International Research Journal of Finance and Economics ISSN 1450-2887 Issue 52 (2010) © EuroJournals Publishing, Inc. 2010 http://www. eurojournals. com/finance. htm Does Education Alleviate Poverty? Empirical Evidence from Pakistan Imran Sharif Chaudhry Associate Professor of Economics. Bahauddin Zakariya University Multan, Pakistan E-mail: imranchaudhry@bzu. edu. pk Shahnawaz Malik Professor of Economics, Bahauddin Zakariya University Multan, Pakistan E-mail: shahnawazmalik@bzu. edu. pk Abo ul Hassan Ph.
D Research Fellow, Department of Economics, Bahauddin Zakariya University Multan, Pakistan E-mail: adam_smith_17@hotmail. com Muhammad Zahir Faridi Lecturer, Department of Economics, Bahauddin Zakariya University Multan, Pakistan E-mail: zahirfaridi4u@yahoo. com Abstract Poverty has become a sensitive and ever remained issue almost in all developing countries of the world. Education plays a vital role in poverty alleviation. Therefore, it is important to investigate that whether different levels of education or literacy cause to alleviate poverty.
The major objective of this study is to evaluate the effects of different levels of education and literacy on the incidence of poverty in Pakistan. Our results suggest that poverty alleviation process would be accelerated if resources are targeted at education sector especially in higher education. Pakistan presents a paradoxical situation. Until the late 1980s Pakistan had achieved a spectacular record of economic growth and reduced incidence of poverty remarkably, but the country had horrible social indicators.
However when social indicators began to improve in the 1990s for a variety of reasons, both internally and externally driven, the average rate of economic growth declined. Contrary to the said situation, the general perception about Education is that the role of education in poverty alleviation, in close co-operation with other social sectors, is crucial. This paper is mainly intended to explore the reality that to what extent education is affective in poverty alleviation in Pakistan. In addition, some important macroeconomic variables have also been taken understudy to find out the reality of the problem.
Keywords: Education; Poverty; Inflation; Economic Growth; Openness; Pakistan International Research Journal of Finance and Economics - Issue 52 (2010) 135 I. Introduction Poverty is a multidimensional phenomenon, encompassing inability to satisfy basic needs, lack of control over resources, lack of education and skills, poor health, malnutrition, lack of shelter, poor access to clean water and sanitation, vulnerability to shocks, violence and crime, lack of political freedom and voices. The poor are the true poverty experts.
They assert on material well being, physical well being, social well being, security of food, security of law and order, public safety, safety from violence and civil conflicts, freedom of choice and action, being a part of the decision making body rather to be a victim of decision making body and the security of jobs. Poverty can be looked at from different angles and depending upon the perspective one adopts definitions of poverty may vary. It differs from country to country and from context to context. Poverty may be absolute or relative.
Absolute poverty can be eradicated but relative poverty cannot. Relative poverty is a dynamic concept because it involves comparison between groups. It exists in all parts of the world, either in packets or on a much larger scale. In Pakistan both absolute and relative poverty exists normally, poverty is measured in monetary terms. The causes of poverty are also multidimensional. 1 There is no single cause that can explain it fully. Poverty is often related to a number of factors: physical, psychological, economic and sociocultural.
Among the physical factors accounting for poverty are an unfavorable natural environment and lack of basic physical and economic infrastructure. These may also relate to poor health and malnutrition. Psychological factors refer to feel of hopelessness, helplessness, lack of confidence in one’s self and poor self-image resulting from inappropriate value system, cultural deprivation and undeveloped potential. These factors may also be related to an inability to participate in democratic processes and behavioral inadequacies aggravated by low levels of literacy and education.
Education is the most important factor that distinguishes the poor from the non-poor; according to Pakistan’s Interim Poverty Reduction Strategy Paper 2001, the percentage of literate of households heads is 27 in poor households while for non-poor households it is 52. Though the origins of human capital theory can be traced to the earlier economists – from Adam Smith (1776) to Alfred Marshall (1920) – it is Theodore Schultz (1961) who created a ‘human investment revolution in economic thought’ by emphasizing the role of human capital in economic growth.
Schultz (1961), Gary Becker (1964), Jacob Mincer (1972) and many others with their voluminous pioneering contributions placed education at a high pedestal in the theories of economic growth. Amartya Sen (1999) rightly argues that education constitutes a part of human freedom and human capability. . Over the period under study many important factors like unemployment, current account deficit and services growth rate have been contributed to why poverty is increasing even though education has increased consistently.
We have tried to give a brief description of the debate of researchers that if increased education has significant impact on income and thus poverty or not or whether there are other factors mitigating or attenuating the impact of education on poverty. However in our analysis, the central focus has been on the role of education in poverty alleviation. Education has important implications for the analysis of changes in a poverty profile in a country. Keeping in view the issues high lighted above, this paper tries to answer following related questions.
Does education play its role to alleviate poverty? What is the role of other key macroeconomic variables in poverty alleviation? What can be generalized about the impact of education on poverty? What are the important policy implications? These questions keep their extreme importance as answering the said questions will bring a solution to the hitherto puzzle that’s why Pakistan is lagging behind on the development path as compared to some developed countries who got independence later than us. 1 Technical consultation on literacy as a tool for the empowerment of the poor, Lampang, Thailand, 1997. 36 International Research Journal of Finance and Economics - Issue 52 (2010) To pursue the problem understudy, this paper is technically divided into several parts. Firstly we have attempted to explain the conceptual and theoretical framework of education and poverty alleviation. So far as the empirical analysis is concerned, we have divided it into two portions. The first portion presents the descriptive analyses and the second portion presents the econometric analysis which has been undertaken by considering autoregressive regression equations. II.
Education and Poverty: A Theoretical Framework The economists often define education as having ‘direct effects’ and ‘indirect effects’. The direct effects of education are the imparting of knowledge and skills that are associated with higher wages. The indirect effects, also often referred to as external benefits, include fulfillment of basic needs, higher levels of democratic participation, better utilization of health facilities, shelter, water and sanitation and the additional effects which occur in woman’s behavior in decisions relating to fertility, family welfare and health.
The relationship between education and poverty can also be examined by rate of return analysis, and production function analysis – at individual as well as social/national levels. Rates of return are estimated using either Mincerian earnings function (Mincer, 1972), or using the concept of marginal efficiency of capital that relates costs of education to the lifetime benefits, essentially earnings associated with education. III. Data and Methodological Issues In order to study the impact of education on poverty, the study chooses time series data, for thirty five years (1972-2007) for Pakistan.
The poverty data sets are collected mainly from Malik (1988), Amjad and Kemal (1997), Jamal (2003) and various issues of Pakistan Economic Survey since 2005, while the data on other variables is collected from World Bank, World Development Indicators (WDI), April 2008, ESDS International, (Mimas), University of Manchester. To make time series data on poverty incidence, a linear interpolation technique is employed. The selected time period presents the paradoxical situation of Pakistan as both growth and social indicators move in opposite directions.
That is why it is selected to understand this paradoxical situation. Thirty five years time period is long enough to capture long run effect of most of the variable constructed in this study. We have tried to keep in view the problem of endogeniety while selecting the explanatory variables for our analysis. The study chooses the absolute poverty (poverty headcount index), education literacy rate, primary school level enrollment rate, middle school level and the university level enrollment widely used proxies for education) as the key variables.
In addition, some useful variables (Growth rate, inflation rate, and Trade openness) have also been included in our model. In this study, autoregressive models are employed for econometric empirical investigation. In our first poverty autoregressive regression model, growth, literacy rate, CPI, and hcr(-1) are used to analyze while in the second model, some enrollment rates at various levels are considered. In order to achieve the objectives of the study, trade openness is also considered to check the robustness of globalization. Log values of the variables are used in the analysis.
We postulate that the incidence of poverty prevailing in the economy is significantly dependent on higher education level. International Research Journal of Finance and Economics - Issue 52 (2010) 137 IV. Results and Discussions a) Descriptive Analysis Our complete data set consist of 35 years of annual observations from 1973-2007 on the selected variables. The descriptive statistic is reported in table 1 which states that the average of head count ratio (HCR) for our study period is 27. 63% with a standard deviation (SD) of 6. 74. The average of primary school enrollment rate is 11316. 8 with 6204. 18, the value of its standard deviation (SD). Middle school enrollment is 2667. 611 on an average and with standard deviation (SD) 1326. 06. The average values for university enrollment rate, real gross domestic product (RGDP) and openness are 83045. 19, 22879. 24, 33. 81 with the value of standard deviations 65444. 71, 5756. 76, 3. 18 are given accordingly. As far as skewness of variables is concerned head count ratio (HCR), primary school enrollment rate, middle school enrollment rate and university enrollment rate are skewed on the rightward whereas openness is skewed leftward.
All the variables are skewed a little. Table 1: Descriptive Statistics HCR 27. 63 25. 20 45. 75 20. 71 6. 74 1. 04 3. 26 6. 64 0. 04 LITR 36. 93 34. 35 55. 00 22. 10 10. 92 0. 24 1. 56 3. 47 0. 18 MIDDLE 2667. 61 2350. 00 5368. 00 963. 00 1326. 06 0. 36 1. 83 2. 84 0. 24 PRIMARY 11316. 78 9827. 00 24465. 00 4210. 00 6204. 18 0. 57 2. 02 3. 36 0. 19 UNIV 83045. 19 65642. 00 296812. 00 17507. 00 65444. 71 1. 76 5. 59 28. 74 0. 00 OPEN 33. 81 34. 35 38. 91 27. 72 3. 18 -0. 30 2. 19 1. 53 0. 47 RGDP 22879. 24 23859. 71 33820. 04 14033. 11 5756. 76 -0. 06 1. 86 1. 97 0. 37 CPI 56. 51 39. 73 149. 0 7. 40 41. 73 0. 67 2. 16 3. 77 0. 15 Mean Median Maximum Minimum Std. Dev. Skewness Kurtosis Jarque-Bera Probability Kurtosis is a measure whether the data set is peaked or flat relative to a normal distribution. Kurtosis statistic of the variables shows that only HCR and university enrollment is Leptokurtic (long tailed or high peakedness) and all other variables are Platykurtic (relatively narrower tailed then the normal curve. However the value of HCR is though high compared to the value of Meso-kurtic curve but it is not too high from the value desired for a normal distribution.
The Jerque-Bera (JB) test of normality gives joint hypothesis of skewness and kurtosis. Jerque-Bera test of normality suggest that if the computed P-value of JB-statistic of university enrollment rate is sufficiently low as the value of the statistic is very different from zero, we state that the residuals for university enrollment rate is not normally distributed. For all other variables included in the present study, it is concluded that residuals for these variables are normally distributed. Table 2: Correlation Matrix HCR 1. 00 -0. 35 -0. 37 -0. 28 -0. 30 -0. 9 -0. 53 -0. 27 LITR 1. 00 0. 99 0. 98 0. 84 0. 25 0. 97 0. 98 MIDDLE 1. 00 0. 99 0. 86 0. 28 0. 97 0. 98 PRIMARY UNIV OPEN RGDP CPI HCR LITR MIDDLE PRIMARY UNIV OPEN RGDP CPI 1. 00 0. 89 0. 20 0. 95 0. 99 1. 00 0. 18 0. 84 0. 91 1. 00 0. 39 0. 17 1. 00 0. 94 1. 00 The degree of the relationship of the variables is also estimated and reported in table 2. All the variables are negatively correlated with each other. The results state that openness is highly correlated and primary, middle, university enrollment rates and RGDP are moderately correlated with HCR. 138
International Research Journal of Finance and Economics - Issue 52 (2010) b) Autoregressive Regression Analysis In our analysis, we have used a data set using time series ranging from 1973-2007. To investigate the significance of education (literacy) on the incidence of absolute Poverty, we have following autoregressive regression models. The robustness of the models is examined by including and excluding some important macroeconomic variables in our analysis. The model is given as below: The Poverty Autoregressive Regression Model- 1 LHCR = ? 0 + ? 1 LRGDP + ? LLITR + ? 3 LCPI + ? 4 LOPEN + ? 5 LHCR (? 1) + ? i Table 3 presents the estimation results in which head count index (HCI) is the dependent variable and the variables such as growth rate, literacy rate, consumer price index (CPI) and head count index (HCI) for the previous year are all explanatory variables in the present analysis. The value of adjusted Rsquared is 94. 5%, implying that 94. 6% of the variation in the dependent variable is explained by the independent variable. The value of R-squared clearly shows robustness of our results. The value of hstatistic is 1. 8, the results indicates that there is no significant autocorrelation problem in the error. The coefficient for growth verifies our theoretical expectations, implying an inverse relationship between poverty and growth. The coefficient for growth is highly significant putting an immense effect on poverty. The results verify the findings of Sarris who could find that overall economic growth reduces overall poverty. The coefficient for literacy is significant in the poverty regression analysis. However the variable is inversely related with the dependent variable which verifies the theoretical relationship of the two variables.
The above results follow the findings of Dollar and Kraay (2002) who have concluded that growth is a prominent factor in eliminating poverty and that the impact of low level of educational attainment is not so much important. The coefficient of the consumer price index (CPI) having an expected theoretical sign, implies a positive relationship with poverty. However coefficient is not statistically highly significant. Our results also second the findings of Romer and Romer who believed that an increase in inflation will be associated with a decline in the unemployment in the short run that may well relatively benefit the poor.
The findings of Agenor (1998) also strengthen our faith on the outcome of our analysis implying the fact about the poverty rates to be positively related with inflation. The previous year’s poverty is highly significant with the incidence of poverty. The coefficient of the variable is keeping a postulated positive sign. The best justification of the result is given by the Ragner Nurkse who could observe that a “country is poor because it’s poor. ” Although the theoretical expectations of our present study are fulfilled yet we have included some more important variables pertaining to the human capital.
We have included primary, middle and university enrollment rates instead of the literacy rate in our model. In order to check the impact of globalization on the incidence of poverty, we have included the trade openness in our analysis. The coefficient of openness is negative and insignificant. Table 3: Estimates of the Model-I Coefficient 5. 77051 -0. 62553 0. 512801 0. 004567 -0. 123046 0. 713883 0. 94 0. 93 1. 58 Std. Error 2. 62493 0. 300753 0. 263391 0. 085448 0. 137595 0. 094954 t-Statistic 2. 198348 -2. 079882 1. 946923 0. 053446 -0. 89426 7. 518185 F-Stat Prob Prob. 0. 0361 0. 0465 0. 0613 0. 9577 0. 3785 0. 0 99. 93 0. 00 Variable C LLGDP LLITR LCPI LOPEN LHCR(-1) R Squared Adj R Squared h-Statistic International Research Journal of Finance and Economics - Issue 52 (2010) 139 The Poverty Autoregressive Regression Model-2 It is a vivid fact that a problem like poverty cannot be eradicated at all. Owing to the said fact study is intended to explore the answer of the question “Does education alleviate poverty? ” To investigate the query, we have followed the regression model. We have developed the poverty regression model. Primary, middle and university enrollment rates as a proxy for education are used in our model.
The model is given below: ? ? 0 + ? 1 LRGDP + ? 2 LPRIMARY + ? 3 LMIDDLE + ? 4 LUNIV + ? Poverty = ? ? ? ? 5 LCPI + ? 6 LOPEN + ? 7 LHCR(? 1) + µ i ? Table 4 presents the estimation results for the poverty regression analysis where the dependent variable is the poverty had count index (HCI) and remaining seven variables namely log of real gross domestic product, log of primary school enrollment, log of middle school enrollment, log of university enrollment, log of consumer price index, log of openness and the log of head count ratio of the previous year are all independent variables.
Note that the adjusted R-squared is 95. 9% implying that the approximately 95. 9% variation in the dependent variable is explained by the independent variables. The coefficient for LRGDP is keeping a negative sign implying the inverse relationship of LRGDP with the incidence of poverty. The theoretical relationship of LRGDP and LHCR also supports the negative relationship of these two variables. But the coefficient for LRGDP is statistically insignificant pervading a little effect on the incidence of poverty.
The coefficient for log of primary enrollment rate and log of middle enrollment rate both keep a positive relationship with the incidence of poverty implying that both the standards minutely aggravate the incidence of poverty. The coefficients for both the levels are statistically insignificant which shows lesser nuisance value of primary and middle standards of education. The results also match with the findings of Rodriguez K Smith (1994) and Coulombe and Mckay (1996) who believe that the likelihood of being poor is higher for the lower levels of education.
The coefficient for the log of university enrollment rate is statistically highly significant in the poverty regression analysis as shown in the table 3. The variable is inversely related with the dependent variable which verifies the theoretical relationship of the two variables. The estimation results verify the findings of all those who believe in an effective role of human development of poverty alleviation. The estimation results stay in line with the findings of Tilak (1994) which emphasize on the role of education.
The results also explain that higher education is one of the most powerful means to reduce poverty. Our results also match with the findings of King (2005) who has argued that the agenda of the millennium development goals for education cannot be achieved without giving right consideration to higher education. All the prominent approaches of development like the human capital approach, the basic need approach, the human development approach and the capability approach which recognize the inverse relation of education and human poverty stay in line with our results.
The coefficient for inflation rate in the poverty regression analysis for log values has become significant statistically and it is positively related with the poverty head count index. The postulated positive sign of inflation portrays the fact that inflation is regarded as more of a problem by the poor. The fact was also found by William Easterly and Stanlay Fischer (2001). According to them the rich are better able to protect themselves against, or benefit from; the effects of inflation then are the poor.
The coefficient of openness is keeping a postulated negative sign, implying an inverse relationship between the incidence of poverty and openness. The estimation result shows that openness is powerfully influencing the poverty head count index as the coefficient of openness is found highly statistically significant. The results match with the findings of Derek H. C. Chen, Thilak Ranawera and Andriy Storozhuk who argue that high level of globalization, globalization would tend to increase poverty. The coefficient for the poverty of previous year is statistically highly significant, keeping a positive relationship with poverty. 40 Table 4: International Research Journal of Finance and Economics - Issue 52 (2010) Estimates of the Model-2 Coefficient 3. 707976 -0. 205005 0. 060653 0. 042189 -0. 154165 0. 127132 -0. 186327 0. 796384 0. 96 0. 95 -1. 68 Std. Error 1. 937434 0. 246698 0. 1637 0. 190211 0. 04069 0. 0777 0. 110726 0. 081578 t-Statistic 1. 913859 -0. 830995 0. 370514 0. 221801 -3. 788787 1. 63619 -1. 682781 9. 762301 F-Sat Prob Prob. 0. 0663 0. 4133 0. 7139 0. 8261 0. 0008 0. 1134 0. 1039 0. 00 114. 37 0. 00 Variable C LLGDP LPRIMAR LMIDDLE LUNI LCPI LOPEN LHCR(-1) R Squared Adj R Squared h Statistic
V. Conclusion and Some Policy Recommendations In this paper, we addressed a key issue in the current debate on economic development: the role of education in poverty alleviation. We have reviewed the empirical evidence on the relationship between education and poverty. The link of education to poverty is one of the most important dimensions of policies towards poverty. Education may affect poverty in many ways. It may raise the incomes of those with education. It may in addition, by promoting growth in the economy raise the incomes of those with given levels of education.
To measure education we used, among others, the literacy rate, primary education level, middle education level and university education level as proxies for education. To measure poverty, we emphasized on the concept of absolute poverty, using the poverty headcount index and as a proxy for relative poverty. We have used the econometric techniques to sketch a few stylized facts in a very complex framework of relationship. The present study incorporates macroeconomic, structural and policy variables to poverty headcount index and education.
More specifically, the poverty equation links the incidence of poverty to CPI, growth, literacy rate, primary school education, middle school education and university education level and openness. The said relationship thus enables the changes in poverty due to the changes in macroeconomic or policy variables to be projected. The relationship is empirically estimated using time series regressions, based on thirty five years data of Pakistan from 1973 to 2007, which determined the magnitudes of the effects of the above mentioned macroeconomic, structural and policy variables on poverty.
The results from the empirical analysis indicate that the university education significantly alleviates the incidence of absolute poverty. It is concluded that university education comes up with a powerful tool for poverty alleviation, keeping an inverse relationship with the dependent variable. As the higher education increases, the level of poverty decreases in the country. This result confirms the expectations that poverty is highly influenced by education. Local universities help developing countries in improving the skills of human capital which ultimately become helpful in poverty alleviating.
University graduates have the specialized skills to earn a living and infuse their sector of employment- whether in the private industry, the public sector or civil society-with the enterprise that underpins success. Getting universal primary education, one of the millennium development goals, without the higher education would simply mean increasing the burden of unskilled population on the economy. Some people consider university education a luxury for developing countries. It is not a luxury, it is essential.
Our estimation results confirm the best known approaches like the human capital approach, the basic needs approach, the human development approach and the Sen’s capabilities approach as all four approaches mainly emphasize on the attainment of education for economic development. Our estimation results carry an important policy implication-namely that the spread or the distribution of higher education among the population can have a powerful impact on their welfare. A household with no education among any of its members may benefit from even one member gaining access to
International Research Journal of Finance and Economics - Issue 52 (2010) 141 education, beyond the immediate gains to that particular individual. And this is not only the case when an improvement in the education of the family’s children, but also it becomes the better and immediate source of earning opportunities for other members. Our empirical results confirm that education plays an effective role in poverty alleviation. Accordingly, a focus of economic policies on education in order to reduce poverty and to speed up development appears to be justified.
Inflation also becomes the cause of poverty while trade openness reduces poverty significantly. Nevertheless, it is recommended that inflation controlled and trade opened policies will definitely and significantly address this issue of poverty alleviation in Pakistan. References [1] [2] [3] [4] Agenor, Pierre-Richard (1998). Stabilization policies, poverty and the Labour Market, Mimeo, IMF and World Bank. Amjad, Rashid, and Kemal, A. R. (1997). Macroeconomic policies and their impact on poverty alleviation in Pakistan. The Pakistan Development Review, 36(1), 39-68. Becker, Gary S. (1964).
Human Capital. New York Colombia University Press for NBER Chen, Derek H. C. , Ranaweera, Thilak and Storozhuk, Andriy, (2004). The RMSM-X+P: A Minimal Poverty Module for the RMSM-X (May 11, 2004). World Bank Policy Research Working Paper No. 3304. Available at SSRN: http://ssrn. com/abstract=610349 Dollar D, Kraay A (2002). Growth is good for the poor. Journal of Economic Growth, 7,195-225. Irfan, Muhammad (2001). Global Trends on Education. The Oxfam Education Report (2001), Chapter 2. Jamal, H. (2003). Poverty and inequality during the adjustment decade: Empirical findings from household surveys.
The Pakistan Development Review, 42(2), 125-136. Khan, Mosin (1990). Macroeconomic Policies and the Balance of Payments in Pakistan. 197286, IMF Working Paper /90 /78, Washington D. C. Malik, Muhammad Hussain (1988). Some New Evidence on the Incidence of Poverty in Pakistan. The Pakistan Development Review, 27(4), 509-516. Marshall, Alfred. Principles of Economics. London: Macmillan 1890 (1st edition); (1920) (8th edition). Book VI, Ch. IV, 2, 3 and 4 (pp. 560—566); Chapter XI, 1 (pp. 660—661) and Chapter XII, § 9 (pp. 681—684). Mincer, Jacob (1972). Schooling, Experience and Earnings.
New York: NBER. Mincer, Jacob (1958). Investment in Human Capital and Personal Income Distribution. Journal of Political Economy, 66. Ministry of Finance Pakistan (2001). Combating Poverty in Pakistan, www. finance. gov. pk Pakistan Economic Survey (various Issues), Ministry of finance, Government of Pakistan, Islamabad. Romer, Christina and David Romer (1998). Monetary Policy and the Well-Being of the Poor. National Bureau of Economic Research Working Paper 6793, November Sarris, Alexender H. (2001). The Role of Economic Development and Poverty Reduction: An Empirical and Conceptual Foundation.
University of Athens, Athens. Schultz, Theodore W. (1961). Education and Economic Growth. In N. B. Henry (Ed), social factor influencing education. Chicago: University of Chicago Press. Sen, Amartya (1999). Development as Freedom. New Delhi: Oxford University Press. Smith, Adam (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. London. (First edition). London: George Rutledge & Sons. 1903. pp. 78-79. World Bank, World Development Indicators (WDI). (April 2008). ESDS International, (Mimas): University of Manchester. [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]
