 2007. Indeed, their growth was the fastest during periods of expansion and their contraction brought down the value of total trade. Their share in total exports has been growing since 1994, peaked at 88 percentage points in 1997 fell to 83 percent in 2003, and climbed to 91 percent in 2007 and grew further to 92 percent in 2008. Although net exports of natural resourcebased products remain high, as

These amounts were below the levels achieved in the 1990s, except for net exports of skins and fur skins and soft wood.

40

56 | P a g e the value of these exports was more than twice that of imports, the trend is unidirectional: exports in terms of imports fell from the average of 510 percent in 199497 to 164 percent in 2008 (Table 25). The narrowing of the gap between Zimbabwe's South Africa and EUdestined exports in terms of factor intensities also underlies a growing reliance on natural resources. While the share of natural resources in EUoriented exports was on average 23 percentage points larger than in Zimbabwe's exports to South Africa in 199497, the difference fell to an average of 10 percentage points in 200207. By 2008, the difference dropped to 3.3 percentage points, and two export baskets have almost fully converged (Table 25).

Table 25: Change in factor content of Zimbabwe's trade in goods in 199497, 19972003 and 200207

LSG rate in 199497 199702 11.0 13.0 1.8 1.7 2.9 1.1 0.8 3.0 Annual Average share in total exports 2008 199497 199702 200207 6.7 19.3 13.1 21.4 21.5 0.7 16.5 15.2 83.8 7.5 3.0 5.6 85.7 5.6 3.1 5.6 85.9 4.5 3.2 6.4 Share in 2008 92.0 2.7 2.1 3.2

Exports

Natural Resources Unskilled labor Capital intensive Skilled labor Imports Natural Resources Unskilled labor Capital intensive Skilled labor Memorandum: Natural Resources Unskilled labor Capital intensive Skilled labor

200207 7.6 4.3 4.0 2.9

14.0 4.1 17.4 18.1 14.2 0.6 10.8 18.7 15.4 9.6 15.4 11.4 Exports in percent of imports 199497 199702 200207 510 82 6 15 441 84 9 26 224 80 10 26

14.6 24.0 38.0 47.0 8.1 8.1 6.0 3.7 44.0 42.2 32.7 28.5 33.3 25.7 23.3 20.8 Absolute difference: S. Africa and EU 2008 199497 199702 200207 2008 164 60 6 13 22.9 6.4 5.3 11.2 21.7 6.5 5.4 9.8 10.1 4.9 0.7 4.5 3.3 1.6 0.6 1.1

Source: Based on partners data from UN COMTRADE Statistics.

The rise to prominence of natural resource intensive exports in Zimbabwe's export bundle has not been the result of different rates of expansion of products with different factor endowments but, rather, shrinking exports in other categories including those of unskilledlabor intensive products. The latter development is particularly worrisome considering that Zimbabwe has a large pool of unemployed labor. Paradoxically, exports of these products experienced the second largest decline: the negative LSG rate of 4.3 percent was larger than the pace of decline of capital intensive exports of 4.0 percent but lower than that of skilled labor intensive products. However, exports of skilled labor intensive products recorded the largest decline in 2008. Their share in total exports fell from an average of 6.4 percent in 200207 to just 3.2 percent in 2008. Developments in Zimbabwe's exports have contributed to the increase in unemployment, as exports of laborintensive products stagnated or contracted. Zimbabwe's revealed comparative advantage (RCA)41 in world markets has been increasingly concentrated in products with low level of processing and high natural resource content. Since the

The RCA is the ratio of the share of a given product in BH's total exports to the share of that same product in the world's total imports. The RCA with a value that exceeds unity suggests a strong specialization in the product (Balassa, 1965).

41

57 | P a g e classification used above to identify exports in terms of their factor content does not distinguish between levels of technology and assigns most agricultural products to resourcebased products,42 a better filter to account for the dominance of natural resourcebased products is the simple classification developed by Landesman and Stehrer (2003). It distinguishes among three broad categories of production activities: (1) Low technology and labor intensive activities, (2) resource intensive activities, and (3) medium to hightechnology production activities. Low technology and labor intensive activities include, among others, agricultural foods and feeds, some animal and vegetable oils, simple manufactured goods, textiles and clothing. Resource intensive activities, accounting now for around twothirds of Zimbabwe's total exports, cover such sectors as mining, steel and iron, chemical industries, and simple industrial products based on intensive use of natural resources (e.g., wood materials, cement, alloys, etc). Medium to hightechnology intensive products include machinery and transport equipment as well as some miscellaneous manufactures such as medical furniture parts or instruments. The long term picture that emerges from examining the development of exports in terms of technology content can be summarized as follows: Zimbabwe's exports of medium to hightechnology products, in terms of value, were flat in 19942008; exports of low technology and labor intensive products fell in 19972008; whereas those of resource intensive products increased dramatically in 200207. Zimbabwe's exports over 19972008 switched from a dominance of products characterized by low technology labor intensive activities to resource intensive ones while exports of medium to high technology products stagnated. Their value varied between US$70 million in 2008 and US$112 million in 19942008. In 2004, resource intensive exports exceeded lowtech labor intensive exports (Figure 8). Zimbabwe's major exports include nickel, ferro alloys, tobacco, raw cotton, cement, precious metal scrap, cut flowers, and citrus fruits. These products account for the bulk (around 70 percent) of Zimbabwe's total exports. They fall within two groups of industrial activities: resource intensive and low technology labor intensive products. The major product in the latter group is tobacco, which accounted for around 50 percent of these exports in 200208. Nickel ores and nickel alloys emerged as Zimbabwe's major exports in the 2000s (Table 26). Their aggregate share in resource intensive exports rose from 22 percent in 2002 to 60 percent in 2007 and then fell to 49 percent in 2008 almost exclusively due to the fall in their world prices over this period (see Section 2.4).

The goodness of results obtained hinges critically on the quality of a classification used to examine export baskets over time by factor mix. The choice is always controversial. There are woeful difficulties to define and measure factor intensity, and trade theorists have long wrestled with it. Special problems emerge when a classification aims to capture "quality" of factors involved (Winters 1997). Some definitions of the groups of goods by factor intensity are overlapping and nonexhaustive. Definitions used here do not suffer from these shortcomingsall industries are taken into account and an industry appears only in one classification; and the classification distinguishes among four types of factors. Since some industries are intensive in terms of more than one factor, the results may be distorted. But even assuming that the initial classification captures adequately factor proportions at a given point of time, with the passage of time it may provide a distorted picture. Some industries may become more capitalintensive or less active in technological terms.

42

58 | P a g e

Figure 8: Exports of lowtechnology labor intensive products, resource intensive and medium to hightech products in 1994 2008 (in millions of US dollars)

1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 Lowtech, labor intensive Resource intensive Medium to Hightech

Source: Own estimates based on data from the UN COMTRADE database Table 26: Exports in terms of technology intensities in 1997 and 200208 (in millions of US dollars and percent) 1997 Lowtech, labor intensive 1,240 of which: tobacco (121) 692 Resource intensive 732 of which: nickel ores (28) 4 nickel alloys (68) 117 Medium to Hightech 106 Total reported 2,078 Lowtech, labor intensive of which: tobacco (121) Resource intensive of which: nickel ores (28) nickel alloys (68) Medium to Hightech 60 56 35 1 16 5 2002 1,007 627 604 55 79 101 1,712 59 62 35 9 13 6 2003 2004 2005 934 822 708 544 400 284 805 1,010 1,141 132 181 221 137 158 227 112 73 108 1,851 1,904 1,956 (in percent) 50 58 43 16 17 6 43 49 53 18 16 4 36 40 58 19 20 5 2006 685 235 1,387 343 340 112 2,185 31 34 63 25 24 5 2007 667 290 1,633 436 537 82 2,381 28 43 69 27 33 3 30 47 67 34 15 3 2008 638 297 1,416 479 219 71 2,125

Source: Based on partners data from UN COMTRADE Statistics.

In contrast, the value of exports of medium to high technology intensive products was stuck between US$70 million (in 2001) and US$112 million (2006) and their degree of concentration was lower than in other product categories. Ceramic plumbing (8122), automotive parts (7812), electrothermic equipment (7758) and wood furniture (8215) were major exportables in 2008. These five threedigit SITC sectors contributed 42 percent to total exports of medium to high technology intensive in 2008. By this measure, the level of concentration was 63 percent in low technology labor intensive products and 79 percent in a resource intensive group. The combination of falling exports of lowtech labor intensive products and expanding exports of resource intensive products in 200207 has been responsible for a dramatic shift in technology content of Zimbabwe's exports. The contraction of exports of lowtech labor intensive products in terms of value

59 | P a g e since 1997 has been nothing short of dramatic. They have been consistently falling since 1997: in 2008 they stood at 51 percent of their peak value in 1997. The surge of resource intensive products coincided with the commodity boom of 200307. Indeed, the share of resource intensive products in total exports increased from 35 percent in 2002 to 44 percent in 2003 and 58 percent in 2006. It stood at 67 percent of total exports in 2008, while the share of lowtech labor intensive fell from an average of 60 percent in 199497 to 30 percent in 2008 (Table 27). Although resource intensive exports grew fastest during the growth phase in 199497, exports of other product groups also experienced a strong growth. During the next two phases, resource intensive exports contracted the least in 19982002 and grew the fastest in 200307 whereas exports of other products experienced negative growth.

Table 27: Dynamics of Zimbabwe's trade in terms of technology and factor intensities in 199497, 19982002, 200307, and 2008 (in percent and millions of dollars)

Exports Lowtech, labor intensive Resource intensive Medium to Hightech int. Imports Lowtech, labor intensive Resource intensive Medium to Hightech int. Memorandum: Lowtech, labor intensive Resource intensive Medium to Hightech int. 199497 533 89 9 21.1 10.6 10.1 6.5 6.3 22.3 199702 532 117 16 8.3 17.4 11.2 200207 238 144 17 69.5 25.4 4.3 2008 97 173 8 10.7 36.1 53.2 199497 883 92 960 15.4 42.4 42.2 199702 848 91 507 19.8 47.1 33.1 200207 418 343 491 28.4 35.2 36.4 2008 19 600 774

Least Square Growth rate

199497 8.2 12.2 9.3 19972002 3.1 2.4 3.3 200207 9.0 19.2 1.8

Annual

2008 4.3 13.3 14.4 199497 60.0 34.6 5.4

Average share in

19982002 58.8 36.5 4.7 200207 37.8 57.4 4.8

Share in

2008 30.0 66.7 3.3

Exports in percent of imports

Trade Balance (US$ millions)

Source: Own estimates based on data from the UN COMTRADE database

Developments in imports coverage by exports by technology, natural resource and labor intensities paint a familiar picture: net exports of both high and medium technology intensive products and low technology labor intensive products continued to fall in 200608 while those of resource intensive products kept increasing. The collapse of net exports of low technology intensive products, mainly of agricultural products, resulted in the deficit in the balance of trade in these products of US$19 million in 2008.

3.5. Victims and survivors: new and vanishing specializations

This section takes a broad look at the change in the composition of Zimbabwe's exports at the level of fourdigit SITC sectors. In order to eliminate single, random commercial transactions, we average the exports data for three periods identified by the earlier examination of export dynamics: the 199407 expansion; contraction in 19982002; and modest rebound in 200307. Average exports of fourdigit SITC in terms of value in each period can be then compared with each other or from a vantage point in a selected year. The broad picture that emerges from this analysis can be summarized by the following three observations: The range of fourdigit sectors involved in exports significantly declined in 200308;

60 | P a g e The losses were more widespread in term of the number of product lines and their values of exports more uniformly distributed while increases in the value of exports were limited to a smaller number of product lines; Survivors included exporters of minerals and other lowprocessed raw materials, which registered modest gains in terms of value, whereas victims included exports of agricultural products and those of unskilled labor intensive clothing.

Zimbabwean firms turned inward or went out of business in the 2000s. As a result, Zimbabwe's export basket became less diversified in terms of the number of fourdigit SITC sectors especially in 200508. During this period, the number of sectors that were not involved in exports (the value of exports equal zero) dramatically expanded indicating either a shift to the domestic market or exit from production altogether. Out of a total of 1,018 fourdigit SITC sectors, 49 sectors were not involved in exports in 199497, this number dropped to 31 in 19982003, and increased to an average of 63 in 200307. But, the average conceals developments in 2007 when the number of zeroexported products increased to 227 and in 2008 their number further increased to 272 sectors with zeroexports up from 159 in 2002. Some further insights can be derived from the differences between exports of fourdigit SITC items in 2008 and their average values in 199407, 19982002 and 200307. Although these differences are in current dollars, i.e. they do not take into account changes triggered simply by falling or rising prices, they identify sectors that experienced the largest changes in exports in 2008 visàvis averages for earlier periods. Table 28 provides information on the number of fourdigit SITC sectors with exports higher (survivors) or lower (victims) in 2008 than in a previous phase; the share of top ten in total gains or losses; and the top five fourdigit SITC items with the largest positive or negative differences in the values of exports in 2008 and respective phases. Three observations can be derived from eyeballing the data in Table 28. First, the gains appear to have been highly concentrated, i.e. limited to around ten fourdigit SITC sectors. The 2008 export basket, when set against averages during the three periods, shows little change in terms of numbers of sectors that expanded their exports and their share in total gains. Note that the number of sectors referred to as survivors varies between 205 and 212, while the share of the top ten in the total increase of exports varies between 80 percent and 85 percent. Second, producers of minerals and their products have displayed the largest resistance to economic adversities while exporters of agricultural products were the main victims, albeit there were some exceptions also among manufactured exporters. Except for raw cotton, Portland cement and oil seed, mineral exporters figured among the top five export survivors with nickel, both ores and mattes, topping the list in each case. New minerals became available for exports in the 2000s: nickels (SITC 284) top them all. The value of these exports never exceeded US$10 million before 2002 when their value reached US$55 million and grew each year to reach US$479 million in 2008 accounting for 22 percent of total exports. Other examples, admittedly much less spectacular than nickel, include copper (both ores and refined), diamonds, and common salt. There also some examples of emerging manufacturing exports: exports of ceramic plumbing fixtures (SITC 8122), which over 19942007 only once reached US$90,000 in 1994, rose to US$11 million and so did exports of steel structures and acyclic mono hydrogen alcohols. Among agricultural products, two sectors deserve mention: soya beans, whose

61 | P a g e exports collapsed in 200107, rebounded in 2008 to US$12 million; and oil seeds, whose exports soared from an of average US$50,000 in 19942007 to US$12 million in 2008 placing this sector among the top five survivors (see Table 28).

Table 28: Victims and survivors in 2008 against the average in 199497, 19982002 and 200307 (in millions of US dollars and fourdigit SITC product lines)

2008 vs.199497 # of sectors 212 10 10 766 10 10 240 234 125 110 32 135 85 49 38 37 383 2008 vs. 19982002 # of sectors Loss/ Gain 205 10 10 790 10 10 242 223 156 133 28 215 60 35 31 26 421 1,160 942 81% 739 456 62% 2008 vs. 200307 # of sectors 219 10 10 744 10 10 168 122 48 17 12 61 43 20 16 14 159

Loss/ Gain 1,150 925 80% 767 447 58%

.

Loss/ Gain 586 500 85% 427 200 47%

Survivors top ten share in total gain Losers top ten losers share in total loss

Survivors top ten share in total gain Losers top ten losers share in total loss

Survivors top ten share in total gain Losers top ten losers share in total loss Nickel mattes/sinters (2482) Other ferro alloys (6715) Nickel ores (2841) Prec. metal waste (2892) Oil seed/etc (2239)

top five survivors

Nickel mattes/sinters (2482) Nickel ores (2841) Other ferro alloys (6715) Nickel/alloys unwrought (6831) Portland etc cements (6612)

Nickel mattes/sinters (2842) Nickel ores/concentrates (2841) Other ferro alloys (6715) Nickel/alloys unwrought (6831) Raw cotton, excl linters (2631) Tobacco stripped/stemmed (1212) Tobacco, not stripped (1211) Asbestos (2784) Maize ex sweet corn nes (0449) Beef, fresh/chilled (0111) Memorandum: difference in total exports

top five losers

Tobacco stripped (1212) Tobacco, not stripped (1211) Cut flowers/foliage (2927) Asbestos (2784) Gold non monetary (9710)

Nickel/alloys unwrought (6831 Tobacco stripped/stemmed (1212) Asbestos (2784) Cut flowers/foliage (2927) Raw cotton, excl linters (2631)

Source: Derived from trade statistics reported to the UN COMTRADE by Zimbabwe's trading partners.

On the other hand, exports of agricultural products overshadowed other sectors in terms of contraction but almost exclusively when exports in 2008 are compared with averages during the phase of export growth in 199497 with the exception of tobacco (both stripped and not stripped). Tobacco was the major victim accounting for 29 percent, 37 percent and 10 percent of losses in 2008 calculated against average exports in 199497, 19982002 and 200307, respectively. Maize and beef were among big "absentees" in 2008 exports when cast against exports in 199497 but not against those in 19982002 as they became so miniscule that they failed to qualify as big losers when their exports in 2008 are compared to exports in the subsequent two phases. But cut flowers did: and they took a hit in relation to their exports during both the contraction and the moderate rebound phase.43 As for manufactured goods, there were plenty of losers, i.e. whose exports either completely disappeared or were less than 50 percent in 2008 of their peaks in 19942007. Exports of furniture (SITC 8215) peaked in 2005 and then fell precipitously. So did exports of precious metals and jewels (SITC 8973), footwear (SITC 8514) and industrial washing dryers (SITC 7452). Although exports of clothing,

Exports of cut flowers were affected not only by the fall in volumes but also by flat world prices in the 2000s (see Table 11 in Section 2.4 of this report).

43

62 | P a g e mainly those of mens/boys trousers (SITC 8414), also dramatically declined from US$23 million in 1994 to US$7 million in 2008, their collapse cannot be attributed to an unfriendly economic environment in 200008 but rather the result of much earlier declines in competitiveness due to increasing labor costs and increased competition from large Asian producers. Exports of clothing (SITC 84) have declined in all phases: they fell from US$52 million in 1994 to US$39 million in 1997, US$29 million in 2002 and US$13 million in 2008. Third, the number of sectors that experienced contraction by far exceeded those that registered increases in their exports. Moreover, since this number displays a relatively low variation (with the number of declining sectors ranging between 790 in 19982002 and 744 in 200307) this suggests a consistent pattern of export contraction. In other words, some sectors were exporting less and less of the same products. Returning to an earlier question whether the fall in the number of sectors suggests an increase in specialization and greater involvement in global division of labor, the answer is unambiguously negative. The developments in Zimbabwe's exports in 200308 point to a significant retrenchment in export activities caused either by external developments or by domestic policies or both. The exit of firms from export activities and fall in exports of around 70 percent of all fourdigit SITC products more than offset expanding exports by the remaining sectors of the economy. Hence, the fall in the number of products in Zimbabwe's export basket was an indication of a shrinking export base in 200308.

3.6. Revealed comparative advantage: predictable shifts from agriculture to extractive sectors

Despite a weak overall export performance in 19982008, some of Zimbabwe's firms still succeeded to increase their presence in world markets. As can be expected from the previous discussion, the loss of external markets was the most serious for agricultural products. But other sectors, including both low technology laborintensive and medium to hightechnology intensive products, have also lost their competitive edge. Yet, the total number of threedigit SITC sectors with revealed comparative advantage (RCA) above unity did not change between 1997 and 2008: in both years there were altogether 42 product groups. But the composition did change reflecting developments in Zimbabwe's exports with the number of resource intensive products increasing and the number of both low technology, unskilled laborintensive and medium to hightechnology intensive products slightly falling. Although the number of threedigit SITC sectors with comparative advantage in world markets for low technology, labor intensive products only slightly fell from 21 to 20, there was a change in their composition (Table 29). Ten out of 22 product groups had comparative advantage in both 1997 and 2008. New comers in 2008 i.e. 12 product groups, were exporterd in both 1997 and 2008. Products that had RCA exceeding unity in 1997 but not in 2008 were such food products as maize, meat, wheat, and cereal. Other products included gold, cotton, tea and charcoal. Products that gained comparative advantage in 2008 included among others eggs, confectionery sugar, oil seeds, textile yarn, animal feeds, and phenols. Although driven by the same products, the expansion in natural resource exports in 20012008 triggered an increase in the number of sectors with revealed comparative advantage from 20 in 1997 to 22 in

63 | P a g e 2008. Among 22 sectors in 2008, there were 13 products that made the list in both years (Table 30). Exports of nickel and raw cotton were also the largest in this category: combined accounting for 58 percent of resource intensive exports in 2008.

Table 29: Low technology labor intensive products with the values of RCA exceeding unity in 1997 and 2008

Products (SITC. Rev. 3) 1997 2008

121 Tobacco, raw and wastes 224 121 Tobacco, raw and wastes 202 075 Spices 17 061 Sugar/mollasses/honey 21 044 Maize except sweet corn. 17 211 Hide/skin (ex fur) raw 18 245 Fuel wood/wood charcoal 10 074 Tea and mate 11 074 Tea and mate 8 122 Tobacco, manufactured 7 061 Sugar/mollasses/honey 7 025 Eggs, albumin 7 011 Beef, fresh/chilled/frozen 7 075 Spices 5 046 Flour/meal wheat/meslin 6 611 Leather 3 071 Coffee/coffee substitute 5 057 Fruit/nuts, fresh/dried 2 971 Gold nonmonetary ex ore 4 062 Sugar confectionery 2 017 Meat/offal preserved 4 222 Oil seeds etc soft oil 2 091 Margarine/shortening 4 651 Textile yarn 1 054 Vegetables, fresh/chilled/frozen 3 091 Margarine/shortening 1 611 Leather 3 054 Vegetables, fresh/chilled/frozen 1 211 Hide/skin (ex fur) raw 3 017 Meat/offal preserved n.e.s 1 652 Cotton fabrics, woven 2 081 Animal feed ex unml cer. 1 057 Fruit/nuts, fresh/dried 2 841 Men's/boys wear, woven 1 841 Men's/boys wear, woven 2 059 Fruit/vegetable juices 1 122 Tobacco, manufactured 1 512 Alcohols/phenols/derivatives 1 048 Cereal etc flour/starch 1 Note: Products marked in bold had values of RCA exceeding unity in both 1997 and 2008. Those marked in italics had comparative advantage in 1997 but not in 2008 Source: Own calculations based on partners' data reported to the UN COMTRADE database.

Table 30: Resource intensive products with values of RCA exceeding unity in 1997 and 2008

Products (SITC. Rev. 3) 1997 Products (SITC. Rev. 3) 2008

683 Nickel 53 284 Nickel ores/concentrates/etc 251 671 Pig iron etc. ferro alloy 46 683 Nickel 69 263 Cotton 35 263 Cotton 67 325 Coke/semicoke/retort c 24 671 Pig iron etc. ferro alloy 37 278 Other crude minerals 18 223 Oil seedsnot soft oil 34 532 Dyeing/tanning extracts 15 273 Stone/sand/gravel 21 292 Crude vegetable materials 10 532 Dyeing/tanning extracts 15 289 Precious metal ore/conc. 9 661 Lime/cement/const. materials 9 284 Nickel ores/concentrates/etc 4 289 Precious metal ore/conc. 9 273 Stone/sand/gravel 4 278 Other crude minerals 8 689 Misc nonferrous base metal 3 292 Crude vegetable materials 8 288 Nonmanufactured base metal waste 3 325 Coke/semicoke/retort coke 7 223 Oil seedsnot soft oil 2 283 Copper ores/concentrates 4 693 Wire prod exc. Install. electrical 2 248 Wood simply worked 3 635 Wood manufactures. 2 287 Base metal ore/concentrates 3 248 Wood simply worked 2 261 Silk 3 277 Natural abrasives 1 642 Cut paper/board/articles 1 697 Base metal household equipments 1 247 Wood in rough/squared 1 345 Coal gas/water gas/etc 1 682 Copper 1 678 Iron/steel wire 1 691 Iron/steel/alum structures 1 667 Pearls/precious stones 1 697 Base metal household equipment 1 Note: Products marked in bold had values of RCA exceeding unity in both 1997 and 2008.Bold italics denote products with RCA exceeding unity in 1997 but not in 2008 Source: Own calculations based on partners' data reported to the UN COMTRADE database

64 | P a g e The decrease in the diversity of exports in terms of a number of exported products with the value of RCA exceeding unity has been the result of Zimbabwe's export performance in two of its most important markets South Africa and European Union. In both, diversification has decreased and the profile of specialization shifted away from low technology, labor intensive products. The number of lowtech labor intensive sectors that outperformed other suppliers, as measured by the RCA's equivalent, the ESI (export specialization index),44 has contracted in both markets indicating a loss of competitiveness (Table 31). The change in the specialization profile of Zimbabwean exports to the EU and South Africa reflects transformation in its export basket in terms of concentration and commodity composition. The contraction in the number of products with ESI above unity comes as no surprise in light of the earlier discussion on export concentration (see section 2 of this chapter). In 1997, Zimbabwe had 70 threedigit SITC product groups with the values of ESI exceeding unity, including 30 representing lowtech labor intensive activities and ten representing medium and hightechnology intensive activities. In 2008, the number of products enjoying specialization advantage dropped to 29 product groups in 2008 (Table 31). All products enjoying specialization advantage dramatically fell across each group with the largest decline of medium to high tech intensive products. The number of threedigit SITC items with ESI above unity was smaller in EU markets and did not change significantly, although its composition did in a predictable manner. The number of lowtech, labor intensive products fell while that of resource intensive products increased.

Table 31: Export specialization in South African and EU markets in 1997, 2002 and 200708: number of SITC threedigit products with values of ESI exceeding unity in terms of technology, natural resource and labor content

Lowtech labor intensive Resource intensive Medium and high tech intensive Total

South Africa

1997 33 27 10 70 2002 31 20 5 56 2007 18 17 2 37 2008 15 13 1 29

European Union

1997 21 9 0 30 2002 15 11 0 26 2007 15 14 0 29 2008 14 12 0 26

Source: Own estimates based on partners' data from the UN COMTRADE database

Zimbabwe's revealed export specialization in both the EU and South African markets has become increasingly similar, focusing on resource intensive products. This appears to have been the result of Zimbabwe's South Africandestined export basket converging towards the composition of Zimbabwe's exports to the EU. Considering the proximity of South African markets and their knowledge among Zimbabwe's exporters, this may point to internal supply constraints rather than the disappearance of sales opportunities in South Africa.

The export specialization index (ESij) for a product j of country i, is specified here as follows: ESij = (xij/Xi) /(mj/M), where: xij is country i's exports of product j to South Africa or the EU; Xi = j xij is country i's total exports to South Africa or the EU; Mj = j xij is South Africa's or EU's total `external' imports of a product j; M = i j xij is South Africa's or EU's total external imports. A value for this index below unity indicates a comparative disadvantage. If the index takes a value greater than unity, the country is considered to have a "revealed" comparative advantage in the product. In this particular case, Zimbabwe has a revealed comparative advantage in a product if its export of that item as a share of its total exports exceeds South Africa or the EU imports of the item as a share of South Africa or EU total imports.

44

65 | P a g e Falling exports of both agricultural and manufactured products have driven the changes in patterns of Zimbabwe's revealed comparative advantage in world markets and export specialization in EU and South African markets. As we have seen, the largest decline was in exports of agricultural foods and feeds. The share of food and live animals in Zimbabwe's total exports fell from 15 percent in 1997 to 14 percent in 2002 and 9 percent in 2008. The value of their exports fell 34 percent from US$251 million to US$165 million. It stood in 2008 at 49 percent of its peak value in 1994 during the 19942008 phase of exports expansion. Products of this group recorded also the largest change in terms of lost revealed comparative advantage. Food and live animals products is the only category of agricultural products whose values of RCA fell below unity indicating a loss of international specialization in 200208. The number of threedigit SITC food and live animals items with values of RCA exceeding unity fell from 21 in 2000 to 11 in 2008. So did their exports, which fell from US$189 million in 1997 to US$32 million in 2002 and 9 million in 2008 (Table 32). The largest contributor to the contraction of exports of this group was maize (SITC 044), whose exports in current prices fell from US$71 million in 1997 to less than US$1 million in 2008. It accounted for 40 percent of the difference between exports in 1997 and 2008 of these products. The second largest `losers' were live animals (SITC 001) and beef (SITC 011): their aggregate contribution to the difference between 1997 and 2008 amounted roughly to another 40 percent.

Table 32: Agricultural `losers:' values of RCA and exports in 1997, 20022008 (in millions of US dollars)

SITC

Products

1997 0.6 6.8 16.8 0.3 5.8 0.1 1.1 0.4 0.4 5.4 1997 33.5 35.7 71.4 0.2 3.8 0.0 5.8 1.7 0.5 36.7 188.9 63.2 9.0

Values of RCA indices

2002 1.0 0.3 2.3 1.6 0.2 0.3 3.3 2.3 0.6 4.1 2002 2.9 1.0 6.7 0.7 0.1 0.0 14.2 6.6 0.5 9.8 42.3 20.0 2.5 2003 0.4 0.4 0.4 0.4 3.6 0.1 4.4 2.4 0.6 3.2 2003 1.3 1.8 1.3 0.2 1.3 0.0 21.3 7.3 0.6 8.6 43.7 21.1 2.3 2004 2005 2006 2007 2008 Int. 0.5 0.0 0.1 0.4 0.1 0.2 2.0 2.5 0.6 3.9 0.3 0.0 0.1 0.9 0.0 0.1 1.8 0.7 0.5 2.8 1.1 0.0 0.0 0.1 0.0 0.0 1.4 0.5 0.7 2.0 0.8 0.0 0.1 0.2 0.0 0.1 1.2 0.4 0.4 1.1 0.7 0.0 0.0 0.2 0.0 0.0 0.7 0.2 0.3 0.9 LTL LTL LTL LTL LTL LTL LTL LTL LTL LTL Int. LTL LTL LTL LTL LTL LTL LTL LTL LTL LTL

001 Live animals except fish 011 Beef, fresh/chilled/frozen 044 Maize except sweet corn. 045 Cereal grains 046 Flour/meal wheat 047 Cereal meal/flour 048 Cereal etc flour/starch 056 Vegetable root/tuber prep/pres 058 Fruit preserved/fruit preparations 071 Coffee/coffee substitute SITC Products 001 Live animals except fish 011 Beef, fresh/chilled/frozen 044 Maize except sweet corn. 045 Cereal grains 046 Flour/meal wheat 047 Cereal meal/flour, nes 048 Cereal etc flour/starch 056 Vegetable root/tuber prep/pres 058 Fruit preserved/fruit preparations 071 Coffee/coffee substitute Total above Share in exports of food and live animals Share in total exports

Exports (millions of US dollars)

2004 2005 2006 2007 2008 1.0 0.6 2.9 0.0 0.0 0.1 0.0 0.0 0.0 0.0 0.4 0.2 0.0 0.2 0.2 0.1 0.3 0.0 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 9.3 8.5 6.9 6.9 3.9 7.5 2.1 1.6 1.2 0.7 0.7 0.6 1.0 0.5 0.2 10.0 8.5 6.7 4.4 3.6 29.1 20.8 19.2 13.1 8.7 13.0 1.5 10.2 1.0 9.5 0.9 7.0 0.6 5.3 0.4

Source: Based on partners data from UN COMTRADE Statistics.

66 | P a g e The contraction of these exports has had significant negative impacts on employment and poverty reduction, as all of these products represent low technology, labor intensive activities andcreate employment opportunities in poor rural areas. The loss of Zimbabwe's competitive edge has not been restricted to the agricultural sector but has also occurred in manufacturing activities. Several manufacturing sectors have ceased exporting, either partly or totally, over the last fifteen years. This list of `losers' includes a variety of sectors ranging from chemicals, iron and steel, miscellaneous manufactured products and agricultural equipment. Leaving aside the loss of comparative advantage in world markets, they share another rather surprising feature: except for three product groups, all others represent resource intensive activities. Three exceptions are manufacturing activities falling within the medium to hightechnology intensive groups (see Table 33).

Table 33: Exports and values of RCA of threedigit SITC manufactured products that lost comparative advantage in world markets in 199708 and their technology and factor content SITC 554 635 672 676 677 678 693 721 821 897 554 635 693 721 821 897 Products Soaps/cleansers/polishes Wood manufactures Primary/prods iron/steel Iron/steel bars/rods/etc Iron/steel railway materials Iron/steel wire Wire prod excl. installations electrical Agric machine ex tractors Furniture/stuff furnishing Jewellery Soaps/cleansers/polishes Wood manufactures Wire prod excl. installations electrical Agriculture machines excl. tractors Furniture/stuff furnishing Jewellery 1997 2002 2003 2004 2005 2006 2007 2008 Intensities 0.84 1.99 0.48 0.64 0.06 1.08 2.05 0.48 0.83 4.08 1.94 1.56 2.98 0.32 0.01 3.84 3.14 0.88 0.75 4.45 3.08 1.54 2.18 0.78 0.09 6.72 6.32 1.44 0.94 1.90 1.69 1.40 3.25 1.30 0.05 4.92 3.78 1.15 0.81 1.04 0.98 1.00 1.02 1.07 0.04 2.12 3.07 1.08 1.36 0.55 0.49 0.96 0.19 0.81 0.19 0.58 1.35 0.69 0.74 0.60 0.16 0.73 0.00 0.45 0.26 0.40 0.73 0.71 0.45 0.43 0.11 0.71 0.01 0.35 0.00 0.23 0.24 0.96 0.40 0.23 RI RI RI RI RI RI RI MH MH MH

1997 2002 2003 2004 2005 2006 2007 2008 4.9 9.0 3.7 2.2 15.2 26.2 61.2 21% 9.8 6.0 3.9 2.9 13.8 27.3 63.8 22% (in millions of US dollars) 16.3 12.8 5.1 4.8 6.3 5.7 3.8 4.1 8.7 5.9 5.0 2.6 5.2 4.2 4.0 2.8 18.9 16.3 26.5 15.8 12.3 6.8 3.7 4.6 67.8 20% 51.6 19% 21.2 7.8 0.0 7.2 36.2 17% 48.1 14% 7.1 5.7 0.0 3.1 16.0 7% 34.7 10% 2.5 7.8 0.1 0.9 11.3 5% 2.4 3.3 1.6 3.3 10.7 3.6 25.0 8% 1.9 2.6 0.5 4.7 7.8 1.7 19.1 8% RI RI RI MH MH MH

Total above Share in manufactured exports

Iron and steel products

(in millions of US dollars)

1.5 5.2 0.2 0.7 7.6 4%

1.8 5.0 0.0 0.4 7.2 3% RI RI RI RI

672 Primary/prods iron/steel 3.0 11.3 10.9 676 Iron/steel bars/rods/etc 1.3 1.4 3.4 677 Iron/steel railway materials 0.0 0.0 0.0 678 Iron/steel wire 1.7 4.3 8.2 Total above 6.0 17.0 22.5 Share in exports of iron and steel 3% 13% 15% Source: Based on partners data from UN COMTRADE Statistics.

67 | P a g e In contrast to products from other sectors of the economy where the loss of comparative advantage was associated with falling exports of the sector as a whole,45 the iron and steel sector stands apart as a success story. The weak export performance and the loss of comparative advantage have been limited to selected products. Exports of other products of this sector have more than offset the contraction in sales of steel products listed separately in Table 33. Since 2003, Zimbabwe's exports of iron and steel (SITC. 67), taking advantage of the boom in prices in world markets, have been growing fast, albeit erratically. They leaped 46 percent in 2004 and 42 percent in 2008 and contracted 4 percent in 2007: in other years, annual growth rates were between 14 percent in 2003 and 2 percent in 2006. With exports worth of US$308 million in 2008, the share of iron and steel products in total Zimbabwe's increased from 8 percent in 2003 to 15 percent in 2008. But developments in exports of iron and steel are not an unambiguous success story. It has been driven exclusively by foreign sales of other ferro alloys (SITC. 6715). Their share in total exports of iron and steel products increased from 82 percent in 2003 to 97 percent in 2007 and 98 percent in 2008 (Annex Table 7). This spectacular rise was not only the result of an increase in exports, although their value almost tripled (2.6 times) in 2008 over 2003. This was also the result of falling exports of other iron and steel products: their aggregate value fell from US$38 million in 2004 to US$31 million in 2006 and US$7 and US$6 million in 2007 and 2008, respectively. The number of subsectors with exports exceeding US$1 million fell from seven in 200405 to four in 2006 and two in 200708. While, without an uptodate industryoriented study, it would be difficult to draw any definite conclusions about broader implications of this increasing reliance on a single product, these developments raise concerns about the future sustainability of the iron and steel sector in Zimbabwe. They may imply growing obsoleteness and falling competitiveness of many steel plants starved of investment and access to modern technologies. Whatever the explanation, this is an another indication of the growing concentration of Zimbabwe's exports. The case of products whose values of RCA exceeded unity in the 2000s is not only less ambivalent but their aggregate values of exports are considerably lower than those of ferro alloys. Yet, these products are of interest as they may point to a future evolution of Zimbabwe's exports basket. There are 16 tripledigit SITC product groups, which represent specializations developed in 1997 2008 and retained in 2008. Data on these exports and their technological content are tabulated in Table 34. Several observations can be derived from eyeballing the data. First, while there is significant variation in rates of growth of individual product groups in 200208, total exports in this group have grown very rapidly since 2002: their LSG growth rate was significantly larger than that of total exports with their share in total exports having had more than doubled since 2002 and almost tripling since 1997. There are two exceptions: textile yarn (SITC 651) and pearls and precious stones (SITC 667). Despite a sluggish growth performance, they retained comparative advantage as the share of world imports of these products in total world imports fell more than their respective shares in Zimbabwe's total exports.

The decline of exports of furniture furnishing (SITC. 821), and associated loss of comparative advantage, was experienced by all other major products of this sector. Exports of furniture (SITC. 82) peaked in 2005 at US$23 million and fell to US$11 million in 2007 and US$8 million in 2008 (partners' data in the UN COMTRADE database).

45

68 | P a g e Second, these new entrants with values of RCA indices above unity do not necessarily signal any serious reshuffling in the composition of Zimbabwe's exports. The largest increase in exports was recorded by medium to hightechnology intensive products, albeit from a very low level and those relying on resource intensive activities continue to tower over other product groups. This is a rather surprising development, which resulted from a sudden surge in exports of ceramic plumbing heating fixture (SITC 8122) from US$14 thousand in 2007 to US$11.4 million in 2008. Considering that sales of these products averaged US$50 thousand over 19942007 and only once exceeded US$100 thousand in 2001 (US$112,000), its exports in 2008 were either a statistical error or some new investments. Neither did statistical services of South Africa nor customs in the EU report imports of SITC 812 products over 1994 2008. The latter reported imports of nonelectric boilers (8121) to the tune of US$2 million in 1999 but nothing in other years. South Africa reported these imports only in 2001 and 2006 worth US$13.5 thousand and US$4.1 thousand respectively. Hence, it seems to be a statistical misinformation as no new investments have been made in this industry. If this was indeed the case, then resource intensive exports had the best performance: their share in `rising' exports increased from 51 percent to 67 percent and that of low technology labor intensive products fell from 49 percent to 33 percent in 2002 08.

Table 34: Exports of threedigit SITC sectors that acquired revealed comparative advantage in 200708 (in millions of US dollars in 1997 and 200208) LSG Tech SITC 1997 2002 2003 2004 2005 2006 2007 2008 0208 cont. 025 Eggs in shell 0.4 2.1 2.3 3.4 3.9 3.6 20.6 3.4 21.3 LTL 059 Fruit/vegetable juices 0.3 0.8 0.4 0.4 0.8 2.9 2.7 2.3 31.7 LTL 062 Sugar confectionery 0.6 2.9 2.2 3.0 3.6 4.2 4.1 2.2 2.7 LTL 081 Animal feed 4.6 4.8 2.0 4.0 4.1 2.2 5.7 8.7 11.7 LTL 222 Oil seeds etc soft oil 5.6 1.1 0.3 0.9 1.4 1.1 1.0 13.1 37.2 LTL 247 Wood in rough/squared 2.4 2.6 1.9 1.6 1.7 2.2 2.9 3.1 6.2 RI 261 Silk 0.0 0.0 0.0 0.01 0.0 0.0 0.0 0.2 21.9 RI 283 Copper ores/concentrates 1.2 9.2 3.4 11.2 12.1 33.0 26.9 23.2 28.7 RI 287 Base metal ore/concentrates 1.3 3.5 0.5 0.6 0.5 0.7 11.8 14.6 38.9 RI 13.3 512 Alcohols/phenols/derivatives 0.8 4.5 4.5 3.2 6.6 6.4 9.4 7.5 RI 9.9 642 Cut paper/board/articles 3.0 4.7 9.8 5.9 9.0 10.4 8.1 10.2 RI 651 Textile yarn 12.6 15.0 12.1 16.3 11.1 15.1 13.4 8.5 5.7 LTL 667 Pearls/precious stones 1.1 0.3 0.9 0.1 0.8 1.1 0.2 0.1 15.7 LTL 682 Copper 12.3 1.8 2.0 5.6 29.2 44.3 51.4 21.2 56.7 RI 691 Iron/steel/aluminum structures 0.3 2.2 1.9 2.1 2.2 3.5 3.7 8.1 21.1 RI 812 Sanitary/plumb/heating fixtures 0.1 0.1 0.2 0.0 0.0 0.0 0.0 11.4 47.8 MHT TOTAL ABOVE 47 55 44 58 87 131 162 138 21.9 Share in total exports 2.2 3.2 2.4 3.1 4.4 6.0 6.8 6.5 Source: Based on partners data from UN COMTRADE Statistics.

Third, despite a weak growth of low technology, labor intensive products, the level of processing embodied in `newly' specialized products appears to have increased. The share of traditional inputs fell from 76 percent in 2002 to 65 percent in 2007 and 67 percent in 2008. The largest exports of low technology labor intensive products were in oil seeds: its exports increased from an average of around US$1 million in 200207 to US$13 million in 2008.

69 | P a g e

3.7. Conclusion

Zimbabwe's exports were, on the one hand, "frozen in time" and, on the other hand, underwent huge transformation as a result of the collapse of agricultural production and the lack of new investment activity. They remained "frozen" only in terms of the degree of processing embodied in exports. With exports of agricultural products falling and industrial raw materials expanding, the share of low processed primary inputs--the total of agricultural and mineral exports--has remained largely unchanged. So has the share of other enduse product groups. But on all other counts, almost a decade of economic decline has transformed Zimbabwe's foreign trade not only in terms of its geographic pattern but also in terms of its composition with significant implications. On the import side, the level of processing embodied in this trade has fallen. The share of resource and unskilled laborintensive products has increased while imports of medium to high technology intensive products have fallen. So has the share of nonagricultural consumer goods. Both are indicative of falling standards of living and depressed investment activity. On the export side, the shift from agricultural to mineral products has amounted to a shift from unskilled labor, low technology intensive activities to capital and naturalresource intensive activities significantly suppressing demand for labor in the economy and contributing to unemployment. To sum up, the main conclusions drawn from the empirical analysis may be summarized as follows: First, the level of processing embodied in Zimbabwe's exports has remained unchanged since 1994. The share of primary inputs has remained at slightly below threequarters of total exports independent of their levels; Second, exports of foods and feeds, iron and steel and consumer goods were the main levers of export growth during the growth phase in 199497 but the picture changed during the period of export recovery in 20032007 in two important respects: first, food and feeds contracted, and second, consumer goods ceased to be a driver in terms of exports in current prices; Third, except for iron and steel, positive net exports across all enduse product categories have been falling. Not surprisingly, the largest decline has been in food and feeds and industrial raw materials. Fourth, factor and technology content changed with resource activities raising to prominence at the expense of falling exports of lowtech, unskilled labor intensive products. Fifth, the gamut of products exported to South Africa has narrowed and revealed comparative advantage shifted away from labor intensive products. Sixth, the shift away from agricultural products has been dramatic with negative impacts on employment and poverty reduction.

In all, developments in foreign trade in the 2000s have also created a huge gap between Zimbabwe's endowment of skilled and disciplined labor and its revealed comparative advantage shifting towards naturalintensive minerals. Yet, a detailed analysis of exports suggests that not all producers of laborintensive goods have been wiped out. Against the background of lackluster export performance and falling diversification in export offer, there are some bright spots rising hope Zimbabwe's export offer may move more in line with its

70 | P a g e endowments with a large pool of cheap and relatively well skilled labor force. There are 16 tripledigits SITC product groups, which represent specializations developed in 19972008 and retained in 2008. While their aggregate exports remain relatively low, they contribute significantly to employment. Whether they expand in years to come, it will hinge critically on improving business climate together with foreign trade regime. The importance of the latter has become critical since the largest barrier to foreign trade has been removed with the switch to "dollarization" and "randarization" of the economy in early 2009. These are the issues addressed in the next chapter.

71 | P a g e

4. Moving forward: how to revive exports?

Zimbabwe's foreign trade performance has been lackluster from the perspective of both its past performance and that of its neighbors. This has not been the result of adverse external developments as these were highly favorable 19942008, so disappointing export performance has been exclusively homegrown. Exports, most negatively affected, were agricultural and simple industrial products. Exports of these low technology, labor intensive products have been consistently declining since 1997. They have been largely responsible for Zimbabwe's disappointing overall exports performance. While large mining and farming (tobacco) operations tend to be more immune to unfriendly business climates, small scale agricultural and industrial activities are highly sensitive to illdesigned economic policies and falling quality of economic governance. Their impact has been particularly visible in terms of Zimbabwe's export performance in the South African market. Consider the following: First, falling exports of small scale agricultural and industrial products have shaped Zimbabwe's export profile there. The number of products with comparative advantage has dramatically contracted with the largest losses incurred by products characterized by high to medium technology and low technology labor intensive intensities. Zimbabwe's export basket has become extremely concentrated exceeding the levels of concentration in Zimbabwe's exports to the EU. It has also become similar in terms of revealed profiles of export specialization increasingly limited to a falling number of natural resourceintensive products. Second, South Africandestined export expansion has also revealed supply limits to Zimbabwe's natural resourceintensive sectors. Except for industrial raw materials, the values of other exports to South Africa in terms of enduse in 2008 were either below their values in 1996 or less than US$1.5 million above (steel and iron by US$1.4 million, fuels by US$1.3 million and machinery by US$0.9 million). The value of exports of foods and feeds was US$50 million below the value in 1996. Despite the fall in agricultural exports, the share of industrial raw materials intended for further processing increased to 92 percent in 2008 from 56 percent in 1996. Hence, minerals, iron and steel and other industrial raw materials have been the levers of export expansion to this market. Third, the expansion of South Africadestined exports appears to have been at the expense of exports to the EU and have not led to an increased presence of Zimbabwean exporters in the world market. The share of South Africa in total Zimbabwe exports increased from around 10 percent at the turn of the century to 37 percent in 2008 while over the same period the share of the EU fell from 16 percent to 7 percent. Yet, the share of Zimbabwe in South African total imports contracted in 2008 to the level preceding its South Africa export boom in 200107. Zimbabwe has also failed to increase its export presence relative to other suppliers in other neighboring markets as well as in the EU, its historically largest market. Zimbabwe's foreign trade performance can be explained in terms of a contracting supply base as a result of illdesigned policies and political instabilities. Foreign trade policies and institutions have been largely irrelevant to the outcome. Even the best policies would have failed to elicit a supply response in the absence of a friendlier business climate and political instabilities. International experience demonstrates that improvements in trade policy are only effective when supplemented by improvements in the business environment. This should not suggest that barriers to foreign trade, as exemplified by very high

72 | P a g e transaction costs of conducting foreign trade operations in Zimbabwe, should not be addressed and removed but indicates that their impact on foreign trade would be limited unless other barriers restraining a supply response are also removed. These comments notwithstanding, steps should be taken to streamline and lower the transaction costs of foreign trade while simultaneously taking measures to remove barriers faced by the private sector in moving towards higher levels of processing of domestically available natural resources and agricultural foods and feeds. Three sectors have traditionally underpinned Zimbabwe's external performance: agriculture, mining, and tourism. While developments in 19992008 have undercut each of them, they are not completely wiped out and some have survived in better shape than others. By the same token, the capacity to recover also varies across them with mining probably being the easiest to revive. Agriculture and industries based on agricultural inputs appear to have been the largest loser of all of them. Yet, with the right policies in place, all can be rebuilt relatively quickly. Reviving exports of labor intensive products is critical for Zimbabwe to enter a path of sustained economic growth. The alternatives are high unemployment, poverty and a high level of aid dependence, as sales of minerals, including diamonds, may not generate sufficiently large earnings to pay for rapidly growing imports. The purpose of this chapter is to take stock of the issues to be addressed in the context of the earlier findings of this study. In contrast to a lot of other developing countries, Zimbabwe's developmental challenges can be addressed by measures that are relatively cost free in economic terms and could be implemented quickly, provided the government is genuinely committed to a progrowth agenda. The potential for growth derives not only from Zimbabwe's rich endowments in natural resources, human capital and geographic location but also the fact that the economy has been suppressed by government policies neither external nor domestic physical constraints. Zimbabwe has several assets. Its labor force is among the most educated in subSaharan Africa and, as the experience of Zimbabwean emigration shows, one of the most disciplined and hard working. Its natural environment is not only attractive to tourists but also uniquely suited for agricultural production. Its fertile soil, water and climate are highly favorable to agricultural production and once made Zimbabwe one of the largest net exporters of foods and feeds in Africa. It is also amply endowed in mineral resources including recently discovered diamonds. Despite being landlocked, it has access to high quality infrastructure in neighboring South Africa. In a nutshell, Zimbabwe has a solid `hardware' for economic development. While in some sectors of the economy this potential can be relatively quickly activated by the removal of policyinduced barriers to economic activity, in other sectors--staved off capital investment for many years--this will take time and will require foreign investment and technology. While we do not have reliable data about capital formation, many indicators suggest that the level of investment has been very low for the last decade or so. FDI inflows have been negligible: the total amount invested over 19992008 amounted to US$389 million compared to US$4.5 billion in neighboring Zambia. Although the FDI inflows more than doubled in 2009 from US$52 million a year earlier to US$105 million, they are estimated to have fallen to US$85 million in 2010. Another indicator of the investment challenge is imports of capital equipment excluding motor vehicles; their share in total imports of goods fell from an

73 | P a g e average of 32 percent in 199598 to 22 percent in 200408. While in 199598 the value of these imports into Zimbabwe was 2.8 times larger than into Zambia, in 200408 these imports were 50 percent lower. The remainder of this chapter briefly discusses the restoration of macroeconomic stability and its impact on economic activities as well as implications for foreign trade policy. It also identifies measures that could improve the business climate, increase investment and boost exports. 4.1. Restoration of macroeconomic stability: its initial impact and the challenge ahead Two events--the Global Political Agreement (GPA), signed on September 15, 2008, and the introduction of a multicurrency regime in April 2009--set the stage for the recovery of the Zimbabwean economy. The former has provided a modicum of political stability, whereas the latter has restored macroeconomic stability. Dropping the domestic currency was not a component of the reform package except for getting rid of hyperinflation as well as the host of administrative measures earlier introduced in futile attempts to combat it. Put differently, the quality of the business environment has improved but only marginally as other barriers that put Zimbabwe at the bottom of several international rankings have remained in place. There has been no improvement in its investment climate, despite the removal of some constraints on current business activities (e.g. the removal of surrender requirement, foreign exchange controls). While too short a period of time (15 months at the time of this writing) has elapsed since the restoration of macrostability to assess its longer term impact on investment and output, several improvements can be easily identified. For starters, inflation fell from an astronomical 500 billion percent in 2008 to 3.6 percent yearonyear as of August 2010 and its immediate result has been a revival of economic growth. Zimbabwe's economy grew five percent in 2009 and the outlook for economic growth--with projected GDP growth of 5.4 percent in 2010 remains positive. Considering that inflation had not been below doubledigit rates since the early 1990s and GDP had been falling every year since 1998, these are remarkable accomplishments. Second, thanks to restored macroeconomic stability and the removal of `antihyperinflation' measures, chronic food and fuel shortages experienced in 2008 have completely disappeared. Zimbabwe has now ceased to be supply constrained: it has become demand constrained as domestic markets have extended their reach to products that previously were subject to direct government controls. The removal of direct state controls over foreign exchange, prices of basic commodities and foreign trade flows have removed perverse incentives and gone some way towards restoring domestic production and exports, at least those that are less sensitive to the quality of business environment. The volumes of exports of some minerals, both mining--platinum group metals, gold, nickel, and diamonds --and value added activities such as high carbon ferro chrome, iron and steel products, and agricultural products are projected to significantly increase in 2010 (Table 35). The growth in volumes exported is projected to be quite substantial reaching very high levels for raw sugar and gold. Due to rising prices, increases in the values of exports will be significantly higher than increases in their physical volumes. The largest price effect will be for copper followed by cut flowers and nickel. Among agricultural products, tobacco regained in 2010 its traditional top position in Zimbabwe's total exports of goods that it lost back in 2004, first to gold and then to platinum group

74 | P a g e metals products, as the volumes of gold exports fell in 200508. The share of tobacco in total exports is estimated to have reached 20 percent in 2010. In addition to tobacco that recorded strong growth recently, other encouraging signs of recovery in the agricultural sector are increases in output of maize and sugar but also indications of increases in the output of agroprocessing. While data for 2009 and first two quarters of 2010 are still not fully available to assess developments following the switch to dollars and rands, some food processing companies are reporting increases in sales and profits. The milk processing firm, Daribord, has announced a 50 percent increase in sales and US$3.1 million profit for the halfyear. This was made possible by a 30 percent increase of milk production. An impressive achievement, although the 3.7 million liters per month now produced is still below the domestic demand estimated at 5.5 million liters.46 In all, agriculture, which contributes 16 percent to GDP, is expected to grow by almost 19 percent in 2010.47

Table 35: Projected exports of selected agricultural and nonagricultural commodities in current prices and volumes in 2010 (in percent) Projected growth in 2010 over 2009 Volume Value Share in total exports

Fluecured tobacco 28% 32% 20% Sugar raw 85% 90% 3% Sugar refined 75% 83% 0% Flowers 10% 20% 1% Gold 59% 80% 14% Platinum Group Metals (PGM) 3% 3% 19% Diamonds 15% 25% 2% Copper 5% 156% 1% Nickel 11% 20% 2% High Carbon Ferro Chrome 4% 6% 7% Cotton Lint 0% 0% 9% TOTAL EXPORTS n/a 21% 79% Source: own calculations based on data provided by Reserve Bank of Zimbabwe, Harare, August 2010.

Yet, the recovery has so far fallen short of expectations considering both Zimbabwe's developmental potential and the depth of the economic contraction experienced over 19982008. While the cumulative GDP growth rate in 200910 of slightly above 10 percent is impressive in itself, this has to be cast against the cumulative contraction of 63 percent in terms of real GDP rates over 19992008. Despite projected export growth of 21 percent in 2010, exports of most major commodities in terms of volume are still below the levels achieved in the past. Copper, cotton lint, together with platinum group metals and raw sugar remain the only major products whose exports both in terms of value and volume exceeded in 2009 their respective levels in 2000. Exports of other traditional products were in 2010 around 70 percent below their volumes in 200001 with nickel and fluecured tobacco 60 percent below their respective volumes in these years. There are also signs that the recovery may be faltering as jobs are being shed and productive capacity remains idle.

Madera, Bright. 2010. "Firms struggle to recapitalize." The Herald Business, Harare, August 18 (B3) According to Business Reporters Bloomberg 08/18/2010, quoted in The Herald, Harare, the agricultural output is expected to grow 20 percent in 2010 driven by tobacco whose production is estimated at around 115,000 tons. This is still more than 50 percent below the 2000 level but almost twice above the output in 2003.

47

46

75 | P a g e Indeed, as captured by export data, the contribution of labor intensive activities to total exports does not appear to have increased following the restoration of macroeconomic stability. The major drivers of export growth in 2010 remain tobacco and minerals. The share of products listed in Table 35 above is estimated to be 79 percent in 2010 up from an average of 73 percent in 200608. This means that the share of other products fell from 27 percent to 21 percent during this period. There is some consolation due to the fact that the increase of exports of laborintensive tobacco has, alone, driven this change: its share increased from an average of 13 percent in 200608 to 20 percent in 2010. The key challenge therefore remains in reviving exports of other labor intensive manufactured goods as well as exports of naturalresource products. This is especially important not only for job creation but also considering that the recovery has triggered increased imports while exports have been falling behind resulting in growing deficits in the balance of trade. Although export earnings are projected to increase to US$1.9 billion in 2010 from US$1.59 in 2009, imports are expected to grow even stronger from US$3.2 billion in 2009 to US$3.9 billion in 2010. So the trade deficit is projected to grow 24 percent in 2010 from US$1.6 billion to US$2 billion. Ultimately, sustaining expanding import demand calls for a significant expansion of exports of goods and services. Since there was little, if any, significant investments into the economy in 19992008, expansion of exports calls for a radical increase of investment, domestic and foreign alike. Contrary to expectations, the restoration of macroeconomic stability in 2008 has so far had no impact on FDI inflows as the fears of property confiscation under the Indigenization and Economic Empowerment Act of 2007 have likely discouraged foreign investors from getting engaged into investment activities in Zimbabwe. Uncertainty created by the Indigenization and Economic Empowerment Bill of 2007 remains an effective barrier to both domestic and foreign investments. Time works against any delay in modernizing industrial capacities. As was shown earlier, some export capacities in agrobusiness and the agricultural sector appear to have survived both hyperinflation and the policy shift from wealth creation to wealth confiscation and redistribution in the 2000s. They appear to have somehow withstood adversities of the economic environment. Although net exports have dramatically fallen, exports still exceed imports and some capacities can be restored as they were likely temporary victims of hyperinflation, the wrongheaded exchange rate policy, and other direct state controls. Indeed, the largest decline took place in 200508: surpluses were between US$800 and US$1.1 billion in 19942002, fell to around US$0.7 billion in 200304, about US$0.4 billion in 200507, and US$150 million in 2008. With the passage of time, some of these capacities still existing in the processed food sector but may still be inadvertently wiped out. Considering high unemployment and the fact that the importance of this sector stems not from its contribution to foreign currency earnings but to employment creation, it would be a great loss to the country's welfare if they did. 4.2. A new setting for foreign trade policy: implications of the multicurrency regime The restoration of macroeconomic stability that has come with the dumping of the domestic currency has created new challenges for enhancing competitiveness of Zimbabwe's products in international markets. The increased exposition of domestic firms to competition from imports calls for an overhaul of various regulations that unnecessarily increase transaction costs and reduce incentives to invest by domestic and foreign firms alike. The hassle cost of doing business or investing in Zimbabwe towers over

76 | P a g e other impediments to expanding exports that is needed to put Zimbabwe back on a path of sustained economic growth based on the best use of its unique endowment of natural resources and human capital. The introduction of the multicurrency regime in April 2009 has drastically altered the parameters underlying the conduct of foreign trade operations and the overall thrust of foreign trade policy simple because getting rid of the Zimbabwean dollar has also automatically removed exchange controls as well as put an end to confiscatory taxes on exports, that is, surrender requirement on export earnings.48 During the tumultuous years preceding the dumping of the Zimbabwean dollar, imports became increasingly subject to control through nontrade measures (IMF 2009). While duties and various border charges might have raised the prices of imports, this was of little relevance for two reasons. First, with constant devaluations of the Zimbabwean dollar, imports became more and more expensive. The devalued currency provided protection unmatched by tariffs or NTBs: the Zimbabwean dollar (Z$) quickly lost value throughout the 2000s. In May 2002, one US$ was equal to Z$350, by the end of July US$1 would command Z$600 and on December 20, 2002, the exchange rate was Z$1,350. The pace accelerated subsequently and in May 2007 the exchange rate reached Z$60 million (Sandawana 2008). This alone created barriers to imports. Second, imports of many goods were indirectly controlled by the state through the direct allocation of foreign exchange or offering preferential exchange rates for some purchases abroad. In a nutshell, successive devaluations and foreign exchange controls made foreign trade policy tools irrelevant in shaping flows of imports. Yet, three things have not changed: the large share of customs revenues in total government revenues; commitments related to participation in regional trade agreements; and continued instability of tariff policy. They remain constant fixtures of Zimbabwe's economic landscape. Taxes and border charges collected by customs at the border still remain a very important source of government revenue. Revenue from customs duties levied on imports amounted to US$132.8 million or 14.3 percent of government revenue in JanuaryJune 2010 down from 31.5 percent in the same period in 2009 (MoF 2010). The decline in the share was simply due to the surge in revenue from other taxes triggered by economic recovery and improved collection. Hence, revenue considerations are likely to remain an important factor in shaping future tariff policy. The freedom in setting tariffs is restricted by bilateral PTAs (preferential trade agreements). Zimbabwe is party to a host of PTAs of which the most comprehensive are the SADC Trade Protocol and the COMESA free trade area. Before these went into effect, Zimbabwe also signed and implemented bilateral trade agreements with Botswana, Malawi, Namibia, and South Africa (Gilson 2010). According to an official Zimbabwe Government document (MoF 2010),49 goods originating from the COMESA and SADC

Although this was a commendable decision, it was flawed. Instead of two currencies, which is reminiscent of running concurrently gold and silver, one currency should have been chosen as an official means of payments. The introduction of two currencies creates problems if their respective exchange rates move in different directions vis àvis each other. The appreciation of a currency triggers hoarding of this currency. Weaker currency pushes stronger one from circulation. 49 "The Free Trade Area Protocols provide for duty free importation of goods from COMESA and SADC Member states, provided such goods meet the set criteria on the rules of origin" (MoF 2010, p. 172).

48

77 | P a g e Member States are imported duty free. Since preferential partners account for around threequarters of Zimbabwe's total imports, this implies that only one quarter of imports is subject to tariffs. Yet, PTAs do not seem to be a binding constraint. First, in fact, not all imports originating in preferential partners appear to enter Zimbabwe dutyfree as there are exemptions listed as sensitive products. The list is extensive encompassing almost all sectors of the Zimbabwean economy that have been engaged in exports. Sensitive products account for between 22 percent and 39 percent of total imports excluding minerals, i.e., industrial raw materials.50 Their share in total imports has significantly decreased since 2002 when it was almost 40 percent. Their contraction reflects falling agricultural output as sensitive products are mostly agricultural products (Table 36).

Table 36: Zimbabwe's exports and imports of sensitive agricultural and nonagricultural products in 200209 (in percent) Agricultural products Nonagricultural products Share in total imports Agricultural products Nonagricultural products Share in total exports excluding industrial raw materials Zimbabwe's imports (in percent) 2002 2003 2004 2005 2006 2007 2008 2009p 54 48 47 79 48 50 60 81 46 52 53 21 52 50 40 19 5 5 3 8 3 3 4 6 Zimbabwe's exports (in percent) 95 97 97 96 95 98 98 99 5 3 3 4 5 2 2 1 39 34 26 20 17 23 21 22

Note: ** 2009 data is very preliminary estimates due to some countries reported to UN COMTRADE System late or still missing values (at least 10% below at expected world level). Source: Based on mirror data from UN COMTRADE Statistics.

Second, even imports of nonsensitive products originating from preferential trading partners do not appear to be spared from tariffs even, in one case, exceeding a corresponding applied MFN tariff rate. As of August 1, 2010, the government raised tariff rates on "selected finished products from the SADC region" (MoF 2010, p. 174). While the preferential rates on five eightdigit CN products were increased from 0 percent to the MFN tariff rates on these goods ranging between 15 percent and 20 percent ad valorem, the tariff rate on SADCoriginating imports of other food preparations (CN 2106. 9090) was set at 10 percent whereas imports from MFN sources pay a tariff of 5 percent ad valorem. This is a unique case of reverse discrimination going afoul. Another element of Zimbabwe's foreign trade regime is the lack of stability in its tariff policy. Tariff rates are often changed: interestingly, changes go in both directions as some tariff rates are increased and others reduced. Since the introduction of the multicurrency regime, all customs duties in Zimbabwe are now levied in foreign exchange only. Partly as a result of this, tariff rates have generally been reduced across the board, although the Zimbabwean authorities have recently taken steps to reestablish import duties on the list of basic commodities for which revenue collection has suspended since 2008 under the Short Term Emergency Recovery Program. Needless to add, some modicum of stability and predictability of changes is important for investment and other business decisions.

50

Although one might ponder over the wisdom of protecting internationally competitive sectors, note that their share in total exports has been falling since at least 2002. This suggests the loss of competitiveness in activities in which Zimbabwean producers have comparative advantage. Hence, as a measure supporting reconstruction, it does not raise much concern.

78 | P a g e The adoption of US dollar or rand has had benefits and costs relating to foreign transactions. Benefits include enhanced macroeconomic stability and lower country risk; lower transaction costs for firms; elimination of exchange rate risk cancelling the related premium in domestic interest rates; and stronger competition. The downside is the loss of the ability to use interest rates and the exchange rate for macroeconomic stabilization purposes, which, in the case of an asymmetric shock can lead, all else equal, to higher variability in output and employment. With the formal `dollarization' of the economy, trade policy is now the only line of protection for domestic producers as there is no resort to devaluation or revaluation of the domestic currency. World prices are directly transferred into the economy. A sharp appreciation of the US dollar against the currencies of Zimbabwe's regional trading partners may undercut competitiveness of Zimbabwean firms visàvis other regional producers in regional markets and make imports from local partners cheaper, possibly putting domestic industries under stress.51 In response, these industries may lobby the government officials for protection from competing imports. On the other hand, the effect of a falling US dollar may be offset by Zimbabwe's partners' devaluing their currencies, although devaluations are unlikely to be introduced to erode the competitiveness of Zimbabwean imports. Hence, the adoption of the US dollar is likely to lead to a growth in foreign trade thanks to the combination of lower transaction costs, disappearance of exchange rate risk, and increased price transparency. But at the same time, it may lead to increased shopping for protection as a weak domestic currency can no longer offer protection from competition from imports. This calls for considerable improvements in Zimbabwe's business climate to lower the cost of trade and increase competitiveness rather than ad hoc trade policy interventions as exemplified by the recent introduction of a 15 percent duty on exports of chrome ores and fines ostensibly designed to increase value added.52 In other words, the benefits of the multiplecurrency regime will not come by default. Moreover, their attainment only marginally depends on foreign trade policy, although ad hoc protectionist measures raising transaction costs may prevent their coming to fruition. As for transactions costs, Zimbabwean firms face high formal barriers to doing business including those associated with foreign trade that are not so much related to weak infrastructure but stem from burdensome regulations.53 As such these could be quickly addressed and removed. Zimbabwe has one of the highest hassle costs of doing business related to its regulations. In trading across borders, Zimbabwe was ranked 167th out of 183 countries or 11 ranks below the average for its neighbors and its ranking is the lowest among comparators in 2009. Since regional benchmarks are not very demanding in terms of transaction costs associated with foreign trade activities, Zimbabwe's performance offers plenty of space for improvement.

The case in point is the loss of market share of Argentina's exporter to those from Brazil, its preferential trading partner from Mercosur in the late 1990s and early 2000s: under the pressure of balance of payments crisis, Brazil devalued its currency whereas Argentina was on a currency board with the US dollar as an anchor, which strongly appreciated at the time. 52 This measure reflects negatively upon the quality of business environment: in countries where capital flows freely to the best use, this type of tax is not only counterproductive but also redundant, as the international experience amply demonstrates. 53 For a detailed discussion, see Appendix 2.

51

79 | P a g e Zimbabwe's low ranking in terms of exploiting opportunities offered by foreign tradeled economic development stems not from its geographical location but selfinflicted policyinduced costs. Zimbabwe has the third highest cost to import and to export in the world after Chad and Central African Republic. While it would be tempting to associate these high costs with those associated with transportation, exacerbated by the lack of direct access to cheaper maritime transport, being landlocked is only part of the explanation as Botswana, Malawi, Rwanda and Zambia, which are also landlocked, have lower costs of both imports and exports Both exporters and importers from Zimbabwe need significantly more time to complete both export and import transactions: they also spend more money. The barriers are therefore more policy rather than physically induced. Bureaucratic procedures consume more time than dealing with infrastructureimposed or technical constraints while the latter account for most of the total transaction cost incurred by exporters and importers alike. Meeting bureaucratic requirements accounts for the bulk of the time needed to complete both export and import transactions. Lower infrastructurerelated costs in Zambia than in Zimbabwe raise suspicion that they are under stating the latter since Zambian exports often transit Zimbabwe. But this does not seem to be the case, as shipments from Zambia can also go directly to ports in Mozambique bypassing Zimbabwe altogether. Furthermore, the cost of delivering shipments and getting them ready for shipping by sea for Malawi's firms is much lower than for firms in Zimbabwe: the cost of ports and terminal handling and inland transportation is US$1,240 for Malawi as compared with US$2,800 and for imports it is US$2,140 against US$4,349.54 In all, foreign trade transaction costs of Malawian firms are roughly half of what their Zimbabwean counterparts have to pay (Appendix Table 10). Hence, geography and infrastructure are not huge barriers to foreign trade but governance is and the main reason why the Zimbabwean economy is unable to fully tap opportunities offered by global markets. Within governance, the crucial issue is that the government lacks credibility in protecting private property rights. 4.3. Expanding the export base calls for new investments but barriers persist Expanding exports seems to be one of the most pressing tasks if Zimbabwe is to avoid an aid dependency trap and enter the path of sustained economic growth. Taking into account the swing in its foreign trade balance in goods from historically a traditional surpluses (including surpluses on the current account) to deficits beginning in 2006 and increasing each year over 200710, the challenge facing policy makers is to find ways of financing `excessive' imports of goods and therefore current account deficits.55 Leaving aside foreign borrowing, assistance and slashing imports, which would undercut economic growth, this can be achieved by expansion of exports of goods and services as well as by attracting FDI inflows. The latter, if not lured by special concessions and political deals, may boost exports of both goods and services by supplying capital and knowhow.

This is also surprising that, as Gilson (2010) notes, trucks registered in Zimbabwe and in South Africa are allowed on a reciprocal basis to transport goods between two other countries even if the third country is used as a point of transit. 55 The Reserve Bank of Zimbabwe projects the current account deficit of US$1,363 million and of goods foreign trade of US$1,765 million in 2010. The corresponding deficits in 2009 were US$928 million and US$1,622 million.

54

80 | P a g e While in publications and discussions of Zimbabwe low capacity utilization is a dominant theme, the crux of the matter is not excessive capacity but rather its absence. Lack of maintenance and investment has resulted in a significant reduction in electricity generation capacity, collapse of water supply, and major disruptions to railway services. The contraction in the volumes of exports of capitalintensive mineral products points to many mines being undercapitalized. Judging also by the significant decrease in diversity of Zimbabwe's exports as measured, among others, by the number of exported goods exceeding US$1 million, many firms have either disappeared or limited their sales to domestic markets. So it is clear that without infusion of new investment Zimbabwe's exports capacity will simply not be rebuilt. Yet, some sectors probably can rebound even without significant investments or improvements in the business climate. As was argued earlier (see Section 3.3), some export capacity in agrobusiness or the agricultural sector appear to have survived adverse economic policies. Agrobusiness still appears to be alive: several developments point also to the potential for sustained growth. Although net exports have dramatically fallen, exports still exceed imports and some capacities can be restored as they were likely temporary victims of hyperinflation, wrongheaded exchange rate policy, and other direct state controls. Indeed, the largest decline took place in 200508: surpluses were between US$800 million and US$1.1 billion in 19942002, fell to around US$0.7 billion in 200304, about US$0.4 billion in 200507, and just US$150 million in 2008. Preliminary estimates suggest that they increased in 2009. Although significantly depressed, some investments have still taken place in Zimbabwe's mining and industrial sector offering hope for a relatively quick rebound. Furthermore, diamonds, whose extraction does not require significant capital outlays, from the Marange field could reportedly generate output worth U$1.7 billion a year.56 If this happens and the Kimberley Process requirements are satisfied, this would almost double the value of projected total exports and close the deficit in goods trade in 2010. This would be mixed news as it would reduce pressures for economic reforms which are indispensable to increased employment and sustained economic growth. Similarly, a significant increase in tobacco output could also be achieved without large investments. While the policy of confiscation of commercial farms "... is likely the main reason that maize production fell by threequarters" (Clemens and Moss 2005) in 200004, other factors also contributed to the contraction. These include inflation and monopolistic powers of the staterun Grain Marketing Board designated as the sole buyer and distributor of meat and wheat as well as a pricesetter in Zimbabwe. With prices of all other goods increasing faster than prices set by the Grain Marketing Board, profitability of growing maize fell and farmers shifted to crops that were not subject to state price controls. Hence, production of maize should expand now macroeconomic stability has been restored. Indeed, this has already happened as exports increased 28 percent in terms of volume and 32 percent in 2010 (see Table 35 above in Section 4.1): the amounts involved are still, however, almost 60 percent below the 2001 level indicating a large scope for improvement. But other industries as well as mining need much more substantive capital outlays in order to address years of neglect and underinvestment. Those firms that have survived the economic chaos of 200008 are in desperate need of modernization. But barriers to achieve this goal are significant. For starters,

56

See http://www.bbc.co.uk/news/business10945366 accessed on August 11, 2010.

81 | P a g e despite some improvements and greater political stability, Zimbabwe remains one of the worst countries in the world in terms of the quality of economic governance and hassle cost of doing business. While large firms may find ways to navigate murky waters of doing business there small firms--critical to reducing persistently high unemployment rates--experience many more difficulties. As compared to its neighbors, Zimbabwe's taxation regime stands out as being particularly oppressive (for a detailed discussion, see Appendix 3): while this may be of little relevance to mining companies, although it often is,57 it is potentially a huge barrier to small businesses often engaged in lowtechnology, laborintensive activities. But this should not suggest that a reform should be confined to the taxation regime. What is badly needed is an overhaul of the existing economic regime and policies. Second, the single largest obstacle to economic recovery is the Indigenization and Economic Empowerment Act of 2007, whose administrative implementation framework, i.e., Indigenization and Economic Empowerment regulations, were issued in 2010. They identify sectors where only `formerly disadvantaged Zimbabwean citizens" are allowed to operate and those where they have to have equity in both existing and future commercial and industrial establishments. As chairman of Zimplats, which-- with US$1 billion invested over 200210--is by far the largest foreign investor in Zimbabwe, emphatically stated in his expose to shareholders:

"The Indigenization and Economic Empowerment Regulations that provide the administrative framework for the implementation of the Indigenization and Economic Empowerment Act were gazetted during the year. As shareholders are aware, this act seeks to ensure that all companies operating in Zimbabwe are majority owned by local Zimbabweans. It is important for me to highlight that your board understands the need for and fully supports the meaningful involvement of Zimbabweans in the country's economy. It is however our firm belief that at this time, Zimbabwe's greatest need is for increased levels of foreign direct investment to create more employment in the country. We therefore urge the authorities to implement this law in a way that does not compromise Zimbabwe's desire to be seen as a preferred investment destination." (Zimplats 2010, p. 5).

The Act, among other actions, is responsible for the government's lack of credibility among investors, foreign and domestic alike. With investors remaining skeptical about the evolution of the political situation since the coalition government took office in February 2009, it should come as no surprise that FDI inflows remain negligble. As Zimplats' annual report diplomatically asserts, "international investors and the donor community whose support is critical for the country's economic recovery remain skeptical of the political processes that have been underway for the past 18 months, as evidenced by the lack of funds flow" (Zimplats, p. 6), But the Indigenization and Economic Empowerment Act not only effectively keeps foreign investors at bay, it also discourages expansion of domestic firms and leads to capital flight and transfers of profits abroad. Uncertainty about ownership status pushes firms to reduce their operations in Zimbabwe to a minimum. Uncertainty created by the Indigenization and Economic Empowerment Bill together with capital scarcity bodes ill for the prospect of reindustrialization of the Zimbabwean economy. Businesses complain that real interest rates even on shortterm loans running at doubledigit levels raise the bar of

57

See Zimplats 2010

82 | P a g e profitability extremely high. Furthermore, longerterm loans are almost impossible to secure. The combination of the absence of collateral and lax enforcement of private property rights by the judicial system appears to be one of major reasons for the high cost of capital there. =Expanding the exportbase calls for new investments. But as long as the investment climate is not overhauled, these investments will not come. And without them, the prospects for a significant increase in exports are doomed to failure. 4.3. Conclusion While a decade of economic chaos has exacted a heavy toll on Zimbabwe's industries, the problem is that policy changes introduced so far have done little to address major weaknesses in the business climate and remove the threat of confiscation of foreignowned firms (or more precisely firms not owned by formerly `disadvantaged people') by the state. It seems that the government has not taken advantage of the window of opportunity created by the decision to move to a multiple currency regime and overhaul its economic regime.58 Instead, the reform package was confined to the removal of regulations introduced earlier as part of government effort to combat hyperinflation through such administrative measures as price controls and foreign currency surrender requirements. The absence of measures that would make the investment environment friendlier to private business activities will continue hampering the recovery. Except for the removal of some direct price controls, no steps have been taken to improve the notoriously poor business climate. In contrast to, for instance, Poland's stabilizationcum transformation program from central planning, which included measures such as liberalizing foreign trade (suspension of import duties together with the removal of other measures raising the cost of imports and exports) and other regulations relevant to the cost of doing business, Zimbabwe's decision to replace its domestic currency with US dollars and the South African rand was not part of an indepth structural reform program that would address weaknesses in its investment and business climate. Other policy measures that accompanied the abandoning the domestic currency included price liberalization and discontinuation of the foreign exchange surrender requirements. Poland's measures went further and deeper. Exposition of domestic producers to competition from imports, easy access to imported products and services, and the creation of an attractive investment climate for domestic and foreign direct investment were all critical to a powerful supply response and Poland's entry on the path to sustainable economic growth. These ingredients are still missing in Zimbabwe's economic regime, as vividly illustrated by the shift in the composition of exports towards "bigitems," whose production has been relatively immune, albeit not completely, to weaknesses in the business environment. Reviving private business activity and the concomitant recovery of exports of both goods and services through measures including the removal of administrative barriers to conducting business operations and investment, improving government's capacity to enforce contracts and private property rights and targeted interventions to increase competitiveness should be a top priority for the government. This

Even this decision was flawed, as one currency should have been chosen as an official means of payments. The introduction of two currencies creates problems if their respective exchange rates move in different directions vis àvis each other. The appreciation of a currency triggers hoarding of this currency. Weaker currency pushes stronger one from circulation.

58

83 | P a g e calls for the development of a comprehensive program of economic reforms that would address all aspects of Zimbabwe's economic policy regime. Without significantly reducing the regulatory hassle and fiscal burden of conducting business in Zimbabwe, the revival of laborintensive exports will be impossible. But this is not only needed for job creation but also to alleviate the very real risk that the country will become even more dependent on foreign aid and continue underperforming relative to its significant potential.

84 | P a g e

References

ADB 2009: Africa Competitiveness Report, African Development Bank. Arvis, JeanFrançois, Gaël Raballand, and JeanFrançois Marteau. 2010. "The Cost of Being Landlocked: Logistics Costs and Supply Chain Reliability," Directions in Development: Trade, World Bank, Washington D.C. Baumol, W.J., R.E. Litan and C.E. Schramm. 2007. Good Capitalism, Bad Capitalism and the Economics of Growth and Prosperity. Yale University Press, New Haven and London. Balassa, Bela. 1965. "Trade Liberalization and 'Revealed' Comparative Advantage," Manchester School 33, pp. 99123. Clemens, Michael and Todd Moss. 2005. "Costs and Causes of Zimbabwe's Crisis," CDG Notes, Center for Global Development, Washington D.C. Collier, Paul. 2007. The Bottom Billion. Why the Poorest Countries Are Failing and What Can Be Done About It, Oxford University Press, Oxford, New York. Djankov, Simeon, Caroline Freund and Cong S. Pham. 2008. "Trading on Time." Review of Economics and Statistics, November. Gilson, Ian. 2010, "Deepening Regional Integration to Eliminate the Fragmented Goods Market in Southern Africa," mimeo, World Bank, Washington D.C. IMF 2009: "Zimbabwe: 2009 Article IV Consultation--Staff Report; Public Information Notice on the Executive Board Discussion; and Statement by the Executive Director for Zimbabwe." IMF Country Report No. 09/139, International Monetary Fund, Washington D.C., May. Implast 2009: Mineral Resource and Mineral Reserve Statement 2009, Impala, Rustenburg. Kaufmann D., A. Kraay, and M. Mastruzzi. 2007. Governance Matters VI: Governance Indicators for 1996 2007 at http://info.worldbank.org/governance/wgi2007/ Landesmann, M. and R. Stehrer. 2003. "Structural Patterns of EastWest European Integration: Strong and Weak Gershenkron Effects," in WIIW Structural Report 2003 on Central and Eastern Europe, Vol. 1, The Vienna Institute for International Economic Studies, Vienna. MoF 2010: The 2010 MidYear Fiscal Policy Review, presented by the Minister of Finance, Hon. T. Biti, Harare, July. Muñoz, Sònia. 2006. "Zimbabwe's Export Performance: The Impact of the Parallel Market and Governance Factors." IMF Working Paper WP/06/28, African Department, International Monetary Fund, Washington DC, January. Ng, Francis and M. Ataman Aksoy. 2008. "Who are the net food importing countries?" Policy Research Working Paper 4457, World Bank, January. Power, Samatha. 2003. "How to Kill a Country: Turning a breadbasket into a basket case in ten easy steps--the Robert Mugabe way," Atlantic Monthly, December. Raballand, Gaël, Charles Kunaka, and Bo Giersing. 2008. "The Impact of Regional Liberalization and Harmonization in Road Transport Services: A Focus on Zambia and Lessons for Landlocked Countries." Policy Research Working Paper 4482. World Bank, Africa Transport Department, Africa Sustainable Development Division, January.

85 | P a g e Richardson, Craig. 2005. "How the Loss of Property Rights Caused Zimbabwe's Collapse," Economic Development Bulletin No. 4, CATO Institute, Washington DC, November 14. Rotberg, Robert I. and Rachel M. Gisselquist. 2009. Strengthening African Governance: Index of African Governance, Results and Rankings 2009, Cambridge, MA. Sandawana. 2007. Africae Sandawana 2002--2007, The Sandawana Column, The New Zanj Publishing House, Second Edition, Harare Stone, J. I. 2001. Infrastructure Development in Landlocked and Transit Developing Countries: Foreign Aid, Private Investment and the Transport Cost Burden of Landlocked Developing Countries. UNCTAD/LDC/112. Geneva: UNCTAD. Weiner, Dan, Sam Moyo, Barry Munslow, and Phil O'Keefe. 1985. "Land Use and Agricultural Productivity in Zimbabwe," The Journal of Modern African Studies 23.2., 25185. Winters, L Alan. 1997. `The Economics of "Catching up" Revisited', The Vienna Institute for Comparative Economic Studies, Reprint Series, No. 168, June. WB 2010: "Note on foreign trade reporting in Zimbabwe: surprisingly decent quality of import statistics," World Bank, mimeo, April 2010. WB 2009: Global Economic Prospects 2009, World Bank, Washington D.C. WEF 2009: Global Competitiveness Report, 200809, Michael E. Porter and Klaus Schwab, eds., The World Economic Forum, Geneva, Switzerland. Wood, A., and K. Jordan, 2000, "Why Does Zimbabwe Export Manufactures and Uganda Not? Econometrics Meets History," Journal of Development Studies, Vol. 37, Number 2. Zimplats 2010: Zimplats Holdings Limited Annual Report 2010, Harare, 2010; accessed on October 23, 2010, available at http://www.zimplats.com/pdf/Ar2010.pdf ZTA 2008: Tourism trends and statistics annual report. Zimbabwe Tourism Authority, Harare, Zimbabwe.

86 | P a g e

Appendix Tables

Appendix Table 1: Imports according to various sources in selected years over 19942008 (in US$ millions and percent) .................................................................................................................................. 90 Appendix Table 2: Mirror imports in percent of official imports: totals and in trade with Zimbabwe in 200008 ....................................................................................................................................................... 91 Appendix Table 3: Exports in official and mirror statistics in 200102 and 200408 (in millions of US dollars and percent) .................................................................................................................................... 94 Appendix Table 4: Exports in percent of mirror exports in total trade and bilateral trade with Zimbabwe of Zimbabwe's major export markets in 2000 08 ...................................................................................... 95 Appendix Table 5: Difference between official and mirror statistics in percent in 200208 ...................... 96 Appendix Table 6: Directions of imports according to official and partners' imports statistics in 200008 (in percent) ................................................................................................................................................. 97 . Appendix Table 7: Trading across borders against overall ease of doing business in Zimbabwe and other comparator countries in 2010 .................................................................................................................... 99 Appendix Table 8: Formal cost of foreign trading in Zimbabwe and selected SADC countries in 2010 (in days and US dollars) .................................................................................................................................. 100 Appendix Table 9: Policy and infrastructureinduced costs of trading across borders in Zimbabwe and its neighbors in 2010 (in days and US dollars) ............................................................................................... 101 Appendix Table 10: Elements of the cost of cross border trading in Zimbabwe as compared with Malawi in 2010 ...................................................................................................................................................... 101 Appendix Table 11: Fiscal and administrative burden of paying taxes and contributions in Zimbabwe and neighboring countries in 2010 .................................................................................................................. 104

87 | P a g e

Appendix 1: Note on foreign trade reporting in Zimbabwe

INTRODUCTION

The objectives of this note is to perform a standard assessment of foreign trade statistics based on the data collected by the Customs administration and those collected by Zimbabwe's trading partners' statistics (mirror trade statistics). This note shows that Zimbabwe's authorities have been up to the task monitoring and reporting across border imports flows while exports have been rather sloppily caught in respective national statistics. This finding is quite surprising considering high levels of the incidence of corruption and low level of the quality of economic governance as perceived by foreign investors and analysts. Countries ranked low in surveys of Transparency International or the World Bank's quality of governance usually hugely under report both exports and imports. A decent degree of congruence of official imports/exports and partners' exports (mirror imports/mirror exports) to Zimbabwe suggest that smuggling and tax or duty evasion at the border is rather limited with one caveat. The portion of trade going through unofficial channels would neither be recorded in a country of origin nor in an importing country. We have no estimates of this unofficial trade. The following two observations concerning the use of national trade statistics in analytical work can be derived from this analysis: First, official imports statistics do not raise any substantive concerns that would make their use inappropriate. Second, as for analysis of export performance especially in the 2000s, mirror statistics provide much better approximation of developments in the real world. For these reasons, mirror statistics provide better insights into composition and dynamics of Zimbabwe's trade. Furthermore, the quality of foreign trade reporting of Zimbabwe's major trading partners--SACU and the EU--is as good as it can get.

MIRROR STATISTICS: CONCEPTUAL REMARKS

Mirror statistics, i.e., those derived from statistics of foreign trading partners, provide a useful check on the quality of foreign trade statistics of a country. Since most countries, except for the US,59 report exports as FOB (free on board) and imports as CIF (cost, insurance, freight), a perfect match between officially reported imports and imports derived from partners' exports, thereafter referred to as mirror imports, is not possible, as it would imply that freight and insurance are costless. Therefore, one would expect that the value of official imports should be significantly larger (smaller) than that of mirror imports (exports). For landlocked countries with imports concentrated in bulky primary commodities, the difference may be huge amounting even to 5060 percent. But there are other reasons contributing to the impossibility of a perfect match between country's official statistics and mirror statistics.

Total exports in the United States are valued free alongside, or FAS. FAS figures provide valuation of exports at the port of exportation and thus exclude charges for loading onto the vessel, the transportation itself, insurance, unloading, and foreign transportation. FAS is similar to FOB valuation of exports (see J.M. Donnelly, U.S.World Merchandise Trade Data: 19482006, CRS Report for Congress, Washington D.C, February, 2007, available at http://fpc.state.gov/documents/organization/81946.pdf.

59

88 | P a g e Officially disclosed import data may differ from imports into a country derived from partners' exports for various reasons related, among others, to differences in conventions (including or not including cost of delivery), smuggling, underinvoicing or misclassification, differences in exchange rates, extra budgetary arrangements for collecting customs revenues.60 Other source of differences relate to the fact that exports tend to be less thoroughly monitored than imports for fiscal reasons and that there are inherent difficulties in identifying reexport or entrepôt activity. Reexports by small traders, in particular, may be difficult to capture in official statistics. Last but not least, the quality of statistical services of some trading partner may be lacking; not to mention that some of them may choose not to publish foreign trade data. These caveats, however, completely lose their relevance if mirror imports systematically exceed the officially reported data. This is usually a strong indication of convoluted regulations governing access to domestic markets leading illegal activities such as smuggling, reverse discrimination or preferential margins (differences between MFN applied tariff rates and preferential tariff rates) implied by free trade agreements, misclassification of goods and underinvoicing of imports, albeit with a caveat. Liberal arrangements for imports for personal use may also lead to underreporting as customs simply do not register them. Furthermore, some governments use customs to collect revenue, which, although controlled by central authorities, is not part of the official state budget. Whatever the immediate reasons are behind negative discrepancies between official imports and mirror ones, they always indicate dramatic departures from the best international practice. Assuming that most trading partners disclose their foreign trade statistics (submit them to the UN COMTRADE database), the total value of their exports to a country (also referred to as mirror imports) should be lower than the total value of a country's reported imports (from these countries) by an equivalent of CIF. Ideally official imports (OM) should exceed mirror imports (MM) by transportation cost or OM = MM + CIF. The discrepancies between OM and MM (mirror trade gaps) provide information about the quality of foreign trade statistics collected by Customs. There are two cases: If the value MM > OM or the ratio of MM/OM*100 > 100, the difference (or the ratio) between the value of mirror imports and official imports, defined as a mirror trade gap, amounts to the minimum value of imports that go unreported into a country. It points to illegal activities at the border: undervaluation of imports to evade payments of duties and other charges or simply smuggling operations. If the value MM < OM and the ratio is lower than 100 by at least, as a rule of thumb, 10 percent, than there is no systemic problem at the Custom, The ratio is more than 10 percent lower for OECD economies as well as, among others, China and India.

Similar criterion can be used to assess the quality of exports statistics: official exports (XO) should be lower than the value of reported imports or mirror exports (XM) by the value of CIF. Here, however, one of the sources of discrepancies may be unrelated to the quality of monitoring goods outflows from a

For a detailed analysis of mirror trade statistics, see Zarubin, G and Bartlomiej Kaminski, "The Limits of Mirror Statistics of Foreign Trade" in M. Belkindas and O. Ivanova, eds., Foreign Trade Statistics in the USSR and Successor States, Studies of Economies in Transition 18, The World Bank, Washington DC, 1995.

60

89 | P a g e country: this may simply reflect underreporting in partner countries. For this and other reasons,61 mirror export statistics do not provide as strong indication of the quality of reporting of exports as an analysis of imports data does. These stipulations notwithstanding, underreporting of exports is relatively common in many developing countries. By the same token, an assessment of the extent to which official statistics do not capture them is important for foreign trade performance analysis. In addition to the COMTRADE database, another source of statistics on foreign trade is the IMF DOT (direction of trade) database. These statistics rely on both national reporting and reconciliation of them against statistics available from other countries. In consequence, the final data are closely aligned, albeit not identical as not all countries report foreign trade data to the UN, with mirror statistics that can be derived from the COMTRADE database.

IMPORTS: IMPRESSIVE ACCURACY

The statistics shown below compare imports totals drawn from statistics Zimbabwe reported to UN COMTRADE with corresponding mirror imports, i.e., exports to Zimbabwe reported by Zimbabwe's trading partners, on the one hand, import statistics published in the most recent Direction of Trade Annual with those derived from export statistics of countries exporting to Zimbabwe. The picture that emerges from this statistics is that there is no systemic problem with the quality of official import statistics as deviations do not point to positive trade gaps consistent over time, although there are some indications of underreporting in recent years. The differences between official and IMF statistics do point to any systematic problems with official reporting. The discrepancies between the IMF and COMTRADE import totals have been moving from positive to negative territory. Their total over 19992008 amounted to US$90 million suggesting that differences have tended to offset each other over time. By the IMF statistics benchmark, imports appear to have been underreported only in 199697, 2006, and 2008 (Appendix Table 1). But they were `over reported' in 200204 and 2007. Turning to mirror import statistics, i.e., based on exports into Zimbabwe reported by countries to the UN, and set against the IMF data the pattern appears to be consistent with adequate trade reporting. Since imports in the IMF statistics include CIF, one would expect they would be 1020 percent higher than mirror FOB imports reported in the UN COMTRADE. But except in 2005, 2006 and 2007 when IMF imports were either 20 percent above mirror imports (in 2006 or 6 percent in 2005 and 2007, the difference was much in other years. The average for 200108 was 32 percent indicating that mirror imports were underreported. What explains this gap? We shall return to this question once Zimbabwe's official imports are tested against mirror imports.

Conversely, official exports should be lower than their respective imports, i.e., mirror exports, reported in partners' statistics. However, since governments have no fiscal stake in exports, as most countries do not levy charges on them, customs services are not vitally interested in verification of exports invoices and precisely reporting them. They are relevant only for economic analysis but do not shed light on the quality of border controls of an exporting country.

61

90 | P a g e

Appendix Figure 1: Imports into Zimbabwe in current prices according to various sources in 19902008 (in millions of US dollars)

4,500 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 0

1990 1991 1992 1993 1994 1995 1996 1997 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Mirror imports

Official imports

IMF imports

Source: IMF DOT and UN COMTRADE databases and Central Statistical Office, Harare, for 2000 and 2003 when no data were reported to the UN

In fact, more pertinent to our analysis is an assessment of mirror trade gaps between official imports and mirror imports derived from the UN COMTRADE statistics. Surprisingly for a country at a low level of the GDP per capita, official imports have consistently exceeded the values of mirror imports indicating a rather impressive absence of positive trade gaps associated with underreporting. Trade gap was negative in every year over 19922008, although it has been declining since 2004. The reasons for its contraction are not clear.

Appendix Table 1: Imports according to various sources in selected years over 19942008 (in US$ millions and percent)

(IMF) (UN) Difference Difference (in percent) Mirror (UN MM) Difference: IMFUN MM Difference IMF vs. UN MM in percent UN Ratio: MM to MO 1994 2,241 2,241 1 0 1,486 755 1995 2,726 2,659 68 3 2,220 507 1996 4,124 2,813 1,310 47 2,290 1,833 1997 3,278 3,092 186 6 2,146 1,132 1999 2,127 2,126 1 0 1,651 476 2001 1,718 1,715 3 0 1,160 558 2002 2,326 2,467 141 6 1,227 1,099 2004 2,153 2,204 51 2 1,702 451 2005 2,072 2,072 1 0 1,954 117 2006 2,812 2,577 236 9 2,348 465 2007 2,681 3,442 761 22 2,526 155 2008 3,457 2,832 625 27 2,680 777

51

66

23

83

80

81

53

69

29

78

48

68

90

50

27

77

6

94

20

91

6

73

29

95

Notes: (IMF) ­ imports data derived from the IMF DOT; UN stands for exports reported to the COMTRADE database; MM ­ mirror imports (exports to Zimbabwe reported by its trading partners; Source: IMF DOT and UN COMTRADE databases.

91 | P a g e One possible explanation is the improvement in the quality of statistics of trading partners that has led to the extension of coverage of Zimbabwedirected exports.62 However, a closer look at the origins of imports suggests that this would be an unlikely explanation as reliable reporters (EU27, South African Customs Union, US, China, Mozambique and Zambia) accounted on average in 200408 for 76 percent of total official imports and 91 percent of Zimbabwe's total mirror imports (Appendix Table 2).

Appendix Table 2: Mirror imports in percent of official imports: totals and in trade with Zimbabwe in 200008 SACU total SACU ZWE Zambia total Zambia ZWE Mozambique total Mozambique ZWE EU 27 total EU 27 ZWE China total China ZWE USA total USA ZWE Memorandum: Share of the above in ZWE total imports Share of the above in ZWE mirror imports Mirror imports in % of ZWE imports from the above 2000 2001 2002 2003 2004 2005 2006 2007 2008 80 N/AV 117

114

Average 200408 85 141 73

114

83 83 83

74

79 81 57 N/AV 98 57

N/AV 120

77 84 58

97

89 324 58

118

91 121 118

74

85 80 67

103

85 96 64

102

95 N/AV 85 N/AV 61 91 96 N/AV . N/AV 96 N/AV

85 41 83 102 62 88 97 65 71 92 126

86 97 120 N/AV 81 82 41 N/AV 65 69 88 88 96 96 57 N/AV 80 N/AV 91 89

93 81 93 77 92 67 22 38 59 100 80 84 83 82 84 95 106 80 69 88 71 74 76 79 83 84 85 86 83 79 97 99 101 101 103 114 152 20 85 82 76 90 96 45 94 87 86 92 58 82 88 68 92 93 103

87 57 83 87 77 83 100 91 76 91 86

65 N/AV

Notes: ZWE stands for Zimbabwe; N/AV--not available; SACU (Southern African Customs Union) includes the following states: Botswana, Lesotho, Namibia, South Africa and Swaziland. Source: Derived from data in UN COMTRADE database.

The quality of their statistics does not leave much to be desired as the data tabulated in Appendix Table 2 show at least in terms of their respective total world imports. The ratios of mirror imports to official imports for total trade are stable and below 100. The United States is an exception mainly because of unique practice of not including CIF in total values of imports. So is Zambia but only in two years--2000 and 2006. But the values of this ratio often exceed 100 percent in trade with Zimbabwe indicating Zimbabwe's underreporting of imports. Except for trade with SACU, other discrepancies can be explained by reporting in different years than in exporter countries or rather commonplace technical issues always distorting trade statistics.63 SACU's

62

Note that while there are often strong incentives to underreport and undervalue the value of imports, there are rarely any incentives to distort exports data. Hence, even in countries where there is significant under reporting of imports, export statistics tend to have better coverage. 63 As convincingly demonstrated in an empirical study of trade statistics worldwide, no full conformity can be achieved even with statistics of highly developed countries. See Jerzy Rozanski and Alexander Yeats (1994). `On the

92 | P a g e Zimbabweoriented exports of US$1.4 billion in 2005 and US$1.7 billion in 2006 exceeded Zimbabwe's official imports from SACU by US$935 million and US$300 million in these years. Since in other years there were no under or overreporting, it is not clear what triggered these differences. Other cases can be easily explained: for instance, imports from the US were underreported to the tune of US$6 million and US$15 million in 2004 and 2005 in Zimbabwe's statistics but in 2006 the value of these imports was "overreported" by almost US$200 million. On the other hand, a seesaw pattern of the values of trade gaps with Zambia reflects technical issues related to timing and currency conversions rather than to systemic weaknesses in border controls. Another possible explanation of the significant increase in the ratio of mirror imports to official imports since 2005 may have something to do with the contraction of transportation costs of imports due to the increased share of neighboring countries in total imports. Indeed, the shift in geographical patterns of imports towards neighboring countries, which has been most likely accompanied by the decline in transportation costs, has been huge. The share of Zimbabwe's adjacent countries--Botswana, Mozambique, South Africa, and Zambia--significantly increased from an average of 43 percent in 1995 99 to 63 percent in 200607 and surged to 75 percent in 2008. Simultaneously, the share of the EU fell from an average of around 23 percent in 199599 to an average of 8 percent in 200408.64 It is obvious that the cost of transport from neighbors is a fraction of the cost of bringing the same goods from Western Europe. It appears that the bulk of exports recorded in respective countries of origin have been also caught and reported by Zimbabwe's customs administration. Although the ratio of mirror imports to official imports has been moving towards the value of 100 percent, it has never exceeded this level. There is one exception: the value of mirror imports in trade with neighbors (SACU plus Mozambique and Zambia) and other major trading partners identified in Appendix Table 2 (EU27, China, and USA) exceeded that of official imports in 2008 indicating underreporting. However, there may be some other indications of some imports not reported but the quantities involved remain very low. Overall, the conclusion relevant for the use of official import statistics is that they provide surprisingly good information about actual flows except for some years, which can be easily identified. In contrast to trade statistics in many developing countries, there are no indications of imports moving across borders without being reported to the authorities including statistical services.

EXPORTS: PUZZLING OVERREPORTING IN THE 2000S

Against the background of a good coverage of imports' reporting in Zimbabwe's official statistics, data on exports strike one as rather sloppy and greatly inaccurate especially over the last decade. Figure 2 presents the time profiles of exports in terms of value as reported in official statistics, the IMF DOT database and the UN COMTRADE database as derived from world imports from Zimbabwe, i.e., Zimbabwe's mirror exports. In contrast to mirror exports, which have been characterized by low dispersion along the longterm trend, IMF and official exports data have displayed strong volatility since

(in) accuracy of economic observations: An assessment of trends in the reliability of international trade statistics', Journal of Development Economics, 44(1), pp.103130. 64 Based on data reported by Zimbabwe to the UN COMTRADE database.

93 | P a g e 2000. In 199099, three export statistics were moving in tandem: an average difference between the IMF and official data was one percent and that between the latter and mirror statistics amounted to 5 percent. In 200008, the IMF estimates and official data have been moving at different rates and in opposite directions: according to one set exports were falling and according to the other they were raising in all years except in 200001 and 2008. While the average rate of growth of IMF exports of 6.9 percent in 19911997 was the same as that of official exports and both below the average of 10 percent for mirror exports,65 in 200008, IMF and official exports grew at different rates with the IMF data yielding an average rate of growth of 18 percent and the official data an average rate of 32 percent. While mirror data also displayed volatility, the average rate of growth of 2 percent was dramatically lower than for other statistics.66

Appendix Figure 2: Exports in terms of value according to various sources in 19902008 (in millions of US dollars)

7,000 6,000 5,000 4,000 3,000 2,000 1,000 1990 1991 1992 1993 1994 1995 1996 1997 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 EXPORTS OFFICIAL IMF EXPORTS MIRROR EXPORTS

Source: IMF DOT and UN COMTRADE databases.

However, the most surprising is an enormous surge in official exports in 2006 not reflected in other statistics: the value of exports of US$6.5 billion reported in 2006 is clearly out of line begging an explanation.67 While this would go beyond the format of this note, let us note that this is a statistical fantasy as exports would amount to 173 percent of the GDP, which was US$3.7 billion in 2006,68 up from 41 percent in 2005 and trade surplus of US$3,851 million would amount to 103 percent of the GDP. That would be the only time that the balance of trade in goods had surplus over 19902008. Furthermore, Zimbabwe's exports to Zambia would be an equivalent of 47 percent of their total imports up from 4 percent a year earlier in 2005.

Since no official data on exports in 1998 are available, it is impossible to calculate growth rates in 1998 and 1999. Coefficient of variation (ratio of standard deviation to the average) was high: 4.2 for the official data and 4.0 for the IMF statistics. The value of coefficient of variation of mirror exports was also high but at 3.8 significantly smaller than for the IMF and official data. 67 As African countries tend to underreport exports (see Francis Ng and Alexander Yeats, "Kenya: Export prospects and problems," Africa Region: Working Paper Series No. 90, World Bank, Washington DC, October 2005), this is rather an unusual development. 68 GDP in current dollars extracted from the World Bank WDI database.

66 65

94 | P a g e Some further insights can be derived from the data tabulated in Appendix Table 3 comparing 1994 to 2008 export totals drawn from statistics that Zimbabwe reported to UN COMTRADE with corresponding export statistics in the IMF DOT database, on the one hand, and partner's data on imports from Zimbabwe, i.e., mirror exports. As noted above, the IMF data present four different pictures depending on the period: rather precise reporting in 199299; huge underreporting in 200001; overreporting in 200207; and underreporting in 2008. In 199299 exports reported by Zimbabwe to UN COMTRADE were almost fully in line with the IMF DOT statistics running on average at 98 percent of the IMF's estimates. In 200001 this ratio dropped to 59 percent and 54 percent with the total aggregate under reported exports amounting to US$2.4 billion. The 200207 witnessed a switch in the opposing direction: exports were hugely overreported especially in 2006 by US$5.5 billion or 6.8 times the value of exports estimated in the IMF DOT statistics. The value of official exports increased in 2006 almost fivefold over its level in 2005 (Appendix Table 3). In 2007, official exports were around 30 percent above the IMF's estimate. In 2008, the pendulum went in the opposite direction: as the value of IMF exports increased almost three times and official exports fell to 51 percent of its level in 2007, the value of nonreported exports reached US$701 million or 71 percent of the IMF's DOT estimate.

Appendix Table 3: Exports in official and mirror statistics in 200102 and 200408 (in millions of US dollars and percent)

Exports IMF Exports UN (XO) Difference: IMFUN UN in % of IMF Mirror exports UN (XM) Difference: XO ­ XM Official exports in % of mirror exports 1994 1,971 1,967 4 100 1,607 361 122 1995 1,901 1,846 55 97 1,712 134 108 1996 2,242 2,122 120 95 1,986 136 107 1997 2,146 2,128 18 99 2,095 32 102 1999 1,908 1,887 20 99 1,778 109 106 2000 3,283 1,925 1,358 59 1,855 70 104 2001 2,222 1,207 1,015 54 1,809 602 67 2002 2,009 2,327 318 116 1,728 599 135 2004 1,769 1,926 158 109 1,924 2 100 2005 1,395 1,394 1 100 1,974 581 71 2006 941 6,427 5,487 683 2,227 4,201 289 2007 2,506 3,308 803 132 2,414 895 137 2008 2,400 1,694 706 71 2,222 528 76

Source: Derived from trade statistics reported to the UN COMTRADE and IMF DOT databases. No official exports data available for 2003.

Before turning to an examination of the data on mirror exports against official exports, the point to be borne in mind is that, as noted earlier, mirror exports should be higher than exports reported by Zimbabwe by the cost of transport and insurance since imports include CIF. It is generally assumed that this adds about 10 percent of the value of shipments. For a landlocked country, the cost of transportation may be even higher. Except for 2001, 2005 and 2008, official exports exceeded mirror exports: the difference, however, varied rather significantly especially in the 2000s. There was huge underreporting of exports in 2005 and 2008 and enormous overreporting in 2006. Since distortions in mirror export statistics heavily depend on the quality of statistical reporting in partner countries, their examination shed some extra light on their reliability in assessing official data. The reliability of mirror export data is larger if official exports are 1015 percentage points below mirror

95 | P a g e exports. By this measure, mirror data appear to reflect adequately exports data of SACU and Mozambique but not of Zambia and the EU (Appendix Table 4).69

Appendix Table 4: Exports in percent of mirror exports in total trade and bilateral trade with Zimbabwe of Zimbabwe's major export markets in 2000 08 SACU total SACU ZWE Zambia total Zambia ZWE Mozambique total Mozambique ZWE EU 27 total EU 27 ZWE Memorandum: share of the above in official exports share of the above in mirror exports 2000 2001 2002 2003 2004 2005 2006 2007 2008 84 143 140 121 95 511 101 59 49 58 58 59 76 62 146 12 85 43 103 74 68 79 192 N/AV 128 128 222 N/AV 86 97 465 N/AV 106 106 87 . 59 N/AV 59 63 78 126 103 58 93 205 108 73 80 79 79 76 108 186 158 106 72 136 131 155 70 938 111 65 81 93 77 92 252 1,519 3,468 271 108 108 108 112 47 310 103 79 70 61 77 66 78 66 81 64 Average 200408 79 137 120 249 87 1,143 109 122 73 64

59 62

Source: Derived from trade statistics reported to the UN COMTRADE.

But the quality of reporting looks different in their trade with Zimbabwe. Zimbabwe has tended to over report its exports to all its major trading partners with the exception of the EU in 200005 and 2008 and pervasive corruption in these countries cannot explain it. The extent of overreporting was particularly high in trade with Zambia in 2006 and Mozambique in 2000 and 20012008 with skyhigh levels having been reached in 200607. Zimbabwe's exports to Mozambique were 35 times higher than its imports from Zimbabwe in 2007. In all, the quality of exports statistics significantly deteriorated in the 2000s. Mirror exports were around US$0.6 billion higher than official exports in 2005 and 2008 but US$4.2 billion and US$0.9 billion lower in 2006 and 2007 respectively. While underreporting of imports by its trading partners may be responsible for some portion of Zimbabwe's `excess exports,' the differences between IMF and mirror statistics suggest this has not been a significant factor. .

OFFICIAL AND MIRROR STATISTICS: IMPLICATIONS FOR DYNAMICS

While both sets of statistics capture similar overall trends in the development of Zimbabwe's trade in 2000s, there are differences in pace of scope of reported change over 200208. Several observations can be derived from Appendix Table 5 comparing official data with corresponding data reported in statistics of Zimbabwe's trading partners. First, geographical concentration of imports is lower according to official than mirror imports statistics, albeit two sets have been converging. Official statistics appear to have been consistently understating imports from SACU, Zambia, and the European Union. But the gap

Since the EU is unlikely to underreport exports, this rather suggests that products imported from the EU enter illegally or are subject of underinvoicing in some countries. The same cannot be said about Zambia: one suspects that exports have been underreported.

69

96 | P a g e significantly declined by 2008 (see Appendix Table 5). The average shares of imports from these partners over 200208 were significantly lower than their respective shares in total mirror imports: for instance, the share of SACU in Zimbabwe's total imports was 55 percent and 66 percent in mirror imports but in 2008 both were very close to each other (70 percent and 71 percent).

Appendix Table 5: Difference between official and mirror statistics in percent in 200208 Shares SACU, of which South Africa Botswana Mozambique Zambia EU27 Value of trade World SACU, of which South Africa Botswana Mozambique Zambia EU27 Dynamic World SACU, of which South Africa Botswana Mozambique Zambia EU27 Imports Average, 200208 27.0 19.3 53.2 36.5 23.4 5.2 Average, 200208 23.5 3.7 10.4 9.4 54.5 10.4 23.9 Least Square Growth, 200108 56.0 20.3 29.2 39.2 8.9 3.9 136.2 2008 1.8 1.5 5.9 5.2 9.3 7.4 2008 5.4 3.6 4.0 0.2 0.4 3.5 12.4 Avg. growth rates 200408 74.0 63.6 61.9 40.4 77.1 49.4 74.5 Exports Average, 200208 15.9 14.2 35.5 82.7 23.9 16.6 Average, 200208 21.7 26.1 24.5 54.5 89.7 62.9 6.3 Least Square Growth, 200108 58.0 33.0 17.8 102.8 84.5 96.3 742.3 2008 28.4 18.9 77.6 71.9 16.0 3.6 2008 31.7 4.0 6.8 70.6 63.1 52.7 27.0 Avg. growth rates 200408 87.4 21.8 91.1 104.2 92.9 98.1 106.8

Source: Derived from trade statistics reported to the UN COMTRADE by Zimbabwe and its foreign trading partners.

So were their respective rates of growth. Total mirror imports grew much faster than official imports with the largest discrepancy for imports from the EU: according to the official data they fell at a LSG rate of minus one percent, whereas mirror statistics pointed to a modest growth of 0.3 percent. The only exceptions were Botswana and Zambia, as official statistics indicated stronger growth. Second, official data tend to overreport exports. In consequence, the difference between respective growth rates amounts to 58 percent for LSG rate (10.4 percent versus 4.4 percent) and to 87 percent for an average annual growth rate (a whopping 44 percent against 5.5 percent for mirror exports). The discrepancy between shares of respective countries and regional groupings has remained very high. For instance, the share of SACU in officially reported exports was 52 percent and in total mirror exports was 37 percent in 2009. Much larger shares of Zimbabwe's two neighboring countries, Mozambique and Zambia, in Zimbabwe's official exports than their corresponding shares raise suspicion that both of them underreport imports from Zimbabwe for tax evasion reasons. Surprisingly, given its high quality of

97 | P a g e governance, the same may apply to Botswana with the official average share of 4.1 percent in 200208 almost 55 percent higher than in mirror exports.

Appendix Table 6: Directions of imports according to official and partners' imports statistics in 200008 (in percent) SACU, of which South Africa Botswana Mozambique Zambia Malawi EU27, of which United Kingdom Rest of the world Memorandum: Neighboring countries SACU, of which South Africa Botswana Mozambique Zambia Malawi EU27, of which United Kingdom Rest of the world Memorandum: Neighboring countries 64.9 2001 2002 2004 2005 2006 2007 2008 Direction of imports according to official statistics 49.7 46.8 2.0 5.2 0.8 0.2 11.6 3.2 29.9 54.9 52.6 1.9 1.9 0.9 0.1 17.0 4.9 19.0 57.1 52.6 4.0 2.3 3.4 0.4 9.0 3.8 22.1 20.1 15.0 5.0 9.9 3.7 0.1 7.1 1.7 52.9 54.1 45.5 8.2 7.7 1.8 0.7 9.0 3.6 10.7 56.7 44.6 11.9 3.6 3.2 5.0 8.7 3.1 12.7 70.0 62.1 7.6 2.9 2.2 0.9 7.7 2.4 6.4 133 116 113 137 78 178 142 45 33 35 115 Index 2008 Aver.200102=100 134 125 392 81 259 601 54 59 26

54.8 57.3 62.3 33.6 63.2 63.3 74.8 Direction of imports according to mirror statistics 60.9 54.6 5.7 3.2 1.4 0.7 17.3 4.5 14.9 62.5 56.4 6.0 4.6 1.3 0.5 14.1 4.2 11.5 68.3 69.5 62.5 54.6 7.6 2.0 5.2 0.5 11.1 2.8 9.3 75.0 69.2 59.4 9.4 2.4 3.8 0.6 8.0 2.1 7.6 61.7 72.1 45.4 10.8 3.2 2.3 0.9 7.8 1.5 21.6 68.1 62.1 47.3 14.7 2.9 3.2 5.2 8.1 1.8 5.7 71.3 63.0 8.0 3.0 2.4 0.8 7.2 1.4 4.6 76.5

Notes: Neighboring countries include Botswana, Mozambique, South Africa and Zambia Source: Derived from trade statistics reported to the UN COMTRADE by Zimbabwe and its foreign trading partners.

Yet, all these differences notwithstanding, the general picture emerging from both sets of statistics has one common feature in common: there has been a rather dramatic reorientation of exports towards neighboring countries as well as in imports, although not to the same extent. Their share in official exports increased from 25 percent in 2000 to 58 percent in 2008 while that in mirror exports from 18 percent to 42 percent over the same period (Appendix Table 6). The respective shares for imports are 55 percent and 75 percent for official imports and 65 percent and 76 percent for mirror imports.

CONCLUSION

Foreign trade statistics of Zimbabwe do not display biases typical for a developing country. There has not been systematic bias of underreporting imports usually indicating evasion of border charges through either smuggling, misclassification of imported goods, and underinvoicing. Neither has been any tendency to underreport exports. If anything, exports reporting has been biased in favor of over reporting them.

98 | P a g e In fact, the puzzle is why exports were so overstated in national statistics in 2006. The direction of exports changed dramatically albeit only for a year with the share of Zambia in total exports rising to 26 percent up from 6 percent a year before and that of South Africa falling from 42 percent in 2005 to 17 percent. It seems that hyperinflation rendered the task of converting local exports invoices particularly challenging. Whatever, the reason, this clearly excludes official exports data for 2006 and 2007 from any discussion of long term foreign trade performance of Zimbabwe. In all, for practical reasons, mirror statistics provide better insights into composition and dynamics of Zimbabwe's trade. Furthermore, the quality of foreign trade reporting of Zimbabwe's major trading partners--SACU and the EU--is as good as it can get.

99 | P a g e

Appendix 2: Note on geography and policyinduced costs of trading across borders

Zimbabwe scores low in the World Bank's Cost of Doing Business comparative surveys seeking to assess formal barriers to conducting business activities by assessing conditions in following areas: starting business, labor market flexibility, registering property, contract enforcement, bankruptcy, protection of investors, foreign trade, and getting credit.70 Zimbabwe was ranked 160th out of 1983 countries and has been among thirty or so countries with the highest hassle cost of doing business related to regulations. In trading across borders, Zimbabwe was ranked 167th or 11 ranks below the average for its neighbors (Appendix Table 6).

Appendix Table 7: Trading across borders against overall ease of doing business in Zimbabwe and other comparator countries in 2010 Ease of Doing Business Botswana Mozambique South Africa Zambia Zimbabwe Memorandum: Malawi Mauritius Namibia Rwanda 132 17 66 67 172 19 151 170 40 2 85 103 (a) 45 135 34 90 159 Trading Across Borders (b) 150 136 148 157 167 Difference between (a) and (b) (c) 105 1 114 67 8

Source: data downloaded from http://www.doingbusiness.org/

Although the deviation of 11 ranks from the average of comparators' rankings in trading across borders is relatively low compared to other dimensions of formal cost of doing business, this is not the result of transparency and simplicity of Zimbabwe's foreign trade regime but it is due to high transaction cost of foreign trading in other countries of the region. All of them score in "trading across borders" well below their average score across all dimensions of the formal cost of doing business with the difference at singledigit level only for Mozambique, Zimbabwe, and Mauritius. Yet, Zimbabwe's rank in trading across borders is the lowest among comparators, which are not demanding at all in terms of transaction costs associated with foreign trade activities (Appendix Table 6). Zimbabwe's low ranking in terms of exploiting opportunities offered by foreign tradeled economic development stems not from its geographical location but selfinflicted policyinduced costs. A closer examination of the components of the transaction cost of completing export or import operation corroborates this observation. Appendix Table 7 presents data on formal costs of trading in Zimbabwe set against some of its SADC partners ranked according to the cost of importing. Zimbabwe had the third highest cost to import and to export in the world after Chad and Central African Republic. While it would be tempting to associate high costs with transportation costs exacerbated by the lack of direct access to cheaper maritime transport, `landlockedness' is only part of an explanation.

Each area has several different indicators estimated by local professionals dealing with issues related to each of these areas that have been surveyed in respective countries in both large cities and rural areas. The Cost of Doing Business' advantage over surveying firms is obvious. Professionals deal with respective issues on a daytoday basis, whereas firms confront the issues only to the extent that the issues directly affect them. The former have an overall view, whereas the latter do not.

70

100 | P a g e

Appendix Table 8: Formal cost of foreign trading in Zimbabwe and selected SADC countries in 2010 (in days and US dollars) Documents to export (number) Time to export (days) Cost to export (US$ per container) Documents to import (number) Time to import (days) 73 35 64 41 51 24 35 30 31 14 Cost to import (US$ per container) 5,101 5,070 3,335 3,264 2,570 1,813 1,807 1,475 1,475 689

Rank world

Country

3 Zimbabwe 7 53 3,280 9 4 Rwanda 9 38 3,275 9 13 Zambia 6 53 2,664 9 14 Botswana 6 30 2,810 9 26 Malawi 11 41 1,713 10 47 Namibia 11 29 1,686 9 49 South Africa 8 30 1,531 9 69 Mozambique 7 23 1,100 10 70 Tanzania 5 24 1,262 7 170 Mauritius 5 14 737 6 Memorandum: average and two countries with highest transportation cost in the world

Average (excluding Zimbabwe) 8 31 1,864 9 36 2,389 Zimbabwe in % of average 93 169 176 104 202 214 1 Chad 6 75 5,497 10 100 6,150 2 Central African Republic 9 54 5,491 17 62 5,554 Note: Inland transportation cost is based on 20-ft container FCL of general cargo that is valued at US$20,000 Source: Based on data in Doing Business 2010: Reforming through Difficult Times, World Bank, Washington D.C. 2010

Geography alone cannot explain much higher costs incurred by Zimbabwe's traders. Note that Botswana, Malawi, Rwanda and Zambia are also landlocked, yet their costs of both imports and exports are lower, albeit the difference is hardly significant for Rwanda. Zambia's exporters paid 20 percent less, Botswana's 14 percent, and Malawi's 48 percent, whereas importers paid 35 percent less in Zambia, 36 percent less in Botswana and 50 percent less in Malawi. Note also that both exporters and importers from Zimbabwe needed significantly more time (about twice as much as compared with the average for selected SADC countries listed in Appendix Table 7) to complete both export and import transactions. Transport is not the only component determining the number of days and amounts spent per container of foreign shipments.71 The breakdown of components identified to produce time and cost estimates allows distinguishing between the policyinduced or procedural aspects and the infrastructuralrelated, although there is an overlap between the two (Appendix Table 8). The latter reflects the fact that poor organization of, for instance, handling operations can dramatically raise the cost even though the port is uptodate in terms of technology and facilities. Costs and times involved stem from bureaucratic procedures in place and, on the other hand, from infrastructural bottlenecks. While not much can be done in a short time perspective about weakly developed infrastructure contributing to high costs through high transportation cost (inland transportation), bureaucratically induced barriers can be easily addressed provided there is political will to do so. Bureaucratic procedures consume more time than dealing with infrastructureimposed or technical constraints while the latter account for most of the total transaction cost incurred by exporters and importers alike (Appendix Table 8). Meeting bureaucratic requirements accounts for the bulk of the time needed to complete both export and import transactions: between 53 percent (Botswana) and 65 percent (Mozambique) of the total time needed to complete a standardized export transaction, and

See Simeon Djankov, Caroline Freund and Cong S. Pham. 2008. "Trading on Time." Review of Economics and Statistics, November.

71

101 | P a g e between 49 percent (Botswana) and 77 percent (Mozambique). Inland transportation and cargo handling at ports account for between 60 percent (Mozambique) and 91 percent (Botswana) of total export transaction costs and between 47 percent (Mozambique) and 85 percent (Zambia and Zimbabwe) of the total import transaction cost.

Appendix Table 9: Policy and infrastructureinduced costs of trading across borders in Zimbabwe and its neighbors in 2010 (in days and US dollars)

Zimbabwe Duration (days) US$ Cost Zambia Duration (days) US$ Cost Botswana Duration (days) US$ Cost Mozambique Duration (days) US$ Cost South Africa Duration (days) US$ Cost

Nature of Export Procedures

Procedures: Documents preparation and customs/technical clearance Infrastructure: Ports and terminal handling and inland transportation Totals:

32

480

39

316

16

260

15

435

19

347

21 53 Duration (days) 46

2,800 3,280 US$ Cost 752

14 53 Duration (days) 37

2,348 2,664 US$ Cost 486

14 30 Duration (days) 20

2,550 2,810 US$ Cost 515

8 23 Duration (days) 23

665 1,100 US$ Cost 775

11 30 Duration (days) 18

1,184 1,531 US$ Cost 472

Nature of Import Procedures

Documents preparation and customs/technical clearance Infrastructure: Ports and terminal handling and inland transportation Totals:

27 73

4,349 5,101

27 64

2,849 3,335

21 41

2,749 3,264

7 30

700 1,475

17 35

1,335 1,807

Source: Derived from data available at http://www.doingbusiness.org/ExploreEconomies/?economyid=172 accessed on June 2, 2010 Appendix Table 10: Elements of the cost of cross border trading in Zimbabwe as compared with Malawi in 2010 Memo: Malawi in % of Zimbabwe Duration (days) 84 67 77 Duration (days) 61 85 US$ Cost 99 44 52 US$ Cost 57 49

Nature of Export Procedures Procedures: Documents preparation and customs/technical clearance Infrastructure: Ports and terminal handling and inland transportation Totals: Nature of Import Procedures Documents preparation and customs/technical clearance Infrastructure: Ports and terminal handling and inland transportation

Zimbabwe Duration (days) 32 21 US$ Cost 480 2,800

Malawi Duration (days) 27 14 US$ Cost 473 1,240

53 3,280 41 1,713 Duration US$ Duration US$ (days) Cost (days) Cost 46 27 752 4,349 28 23 430 2,140

Totals: 73 5,101 51 2,570 70 50 Note: Inland transportation cost is based on 20-ft container FCL of general cargo that is valued at US$20,000 Source: Based on data in Doing Business 2010: Reforming through Difficult Times, World Bank, Washington D.C. 2010.

Zimbabwe, however, beats each neighbor across all discussed dimensions of foreign trade transaction costs in both dollar and time terms except for Zambia where it takes more time to prepare documents necessary for exports, albeit not for imports. Lower infrastructurerelated costs in Zambia than in Zimbabwe raise suspicion that they are under stated for the latter. But this does not seem to be the case, as shipments from Zambia can go directly to

102 | P a g e ports in Mozambique bypassing Zimbabwe. Furthermore, the cost of delivering shipments and getting them ready for shipping by sea for Malawi's firms is much lower than for firms in Zimbabwe: the cost of ports and terminal handling and inland transportation is US$1,240 for Malawi as compared with US$2,800 and for imports it is US$2,140 against US$4,349. In all, foreign trade transaction costs of Malawi firms are roughly half of what their Zimbabwean counterparts have to pay (Appendix Table 9 above). The above discussion provides empirical illustration of the fact that geography and infrastructure is not a barrier to foreign trade of Zimbabwe. Governance is the major impediment and the main reason why Zimbabwean economy has failed to exploit opportunities offered by plugging into global markets.

103 | P a g e

Appendix 3: Taxation as the most binding formal constraint to do business in Zimbabwe

In terms of formal ease of doing business, Zimbabwe has the most hostile regulatory environment as compared to its neighbors, albeit its positive score on enforcement of contract is an encouraging sign for easiness of improvement. Excluding mining, had decisions by foreign investors been made solely on the basis of hassle cost of conducting business operation, Zimbabwe's neighbors would have attracted capital well before their attention would have turned to Zimbabwe. Legal environment for doing business in Zimbabwe imposes much larger burden on firms to comply with regulations than in neighboring countries. Zimbabwe ranked 159th in overall ease of doing business among 183 countries surveyed by the World Bank's cost of doing business in 2010.72 Its rank was well below rankings of its neighbors--Botswana was ranked 45th, Mozambique 135th, South Africa 34th, and Zambia was ranked 90th. Zimbabwe's ranking in ten dimensions of doing business (starting business, labor market flexibility, registering property, contract enforcement, bankruptcy, paying taxes, licenses needed to build a warehouse, foreign trade, and protection of investors) is better than the average ranking of its neighbors only in registering property (11 ranks higher) and enforcing property (17) (See Figure 9). Since enforcing contracts cannot be improved with the stroke of a pen, Zimbabwe's better ranking (78th) than any of its neighbors seems to be encouraging.

Appendix Figure 3: Formal ease of doing business in Zimbabwe as compared to its neighbors in 2010

40 20 0 20 40 60 80 100

Source: data downloaded from http://www.doingbusiness.org/

Difference: neighbors' average minus Zimbabwe rankings

Among the dimensions of formal cost of doing business covered in the World Bank's 2010 survey, the most binding constraint to investment, foreign and domestic alike is taxation regime. The critical weakness of Zimbabwe's formal business climate viewed through the lenses of Zimbabwe's ranking in the context of its neighbors is taxation. Zimbabwe's 130th rank, although better than its overall rank in ease of doing business of 159, is well below ranks of its neighbors of which the worst performer,

Information collected are not based on survey of firms but provided by local professionals dealing with issues related to these areas that are surveyed in respective countries in both large cities and rural areas. They shed light on procedures, laws, and regulations and say nothing about informal constraints such as corruption or predatory administration. For instance, the absence of barriers in starting or closing business may be offset by corruption.

72

104 | P a g e Mozambique, was ranked 97th. As can be seen from Appendix Table 10 summarizing information for Zimbabwe and neighboring countries on the total number of payments per year, the time it takes to prepare, file, and pay (or withhold) the corporate income tax, the value added tax and social security contributions (in hours per year), and the total tax rate, which measures the amount of taxes and mandatory contributions payable by the business in the second year of operation, expressed as a share of commercial profits, Zimbabwe's businesses face the highest burden across the three above dimensions.

Appendix Table 11: Fiscal and administrative burden of paying taxes and contributions in Zimbabwe and neighboring countries in 2010

Botswana Mozambique South Africa Zambia Zimbabwe Zimbabwe in % of average

Total number of payments per year 19 37 9 37 51 200

Time (days) 140 230 200 132 270 154

Total tax rate Memorandum: (% profit) Rank in Paying Taxes 17 18 34 97 30 23 16 36 39 130 161 299

Source: Derived from data in http://www.doingbusiness.org/ExploreTopics/PayingTaxes/Details.aspx?economyid=27

Zimbabwe's mediumsized domestic companies face the highest tax burden and the most demanding administrative requirements to comply with tax regulations against its competitors in neighboring countries, which discourages foreign and domestic investment as well as undercut their ability to compete in neighboring markets. They have to spend 54 percent more time fulfilling tax and mandatory contributions requirements than on average their counterparts in neighboring countries; they have to make twice as many payments per year; and the total tax rate of 39 percent is 61 percent higher than on average in neighboring countries. All other things equal, foreign investor already committed to investing in the region would put Zimbabwe at the bottom of its preferential list. Furthermore, many domestic entrepreneurs will either refrain from establishing business or enter the informal sector. It might be tempting to dismiss Zimbabwe's much higher regulatory hassle cost of doing business than in neighboring countries on the ground that real ease of doing business may be lower. Indeed, formal ease may not be the same as the real ease: for instance, the rankings assign the same weight to each area and corresponding indicators of doing business putting a country with a low number of procedures higher in the ranking despite a very long time needed to complete these procedures. Furthermore, a country may have low tax rates and simple rules of compliance and still informal cost of compliance with a tax code may be huge due to predatory and corrupt administration. On the other hand, the higher tax and compliance burden may be offset by virtual absence or low informal cost of compliance. But this does not seem to be the case as these two stipulations do not fully apply to Zimbabwe: First, Zimbabwe has both complex procedures and long times. Second, considering a very high incidence of corruption, Zimbabwean businesses face both high taxes and significant informal costs, albeit with a caveat. Informal payments may lower taxes paid. This however automatically excludes higher quality foreign investments and encourages the development of informal sector. Hence, there are no reasons to argue that real conditions of doing business are better than formal ones. With taxation regime looming over other dimensions of formal cost of doing business in Zimbabwe as the most binding constraint, tax reform is critical to increase foreign and domestic investments. This, however, will not be enough to trigger FDI inflows, as other constraints are even more binding. The

105 | P a g e Indigenization and Economic Empowerment Act towers them all. The bill, which was signed in 2007 and became law on 17 April 2008, has ever since been poisoning investment climate, in spite of the fact that no regulations on its implementation were announced as of August 2010. The bill requires a 51 percent ownership of all companies operated in Zimbabwe by formerly `disadvantaged' citizens of Zimbabwe. Albeit not yet having been implemented, the ensuing uncertainty over property rights erects a huge barrier to FDI inflows. It also suppresses investments in domestic firms with ownership not meeting the `disadvantaged person' criterion and may even encourage some to transfer profits abroad rather than investing them domestically.

106 | P a g e

Statistical Annex

Annex Tables

Annex Table 1: Values of Transparency International's corruption perception indices for Zimbabwe and its comparators in 1998, 2002, 2007 and 2008 ........................................................................................ 107 Annex Table 2: Revenue from international tourism per arrival in Zimbabwe and Zambia in 19962008 (in US dollars and millions of dollars) ............................................................................................................ 107 . Annex Table 3: Change triggered by change in export prices and exported quantities in 200008 (in percent) ..................................................................................................................................................... 108 Annex Table 4: Top twenty fourdigit SITC exports in "peak years" of 1997 and 2007 and in 2008 (in millions of US dollars and percent) ........................................................................................................... 109 Annex Table 5: Top twenty fourdigit SITC exports in `trough' years of 1994 and 2002 (in millions of US dollars and percent) .................................................................................................................................. 110 Annex Table 6: Exports of iron and steel products in 1994, 1997, 200208 (in millions of current US dollars) ...................................................................................................................................................... 111

107 | P a g e

Annex Table 1: Values of Transparency International's corruption perception indices for Zimbabwe and its comparators in 1998, 2002, 2007 and 2008

Botswana Malawi Mauritius Mozambique (a) Namibia South Africa Tanzania Zambia Zimbabwe Memorandum: Average (b) Zimbabwe in % of average Botswana in % of average Rwanda (c)

1998 6.1 4.1 5.0 3.5 5.3 5.2 1.9 3.5 4.2 4.3 97 141 n/a

2002 6.4 2.9 4.5 2.7 5.7 4.8 2.7 2.6 2.7 4.0 67 159 n/a

2007 5.4 2.7 4.7 2.7 4.5 5.1 3.2 2.6 2.1 3.9 54 140 2.8

2008 5.8 2.8 5.5 2.8 4.5 4.9 3.0 2.8 1.8 4.0 45 145 3.0

Index 2008 2009 1998=100 5.6 95 3.3 68 5.4 110 2.5 80 4.5 85 4.7 94 2.6 158 3.0 80 2.2 43 3.8 93 59 46 149 102 3.3 n/a

Notes: (a) value of CPI in 1999 and 2003, as no data are available for 1998 and 2005; (b) the average includes Botswana and excludes Zimbabwe; (c) classified first in 2005 with a score of 2.6 Source: Extracted from data in the website of Transparency International. Annex Table 2: Revenue from international tourism per arrival in Zimbabwe and Zambia in 19962008 (in US dollars and millions of dollars)

Zimbabwe Receipts (current US$ millions) Receipts (current US$) per arrival Zambia Receipts (current US$ millions) Receipts (current US$) per arrival Receipts Receipts per arrival 1996 232 145 1996 .. ... ... ... 1997 205 153 1997 29 85 707% 180% 395% 68% 1998 158 76 1998 40 110 1999 202 90 1999 53 2000 125 64 2000 133 2001 81 37 2001 80 2002 76 37 2002 64 2003 61 27 2003 88 213 69% 13% 2004 194 105 2004 92 179 211% 59% 101% 43% 2005 99 64 2005 98 146 307% 102% 264% 95% ... ... 2006 338 148 2006 110 145 2007 365 146 2007 138 154 2008 .. ... 2008 146 180

131 291 163 113 Zimbabwe in terms of Zambia 381% 68% 94% 22% 101% 22% 119% 33%

Sources: World Bank WDI database.

108 | P a g e

Annex Table 3: Change triggered by change in export prices and exported quantities in 200008 (in percent) A. Price effect: percent difference between exports in base year expressed in prices in 200001 and value of exports in a base year (in percent) Quantities in P2000 P2001 P2002 P2003 P2004 P2005 P2006 P2007 P2008 in 2000 0 1 30 48 26 40 40 53 77 in 2001 1 0 30 45 24 38 37 47 75 in 2002 22 24 0 12 2 7 5 15 34 in 2003 31 34 12 0 13 4 5 7 22 in 2004 0 2 25 43 0 32 45 55 86 in 2005 26 30 6 9 11 0 0 10 33 in 2006 25 27 3 12 12 0 0 9 31 in 2007 27 30 7 6 17 6 7 0 20 in 2008 36 37 17 6 30 20 22 19 0 B. Quantity effect: percent difference between value of exports of a base volume in successive years over a base year (in percent) Export prices in 2000 in 2001 in 2002 in 2003 in 2004 in 2005 in 2006 in 2007 in 2008 Q2000 Q2001 Q2002 Q2003 Q2004 Q2005 Q2006 Q2007 Q2008 0 11 7 5 9 10 20 27 28 12 0 6 19 22 23 31 37 37 6 6 0 12 15 16 25 32 33 7 19 12 0 1 3 14 22 23 13 25 21 7 0 2 16 25 29 14 27 21 8 3 0 13 23 27 32 44 36 23 21 15 0 13 18 52 64 58 46 41 34 15 0 10 58 75 65 49 53 45 25 8 0

Source: Own calculations based on the data provided by Reserve Bank of Zimbabwe.

109 | P a g e

Annex Table 4: Top twenty fourdigit SITC exports in "peak years" of 1997 and 2007 and in 2008 (in millions of US dollars and percent) Share (%)

25.7 10.4 7.2 6.4 5.6 3.1 2.9 2.8 2.7 1.8 1.7 1.5 1.3 1.2 1.1 1.1 0.9 0.8 0.8 0.8 79.9 43.4 70.4

SITC

1212 6715 2631 1211 6831 0449 2784 2927 9710 0711 0611 0111 8973 1213 0545 3250 8414 7924 6522 0571

Products

Tobacco stripped/stemmed Other ferro alloys Raw cotton, excl linters Tobacco, not stripped Nickel/alloys unwrought Maize ex sweet corn Asbestos Cut flowers/foliage Gold non monetary ex ore Coffee, not roasted Raw sugars Beef, fresh/chilled Precious metal jewellery Tobacco refuse Vegetables nes,frsh/chld Coke/semi coke/retort c Men/boy trouser/etc woven Aircrft nes over 15000kg Woven unb cotton fab nes Citrus fruit fresh/dried

1997

534 217 150 133 117 65 60 58 56 37 36 31 26 24 23 23 19 17 17 16 1,660 901 1,463

SITC

6831 1212 2841 2842 6715 2631 6821 2731 6612 0611 2927 0571 1223 2831 2784 6672 0251 0612 2482 1211

Product Rev. 3

Nickel/alloys unwrought Tobacco stripped/stemmed Nickel ores/concentrates Nickel mattes/sinters/++ Other ferro alloys Raw cotton, excl linters Copper refined/unrefined Quarried stone slabs Portland etc cements Raw sugars Cut flowers/foliage Citrus fruit fresh/dried Manufactured tobacco nes Copper ores/concentrates Asbestos Diamonds unset Eggs in shell Cane/beet sugar nes Softwood,simply worked Tobacco,not stripped

2007

537 266 224 213 207 115 51 38 38 34 34 31 28 27 24 23 21 20 18 14 1,962 1,027 1,757

Share (%) SITC

22.6 11.2 9.4 8.9 8.7 4.8 2.2 1.6 1.6 1.4 1.4 1.3 1.2 1.1 1.0 1.0 0.9 0.8 0.7 0.6 82.4 43.2 73.8 6715 1212 2842 2841 6831 2631 0611 6612 2731 2927 2831 6821 1223 0571 2892 2119 6672 2482 0612 3250

Product Rev. 3

Other ferro alloys Tobacco stripped/stemmed Nickel mattes/sinters/++ Nickel ores/concentrates Nickel/alloys unwrought Raw cotton,excl linters Raw sugars Portland etc cements Quarried stone slabs Cut flowers/foliage Copper ores/concentrates Copper refined/unrefined Manufactured tobacco nes Citrus fruit fresh/dried Prec.metal waste/scrap Hide/skin nes/waste Diamonds unset Softwood,simply worked Cane/beet sugar nes Coke/semi coke/retort c

2008

301 273 245 234 219 117 58 34 33 30 23 21 21 19 17 16 16 15 15 15 1,721 818 1,566

Share (%)

14.2 12.9 11.5 11.0 10.3 5.5 2.7 1.6 1.6 1.4 1.1 1.0 1.0 0.9 0.8 0.7 0.7 0.7 0.7 0.7 81.0 38.5 73.8

Total above Top three Top ten

Source: Own calculations based on partners' data reported to the UN COMTRADE database.

110 | P a g e

Annex Table 5: Top twenty fourdigit SITC exports in `trough' years of 1994 and 2002 (in millions of US dollars and percent) SITC Product Rev. 3 1994 337 105 105 92 71 68 55 52 37 35 27 22 22 20 18 16 14 14 13 12 1,135 546

957

Share (%) 21.3 6.6 6.6 5.8 4.5 4.3 3.5 3.3 2.3 2.2 1.7 1.4 1.4 1.2 1.2 1.0 0.9 0.9 0.8 0.8 71.6 34.5

60.4

SITC

Product Rev. 3

2002 Share (%) 550 108 93 79 67 54 53 42 33 27 25 24 19 18 18 15 15 14 13 13 1,283 751

1,107

1212 Tobacco stripped/stemmed 6831 Nickel/alloys unwrought 6715 Other ferro alloys 0611 Raw sugars 1211 Tobacco, not stripped 2631 Raw cotton ,excl linters 0449 Maize ex sweet corn nes 2784 Asbestos 2927 Cut flowers/foliage 0111 Beef, fresh/chilled 6522 Woven unb cotton fab nes 8414 Men/boy trouser/etc wovn 6513 Cotton yarn nes 1213 Tobacco refuse 0112 Beef, frozen 6114 Bovine/equine leathr nes 8973 Precious metal jewellery 0545 Vegetables nes,frsh/chld 2482 Softwood,simply worked 6821 Copper refined/unrefined TOTAL ABOVE Top three

Top ten

1212 Tobacco stripped/stemmed 6715 Other ferro alloys 2631 Raw cotton,excl linters 6831 Nickel/alloys unwrought 2927 Cut flowers/foliage 2841 Nickel ores/concentrates 1211 Tobacco,not stripped 0611 Raw sugars 2784 Asbestos 8973 Precious metal jewellery 0571 Citrus fruit fresh/dried 1213 Tobacco refuse 0545 Vegetables nes,frsh/chld 8414 Men/boy trouser/etc woven 2882 Nonfer metal waste nes 2731 Quarried stone slabs 0741 Tea 6513 Cotton yarn nes 0751 Peppers dried/crush/grnd 6612 Portland etc cements TOTAL ABOVE Top three

Top ten

32.1 6.3 5.4 4.6 3.9 3.2 3.1 2.4 1.9 1.6 1.4 1.4 1.1 1.1 1.0 0.9 0.9 0.8 0.8 0.8 74.9 43.9

64.7

Source: Own calculations based on partners' data reported to the UN COMTRADE database.

111 | P a g e

Annex Table 6: Exports of iron and steel products in 1994, 1997, 200208 (in millions of current US dollars)

SITC 6715 6724 6726 6727 6728 6731 6732 6733 6734 6735 6741 6742 6743 6744 6745 6751 6752 6753 6754 6755 6757 6761 6762 6763 6764 6768 6770 6781 6782 6791 6793 6794 6795 Other ferro alloys Ingot/primary iron/steel Semifin iron/steel<.25%c Semifinished iron/steel. >.25%c Semifinish alloy steel Flat rolled steel1 Flat rolled steel2 Cold rolled steel3 Colled rolled steel4 Flat rolled steel nes Zinc coated/plated steel Tin plated/coated steel Steel paint/plastic cover Coated steel plate w>600 Coated/platd steel w<600 Flat siliconelect steel Flat highspeed steel Hot rolled stnless steel Hotrolled alloy steel Coldrol stainless steel Flat roll alloy stl nes Hotr coil bar/rod iron/steel Hotform steel bar/rod nes Coldform air/steel bar nes Iron/steel bars nes Iron/steel angle/shape/sect Iron/steel railway materials Iron/nonalloy steel wire Wire, stainless/alloy steel Iron/steel tube seamless Seamed pipe/tube d>406.4 Seamed tubes/pipes nes Iron/steel pipe fittings 1994 105 0.10 4.35 0.01 0.00 0.98 1.69 0.44 2.24 0.13 1.00 0.47 0.07 0.02 0.65 0.00 0.00 0.00 0.00 0.00 0.00 1.77 1.73 0.15 0.87 1.50 0.09 0.65 0.00 0.35 0.13 1.27 0.30 1997 217 0.01 2.45 0.00 0.00 0.00 0.00 0.00 0.00 0.08 0.04 0.03 0.00 0.03 0.01 0.00 0.00 0.04 0.00 0.00 0.09 0.21 0.07 0.10 0.16 0.75 0.03 1.37 0.28 0.06 0.02 1.09 0.05 2002 108 6.99 0.15 3.76 0.00 0.00 0.00 0.00 0.00 0.07 0.31 0.05 0.02 0.25 0.00 0.03 0.05 0.00 0.00 0.00 0.05 0.34 0.01 0.02 0.05 0.98 0.00 4.22 0.05 0.12 0.00 1.89 0.19 2003 118 0.74 0.16 8.45 0.00 0.00 0.00 0.00 0.00 0.01 0.28 0.16 0.00 0.04 0.00 0.00 0.00 0.02 0.00 0.00 0.05 1.31 0.09 0.02 0.12 1.82 0.04 8.07 0.09 0.08 0.00 4.92 0.34 2004 173 3.71 0.57 15.30 0.00 0.00 0.00 0.00 0.00 0.05 0.27 0.02 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.00 0.02 1.21 0.01 0.26 0.18 6.14 0.02 7.13 0.05 0.01 0.00 3.28 0.24 2005 202 2.39 2.12 1.57 0.00 0.00 0.00 0.00 0.00 0.16 0.07 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.98 0.02 0.03 0.03 4.63 0.02 3.07 0.04 0.15 0.03 1.91 0.28 2006 191 0.24 0.49 0.42 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.00 0.00 0.13 0.00 0.00 0.00 0.04 0.00 0.00 0.04 1.09 0.20 0.00 0.05 6.45 0.10 0.91 0.00 20.54 0.00 0.62 0.08 2007 207 0.02 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.04 0.06 0.04 0.00 0.01 0.00 0.00 0.00 0.04 0.01 0.00 0.12 0.90 0.30 0.25 0.07 3.68 0.17 0.73 0.01 0.01 0.02 0.32 0.04 2008 301 0.05 0.00 0.00 0.04 0.00 0.22 0.00 0.00 0.01 0.07 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.35 0.34 0.04 0.97 3.35 0.00 0.40 0.00 0.00 0.00 0.09 0.17

TOTAL ABOVE Memorandum:

Share of ferro alloys (SITC. 6715) in percent

125 224 127 145 211 219 223 214 307 86 34 4 97 28 2 98 37 2

83 97 85 82 82 92 RCA of ferro alloys (SITC. 6715) 29 46 38 34 32 35 Number of exports above US1 million exports 8 4 5 6 7 7 Source: Own calculations based on partners' data reported to the UN COMTRADE database.

Information

Microsoft Word - BK_ZWE 02 2011 final.docx

112 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

363075


You might also be interested in

BETA
untitled
The Mineral Industry of Zimbabwe in 2000
Microsoft Word - BK_ZWE 02 2011 final.docx
The Mineral Industry of Zimbabwe in 2002
Policy Series cover 17.qxd