Trends in inequality in Pakistan between 1998-99 and 2001-02.

Anwar, Talat

There have been various attempts to estimate absolute poverty during the 1990s but there is hardly any attempt to examine the issue of income inequality in Pakistan. The paper examines the trends in inequality from the most recent micro datasets of Pakistan Integrated Household Surveys (P1HS), 1998-99 and 2001-02. The results suggest that inequality has worsened in Pakistan between 1998-99 and 2001-02. While inequality declined in urban areas, inequality increased in rural areas over the period. Although urban inequality declined, the poorest 40 percent income groups in urban areas lost their income share, implying that decline was at the expense of the poor. While inequality declined in Punjab, NWFP, and Balochistan across urban areas, inequality increased significantly in urban Sindh. On the other hand, inequality declined in rural areas in Sindh, NWFP, and Balochistan but increased in rural Punjab over the period. The result suggests that decline in economic growth seems to have resulted in losses of income amongst the poorest groups and thus increased income inequality in Pakistan during the period. While a number of studies have already shown a rising trends in absolute poverty, inequality seems to have worsened also throughout the 1990s. The results suggest that income inequality is higher in 2001-02 than in the previous years of the 1990s. A strand of research shows that high inequality entails a lower subsequent rate of growth in average income and hence a lower rate of progress in reducing absolute poverty. Thus, if government aims to reduce absolute poverty via its growth-accelerating strategy, it should focus fundamentally on reducing high-income inequalities through its redistributive policies of taxes and transfers.

1. INTRODUCTION

Although there has been a much debate on poverty in Pakistan in recent time, the discussion on inequality remained limited. Poverty and inequality are closely linked--for a given mean income, the more unequal the income distribution, the larger the percentage of the population living in income poverty. Thus, incomes at the top and in the middle of the distribution may be just as important to us in perceiving and measuring poverty as those at the bottom. It is, thus, important to monitor the whole income distribution rather than merely the bottom of distribution.

The issue of income inequality in Pakistan has been important in the policy discussions since the early 1960s. Since then, a number of attempts have been made to estimate the income or expenditure inequality using the Household Income and Expenditure Survey (HIES) data. However, a perception of increasing absolute poverty in Pakistan has shifted the focus of studies from inequality (or relative poverty) to absolute poverty. Consequently, a number of attempts have been made by various authors/institutions to estimate the poverty in Pakistan in the 1990s. The debate on trends in poverty during the 1990s--an era of stabilisation and structural adjustment has been wide-ranging in Pakistan. However, there is no discussion on the changes in income distribution from the policy and institutional reforms. World Bank (2003); FBS (2001) and Kemal (2003) are only three exceptions. While the former two studies report Gini Coefficients in their studies on absolute poverty in Pakistan without explaining its variations over time, the latter study is a comprehensive review on the income distribution in Pakistan. It is this context that guided the author to evaluate the trends in inequality in Pakistan using the most recently available household data sets--PIHS 1998-99 and 2001-02. The results for the year 2001-02 are being presented for the first time, which should be useful to assess the impact of various polices pursued by the government during this period. This paper is organised as follows: Section 2 provides a review on the historical trends in income inequality in Pakistan. Section 3 discusses measurement of inequality. Section 4 discusses the data sets that have been used in this study. Section 5 examines the trends in inequality in Pakistan between 1998-99 and 2001-02. Finally, Section 6 draws some conclusions from the analysis.

2. REVIEW OF INEQUALITY

Existing work on inequality shows that a large number of attempts have been made to estimate extent of income inequality in Pakistan during the last four decades. Various attempts on income distribution include Bergen (1967); Azfar (1973); Khundkar (1973); Naseem (1973); Alauddin (1975); Chaudhry (1982); Mahmood (1984); de Krujik and Leeumen (1985); Ahmad and Ludlow (1989) and Malik (1992). The major limitation of the earlier studies was that they were based on published grouped data set of Household Income and Expenditure Surveys (HIES) conducted by the Federal Bureau of Statistics, Government of Pakistan. The grouped data does not allow wide range of adjustments among household income or expenditure. Thus, inhibiting better quality estimates. However, recently when HIES micro data became available in the 1990s, a number of authors used the micro data to estimate the Gini Coefficient. These included Malik (1992a); Anwar (1997) and Abroad (2000). More recently, while FBS (2001) and Word Bank (2003) reported Gini Coefficients for the 1990s in their studies on poverty without explaining its variations, Kemal (2003) presented a comprehensive review on income distribution in Pakistan. Table 1 reports different inequality indices estimated by various studies.

Chart 1 summarises the inequality trends implied by the above evidence from 1963-66 to 1998-99. The evidence suggests a declining trend in income (or expenditure) inequality between 1963-64 to 1970-71. There is only one exception where inequality increased in urban areas between 1963-64 to 1966-67. The evidence in the 1970's and 1980s suggests that inequality seems to have worsened over these two decades. However, only in one case inequality declined in urban areas between 1978-79 to 1984-85, while the overall inequality continued to increase during this period, which is consistent with the long-term inequality trend over the last two decades. It is noteworthy that during the above period, the economy witnessed a remarkable growth rate mainly due to the capital inflow in the form of worker's remittances and foreign aid, which has also increased the real wages over the period. However, the wage increases and worker's remittances were not spread evenly among households, which seem to have increased income inequality during the above period. Thus, it appears that rapid economic growth has increased inequality during the 1970s and 1980s. On the other hand, evidence shows that the rapid growth has reduced' the absolute poverty over the period. It is mainly due to the fact that large remittances from overseas workers to their families increased the income of a large number of people below the poverty line. In addition, the real wages of both skilled and unskilled workers have also increased over the period. The rise in real wages together with remittances resulted in a decline in proportion of households in absolute poverty over the last two decades. Thus, it appears that while rapid growth has worsened income inequality, the rapid growth has reduced the poverty during the 1970s and 1980s.

On the other hand, trends in income inequality during the 1990s are different from the decades of the 1970s and 1980s. The evidence suggests an increasing trend in inequality in Pakistan between 1987-88 and 1992-93. While urban inequality remained stagnant, the rural inequality continued to rise between 1987-88 and 1992-93. Finally, inequality increased in all regions between 1992-93 and 1998-99. It is noteworthy that economic growth has slowed down during the decade of the 1990s, which seems to have affected the income of the poorest segments of the population and thus resulted in higher inequality in Pakistan. Thus, it appears that while rapid growth has worsened the inequality during the 1970s and 1980s, the slow growth has also increased inequality in Pakistan during the 1990s.

3. MEASUREMENT OF INEQUALITY

The most common approach is to select number of inequality measures and compute them to rank the income distribution. A number of different inequality indices have been proposed on different basis. (2) These include Coefficient of Variation, Gini Coefficient, Atkinson index etc. These measures are more sensitive to changes in different parts of the size distribution than others.

The well-known Gini Coefficient is the ratio of twice the area between the Lorenz curve and the diagonal. There are various ways of expressing this ratio--for example

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)

Where [y.sub.i] are arranged in ascending order by their subscripts and thereby created scope for numerous generalisation. The Gini Coefficient is most sensitive to the middle part of distribution because it depends on the rank order weights of income recipients and on the number of recipients within a given range.

Although a number of different inequality indices (3) have been proposed on different bases, an inequality measures ought to satisfy a minimal set of fundamental properties.

They are:

(1) Inequality Aversion; (2) Replication Invariance; and (3) Anonymity.

Inequality aversion is also referred to as Pigou-Dalton Principle of transfer sensitivity. The principle requires that whenever a unit income is transferred from a richer person to a poorer person and such a transfer does not reverse the ranking of the two individuals, then the measure of inequality should decrease. Replication Invariance requires that if several populations identical in every respect were combined, inequality in the combined population would be the same as for the separate ones. Anonymity presumes that appropriate adjustment for differences in needs has been made. Gini Coefficient satisfies these minimal set of properties and is the most commonly used measure of inequality. Therefore, this paper chooses Gini Coefficient to measure income inequality in Pakistan between 1998-99 and 2001-02.

4. THE DATA SET

This paper uses micro data of Pakistan Integrated Household (PIHS) conducted by the Federal Bureau of Statistics (FBS), Government of Pakistan Islamabad for the year 1998-99 and 2001-02. The universe consists of all urban and rural areas of the four provinces of Pakistan defined as such by the 1998 Population Census. The sample of PIHS 1998-99 consists of 14,821 households both rural and urban in all the four provinces of Pakistan. On the other hand, PIHS 2001-02 has 14,831 sample households. The micro data files contain weighting factors, which are designed to obtain the nationally representative estimates of population.

The household is the basic unit for which the information has been collected. It is widely accepted that income components are less reliably reported to surveyors than are expenditure items. Furthermore, incomes of the poor often vary over time, particularly in rural areas where income depends on rain-fed agriculture. This observation implies that consumption expenditure will be better a indicator than income for measurement of living standard. Hence, household consumption expenditure including non-durables and some durables is used as a proxy for 'permanent income' for the measurement of inequality in this paper.

Evidence shows that distributional assessments in Pakistan have been conducted without taking adequate account of differences in needs and economies of scale in household consumption. The common practice in Pakistan has been to divide the household expenditure by household size. This is considered to be rather unsatisfactory. This paper corrects the data for household size and composition using 0.8 for all family members.

5. TRENDS IN INEQUALITY BETWEEN 1998-99 AND 2001-02

This section examines the trends in the distribution of expenditure between 1998-99 and 2001 from two household surveys--PIHS 1998-99 and 2001-02. Table 2 reports Gini Coefficient for Pakistan by regions and province between 1998-99 and 2001-02. The results indicate an increase in inequality in overall Pakistan during the period. The result is also confirmed by the Lorenz curve for Pakistan which indicates that the Lorenz curve for 2001-02 lies below the 1998-99 curve implying that the distribution of 2001-02 is more unequal than the distribution of 1998-99 (See Figure 1). However, trends at the regional level depict a different picture. While inequality has declined in urban areas, it seems to have increased in rural areas during the period.

However, the above results conceal significant differences in changes in different parts of distribution, which are not captured by these inequality measures. Tables 3 and 4 report the percentage share of expenditure as proxy for income between 1998-99 and 2001-02 for overall Pakistan, rural and urban regions. The percentage share of expenditure indicate that while lowest 40 percent lost their income share, the middle and highest 20 percent gained in their income share implying that inequality increased in Pakistan during the period at the expense of the lowest 40 percent group. In rural areas, lowest 20 percent and middle 40 percent to 60 percent also lost their income share, whereas remaining income groups observed an increased in their expenditure (See Table 4). On the other hand, both the lowest 20 percent and the highest 20 percent lost their income share in urban areas but erosion of their income share income share of the lowest 20 percent was substantial as they lost 5 percentage point decline in their income share over the period. These trends are consistent (4) with the trends in absolute poverty as more unequal income distribution resulted in larger groups of people who are excluded from economic opportunities resulting in higher poverty level in 2001-02 compared to 1998-99. Results at province level indicate that two provincial regions shared in the countrywide trends in inequality (See Table 2). While rural Punjab indicates an increasing trend in inequality, Sindh, NWFP and Balochistan showed a declining trend in inequality across rural areas. Similarly, while inequality increased in urban Sindh, it declined in Punjab, NWFP and Balochistan.

[FIGURE 1 OMITTED]

6. CONCLUSIONS

The paper reviewed the historical trends in inequality and examined inequality trends from two household surveys for 1998-99 and 2001-02. The results suggest that inequality as measured by the Gini coefficient has increased in Pakistan between 1998-99 and 2001-02. While inequality rose in rural areas unambiguously, it has declined in urban areas during the period. A comparison of Gini coefficient from Tables 1 and 2 suggests that income inequality is turned out to be the highest in 2001-02 than in the previous years of the 1990s. More unequal societies tend to develop larger groups of people who are excluded from opportunities others enjoy such as a better education, access to loans, or to insurance and who therefore do not develop their full productive potentials. It is thus imperative to reduce income inequality. A strand of research shows (5) that high inequality entails a lower subsequent rate of growth in average income and hence lower rate of progress in reducing absolute poverty. Thus, if government aims to reduce absolute poverty via its growth accelerating strategy, it should focus primarily on reducing high-income inequalities through its re-distributive policies of taxes and transfers.

While World Bank (2003) and FBS (2001) showed a rising trends in inequality between 199:2-93 and 1998-99, it can be concluded together with the finding of this paper that inequality continued to rise persistently throughout the decade of the 1990s. In contrast to the earlier decades when inequality increased due to rapid economic growth, inequality seems to have worsened because of lower economic growth during the decade of 1990s. The lower economic growth seems to have resulted in losses of income amongst the poorest segments of the population.

It is noteworthy that the country had experienced a worsening in its governance profile during the 1990s. Corruption is an aspect of governance that hurt the poor through a variety of channels: lower economic growth, more regressive taxes, lower and more ineffective social spending and disincentives to investment in the human capital of the poor. Corruption also increases income inequality and poverty by perpetuating unequal distribution of assets. Thus, good governance is crucial for reducing income inequality and poverty. Good governance can have strong egalitarian effects. Evidence shows (6) that good governance is associated with higher levels of social spending, more effective targeting of resources to the poor and better quality public services.

The above changes in inequality are the result of various socioeconomic factors. To determine what is driving the increase in inequality in Pakistan throughout the 1990s, one needs to examine various factors, which are discussed here for future research. The most important force driving income inequality changes is the changes in inequality of labour earnings, which may be combined with the non-wage income associated with self-employment and entrepreneurial activities. While government taxes and transfer play an important role, the wage dispersion and non-wage income from self-employment are the most important factors in determining the changes in inequality. While income from self-employment is more unequally distributed, the wage inequality may increase because of changes in returns to education as earning differential between university-educated workers and workers with primary education may have increased over time.

Comments

This paper is very important and timely, and the author has rightly pointed out that too much focus on poverty has overshadowed the issue of inequality. I also appreciate the author's intention to remove ambiguities about the trends in inequalities. However as the discussant of the paper it is my duty to point out some inadequacies that defy the strong conclusions drawn by the author.

The author in his presentation has focused only on the trends in Gini-coefficients, (which differ significantly from those of the coefficient of variation). However in the paper the discussion is focused on the coefficient of variation in conjunction with the three graphs, and on the basis of these the author draws some very strong conclusions. Therefore I would only focus on Table 2, and the three graphs. The first major conclusion stated on page 14 is that the high coefficient of variation during the 2001, in conjunction with the graphs shows "unambiguously" that inequality increased in 2001 according to all standard transfer sensitive relative inequality measures as shown in Table 2.

This is a very strong conclusions because it will be seen from Table 2 that the coefficient of variation (C.V.) is not necessarily high uniformly, as shown in the panels of Table 2. In the first panel high inequality in 2001-02 for Pakistan can be attributed to higher inequality in urban areas because, for the rural areas it shows a considerable decline in 2001. However in the second panel we see that in 2001 the rural areas inequality based on C.V. is only high in rural Balochistan, and in the third panel we see that in the urban areas the C.V. is high for Balochistan as well as for Sindh. Therefore correspondingly, we see that rural inequalities were higher for Punjab, Sindh and NWFP during 1998-99. Similarly urban inequalities were higher for Punjab and NWFP during 1998-99. Thus the conclusion arrived at page 14, that "unambiguously" inequalities increased in 2001 does not hold for all the regions of the country.

The second major conclusion reached by the author is that urban inequality has increased and rural inequality has declined in Sindh. Interestingly, only in case of Sindh the Gini-coefficient and the coefficient of variation follow the same trends. Why is it not so for any other provinces? I have serious reservations about the 2001 data, particularly for Sindh because so far all studies have shown that overall the rural situation in Sindh is more worse be it poverty or income equalities. In this regard I would like to refer the author to the distinguished lecture by Prof. John Mellor delivered at the 17th Annual General Meeting with reference to Ghulam Muhammad Memorial Lecture published in Papers and Proceedings series of the PDR of Winter 2001. On page 388 under the title "Inequalities of Land Ownership", he writes that inequalities in land distribution are worst in Sindh, and than he goes on to emphasis the neglect of education in Sindh and how it retards agricultural growth, and poverty. Only income or consumption expenditure cannot define rural inequalities. For convincing details on this issue, I would urge the author to read Prof. Mellor's lecture.

In this context the second relevant piece of information is the Explanatory Notes on Problems and Issues in PIHS 2001, where problems identified include wrong coverage of variables in wrong PSUs with regard to income levels (page 2), and the first reason for the problems identified is "Poor Quality of Enumeration especially in Sindh". What it implies is that PIHS 2001 data has serious problems and it needs to be used with caution.

Furthermore, it is very surprising that while measuring inequality the author fails to look at the pioneering work of Prof. Kakwani who wrote the first book on measurement of inequalities in 1980s. This book, incidentally is referred to in the methodological framework the author has used.

This brings me to the central issue of methodology used in this paper. I have looked at the articles referred to in the study with respect to the methodology adopted and I think it would be helpful to the readers if the methodology is made more explicit. A good grasp of the methodology would also help overcome the inadequacies in the results.

Three minor points for the consideration of the author. The author prefers micro data without taking into account its limitations. While the seasonality problem in aggregative data is somewhat reduced, it is not so in the case of micro data, [see Kemal (2003)]. There is too much repetition but very limited analysis. Thirdly, an important issue like inequality merits serious consideration in terms of the data being used and the analysis. The paper would be significantly improved with some more effort, and would make a useful addition to the literature as well as an important guideline to the policy-makers.

Faiz Bilquees

Pakistan Institute of Development Economics,

Islamabad.

Author's Note: Thanks are due to Dr Sarfraz Khan Qureshi, formerly Director, PIDE, for his critical comments on an earlier draft of this paper. The author is also grateful to Mr Hammad Ali of the Federal Bureau of Statistics, Islamabad, for providing technical help and guidance for PIHS 1998-99 and 2001. The views expressed are those of the author and do not necessarily reflect those of the State Bank of Pakistan.

REFERENCES

Ahmad, E., and S. Ludlow (1989) Poverty, Inequality and Growth in Pakistan. The Pakistan Development Review 28:4, 831-850.

Ahmad, Mahboob (2000) Estimation and Distribution of Income in Pakistan, Using Micro Data. The Pakistan Development Review 39:4, 807-824.

Alauddin, T. (1975) Mass Poverty in Pakistan: A Further Study. The Pakistan Development Review 14:4, 431-450.

Anwar, Talat (1997) Inequality and Social Welfare in Pakistan. Journal of Social Sciences and Humanities 1:2.

Anwar, Talat, and Sarfraz Khan Qureshi (2003) Trends in Absolute Poverty in Pakistan: 1990-2001, Proceeding of the Papers at the 18th Annual General Meeting and Conference held in January, Islamabad.

Azfar, J. (1973) The Distribution of Income in Pakistan: 1966-67. Pakistan Economic and Social Review 11, 40-66.

Bergan, A. (1967) Personal Income Distribution and Personal Savings in Pakistan, 1963-64. The Pakistan Development Review 7, 160-212 and 17, 63-80.

Chaudhry, M. G. (1982) Green Revolution and Redistribution of Rural Income in Pakistan The Pakistan Development Review 21:2, 173-205.

Cowell, Frank (1993) Measuring Inequality. (2nd editions).

Culyer, A. J., and Adam Wagstaff (1997) The Economic Analysis of Social Policy. Cheltenham: Edward Elgar.

Kakwani, Nanak (1980) Income Inequality and Poverty: Methods of Estimation and Policy Applications. New York: Oxford University Press.

Kakwani, Nanak (1990) Testing for Significance of Poverty Differences with Application to Cote d Ivoire, Living Standard Measurement Study. The World Bank, Washington, D. C. (Working Paper No. 62.)

Kemal, A. R. (2003) Income Distribution Studies in Pakistan and the Agenda for Future Research. In Human Condition Report 2003. Centre for Research on Poverty Reduction and Income Distribution (CRPRID), Islamabad.

Khandkar, R. (1973) Distribution of Income and Wealth in Pakistan. Pakistan Economic and Social Review 11.

de Krujik, Hans, and Myrna van Leeumen (1985) Changes Poverty and Income Inequality in Pakistan during the 1970s. The Pakistan Development Review 24:3&4, 407-422.

Lambert, P. J. (1989) The Distribution and Redistribution of Income: A Mathematical Analysis. Oxford: Basil Blackwell.

Mahmood, Zafar (1984) Income Inequality in Pakistan: An Analysis of Existing Evidence. The Pakistan Development Review 23, 2 & 3.

Malik, Shahnawaz (1992) Rural Poverty in Pakistan. Unpublished Ph.D. Thesis Submitted to the University of Sussex.

Malik, Sohail, J. (1992a) Rural Poverty in Pakistan: Some Recent Evidence. The Pakistan Development Review 31:4, 975-995.

Mauro, Paolo (1998) Corruption and the Composition of Government Expenditure. Journal of Public Economics 69, 263-79.

Morris, N., and I. Preston (1986) Inequality, Poverty and the Redistribution of Income. Bulletin of Economic Research 38:4.

Naseem, S. M. (1973) Mass Poverty in Pakistan: Some Preliminary Findings. The Pakistan Development Review 12:4, 312-360.

Pakistan, Government of (2001) Poverty in the 1990s. Islamabad: Statistics Division, Federal Bureau of Statistics.

Pakistan, Government of (1998-99 and 2001) Household Income and Expenditure Survey (HIES), 1998-99 and 2001 Micro Data Files. Islamabad: Statistics Division, Federal Bureau of Statistics.

Pakistan, Government of (Various Issues) Economic Survey. Islamabad: Ministry of Finance.

Ravallion, Martin (1997) Can High-inequality Developing Countries Escape Absolute Poverty? Economic Letters 56 (1997), 51-57.

Shorrocks, A. F., and J. E. Foster (1987) Transfer Sensitive Inequality Measures. Review of Economic Studies 54, 485-497.

Shorrocks, A. F. (1983) Ranking Income Distributions. Economica 50, 3-17.

State Bank of Pakistan (2003) Special Section on "Making Health Services Work for the Poor in Pakistan: Rahim Yar Khan Primary Healthcare Project" in SBP First Quarterly Report for FY2004.

World Bank (2000) World Development Report, 2000-2001: Attacking Poverty. Washington, D. C.

World Bank (2003) Poverty in Pakistan: Vulnerabilities, Social Gaps, and Rural Dynamics. Washington, D.C.

(1) See Anwar and Qureshi (2003).

(2) For a good discussion of inequality measures, See Kakwani (1980); Cowell (1993); Morris and Preston (1986); Lambert (1989) and Culyer and Wagstaff (1997).

(3) For a good discussion of inequality measures, See Kakwani (1980, 1990); Cowell (1993); Morris and Preston (1986); Lambert (1989) and Culyer and Wagstaff ( 1997 ).

(4) See Pakistan (2003). Also see Anwar and Qureshi (2003), Trends in Absolute Poverty in Pakistan: 1990-2001, Proceedings of 18 th PSDE Meeting and Conference held in January 2003, Absolute poverty increased from 30.4 percent in 1998-99 to 35.6 percent in 2001.

(5) See Ravallion (1997).

(6) Sec Mauro (1998); World Bank (2000).

Talat Anwar is Joint Director, Research Department, State Bank of Pakistan, Karachi.

Table 1

Inequality Trends 1963-64 to 1998-99

 Inequality Trends
 (Gini Coefficients)
 Unit of
Authors Measurement Region 63-64 66-67 68-69 70-71 71-72

Bergan Household Rural .357 -- -- -- --
(1967) income Urban .430 -- -- -- --
 Overall .381 -- -- -- --

Khundkar Household Rural .350 .321 .294 -- --
(1973) income Urban .366 .384 .364 -- --

Naseem Real Rural .299 .299 .262 .262 --
(1973) consumption Urban .331 .371 .361 .352 --
 expenditure
 of
 Household

Alauddin Real income Rural .349 .330 .293 .291 .310
(1975) of Urban .374 .393 .380 .363 .382
 Household

Chaudhry Household Rural .348 .319 -- .219 .308
(1982) income

 Per capita Rural .223 .186 -- .146 .164
 income

Mahmood Household Rural .350 .318 .300 .303 .295
(1984) income Urban .381 .380 .374 .360 .363

 (Other measures of inequality
 based on household income)

 Coeff. of Rural .694 .634 .577 .567 .611
 Var Urban .769 .815 .813 .757 .786

 Atkinson's Rural .098 .081 .072 .069 .075
 index Urban .116 .117 .115 .105 .107
 [epsilon] =
 0.5

 Atkinson's Rural .427 .357 .339 .320 .332
 index Urban .452 .426 .414 .400 .377
 [epsilon] =
 3.0

 Theil's Rural .204 .172 .147 .143 .159
 Index Urban .246 .257 .253 .227 .237

 S.D. of Rural .632 .562 .540 .523 .540
 logs Urban .674 .648 .636 .619 .606

Ahmed and Household Rural -- -- -- -- --
Ludlow expenditure Urban -- -- -- -- --
(1989)

Malik Household Rural -- -- -- -- --
(1992a) expenditure Urban -- -- -- -- --
 Overall -- -- -- -- --

 Inequality Trends
 (Gini Coefficients)

FBS (2001) Rural -- -- -- -- --
 Urban -- -- -- -- --
 Overall -- -- -- -- --

World Bank Rural -- -- -- -- --
(2003) Urban -- -- -- -- --
 Overall -- -- -- -- --

 Inequality Trends
 (Gini Coefficients)
 Unit of
Authors Measurement Region 1979 84-85 87-88

Bergan Household Rural -- --
(1967) income Urban -- --
 Overall -- --

Khundkar Household Rural -- --
(1973) income Urban -- --

Naseem Real Rural -- --
(1973) consumption Urban -- --
 expenditure
 of
 Household

Alauddin Real income Rural -- --
(1975) of Urban -- --
 Household

Chaudhry Household Rural -- -- --
(1982) income

 Per capita Rural -- -- --
 income

Mahmood Household Rural .307 -- --
(1984) income Urban .414 -- --

 (Other measures of
 inequality based on
 household income)

 Coeff. of Rural .658 -- --
 Var Urban .927 -- --

 Atkinson's Rural .085 -- --
 index Urban .141 -- --
 [epsilon] =
 0.5

 Atkinson's Rural .354 -- --
 index Urban .473 -- --
 [epsilon] =
 3.0

 Theil's Rural .179 -- --
 Index Urban .315 --

 S.D. of Rural .565 -- --
 logs Urban .699 -- --

Ahmed and Household Rural .312 .328 --
Ludlow expenditure Urban .404 .392 --
(1989)

Malik Household Rural -- .305 .325
(1992a) expenditure Urban -- .394 .451
 Overall -- .302 .338

 87-88 92-93 98-99

FBS (2001) Rural -- .239 .252
 Urban -- .317 .359
 Overall -- .269 .302

World Bank Rural .240 .252 .251
(2003) Urban .316 .316 .353
 Overall .270 .276 .296

Sources: As cited above.

Table 2

Gini Coefficient for 1998-99 and 2001-02

Region 1998-99 2001-02

Pakistan
 Overall 0.3019 0.3067
 Rural 0.2521 0.2534
 Urban 0.3596 0.3581

Rural
 Punjab 0.2575 0.2699
 Sindh 0.2477 0.2228
 NWFP 0.2390 0.2359
 Balochistan 0.2274 0.2040

Urban
 Punjab 0.3777 0.3475
 Sindh 0.3352 0.3763
 NWFP 0.3535 0.3207
 Balochistan 0.2583 0.2519

Overall
 Punjab 0.3099 0.3059
 Sindh 0.3082 0.3434
 NWFP 0.2684 0.2555
 Balochistan 0.2314 0.2179

Source: Author's computation from PIHS, 1998-99 and 2001.

Table 3

Percentage Share of Expenditure between 1998-99 and 2001-02
in Pakistan

 Percentage Share
 of Expenditure % Change in
Population Expenditure
Income Groups 1998-99 2001-02 Share

Lowest 20% 9.45 9.12 -3.4921
Lower Middle 20% to 40% 13.17 13.16 -0.0759
Middle 40% to 60% 16.34 16.46 0.7344
Upper Middle 60% to 80% 20.88 20.98 0.4789
Highest 20% 40.16 40.28 0.2988

Source: Author's calculation from PIHS 1998-99 and 2001.

Table 4

Percentage Share of Expenditure between 1998-99 and 2001-02
in Rural and Urban Areas

 Percentage Share % Change in
 of Expenditure Expenditure
 Rural Share between

Population Income 1998-99 2001 1998-99 and
 Groups 2001-02

Lowest 20% 10.38 10.26 -1.1560
Lower Middle 20%
 to 40% 14.33 14.35 0.1395
Middle 40% to
 0.6 17.54 17.53 -0.0570
Upper Middle 60%
 to 80% 21.95 21.99 0.1822
Highest 20% 35.80 35.87 0.1955

 Percentage Share % Change in Expenditure
 of Expenditure Share between 1998-99
 Urban and 2001-02

Population Income 1998-99 2001
 Groups

Lowest 20% 8.17 7.7 -5.7527
Lower Middle 20%
 to 40% 11.63 12.02 3.3533
Middle 40% to
 0.6 14.92 15.37 3.0160
Upper Middle 60%
 to 80% 20.24 20.6 1.7786
Highest 20% 45.04 44.31 -.6207

Source: Author's calculation from PIHS 1998-99 and 2001.

Chart 1
Trends in Inequality in Pakistan 1963-64 to 1998-99

Years Rural Urban Overall

1963-64 to 1966-67 [down arrow] [down arrow]
1966-67 to 1968-69 [down arrow] [down arrow] [down arrow]
1968-69 to 1970-71 [down arrow] [up arrow] [down arrow]
1970-71 to 1971-72 [up arrow] [up arrow] [up arrow]
1971-72 to 1978-79 [up arrow] [up arrow] [up arrow]
1978-79 to 1984-85 [up arrow] [up arrow] [up arrow]
1984-85 to 1987-88 [up arrow] [up arrow] [up arrow]
1987-88 to 1992-93 [up arrow] Stagnant [up arrow]
1992-93 to 1998-99 [up arrow] [up arrow] [up arrow]

[up arrow]: An increase in inequality between two years.
[down arrow]: A decrease in inequality between two years.
Source: Various studies cited above.