The International Journal of Economic Policy Studies

Transcription

1 The International Journal of Economic Policy Studies Volume Article 3 Economic Growth, Inequality and Poverty in Developing Countries: Policy Issues and Challenges Abu Girma MOGES Associate Professor Institute of Humanities and Social Sciences, University of Tsukuba, Japan Tennoudai 1-1-1, Tsukuba Science City, Ibaraki moges.abu.gm@u.tsukuba.ac.jp ABSTRACT Poverty alleviation has attracted a considerable degree of policy attention in recent years both at national and international levels. Poverty is no longer accepted as a way of life and coordinated attempts are being made to address the challenge. However, it takes a concerted and deliberate economic policy effort to generate robust and sustainable economic growth and its fair distribution to an increasing share of the population to address the problems of poverty in the long term perspective. This article explores the policy issues and challenges that developing countries face in reducing poverty both from the perspective of sustaining economic growth and empowering an increasing share of the population to actively participate in the growth process and earn a steady improvement in standard of living. The paper argues that economic growth is a necessary but not sufficient condition for poverty alleviation in developing countries. Using a new and nationally representative household survey dataset, we find robust growth elasticity and inequality elasticity of poverty. The study emphasizes that developing countries should pursue both economic growth and income distribution policy objectives in order to bring about poverty reduction because a one sided approach would have limited effectiveness for sustainable poverty alleviation. Key words: economic growth, income distribution, poverty, and public policy. JEL Classification Code: I32, O11 41

2 Vol Economic Growth, Inequality and Poverty in Developing Countries: Policy Issues and Challenges 1. Introduction Poverty has been endemic, widespread, and chronic for much of the economic history of the world and for the majority of the human population. The underlying causes of such a grim situation were mainly economic in nature because average productivity of labor was too low to allow a decent standard of living for the majority of the population. A small segment of the population may be able to escape the poverty trap mainly due to a pattern of income and wealth distribution that secured for them a better level of income and standard of living. Poverty was the general rule than the exception in such an economic environment. This situation of generalized poverty gave way to a steady growth in the productive capacity of national economies, and improvement in the standard of living, first in countries that undertook industrial revolution and then slowly the process spread to the rest of the world. And yet, until the early period of the 20 th century, public policy to address the problem of chronic poverty and destitution was conspicuously absent and poverty was considered a way of life about which little could be done. In recent decades, mass poverty is recognized not only as ethically and politically unacceptable but also as a formidable hurdle for sustainable economic growth (Cornia, 2004; Grada, 2007; Kanbur, 2005; Currie, 2011; Moges, 2013; Persson and Tabellini, 1994). Increasing awareness of the challenge and improvement in the economic capacity of nations to address the problems of chronic poverty has created the necessary environment for governments to undertake poverty alleviation policies. An increasing number of developing countries have adopted poverty reduction strategies and policies with the financial and technical assistance from international development partners and donors. Whereas a number of developed and some developing countries have achieved remarkable economic growth and poverty reduction, chronic poverty and economic stagnation have remained salient features of most developing countries. The effectiveness of poverty alleviation policies largely depends on the responsiveness of the incidence of poverty in developing countries to public policy measures to promote economic growth and reduce inequality. The relationship between economic growth, inequality and poverty in the context of developing countries, however, is imprecise from both theoretical and empirical perspectives. Whereas economic growth is necessary for sustained reduction in poverty, the level of inequality and changes in the pattern of income distribution exert influence on the propensity of economic growth on poverty alleviation. The emphasis on economic growth alone to address the problems of poverty, without due consideration for issues of inequality, however, has limited 42

3 International Journal of Economic Policy Studies effectiveness. And yet, income redistribution policies without robust growth rate may not be sustainable because they may weaken the efficiency of the economy to generate growth and improvement in standards of living. It is therefore important for developing countries to pursue policies that promote sustainable and shared economic growth that would have the capacity to make a dent in the extent and depth of poverty and prove opportunities for poor people to escape the poverty trap. The central theme of this paper is that effective poverty reduction in developing countries requires public policies that promote economic growth and improve the distribution of income across households. The paper argues that addressing the challenges of poverty in developing countries requires coordinated and continuous policy efforts and innovations that would facilitate the accumulation and effective use of productive resources, create equitable opportunities for all economic agents to improve their capacity to increase productivity, pursue prudent social policies to reduce excessive economic disparities across households, and promote economic growth. In developing the evidence in support of its central theme, the study contributes to the theoretical and empirical analysis of the role of economic growth and income distribution in addressing the problems of chronic poverty in developing countries. The study uses a new and comprehensive set of nationally representative household survey data from developing countries to quantify the responsiveness of the incidence of poverty to economic growth and inequality forces. The rest of the paper is organized as follows. Section two addresses the relationship and interaction factors between economic growth and the pattern of income and capability distribution. Section three examines the relative importance of inequality and the distribution of income in addressing poverty and discusses the degree to which policies to reduce inequality could translate into an effective and sustainable mechanism of poverty alleviation in developing countries. Section four focuses on the issues of economic growth and poverty reduction and how these condition the pattern of income distribution. Section five examines, as much as data allows, the quantitative relationships and interactions of the various factors that influence growth performance, inequality and poverty in a sample of developing countries. Section six briefly discusses the policy implications of the analysis and draws conclusions. 2. Economic Growth and Income Distribution Developing countries constitute a diverse set of countries with considerable share of the global population but with relatively weak economic capability. These countries generally are home to the poorest people in the world. In developing countries, poverty is widespread, population growth rate is still fast, and most children are born into poor households. The performance of 43

4 Vol their economy is such that the productivity of labor is relatively low, capital intensity is thin, investment and saving rates are generally low, human capital as well as social capital are quite small. The combination of these factors has resulted in a low standard of living, and long term growth of income has been quite modest. There has been a steady and significant divergence in economic growth and income levels between developing and developed countries resulting in considerable inequality of income across nations. Moreover, there is a high degree of income inequality within national economies. The developing countries, despite their relatively low income level, exhibit a high degree of income inequality. The relationships between the level and growth rate of income on the one hand and the distribution of income on the other have been subjects of both theoretical and policy dispute over a long while. One of the objectives of public policy is to improve the distribution of opportunities and income and yet indicators of inequality in developing countries barely show change over time, or a significant relationship with growth rate. The theoretical relationship between the level and growth of income per capita and the pattern of income distribution has remained a point of contention. At the earlier stage of development, countries tend to have more egalitarian distribution of income because most of the economic agents are poor and the economies are characterized by generalized poverty. When economic activity is predominantly agrarian and most of the labor force is engaged in the primary economic sector, egalitarian pattern of income distribution tends to dominate where income for all is near subsistence. The historical pattern of economic growth indicates that a leading sector with a small segment of the population served as an engine of growth in the development process. This initial progress of growth, unless followed by a rapid catch-up by the rest of the economy, could lead to the emergence of gaps in the economy both in productivity and income across sectors and households, leading to deterioration in the pattern of income distribution. This process could follow a pattern where the leading sector involves the rest of the economy and growth spreads throughout the economy, or the leading sector maintains its lead and builds inertia and protection against other sectors. If the economy manages to develop backward and forward linkages and the growth process continues and spreads, more and more economic agents benefit from the growth process and the overall pattern of income distribution tends to improve. This process closely follows the pattern of economic transformation from agrarian to industrial and post-industrial economic structure. The dual economy dynamics suggest that migration from the agricultural to the industrial sector creates a process in which the new sectors enjoy both higher productivity and a higher share of income relative to the rural sector. The overall pattern of income distribution in this process would tend to deteriorate before improvement is observable at 44

5 International Journal of Economic Policy Studies a higher level of income. However, this process is not necessarily evident in currently developing countries which seem to be caught up in the initial stage of economic transition. The evidence to support what is referred to as the inverted U-shape curve, or the Kuznets hypothesis, requires time series data on an economy that has passed through economic transformation and the development process. Such evidence is hard to come by in developing countries and economists rely largely on cross country observations of developed and a few developing countries to test the hypothesis. Such empirical evidence would hardly be sufficient to suggest that developing countries have to accommodate deterioration in the pattern of income distribution that comes with economic growth. In the context of developing countries, as figure 1 portrays, the hypothesis, despite fitting the general pattern, does not have firm empirical support. The Kuznets hypothesis has been challenged from different perspectives. First, it lacks the theoretical underpinning as to why income distribution deteriorates before it improves along with income levels or growth rate. The unbalanced economic growth strategy provides an explanation of the process but does not constitute a body of theory to set the observation within a consistent theoretical framework. Second, the relationship between the level and growth of income and inequality indicators may not be linear and the causal relation might as well run reverse wise. The initial level of income and subsequent growth processes may exhibit conditional convergence towards a steady state. But this process seems to apply under the condition that countries have already achieved a minimum threshold level of initial income, socio-political development, and appropriate economic policy environment for investment and growth. Third, the fact that developing countries are late comers to industrialization suggests that they can potentially learn from the experiences of developed nations and pursue accommodative policies that minimize the trade-off between equity and growth objectives. Fourth, it is plausible to consider, as the experiences of some late industrializing countries in East Asia indicate, that egalitarian pattern of income distribution might facilitate sustainable and inclusive growth rate (Birdsall et al., 1995; World Bank, 1993, 2005). This might provide further support for public policies to address income inequality so as to create enabling environment for rapid economic growth performance. By implication, initial level of inequality in developing countries might constrain the long run economic growth performance because of lack of capabilities among the masses to sustain economic growth and due to pressure for redistributive policies. The Kuznets hypothesis has been the subject of considerable discussion and criticisms from different perspectives. Whereas it served for summarizing historical tendencies in income distribution and the growth process, limitations abound. First, it lacks a strong theoretical underpinning as to why income distribution deteriorates before it improves along with income 45

6 Vol Figure -1: Initial Inequality and Growth Performance in Developing Countries Average Growth Rate of Income BLR TJK MNE LVA GIN FJI GMB SWZ CAF BRA NAM ROM GEO BLZ BGR Initial Gini Coefficient Source and Note: Computed from the World Bank database (World Bank, 2012). Initial inequality index, as measured by the Gini-coefficient, refers to the value in the first household survey for each country in the sample and average growth rate of income measures the growth rate of income between the first and the latest available household survey for each country. Please refer to the footnote 1 for country codes. levels or growth rate. The experiences of national economies in terms of the distribution of income suggest the role of political economy and institutional factors in shaping the type of public policies that address inequality concerns and undertake concrete and innovative measures (Acemoglu and Robinson, 2002). The unbalanced economic growth strategy provides an explanation of the process but does not constitute a body of theory to set the observation within a consistent theoretical framework. Second, the relationship between the level and growth of income and inequality indicators may not be linear and the causal relation may as well run reverse-wise. The initial level of income and subsequent growth processes may exhibit conditional convergence towards a steady state. But this process seems to apply under the condition that countries have already achieved a minimum threshold level of initial income, socio-political development, and appropriate economic policy environment for investment and growth. Third, the fact that developing countries are late comers to industrialization suggests that they can potentially learn from the experiences of developed nations and pursue accommodative policies that minimize the trade-off between equity and growth objectives. Fourth, it is plausible to consider, as the experiences of some late industrializing countries in East Asia indicate, that an egalitarian pattern of income distribution might facilitate sustainable and inclusive growth rate (Birdsall et al., 1995; 46

7 International Journal of Economic Policy Studies World Bank, 2005). This might provide further support for public policies to address income inequality so as to create an enabling environment for rapid economic growth. By implication, the initial level of inequality in developing countries might constrain the long run economic growth performance because of lack of capabilities among the masses to sustain economic growth and due to pressure for redistributive policies. There is some evidence to suggest that the initial level of inequality could adversely affect the subsequent growth rate performance of national economies (Alesina and Rodrik, 1994; Persson and Tabellini, 1994). This functional implication of inequality on the growth performance of countries would make it important, even if policy makers are focused on accelerating economic growth, to address poverty and overall improvement in living standards. The evidence from household consumption expenditure surveys from developing countries, however, does not seem to lend strong support in favor or against the hypothesis. Both the cross country dataset as well as the full sample of interval analysis for developing countries indicate no statistically significant relationship between the initial inequality, measured by the Gini-coefficient, and the subsequent rate of growth in income. Figure 1 graphically depicts the cross sectional perspective of the relationship between initial level of inequality and subsequent growth rate performance of developing countries. The Kuznets hypothesis remains a hypothesis with a limited theoretical framework to guide policies in addressing the problems that developing countries face between the competing objectives of fast economic growth and egalitarian distribution of income and capabilities. Inequality in developing countries has strong inertia, and the sample countries seem to face strong resistance to change, or have limited political and social will to undertake major reforms related to inequality issues. In our sample of households in developing countries, we find that the average initial inequality index, as measured by the Gini-coefficient of consumption expenditure, was 41.8 percent. Moreover, inequality exhibits, as figure 2 illustrates, resistance to decline even if most developing countries at least nominally pursue policies to reduce inequality. There is a tendency for inequality to get worse in countries with initial inequality index below 36 percent, which could also be caused by emphasis on growth objectives without sufficient consideration to inequality concerns, or effective policy instruments to minimize worsening inequality. At a higher level of inequality, however, income distribution becomes a political and public policy priority and there is a tendency and commitment to address concerns of inequality between households. As a result, inequality tends to decline and even moderate economic growth translates into robust reduction in the incidence of poverty. 47

8 Vol It is plausible that the initial level of inequality and subsequent rate of change in inequality may have a non-linear relationship. Considering a quadratic relationship and fitting to the available data suggests that between 36 and 54 percent levels of the initial inequality index, initial inequality was followed by reduction in inequality. In the sample countries with extremely high inequality, the pressure to reduce inequality gets weaker perhaps due to stiff resistance against the measure by vested interest groups. However, the overall average percentage change in inequality is marginal at 0.23 percent per year in the cross country sample, whereas the average is 0.02 percent for the full sample of inequality intervals. The data indicates that reduction in inequality is a very slow process, and low income developing countries might even face the challenge of rising income inequality. This implies that developing countries with a higher degree of inequality need an even higher rate of economic growth to achieve poverty reduction. It is important for developing countries, especially with high inequality index, to comprehensively explore the hurdles and forces of resistance for reforming the pattern of income distribution. Effective income distribution reforms could be potent instruments in addressing poverty problems in low and middle income countries. The observations across developing countries seem to suggest that inequality level is resistant to significant changes due to lack of effective policy instruments or lack of political will Figure-2: Initial Inequality and Its Rate of Change in Developing Countries Source and Note: Author s computation from the World Bank database (World Bank, 2012). Initial inequality index, as measured by the Gini-coefficient, refers to the value in the first household survey for each country in the sample, and rate of change in inequality measures the average percentage change in the level of inequality between the initial and the latest household survey figures for each country in the sample. Please refer to footnote 1 for country codes. 48

9 International Journal of Economic Policy Studies Figure-3: Initial Inequality and Average Growth of Income between Household Surveys Source and Note: Author s computation from the World Bank database (World Bank, 2012). Initial Gini-coefficient indicates the index of inequality for each round of household survey of countries in the sample and growth rate of income measures the average annual growth between household surveys. to address the situation. A closely related argument concerns the relationship between the level of initial inequality and subsequent economic growth performance of national economies. Does inequality hamper growth performance? In our context, does initial inequality in developing countries put pressure on governments to pursue redistributive policies that might reduce accumulation efforts and hamper economic growth performance? The evidence to test this relation would be to relate the initial measure of inequality to the policy choice and stance of governments in the sample countries towards redistribution and how such policy choices would affect economic growth performance. A substitute test of such a relation would be to examine the relation between a measure of initial inequality, the Gini-coefficient of consumption expenditure, across households and the growth rate of mean survey income over survey intervals. This interval analysis addresses the short-to-medium term interaction between inequality and growth forces in the sample countries. Figure 3 graphically depicts the relation among the sample countries during recent decades. The data shows how much initial inequality at each interval of surveys is related to subsequent growth rate performance. Since the number and interval of household surveys for each country is different, the data contains unbalanced indicators of the two variables. All in all, 550 valid data intervals are computed and its central implication is that the resistance of inequality to change limits the extent to which growth in income could be realized in the sample countries. 49

10 Vol When initial inequality is small, marginal deterioration in income distribution does not seem to prevent households from registering growth in income and consumption as it decompresses its reward system and provides more incentive for economic agents to exert effort in growth. However, the benefit weakens as the degree of inequality increases. The relationship between changes in inequality and change in mean household consumption expenditure is quite weak. As figure 4 shows, households in the sample on average registered positive, even though marginal, growth in their consumption expenditure under different scenarios in terms of inequality indicators. During the entire survey period, 66 percent of the observation intervals indicated increase in the average consumption expenditure of households. On the inequality side, 55.1 percent of the intervals between surveys indicate reduction in the Gini-coefficient and hence improvement in the pattern of distribution of consumption expenditure of households percent of the cases indicate deterioration both in income distribution and in average household income. A relatively significant set of intervals was observed whereby 35.6 percent of cases in which average income increased and income distribution improved as well. This broadly suggests the fact that growth in income can be achieved without compromising the distribution of income. The relationship between income distribution and growth rate of consumption expenditure of households, even if they are not causally and robustly related, is such that stable or improving distribution of income occasions higher consumption capacity for households. Taking a broader scope and one average indicator per sample country, a cross section observation of growth in consumption expenditure of households and the level of inequality in the sample indicates that about 48 percent of the sample countries managed to register positive growth in mean consumption expenditure as well as reduction in inequality indicators. This process facilitates robust reduction in the incidence of poverty. However, about one third of the sample countries had a rise in inequality with positive growth in income. In this case, the effect on the incidence of poverty depends on the relative growth rate of income and inequality. The other dimension of the argument about inequality issues is related to the objective function of the debate. The argument for an improved pattern of income distribution does not necessarily suggest elimination of inequality. There are some inherent differences across households that reflect their productive capacity and income earnings. If public policy is solely concerned with the distribution of economic outcomes without due consideration to the effort that economic agents exert in achieving their income or wealth, equality of outcome instead of equity in opportunities might have consequences that weaken the economy in generating growth, employment and income. Such redistributive policies often have short term effects and are hardly 50

11 International Journal of Economic Policy Studies Figure-4: Growth and Inequality in Developing Countries Percentage change in income PER KGZ VEN BLZ MDA ARG ROM HUN LSO HND KGZ2 HRV KAZ 0 ZMB GEO2 GEO BLR 0.6 Percentage Change in Inequality Source and Note: Author s computation from the World Bank database (World Bank, 2012). Percentage change in inequality is computed as annual average growth in the Gini-coefficient between household surveys, and percentage change in income is computed as annual percentage growth rate of survey consumption expenditure. Please refer to footnote 1 for country codes. sustainable. The alternative approach suggests increasing as much as possible equity in opportunity for economic agents to maximize their productive power and their capacity to generate a decent income and standard of living. Such an approach would focus on the process of capability formation of economic agents, and address constraints in accumulation of productive capacity. If the environment enables economic agents to accumulate productive educational skills and have a healthy life, but the job market creates unequal access, addressing this constraint is more effective and sustainable in generating economic growth and a more equitable distribution of income and wealth. Improving the environment in which economic resource accumulation, production and distribution takes places becomes a more effective public policy instrument in addressing the observed level of inequality within national economies perhaps more than attempts to equalize outcomes through income redistribution policies that might come with distortion of resource allocation and incentive structures. This process might also explain why the pattern of income distribution improves only after a country attains a minimum threshold level of income. As income improves, households would have the capacity and the incentive to accumulate both physical and human capital, leading to higher growth rate in productive capacity among a wide spectrum of economic agents. Once such a process gets underway, it has simultaneous effects on 51

12 Vol the growth rate of productivity as well as benefiting an increasing segment of the population and hence improving the pattern of income distribution. It is therefore necessary to move a step backward and address why income distribution varies across households and how much it reflects variation in ownership of the means of production such as land, labor, human capital, physical capital and entrepreneurship. The implication of such a step-wise analysis of the underlying features of economic inequality of a country is obviously more challenging, and long term policies are necessary to address the distribution of capabilities and ownership of the means of production that are the ultimate causes of inequality of income distribution in a country. It also underlies the limitations of income or consumption redistribution policies and the necessity to reorient public policies to address the issues of income distribution and inequality in a comprehensive way. Inequality in economic outcome such as income or wealth does not necessarily suggest economic inefficiency because such distribution patterns may reflect variations among economic agents in terms of effort in their economic activities. However, initial inequality can prevent economic mobility and deprive those from poor households the opportunity to accumulate productive capabilities and human capital for effective participation in the economy. Policies that may distort such individual and group effort could have adverse effects on the long term economic performance of a national economy leading to sub-optimal outcome for all households and the nation at large. 3. Inequality and Poverty Income poverty is closely related to the level of average incomes and the dispersion of income relative to the mean per capita or per household income. For countries with somewhat similar levels of average income, the distribution of income exerts a direct influence on the incidence and depth of poverty. This relationship influences how the growth rate of income could be used for poverty alleviation objectives. In other words, countries with an initial egalitarian pattern of income distribution could translate their growth performance into robust reduction in the incidence and depth of poverty (Kanbur, 2005). Likewise, the extent of poverty reduction could have accelerated if the pattern of income distribution in developing countries had been more egalitarian (Fosu, 2011). These observations strongly suggest that there are important relationships between the level of income inequality and the extent of poverty. Moreover, there are strong indicators that suggest the critical importance of equitable distribution of income in addressing the challenges of poverty in developing countries. Unlike the conventional view that underdeveloped countries have an egalitarian pattern of income distribution, a large number of developing countries exhibit considerable level of 52

13 International Journal of Economic Policy Studies inequality in income (UNU-WIDER, 2012; Milanovic, 2005). Given the relatively low level of average per capita income in such economies, combined with a high degree of income inequality, it is apparent that chronic poverty is widespread and a large segment of the population is deprived of the opportunity to achieve a productive and decent standard of living. It is also evident that the level of inequality in such highly unequal countries does not seem to improve over time without major reform measures such as redistribution of land ownership or public provision of essential services such as education and health services. In the absence of commitment of such reform measures, inequality might remain high or even deteriorate further. As a result, despite its potentially significant impact on poverty alleviation, the lack of effective policy instruments, and the propensity to resist by powerful members of society, have limited the role of reform measures in reducing inequality and addressing poverty challenges. The prevalence and persistence of severe income inequality suggests limited opportunities for economic mobility for the poor, and perpetuation of unequal access to productive asset accumulation for the majority of economic agents. This would constrain the capacity and the contribution of the majority from active participation in economic affairs and deepening of social tension. In developing countries with Gini index consistently above 40 percent, inequality poses a serious problem by its own right and limits the capacity of economic growth to translate into a better standard of living for the poor. Addressing inequality, and designing effective policies that would provide better opportunities for the poor to accumulate productive assets, could improve both the pattern of income distribution and growth contribution of the poor. Public policies to address the incidence of poverty in developing countries with a high level of inequality are confronted with more challenges than countries where an egalitarian pattern of income distribution prevails. The rich in these societies have the resources and political connections to evade tax obligations and protect their economic interests. It is usually difficult to design policies that would siphon resources from the rich and redistribute to the poor without creating economic distortions. However, in such a political economic environment, there is little political commitment to undertake egalitarian economic policy reforms. Even if the subsequent economic distortions were minimal, it would not be possible to dent poverty on a sustainable basis when the majority of the population lives in chronic poverty. Policy makers in developing countries have limited choices and policy tools to address the problem of inequality, in addition to facing prohibitively costly political resistance where the relatively rich play a disproportionately strong political role and protect their economic interest. Or when they have the political commitment to address the problem, redistributive measures may create disincentives to capital accumulation and weaken growth performance. Moreover, there are serious constraints in 53

14 Vol the implementation capability of the bureaucracy in such poor countries. Developing countries have nominally high and progressive income tax regimes, and yet effective tax progressivity is much weaker than its nominal rate indicates. This constraint coupled with possible distortion in the economic behavior of agents may in effect make income redistributive policies ineffective, and may even lead to a higher incidence of poverty. Some of the poor countries, whose economic base is dominantly agrarian, have a relatively egalitarian distribution of income. However, they are under pressure to decompress their incentive and reward structure. With market oriented economic policies, globalization and openness to international trade, and operation of NGOs in poor countries, the local factors market and incentive structure could not be isolated and maintained for long. Local professionals would seek highly rewarding jobs from the local private sector, NGOs, or even migrate abroad. This pressure for reforming the remuneration system may exacerbate, at least in the short to medium run, the pattern of income distribution in developing countries through high reward packages for skilled manpower. The situation is more complicated when an economy is in generalized poverty where per capita productivity and income are below the threshold of chronic poverty. Under such a daunting situation, income redistribution policies would have limited effectiveness in addressing the problems of poverty unless they are backed by strong economic growth performance. The pathways to escape the generalized poverty situation in low income developing countries need strong economic growth and serious reforms that could initiate and sustain economic growth. However, the quantitative evidence in this study supports the view that policies to reduce inequality, though difficult to implement, are potent in reducing poverty more than economic growth, once the growth process gathers momentum. By the same token, countries that see deterioration in income distribution would find their economic growth less effective in reducing the incidence of poverty. It is therefore apparent that prudent poverty alleviation policies combine both growth and inequality objectives to effectively and sustainably address the challenges of poverty in developing countries. 4. Economic Growth and Poverty Alleviation A robust and sustained rate of economic growth is the most effective means of poverty alleviation. This is a widely shared and empirically supported view, and policy makers in developing countries attempt to improve their growth rate performance in order to raise the standard of living and reduce the incidence and depth of poverty. For a given pattern of income and wealth distribution, policies that promote economic growth have more capacity to reduce poverty and 54

15 International Journal of Economic Policy Studies turn the poor into active participants in economic prosperity. A growing economy also has greater capacity both to address improvement in standard of living and to implement social welfare policies without major distortion in incentives and risking macroeconomic instability. Such social policies in turn reduce tension in society and allow more economic agents to participate in meaningful and productive economic endeavors reinforcing the growth promotion objectives. The effectiveness of economic growth in reducing indicators of poverty is mediated through a number of conditional factors. Countries with more-or- less similar rates of growth in average income may have different rates of poverty reduction for a number of reasons (Goldberg and Pavcnik, 2007; Barro, 1999; Currie, 2011). First, if the pattern of economic growth is such that it benefits a small segment of economic agents partly because growth opportunities come in discontinuous form that might result in some sectors and economic agents benefiting more and faster than others. Such small disadvantages might accumulate over time and create relative competitive advantage for leading sectors and economic agents compared to the average. The pattern of economic growth might have intrinsic features that would have implications for the distribution of benefits, and can effectively limit the capacity of economic growth to benefit most economic agents in the system. Moreover, depending on the linkages that domestic industries and sectors have in the economy, the benefits and growth rate of the leading sectors may not automatically lead to eventual benefits for the rest of the economy. Second, developing countries at the early stage of their economic development ladder may be confronted with low income equilibrium trap that is hostile to robust and sustained economic growth rate. The central problem of such economies is how to break away from the vicious circle of the poverty trap in which low productivity, low human capital, meager savings rate, and overall human deprivation limit the capacity of the economy to initiate and sustain economic growth. Such is the situation where poor countries remain in poverty, and growth, if it happens at all, requires external intervention to break away from the poverty trap through investment or human capital accumulation for sustained economic growth. Third, besides the economic factors, institutions and policies influence how economic growth translates into poverty reduction. Growth can have direct impact on the lives of the poor if they are empowered and productively engage in economic activities. Building the capabilities of the poor and ensuring their empowerment is a critical challenge for developing countries. The fact that economic growth remains the most effective mechanism to address the challenges of poverty in developing countries suggests that it is important to identify the emerging and potential forces of sustainable economic growth in the economy. The recent policy stance of developing countries, by deliberate choice or pressure from their external donor countries, 55

16 Vol emphasizes poverty reduction strategies where pro-poor growth strategies as well as safety nets were used to alleviate the incidence and depth of poverty in developing countries (World Bank, 2005). This is a welcome change in policy priorities that could realize the productive potential of the poor and make the growth process benefit the poor the most. However, there is a strong case for addressing the second generation of growth policies that appeal to and involve an increasingly dominant force of non-poor in developing countries. Economic policies that have limited appeal for the non-poor would be less effective in sustaining economic growth and consistently improving standards of living in developing countries. Developing countries need to consistently evaluate their economic policies in terms of creating more opportunities for those economic agents that have the capability to accelerate growth and create productive employment in the system. Such reform measures would benefit all economic agents including the poor, and would maintain incentives for the emerging middle class members to play important roles in the economy. The quantitative significance of economic growth in reducing the incidence of poverty is an important policy issue that has attracted the attention of poverty related research projects. During the 1990s, estimates of the growth elasticity of poverty in developing countries suggested a magnitude that ranged from -2.0 to -3.0 (Ravallion and Chen, 1997; Adams, 2004). These estimates suggest that a 10 percent increase in average growth rate performance could translate into a 20 to 30 percent reduction in the incidence of poverty. This is remarkable and confirms the potent power of economic growth in poverty alleviation in developing countries. As we observe in the next section, estimates for the growth and inequality elasticity of poverty in developing countries using more recent and comprehensive data reveal departures from earlier estimates, and suggest new policy recommendations. Economic growth remains an important force to reduce poverty provided that developing countries are able to initiate and sustain their growth processes, accompanied by effective measures to promote fair distribution of income and opportunities. 5. Quantitative Issues, Estimation, Discussion and Policy Implications The relative importance of economic growth and inequality in the process of poverty reduction could ultimately be estimated through quantitative evidence and qualitative analysis of the nature and robustness of the relationships between underlying variables. The effect of economic growth on poverty reduction is mediated through the pattern of income and capability distribution. Robust growth in income would have a significant effect on poverty if the prevailing pattern of income distribution is more egalitarian. The role of economic growth and egalitarian distribution of income on poverty alleviation is revealed through quantitative estimation of the relationships 56

17 International Journal of Economic Policy Studies between these variables. The relationship between the two processes is mediated through the prevailing distribution pattern, and for a given pattern of income distribution, faster economic growth leads to a lower incidence of poverty. Moreover, for a given level of economic growth, a more egalitarian distribution of income and opportunities facilitates deeper poverty reduction because growth would benefit most of the population. The first relationship is measured by the growth elasticity of poverty and is expected to have an inverse relationship, whereas the latter process is captured by the inequality elasticity of poverty which has positive relationship. The level of and changes in inequality, however, seem to be related with the level and growth of income. The famous Kuznets hypothesis has been a point of contention in economic analysis and attempts are still made to provide uncontested empirical evidence in relation to the hypothesis and its implications (Bruno et al., 1998; Ravallion and Chen, 1997; Adams, 2004). The emerging view on the issue seems to suggest that economic growth has little impact on inequality due to both the lack of political will to reform the pattern of income distribution or lack of effective instruments to change inequality. This is especially true where poverty is generalized and the mean income falls short of the poverty line. Such egalitarian distribution of income reflects more the sharing of poverty instead of improving the welfare of households. With the rise of income among some segments of the population, inequality starts to get worse as the growth process involves some households only in both the outputs and factors market. As figures 2 and 3 indicate, the different options and experiences that countries have had with respect to the growth in household mean income and changes in inequality, though transitional in nature, need to be managed prudently by public policy to reduce the social and political opposition it might ignite in developing countries. This section examines and reports the quantitative properties of the relationship between economic growth, income distribution and poverty in a large sample of developing countries. This study is unique and differs from previous similar estimation efforts in a number of ways. First, a more comprehensive and up to date dataset on mean consumption expenditure and distribution of income across households, using nationally representative household surveys in developing countries since the 1980s, is used to estimate the critical parameters for our analysis. Unlike previous studies where income data is generated from national income accounts and income distribution data was obtained from household surveys, which are not usually compatible with each other, this study uses the same source of nationally representative household survey data for both average consumption expenditure to reflect income and the Gini-coefficient, reflecting the pattern of income distribution. Second, there is a major revision in the purchasing power parity based international poverty line which made earlier data and studies at least partially obsolete. 57

18 Vol The upward revision of the poverty line from $1 per day per capita to $1.25 per day per capita led to significant re-estimation of the incidence of poverty in developing countries (World Bank, 2012). The revision has led to the reclassification of households into poor and non-poor categories in significant ways (Chen and Ravallion, 2010). This study uses a new set of estimates on the incidence of poverty and generates a more reliable estimate of elasticity of poverty. Third, cross examination of the survey data with the data from national income accounts is conducted to assess the plausibility of the data from both household and macroeconomic perspectives. This unique and most comprehensive household survey based dataset on average household income, inequality, and poverty is used to make quantitative estimates of the relationship between the variables under consideration. The data is generally more recent, consistent, nationally representative and hence more accurate in reflecting the underlying relationships between the main variables. The data is collected from the World Bank World Poverty Monitoring database and World Bank Development Indicators database, and the World Bank managed International Comparison Project (ICP). I also make use of the World Income Inequality Database from UNU- WIDER and the latest purchasing power adjusted national income dataset World Tables 7.1 version from the University of Pennsylvania center of International Comparisons. Out of 157 developing countries, according to the World Bank current classification, 99 countries with at least two nationally representative household survey data for all the variables under consideration are included in our sample 1. This generated 668 observations per variable. From this dataset further elimination of invalid entry values for one or more of the variables was made. On the basis of this data selection, the remaining data was formed into intervals using two data points to account for changes in the underlying variables. As a result, 552 intervals of data points on mean income/consumption expenditure, income inequality, and poverty rate at both 1 The list of countries and their codes is as follows: Albania (ALB), Algeria (DZA), Argentina(ARG), Armenia(ARM), Azerbaijan (AZE), Bangladesh (BGD), Benin (BEN), Belarus (BLR), Belize (BLZ), Bhutan (BTN), Bolivia (BOL), Bosnia & Herzegovina (BIH), Botswana (BWA), Brazil (BRA), Bulgaria (BGR), Burkina Faso (BFA), Burundi(BDI), Cambodia (KHM), Cameroon (CMR), Central African Republic (CAF), Chile (CHL), China (CHN), Colombia (COL), Costa Rica (CRI), Cote d Ivoire (CIV), Croatia (HRV), Dominican Republic (DOM), Ecuador (ECU), Egypt (EGY), El Salvador (SLV), Estonia (EST), Ethiopia (ETH), Fiji (FJI), Gambia (GMB), Georgia (GEO), Ghana (GHA), Guatemala (GTM), Guinea (GIN), Guinea-Bissau (GNB), Guyana (GUY), Honduras (HND), Hungary (HUN), India (IND), Indonesia (IDN), Iran (IRN), Jamaica (JAM), Jordan (JOR), Kenya (KEN), Kazakhstan (KAZ), Kirgizstan (KIR), Latvia (LTV), Laos (LAO), Lesotho (LSO), Lithuania (LTU), Macedonia (MKD), Madagascar (MAD), Malawi (MWI), Malaysia (MYS), Mali (MLI), Mauritania (MRT), Mexico (MEX), Moldova (MDA), Montenegro (MNE), Morocco (MAR), Mozambique (MOZ), Namibia (NAM), Nepal (NEP), Nicaragua (NIC), Niger (NER), Nigeria (NGA), Pakistan (PAK), Panama (PAN), Paraguay (PRY), Peru (PER), Philippines (PHI), Poland (POL), Romania (ROM), Rwanda (RWA), Senegal (SEN), Slovak (SVK), Slovenia (SVN), South Africa (ZAF), Sri Lanka (LKA), Swaziland (SWZ), Tanzania (TZA), Tajikistan (TJK), Thailand (THA), Timor-Leste (TMP), Tunisia (TUN), Turkey (TUR), Turkmenistan (TKM), Uganda (UGA), Ukraine (UKR), Uruguay (URY), Venezuela (VEN), Vietnam (VNM), West Bank and Gaza (WBG), Yemen (YEM), and Zambia (ZAM). 58