Read 01_Trends and Characteristics of the Middle-Income Class in the Philippines_Is it Expanding or Shrinking.doc text version

10th National Convention on Statistics (NCS) EDSA Shangri-La Hotel October 1-2, 2007

Trends and Characteristics of the Middle -Income Class in the Philippines: Is it Expanding or Shrinking? by Romulo A. Virola, Mildred B. Addawe and Ma. Ivy T. Querubin

For additional information, please contact: Author's name Designation Affiliation Address Tel. no. E-mail Co-author's name Designation Affiliation Address Tel. no. E-mail : : : : : : : : : : : : Romulo A. Virola Secretary General National Statistical Coordination Board 403 Sen. Gil Puyat Avenue, Makati City (0632) 897-2171 [email protected] Mildred B. Addawe/Ma. Ivy T. Querubin Statistical Coordination Officer III/Statistical Coordination Officer II Nation al Statistical Coordination Board 403 Sen. Gil Puyat Avenue, Makati City (0632) 897-2170 / (0632) 896-5390 [email protected]/ [email protected]

Trends and Characteristics of the Middle-Income Class in the Philippines: Is it Expanding or Shrinking? by Romulo A. Virola, Mildred B. Addawe and Ma. Ivy T. Querubin1

ABSTRACT It is often said that while the Philippines has experienced some economic growth in the past, this has not been felt by the marginalized sectors of society; that no trickle-down effect has happened. Thus, income distribution continues to be inequitable and poverty persists. Official poverty statistics and the poverty literature in the Philippines generally focus on indicators concerning the low-income classes and their characteristics. However, not much research has been done about the middleincome class, who can actually provide the necessary resource inputs to enhance productivity and stimulate economic growth. Questions have been raised too on whether the pervasive poverty in the Philippines, as compared to our neighbors, has been the result of a shrinking Filipino middle -income class and whether this, in turn, has been caused by the migration of Filipinos to other countries. Also, the definition of the middle-income class vis -à-vis other income classes has yet to be established in the Philippines. This paper explores two alternative definitions of the middle-income class, one based on income alone and the other based on income and other socioeconomic characteristics, to come up with a proposed operational definition of the group. It then presents the growth trends and some characteristics of the Filipino middle-income class over the past decade (1997 to 2003) based on the proposed operational definition.

I.

Introduction In the Philippines, much work has been done to study the low-income classes and

their characteristics as poverty persists in the country. However, not much research has been done about the middle-income class, who can actually provide the necessary resource inputs to enhance productivity and stimulate economic growth. Easterly [3] states that countries with a big middle class have a higher level of income and growth. Societies which are relatively homogenously middle-class have more income and growth because they have more capital and infrastructure accumulation, they have better national economic policies, more democracy, less political instability, more modern sectoral structure, and more urbanization. According to Aristotle, Decornez [2], the best political community is formed by citizens of the middle class, and those states are likely

1

Secretary General, Statistical Coordination Officer III and Statistical Coordination II, respectively, of the National Statistical Coordination Board. The views expressed in this paper are those of the authors and do not necessarily reflect the views of the NSCB. The authors acknowledge the assistance of Redencion M. Ignacio, Noel S. Nepomuceno and Andrea C. Baylon in the preparation of this paper.

to be well administered. Where the middle class is large, there are least likely to be factions and dissensions. Researches have been conducted by market research agencies and other researchers in the Philippines on the socioeconomic classification (SEC) of households. In the 1990s, the Market and Opinion Research Society of the Philippines (MORES) [5], consisting of TNS Trends, AC Nielsen, Synovate, etc., conducted a study to determine which variables are most discriminating in identifying SEC, to develop guidelines for social class inclusions, and to determine the distribution of the population by SEC. In addition, Oblea of the National Statistics Office developed a socio -economic index in classifying households in NCR using the results of the 2004 Annual Poverty Indicators Survey [7]. Some of the SEC indicators currently used by market research agencies in the Philippines include the following: neighborhood, home durability, outdoor quality, indoor quality, occupation of household head, educational attainment of the household head, facilities in the home, and household income [8]. However, as noted in the presentation by MORES in the 28 June 2006 Conference of the Philippine Statistical Association [5]. the SEC discriminating variables have yet to be "officially standardized". There were suggestions to work towards the simplest SEC categories (upper, middle and lower class) and the simplest set of indicators as in the United States and Europe, which use income and occupation, respectively, as determinants of SEC. In the United States, various ideological and economic theories define the middle class as consisting of all those who are neither "poor" nor "rich", or as being a relative elite of professionals and managers, defined by lifestyle and influence [10]. Identifying those attributes that are indicative of the middle class is admittedly a difficult endeavor, as one can define the middle class by means of lifestyle or by using the "statistical middle". The statistical middle class is a term encompassing all those individuals who might at one point or another be identified as middle class, as they are in neither extreme of the socioeconomic strata. The term can also be used to describe those at the actual center of the income strata, who may also be referred to as the middle-middle class. Characterizing the middle class is of course, of interest not only to market research agencies but also to development planners. As has been pointed out, a big middle class in a society brings many advantages and opportunities for growth and development.

An AC Nielsen study estimated the Philippine middle class (Class C) to be a fifth of the country's household population between 1993 and 1998 while the market research agency Sofres estimated the middle class to be no more than 9% between 1993 and 1998 [8]. And while the private sector has been releasing and using statistics on the middle class, to date, there are no official statistics generated in the Philippine Statistical System that iden tify and define who, where and how many are the low, middle, and high-income classes. It is therefore highly desirable to come up with a clearer definition of the middle class and to study whether the middle class is expanding or shrinking in the Philippines, from the point of view of official statistics. This paper intends to define and present the trends and characteristics of the middle -income class from 1997 to 2003 using the results of the Family Income and Expenditures Survey (FIES) of the NSO. The next section discusses the data limitations of the study arising from the use of the triennial FIES and the quarterly Labor Force Survey (LFS). The third section presents the methodology and the results are shown in the fourth section together with the proposed definition of the middle class. The last section presents some concluding remarks and recommendations. II. Limitations of the Study The study made use of the data coming from the 1997, 2000, and 2003 FIES and the January 2001 and 2004 LFS of the NS O. Hence, the variables considered are limited to those included in these surveys. While this study had initially wanted to assess whether the Filipino middle-income class might be shrinking because of the migration of Filipinos to other countries, available information did not allow this. Instead, what was done was to examine the movement of families with OFW members in the socio -economic classification. For this purpose, the paper used the January 2001 and 2004 LFS results, which were merged with the 2 000 and 2003 FIES, respectively. Thus, the movement of families with OFWs was studied only for the 2000-2003 period.. It was also observed that categories for some variables, e.g., occupation of household head and total household members by age group, were not comparable for the three FIES years. The occupation codes for the 1997 FIES followed the 1977 Philippine Standard Occupation Code (PSOC) while the 2000 and 2003 FIES adopted the 1992

PSOC. Also, it should be noted that there is no available longitudinal data from FIES and LFS for the years specified. III. Methodology A. Identifying Middle-Income Class Based on Income Cluster Analysis [4] is performed on the 1997 FIES to determine an a priori grouping of families based on income, which could be used to classify families into the low, middle or high income clusters. Based on the resulting clusters, the lower and upper income limits of the middle class are determined. The minimum and maximum income of the proposed middle class using 1997 data, are then extrapolated to 2000 and 2003 using two series of the CPI, one with base year 1994 and base year 2000 but the results were almost the same Using the income limits for the three years, the middle income class is determined. In order to identify the significant determinants of the total income of the middleincome class, multiple regression analysis was carried out on the families that have been identified as middle class. The dependent variable was the logarithm of the total household income with the following list of independent variables: Household/Household Head Characteristics · · · · · · · · · · · educational attainment of household head occupation of household head marital status of household head age of the household head sex of the household head class of worker of household head type of household total employed household members total household members by age group total non-relative household members square of family size (to capture non-linear relationship between family size and welfare)

·

tenure status of the household

Housing Characteristics · · · · · · · roof materials of the housing unit wall materials of the housing unit type of building/house type of toilet facilities source of water supply facilities in the house location of house (urban or rural)

The independent variables identified were sourced from the possible correlates of poverty in Albert and Collado's Study "Profile and Determinants of Poverty in the Philippines" [1] and the list of significant predictors of income resulting from the NSCB Project on the Estimation of Local Poverty in the Philippines [6]. Furthermore, the indicators from the 5-point system of MORES for identifying SEC of households (Table 1) where households are classified into the following 5 groups were also considered: AB ­ 30-35 points C1 or Upper C ­ 25-29 points C2 or Broad C ­ 20-24 points D ­ 15-19 points E ­ 7-14 points

o o o o o

Table 1. Indicators for the MORES socio -economic classification of households

MORES SEC Indicators Neighborhood Located in generally slum district Mixed neighborhood with predominantly small houses Mixed neighborhood of large and small houses Mixed neighborhood with predominantly large houses Exclusive subdivision, town houses and condominiums Home Durability Temporary structure/ "barong barong" Made of light and cheap material, poorly constructed Made of light and heave materials Points 1 2 3 4 5 1 2 3

MORES SEC Indicators Made of good quality materials Made of high quality materials Outdoor Quality Unpainted and dilapidated Generally unpainted and in need of major repairs Painted but may need some repairs Well-painted but needs some minor repairs Well painted and not in need of repairs Indoor Quality Unpainted and dilapidated Generally unpainted and in need of major repairs Painted but may need some repairs Well-painted but needs some minor repairs Well painted and not in need of repairs Occupation of Household Head Unskilled Blue Collar White Collar Supervisory, Small Business Professional, Managerial, Medium-sized business Educational attainment of household head Some elementary school Some high school Some college Graduate of non -exclusive/state colleges Graduate of exclusive Facilities in the home 0 ­ 1 facility 2 ­ 4 facilities 5 ­ 7 facilities, with or without car 8 ­ 10 facilities, with car > 10 facilities, with new car (5 years or less) Household Income P8,000 and below P8,001 ­ P 15,000 P15,001 ­ P 30,000 P30,001 ­ P 50,000 P50,001 and higher

Points 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5 1 2 3 4 5

This study likewise considered the following characteristics of class C households as reported by Oblea [7]: female-headed household, not living in apartment, with membership in PhilHealth, water supply comes from safe source , use own sanitary toilet facility, floor

area of the house is not within the range of 10 to 29 sq. meters, household head is not in elementary level, household head is not working in government office nor working as laborer, and the household owned the following amenities: car, components, refrigerator, and radio. A list of significant variables of income among the middle class was determined using regression. Regression was applied on the three consecutive FIES years to check the stability over time of the middle-income pre dictors. In addition, the number of middle-income families with OFW members was estimated for the three reference years. B. Identifying Middle-Income Class Based on Socio-Economic Characteristics A second approach for identifying the middle-income class is through socioeconomic characteristics. The significant variables that are found through multiple regression analysis to be predictors of income for families in the middle class are used to screen this income class. Several combinations of variables were tried; only the final list of non-income predictors of this class is shown here. IV. Results The cluster analysis on the 1997 data produced five groups of families from which to identify the low, middle, and high-income classes. Table 2. Annual Family Income of the Low, Middle, and High-Income Classes: 1997

Cluster 1 2 3 4 5 Family Income Mean 64,649 230,190 614,741 1,788,460 5,758,452 Median 57,158 212,338 558,000 1,560,000 5,477,600 Minimum 3,064 148,307 Maximum 148,294 420,542 Percent Families 76.7 Income Class low

20.1 middle 2.9 middle 0.3 0.1 high high

421,000 1,207,122 1,215,509 3,472,515 4,004,148 8,315,380

As can be gleaned from Table 2, cluster 2 alone or clusters 2 and 3 together appear to be reasonable candidates to form the middle -income class. The authors believe that families with income falling under cluster 3 are still middle class and decided to define the middle class based on income to be families belonging to clusters 2 and 3. These two

groups have mean family income of P230,190 and P614,741, respectively. About 23% were estimated to comprise this class in 1997. Table 3 shows the comparative income profile of the middle-income class in 1997, 2000 and 2003 using the two CPI series ( 1994 and 2000 Base Year). It is noted that the proportion of families belonging to the middle class is about the same for the two CPI series. What is noteworthy is the fact that while the middle class shrank only a little bit between 1997 and 2000, there was at least a 2 percentage point decrease in the population share of the middle class between 2000 and 2003. Table 3. Annual Family Income of the Middle-Income Class: 1997, 2000, and 2003

Year Family Income (1994 based CPI) Mean Median Minimum Percent Maximum Families Family Income (2000 based CPI) Mean Median 226,398 268,700 298,380 Minimum 148,307 178,468 203,109 Percent Maximum Families 1,207,122 1,449,295 1,651,632 23.0 22.7 19.9

1997 278,305 226,398 148,307 1,207,122 23.0 278,305 2000 339,902 276,507 181,132 1,474,296 22.2 331,646 2003 382,529 311,184 203,848 1,659,188 19.7 366,329 Notes: CPI (1994=100): 1997 ­ 124.7; 2000 ­ 152.3; 2003 ­ 171. CPI (2000=100): 1997 ­ 83.1; 2000 ­ 100; 2003 ­ 113.8

Table 4 shows that the number of middle income families increased from 1997 to 2000 but decreased from 2000 to 2003. It also shows that the percentage share of both the middle and high income classes shrank between 1997 and 2000 as well as between 2000 and 2003, resulting in an expanding low income class in Philippine society. As of 2003, less than 1 in 100 families belongs to the high income class; about 20 are middle income and 80 are low income. And in a span of 6 years from 1997 to 2003, close to 4 families for every 100 middle income families have been lost to the low income category. Table 4. Distribution o f Families By Income Classes: 1997, 2000, and 2003

Year 1997 2000 2003 Low Level 10,881,914 11,598,258 13,172,033 Percent 76.6 77.0 79.9 Income Class Middle Level Percent 3,260,474 23.0 3,422,524 22.7 3,282,511 19.9 High Percent 0.4 0.3 0.2

Level 50,074 51,160 25,849

Table 5 presents the parameter estimates of the regression model for the determinants of middle-income in the Philippines for the years 1997, 2000 and 2003. Note that the model for 2003 has the highest adjusted r-square, accounting for 53% of the variability of the model. The 1997 and 2000 model, on the other hand accounts for 38% and

44% of the variability, respectively. Also, because the dependent variable is in natural log form, the estimated regression coefficients measure approximately the percentage change in total income of the middle-income class from a unit change in the independent variable. Only four of the independent variables in Table 5 were significant predictors of income for all three years, 1997, 2000 and 2003: ownership of stereo, presence of air conditioning unit, squared value of family size and number of nonrelative member of the household. Seven other variables had negative effect on income in some but not all years: if the household head worked in the trade and related industry, number of household members aged less than 1, number of household members aged less than 7, number of household members aged 1 to 6 years old, number of household members aged 7 to 14 years old, number of household members aged 15 to 24 years old and if the household head is a male. Ownership of air conditioning unit meant an increase in income by 26.9%, 15.9% and 22.0% in 1997, 2000 and 2003, respectively. Ownership of stereo meant respective increases of 7.7%, 4.1% and 26.4% during the three years. A unit increase in the squared family size meant an increase in income of 10.7%, 15.4% and 12.0% while an additional nonrelative member meant an increase of 5.4%, 7.6% and 11.4% . The percentage reduction in income coming from the independent variables which showed negative impact in not very big. The largest decrease comes from the sex of the household head ­ male heads of households meant less income by about 5%. The coefficients for the highest educational attainment of household head are positive and significant in 1997 and 2000. If the household head has a postgraduate degree, the annual family income of the family is expected to increase by a minimum of 25.5% in 1997 and 17.3% in 2000. Household head attainment of a college degree, on the other hand has a smaller effect of 10.8% and 10.5% increase in the 1997 and 2000 total family income. This highlights the importance of (higher) education in the socio-economic status of an individual. In terms of occupation, if the household head works as an administrative, executive or managerial worker, sales worker, official of government, manager or supervisor or a professional, it also increased the annual family income for some years. For example, in 2003, if the household with head worked as an official of government, a manager or a supervisor or a professional, the annual income would increase by 31.6%. However, the

household head working in trade industry had a negative effect of 3.6% on the total family income in 2000. Moreover, all household facilities enumerated in Table 5 are found to give a positive effect on the total family income respectively. As may have been expected, families living in urban areas had income higher by about 6% compared to families in rural areas in 1997 and 2000. Finally, ownership of a house with roof made of strong materials meant an increase in income of 7.3% and 3.8% in 1997 and 2000. at some point in time. Ownership of car, though not statistically significant in 2003, increased income by 20% and 17% in 1997 and 2000,

Table 5. Significant Predictors of the Annual Family Income of the Middle -Income Class 1997 (Adj. R 2000 (Adj. R 2003 (Adj. Rsquare= 0.38) square= 0.44) square= 0.53) Variable Variable Label Coef.

hea_post a dummy variable for whether the head has postgraduate degree a dummy variable for whether the head has college degree a dummy variable for whether the head works as administrative, executive and managerial workers a dummy variable for whether the head works as sales workers a dummy variable for whether the head works as official of government, corporate executive, managers and supervisors a dummy variable for whether the head works as professionals 0.255

P>t

0.0

Coef.

0.173

P>t

0.0

Coef.

P>t

hea_coll

0.108

0.0

0.105

0.0

hea_admi

0.107

0.0

hea_sale

0.046

0.0

hea_gov

0.067

0.0

0.316

0.0

hea_prof

0.058

0.0

Variable

hea_trad

Variable Label

a dummy variable for whether the head works as trades and related workers a dummy variable for whether the household has radio a dummy variable for whether the household has vcr/vtr,vcd/dvd a dummy variable for whether the household has telephone/cellphone a dummy variable for whether the household has computer a dummy variable for whether the household has microwave oven a dummy variable for whether the household has stereo a dummy variable for whether the household has refrigerator a dummy variable for whether the household has freezer a dummy variable for whether the household has air conditioner a dummy variable for whether the household has car a dummy variable for whether the roof is made of strong materials a dummy variable for whether the house and lot is own squared value of family size variable

1997 (Adj. R square= 0.38) Coef. P>t

2000 (Adj. R square= 0.44) Coef.

-0.036

2003 (Adj. Rsquare= 0.53) Coef. P>t

P>t

0.0

fac_rad

0.042

0.0

fac_vcr

0.092

0.0

0.057

0.0

fac_pho

0.143

0.0

fac_mic

0.148

0.0

fac_ove

0.117

0.0

0.300

0.0

fac_ste

0.077

0.0

0.041

0.0

0.264

0.0

fac_ref

0.090

0.0

0.053

0.0

fac_fre

0.146

0.0

fac_air

0.269

0.0

0.159

0.0

0.220

0.0

fac_car

0.204

0.0

0.168

0.0

hou_rost

0.073

0.0

0.038

0.0

hou_own

0.035

0.0

Fsizesq

0.107

0.0

0.154

0.0

0.120

0.0

Variable

less1

Variable Label

total number of household member with age less than 1 year old total number of household member with age less than 7 years old total number of household member with age 1 to 6 years old total number of household member with age 7-14 years old total number of household member with age 15-24 years old a dummy variable for whether the sex of the head is male total number of nonrelative member total number of employed household member a dummy variable for whether the household lives in urban area

1997 (Adj. R square= 0.38) Coef.

-0.029

2000 (Adj. R square= 0.44) Coef. P>t

2003 (Adj. Rsquare= 0.53) Coef. P>t

P>t

0.0

less7

-0.017

0.0

s1121age

-0.033

0.0

s1131age

-0.026

0.0

s1141age

-0.010

0.0

hea_male

-0.055

0.0

-0.044

0.0

non_rel emp_mem

0.054 0.048

0.0 0.0

0.076 0.042

0.0 0.0

0.114

0.0

Urb

0.061

0.0

0.057

0.0

_cons

11.680

0.0

11.806

0.0

11.401

0.0

Looking at Table 6, the mean income of the families with OFW member is much higher by 75% in 2000 and 93% in 2003 than the mean income of all Filipino families. The mean income of middle-income families with OFW member rose by P33,986 or 9.5% from 2000 to 2003. Furthermore, of the total families with OFW members in 2000 and 2003, 56.5% and 57.0%, respectively, are classified as middle - income families (Table 7). It is also worth noting that the middle-income families with OFWs account for 15.8% and 15.7% of the middle -income families in 2000 and 2003, respectively. Also, there is a decrease of 4.58% in the number of middle -income families with OFW member from 2000 to 2003.

Table 6. Annual Income of Families with OFWs: 2000 and 2003

Increase/ Decrease in Mean Income 2000 All Families MiddleIncome Families Families with OFWs MiddleIncome Families with OFWs Mean 145,121 331,646 Median 89,810 268,700 Min 4,273 178,468 Max 8,441,242 1,449,295 Mean 147,888 366,329 Median 95,291 298,380 2003 Min 3,086 203,109 Max 32,256,048 1,651,632 Level Percent 2,767 1.9 34,683 10.5

254,493

199,947

12,223

2,014,196

282,846

227,521

12,595

2,317,360

28,353

11.1

357,782

295,883

178,476

1,392,000

391,768

324,338

203,248

1,651,632

33,986

9.5

Table 7. Middle -Income Families with OFWs: 2000 and 2003

Percentage Increase/ Decrease Percentage Percentage Percentage Percentage Percentage Percentage of MiddleShare to Share to Share to Share to Share to Share to Income Levels Middle Levels MiddleAll Families All Families Families Income Income Families with OFWs Families with OFWs with OFW Families Families members 15,071,941 100 16,480,393 100 2000 2003 3,422,524 22.7 100.0 3,282,511 19.9 100.0

All Families MiddleIncome Families Families with OFWs MiddleIncome Families with OFWs

954,667

6.3

100

903,831

5.5

100

539,799

3.6

15.8

56.5

515,056

3.1

15.7

57.0

(4.6)

B. IDENTIFYING THE MIDDLE CLASS USING NON-INCOME INDICATORS Figure 1 shows the scheme for identifying middle -income families through nonincome characteristics. Descriptive analysis was performed on the significant predictors of middle -income and the following four variables are observed to be the major determinants of this class for the three reference years: ownership of house and lot; housing unit made of

strong roof mate rials; ownership of radio and refrigerator. These are the four variables which showed as the most common characteristics of families classified as middle income in 1997, 2000 and 2003. Note that this set is different from the set of four variables identified in the first part of this section that significantly increased family income in all three years

Figure 1. Identification of Middle-Income Class Based on Socio-economic Characteristics

OTHER INCOME CLASS

NO

NO

NO

NO

Own house and lot?

House made of strong roof?

Own refriger ator?

Own radio?

YES

YES

YES

YES

MIDDLE-INCOME CLASS

Using socio-economic characteristics as basis for identifying middle -income class, a pattern close to that in Table 4 was observed as shown in Table 8. The number of middleincome families increased by 653,921 or by 26% in 2000 but decreased by 263,150 or by 8.3% in 2003. However, under this classification scheme, the proportion of families belonging to the middle income class increased between 1997 and 2000 but decreased between 2000 and 2003. Table 8. Annual Income of Middle-Income Families Based on Other Socio -economic Charac teristics

Year 1997 2000 2003 Families Level 2,526,258 3,180,180 2,917,030 Percent 17.8 21.1 17.7 Mean Family Income 271,454 289,174 289,669

This paper illustrates that whether using income or non-income socio-economic characteristics to define the middle class, the Filipino middle class has shrunk from 2000 to 2003. Using the income definition, the paper also shows a decreasing share of the high income bracket. This indicates an increasing vulnerability of the Filipino families to poverty.

Proposed Definitions of the Middle-Income Class Using the results of this study, the following are the proposed definitions of middleincome class, the first one is based on income while the second is based on other characteristics: (1) The middle-income class could be defined as those families with annual income ranging from P148,307 to P1,207,122 in 1997; P178,468 to P1,449,295 in 2000; and P203,109 to P1,651,632 in 2003. For 2007, the ra nge should be from P251,283 to P2,045,280. (2) A family may be classified as middle-income class if it shows all of the following characteristics: (a) ownership of house and lot; (b) housing unit is made of strong roof materials; (c) ownership of refrigerator; and (d) ownership of radio. V. Concluding Remarks and Recommendations The results of this paper seem reasonable in identifying the Filipino middle class. Nonetheless, further improvements can be made on the proposed definition. For one, the updating of the income limits can be refined. Inclusion of variables which may have been missed by the modeling exercises but which are possible determinants of the middle income class and which have available data, can also be considered. Also, while 74% and 79% of the identified middle class owned a house and lot in 1997 and 2000, respectively, there may be families who rent their house who can be considered as belonging to middle-income class. Therefore, this may be considered for exclusion from the definition. This paper focused mainly on the changes in income and socioeconomic characteristics of the middle -income class. It may be good to study also in the future the

characteristics of the low and high -income class vis-à-vis middle-income class using the FIES and the LFS results. This can refine the classification rules proposed in the paper. Other statistical analysis tools like factor/principal components analysis and discriminant analysis can also be tried. In the meantime, the definition proposed by the paper can be used. Because

collecting income data may involve more measurement problems than the non-income characteristics, and because the non-income variables can be collected more frequently and more cheaply than the income data, the use of the latter definition is preferred. The data can be collected via a rider to the existing LFS of the NSO. Results shown in the paper seem to also indicate an increasing vulnerability of the Filipino middle class which could translate into increased poverty incidence in the future. This surely calls for more aggressive and timely interventions by both the government and the private sector. Higher investments in education that are accompanied by reforms in the education sector can lessen the threat of a shrinking Filipino middle class, as suggested by the results of this paper. Finally, the paper exhibited the power of statistics to inform decisions. These statistics are generated not without cost. It is therefore imperative that both government and the private sector learn to appreciate and to invest in statistics.

References [1] Albert, J. R. G. and Collado, P.M. (2004) Profile and Determinants of Poverty in the Philippines. Proc. of the 9 th National Convention on Statistics Decornez, Shubhasree Seshanna. An empirical analysis of the American middle class (19681992), Ph. D Dissertation, Vanderbilt University. August 1998. Easterly, William. (2001) The Middle Class Consensus and Economic Development. World Bank Hair, Anderson, Tatham, and Black (1998) Multivariate Data Analysis. Prentice-Hall, Inc. U.S.A. Mercado, Judy (2006) Marketing and Opinion Research Society of the Philippines. Where do you belong? Standardizing SEC classification. PSA Conference NSCB (2005) Estimation of Local Poverty in the Philippines Oblea A. F. (2006) Development of Socio-Economic Index in Classifying Households in NCR. Proc. of the 2006 Philippine Statistical Association Conference Roberto, Eduardo (2000). Making Sense of the Major Systems for the Socio-Economic Classification of Philippine Households. The PBS Web Page. 25 June 2004. <http://www.pbs.org/now/politics/middleclassoverview.html > Wikipedia, The Free Encyclopedia Web Page. August 2007. <http://en.wikipedia.org/wiki/American_middle_class>

[2]

[3] [4] [5]

[6] [7]

[8]

[9] [10]

Information

01_Trends and Characteristics of the Middle-Income Class in the Philippines_Is it Expanding or Shrinking.doc

18 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

522560


You might also be interested in

BETA
01_Trends and Characteristics of the Middle-Income Class in the Philippines_Is it Expanding or Shrinking.doc