Macroeconomic Fundamentals of Poverty and Deprivation: an empirical study for developed countries

Macroeconomic Fundamentals of Poverty and Deprivation: an empirical study for developed countries Duarte Guimarães Master in Economics (Supervision by Sandra T. Silva and Ana Paula Ribeiro) September 2011 Macroeconomic Fundamentals of Poverty and Deprivation i Acknowledgements First and foremost I am thankful to my supervisors, Ana Paula Ribeiro and Sandra T. Silva, for their patience and helpful support in all phases of this dissertation, giving me always strength and encouragement. Without such valuable support, this work would have been much harder to accomplish. I am also thankful to Filipa Melo with whom I spent straight hours working side by side, both of us writing our own dissertations. Although working with friends can be very difficult, this was not certainly the case. Last, but not least, I want to thank to my family, particularly to my mother and father, who always provided me with all the necessary material and emotional conditions for a successful student’s life. Macroeconomic Fundamentals of Poverty and Deprivation ii Abstract This study aims at providing a positive contribution to the literature on the macroeconomic determinants of poverty. The literature points, on the one hand, to the evolution of poverty concept from a pure material deprivation to a multidimensional phenomenon, encompassing both physiological and social deprivations. In this regard, most of the applications are targeted to the measurement of poverty in the less developed countries. On the other hand, the research on the role of Macroeconomics in explaining poverty is rather scarce. In this context, this dissertation proposes a composite poverty index that captures seven deprivation dimensions which, relying on the literature and data availability, are important to a comparative assessment of deprivation across developed countries. The sample includes 18 countries of the European Union, from 2005 to 2008. Moreover, relying on the macroeconomic transmission mechanisms that influence poverty, a panel data econometric approach is implemented in order to study the relation between the proposed composite index and macroeconomic variables. Results show that a multidimensional poverty concept is also relevant for assessing deprivation in developed countries and that, in line with the relevant literature, the dynamics of some macroeconomic variables is crucial to deprivation performances. Keywords: Poverty; Deprivation; Macroeconomic transmission mechanisms; Poverty indexes; Panel data; European Union. JEL Codes: I32; C13; C33 Macroeconomic Fundamentals of Poverty and Deprivation iii Table of Contents Acknowledgements ............................................................................................................... i Abstract ................................................................................................................................. ii Table of Contents ................................................................................................................ iii Index of Tables .................................................................................................................... iv Index of Figures ................................................................................................................... v 1. Introduction ..................................................................................................................... 1 2. Poverty: concepts and measurement ............................................................................. 3 3. Poverty fundamentals ................................................................................................... 11 3.1. Economic growth and poverty ................................................................................. 12 3.2. Macroeconomic stabilization and poverty ............................................................... 15 3.3. Institutional framework and poverty ........................................................................ 18 4. Assessing multidimensional poverty across some European Union countries ........ 22 4.1. Measures of multidimensional deprivation .............................................................. 22 4.2. Index of multiple deprivation for developed countries: a proposal ......................... 25 5. Deprivation and macro mechanisms: a panel data analysis ..................................... 32 5.1. A brief review on methodology ............................................................................... 32 5.2. Model specification and results ............................................................................... 34 6. Conclusions .................................................................................................................... 40 References ........................................................................................................................... 42 Annex A: Systematic review on the macroeconomic determinants of poverty/deprivation ........................................................................................... 45 Annex B: Transformed ranking per deprivation dimension ......................................... 50 Annex C: Index of multiple deprivation for developed countries ................................. 54 Macroeconomic Fundamentals of Poverty and Deprivation iv Index of Tables Table 4.1: Structure of the IMD_D ...................................................................................... 26 Table 4.2: Most deprived countries by dimension ............................................................... 27 Table 4.3: IMD_D – Portugal vs EU18 average .................................................................. 29 Table 4.4: MP vs IMD_D .................................................................................................... 30 Table 5.1: Tests on cross-section and period fixed effects .................................................. 36 Table 5.2: The Model ........................................................................................................... 37 Table 5.3: Cross-section fixed effects values per country ................................................... 39 Table A.1: Macroeconomic determinants and mechanisms of poverty/deprivation – a summary of the literature .................................................................................... 45 Table B.1: Income deprivation ............................................................................................ 50 Table B.2: Employment deprivation .................................................................................... 50 Table B.3: Health deprivation and disability ....................................................................... 51 Table B.4: Education, skills and training deprivation ......................................................... 51 Table B.5: Barriers to housing and service .......................................................................... 52 Table B.6: Crime .................................................................................................................. 52 Table B.7: Living environment deprivation ........................................................................ 53 Table C.1: Index of multiple deprivation for developed countries ...................................... 54 Macroeconomic Fundamentals of Poverty and Deprivation v Index of Figures Figure 4.1: IMD_D (2005-2008), EU countries .................................................................. 28 Figure 4.2: IMD_D versus MP ............................................................................................ 31 Macroeconomic Fundamentals of Poverty and Deprivation 1 1. Introduction Poverty is defined nowadays as a multidimensional phenomenon, but it took a lot of time for social, in particular, economic research to attain this stage of maturity. From the beginning of human History until some decades ago, poverty was seen as a neutral phenomenon, particularly related with differentials on earnings or on the quantity of material wealth. Yet, currently, it is recognized as encompassing virtually all features characterizing the human being. Fighting poverty is now at the top of the political agenda of the most relevant international institutions, like the World Bank (WB) and the International Monetary Fund (IMF). Hence, the concept of poverty has been evolving as the following quotations show: “Poverty amid plenty is the world’s greatest challenge” (Wolfensohn, 2005: 240, quoted from Akoum, 2008: 226); “Eradicating poverty is an ethical, social, political and economic imperative of humankind”, (Resolutions adopted by the UN General Assembly, 1996, quoted from Akoum, 2008: 226). Nevertheless, despite this evolution of the concept, some caveats still persist when it comes to the measurement of the different dimensions of poverty and also on how to encompass them. Further discussion on the concept is needed. Some authors (e.g., Agénor, 2005) claim that research relating macro aspects with poverty are rare, and emphasize that, in order to better understand poverty and contribute to its reduction, microeconomic decisions must be encompassed with macro outcomes. So, it is crucial to study, in detail, the macroeconomic transmission mechanisms focused on the poor and the socially excluded ones. Such an analysis demands an incursion into a complex matrix that covers not only economic growth and poverty, but also macroeconomic stabilization and institutions. After a discussion around the concept of poverty, our work goes deep on reviewing the macroeconomic determinants, as well as the corresponding mechanisms, related to the phenomenon. In the second part of our analysis, we proceed with an empirical approach to the subject. Since measurement issues are still highly debated in this field, we implement a review on studies focused on assessing poverty. This review is our departing point to propose our own measure, the Index of Multiple Deprivation for Developed Countries (IMD_D), to assess deprivation in 18 countries of the European Union (EU), from 2005 Macroeconomic Fundamentals of Poverty and Deprivation 2 to 2008. This index aims at encompassing and measuring different deprivation dimensions, specifically for developed countries. Finally, based on our IMD_D, we use an econometric model to analyse if the macroeconomic variables, pointed out in the literature, are able to explain the evolution of IMD_D. The results of this exercise are then compared with the related literature in order to check for the robustness of the most relevant theoretical explanations. This thesis is structured as follows. In Chapter 2 the concepts of poverty and development, and measurement related issues are revised. Chapter 3 offers an encompassing review on the literature about macroeconomics and poverty. The empirical part starts with Chapter 4 that presents our IMD_D index and continues in Chapter 5 with a panel data econometric study. Chapter 6 concludes. Macroeconomic Fundamentals of Poverty and Deprivation 3 2. Poverty: concepts and measurement The definition of poverty is not straightforward. Arthur Shostak shows how problematic this issue is, by claiming that poverty is such a personal experience that only the poor can understand it (in Misturelli and Heffernan, 2008). Despite this complexity, a working definition is required. Before trying to define poverty there are some important issues that should be referred in advance, because the way poverty is understood and represented is of great importance to set the boundaries of the development of the concept. First, if poverty is to be defined as an economic problem, interventions on economic issues will be the main focus. Reciprocally, if the definition is framed as a national-phenomenon, the main focus of interventions will have to be on national issues (Misturelli and Heffernan, 2008). Second, Lomasky and Swan (2009), among others, identify two types of contemporary definitions of poverty. The definitions can invoke absolute or relative measures, i.e., if we have a society in where everyone’s wealth increases the same, the absolute poverty will decline, yet, relative poverty will stay the same. As Mabughi and Selim (2006: 184) mention “[a]bsolute poverty [is] referred to the subsistence below a minimum, socially acceptable living condition, established based on nutritional requirements and other essential goods”. Thus, absolute poverty refers directly and only to the poorer classes: if their wealth increases, the index shows a decline of absolute poverty. Relative poverty compares the situation across different classes of income, increasing or decreasing with streams in the gap between classes. This is a good measure because while differences between individuals persist, this definition is dynamic in nature. Relative poverty is more used in developed countries, but nowadays is also increasingly used in less developed and developing countries (Mabughi and Selim, 2006). Certainly, an important aspect to assess the links between macroeconomic performance and poverty is the way poverty is measured. For example, as DeFina (2002: 44) states, “An understanding of how aggregate labor market changes affect poverty theoretically depends on how poverty is measured”. Distinct official aggregation methods of poverty rates in each country brings different effects from labor market functioning. This happens because some aggregation methods neglect not only important characteristics of the poor population but also the number of poor individuals. Macroeconomic Fundamentals of Poverty and Deprivation 10 by the OECD established that some rich countries have indeed gone richer and some poor countries have gone poorer. Moreover, it is also true that the majority of the population in some countries has gone poorer and only a minority has gone richer. Yet, China and India have shown growth and dragged millions upon millions of people out of poverty. “So whether you are optimistic or pessimistic about what is happening in the world to income inequality and poverty depends on whether you think a glass is half filled or half empty. Both are true.” (OECD, 2008: 15) Some countries such as Germany, Canada, Norway, Italy, United States, and Finland show an increase on income inequality, but others, like Mexico, Greece, United Kingdom, and Australia, show the opposite (OECD, 2008). The recent evolution in the concept of poverty brought not only a measurement problem of how to encompass all the different dimensions of deprivation, but also the need for economic policy to find alternative ways to fight it. Some authors, like Agénor (2005), defend that more macroeconomic-oriented research is essential to a better targeted intervention. In this context, the next chapter reviews the literature on the intrinsic transmission mechanisms that establish the links between macroeconomic performance and poverty. Macroeconomic Fundamentals of Poverty and Deprivation 11 3. Poverty fundamentals Poverty, as ultimately defined in the previous chapter, is a multidimensional phenomenon that besides lack of income, should also account for deprived health, housing, education and other material conditions, as well as personal violence or even natural disasters. Agénor (2005) sustains that works relating macro aspects with poverty are rare, and the papers that do exist are normally underrated because, while focusing on the transmission mechanisms of macro shocks to the poor in developing countries, they fail to capture, for instance, the complex nature of labor markets. Hence, the author points to the need to redirect research on converging macroeconomics with poverty reduction goals. He stresses that microeconomics has been the central scientific approach used to fight against poverty, whereas macroeconomics has been, to this regard, mostly neglected. Moreover, he complains that economists suffer from lack of research interest in poverty subject, being only preoccupied with measurement aspects. However, to reduce poverty, microeconomic decisions must be encompassed with macro outcomes. In order to better understand the links between poverty and macroeconomics, in the short-, mediumand long-run, we rely on a literature review to discuss the intrinsic transmission mechanisms, aggregated into three groups: i) those that have a major influence on poverty through the economic growth channel, ii) the ones capturing the links between macroeconomic stabilization and poverty, iii) and those induced by the institutional environment, influencing both the economic growth and the stabilization mechanisms. In Section 3.1 we will focus on the macroeconomic transmission mechanisms that can influence growth, and on whether they also impinge on poverty. In Section 3.2 the main focus will be on the macroeconomic transmission mechanisms that can influence poverty through affecting the stabilization performance. Finally, Section 3.3 covers the institutional environment-related mechanisms. A reform of the institutional framework, with positive outcomes both on growth and stabilization, is expected to influence poverty positively. A compact summary of the literature survey is presented in Table A.1 in Annex A, which reports the reference entry, main aims and results on the links between Macroeconomic Fundamentals of Poverty and Deprivation 12 macroeconomic variables and poverty measures. The corresponding transmission mechanisms are also briefly reported. 3.1. Economic growth and poverty In order to eradicate poverty, some authors like Epaulard (2003), Agénor (2005), and Akoum (2008), defend ‘pro-poor growth’ policies. This new term is now widely used in both academic and international policy environments and “[a] common view is that growth is pro-poor if it reduces poverty significantly” (Agénor, 2005: 376). Growth, by itself, is seen as the most important characteristic to push a society out from poverty. “Growth is necessary to reduce poverty, and pro-poor macroeconomic policies are those that enhance the efficiency of growth to reduce poverty” (Epaulard, 2003: 21). Epaulard (2003) focus the importance of growth, but also emphasizes the importance of the distributional patterns. The larger median income is the more will be the impact of growth on poverty reduction. Ames et al. (2001) also sustain the importance of the so called ‘growth effect’ in order to achieve poverty reduction and emphasize two important characteristics that can affect the mechanisms through which growth impinges on poverty reduction: distributional patterns and sector composition. In a poverty reduction strategy, growth would be more efficient if distributional patterns were improved at first but, if not, growth will, in the end, push for such improvement, as growth, by itself, improves the distributional patterns (Ames et al., 2001). Enhancing the quality of growth by increasing the growth share to the poor is essential: thus, policies that reform land tenure, change marginal and average tax rates, and increase pro-poor social spending, should be used. The second characteristic emphasized by the authors is normally related to the conventional wisdom that growth strategies, linked with poverty reduction strategies, should be biased towards sectors where poor people are more allocated to. However, as the authors highlight, these kind of actions can actually influence positively the poor’s situation in the short run but, in the long run, they can contribute to increase poverty rather than decreasing it (e.g., if investments are mainly allocated to agriculture, they will have positive influence in decreasing rural poverty in the short run, but the increased dependence on this activity may, over the long run, intensify output variability). So, growth strategies should not be conducted Macroeconomic Fundamentals of Poverty and Deprivation 13 only towards one sector; instead, these strategies should focus on removing distortions that constrain growth in any sector. Nevertheless, some authors disagree on the negative relation between growth and poverty. For instance, Akoum (2008) concludes that although some countries have experienced high growth rates, they have not necessarily exhibited a decrease in poverty. This may be related with macroeconomic instability and/or poverty traps. As Azis (2008: 22) concludes, “the mechanisms by which macroeconomic policy affects poverty are too complex to be generalized”. In fact, poverty traps are usually referred in the literature to explain the difficulties developing countries have to perform the ‘initial jump’ to emerge out from poverty. Developed countries usually use external aid to help inducing growth in developing countries. Agénor (2005) defends that establishing empirically the existence of poverty traps is a crucial step for sensible policy design. Moreover, “[t]he relevance of aid to growth is often statistical insignificant, and when positive and significant, relatively small” (Agénor et al., 2008: 278). In an effort to assess empirically how poverty traps relate with low savings and productivity, Kraay and Raddatz (2007) do not seem to find strong evidence to support this relation, which casts doubts on the underlying theoretical motivations for the existence of poverty traps. It is important to recall that, in general, poverty traps related literature points to low savings rates at low levels of development, a sharp increase at intermediate-development ranges, leveling out at high development rates, and to a sharp increase in productivity once a certain level of development is achieved. Kraay and Raddatz (2007) seem to stress the opposite. The paper identifies sharp increases in savings at very low capital stocks, then a flat section followed by another increase in savings section for high capital stock levels. As for productivity, only constant and moderate increasing returns are found. Therefore, the association of poverty traps with low savings and productivity does not seem to be empirically relevant. Kraay and Raddatz (2007) also reject the relation between aid, investment, and growth, since it finds no evidence that aid will be necessary to influence the ‘initial jump’ to run away from poverty. Moreover, the results also do not support the idea that aid raises investment. Easterly (1999), cited by Kraay and Raddatz (2007), finds that this effect is positive and statistical significant in only 17 out of 88 countries, while no Macroeconomic Fundamentals of Poverty and Deprivation 14 support is found for the relation between investment and growth. This lack of evidence does not mean that aid is not important; only that the relation should be more carefully analyzed by incorporating the quality of public institutions. The issue on institutions will be further discussed in Section 3.3. According to Azis (2008), although many authors argue that growth is by itself one of the most important determinants to reduce poverty, this claim is incomplete. The effects of growth cannot and should not be generalized. Clarifying the specific effects of growth on the poor of each country is essential to choose the right policies. The quality of the distributional patterns is also too general, because this quality problem requires more explanations on how the distributional patterns can be improved while still preserving growth. Tarabini (2010) also argues that economic growth is insufficient for poverty reduction and that education is essential to fight poverty. Being so, a strong investment in education should be a priority in national development strategies. “The great amount of reports and documents published by the World Bank to date contribute not only to developing and to consolidating its ‘new’ top priority of fighting poverty, but also to highlighting the importance of education as one of the key mechanisms in achieving this goal. (…) basic education for poor people is understood as a crucial element for stimulating their empowerment and activation and, consequently, for increasing their capacity to create income and their chances of breaking the intergenerational cycle of poverty” (Tarabini, 2010: 207). Hence, education can positively influence productivity, economic growth and social development. Furthermore, Petrakis and Stamatakis (2002) show that primary education is essential to increase productivity levels and growth in low-income countries, being the importance of the secondary education more moderate, but still high. Instead, higher education levels seem to be more advantageous to growth and development in wealthy developed countries. As the level of development increases, the countries need higher levels of education that will generate higher levels of labor productivity. The two processes, education and development, are, thus, complementary. The investments on pro-poor programs and on efficient delivery of essential public services are also crucial (public education, public health, social welfare, infrastructure, etc.). Moreover, public investment can also enhance private investment (Ames et al., 2001). Macroeconomic Fundamentals of Poverty and Deprivation 15 In the next section, and since macroeconomic stabilization may also be directly or indirectly (as a means to achieve economic growth) related to poverty, we will focus on the main mechanisms that, to this respect, are referred amongst the literature. 3.2. Macroeconomic stabilization and poverty Stability exists when economic relationships are balanced (e.g., domestic demand/output, payments/domestic revenues, savings/investments, etc.). However, stability does not mean that deficits or surpluses cannot exist; instead, it just requires that they are financed in a sustainable manner. Defining an economic situation as stable or unstable is not straightforward, being necessary to look at a combination of key macroeconomic variables (e.g., inflation, growth, public sector deficit and debt, current account deficit, international reserves). Economic instability is normally associated, among others, with stagnant or declining Gross Domestic Product (GDP), double-digit inflation rates, high and rising levels of public debt, and large current account deficits financed by short-term borrowing. Moreover, it has two main sources: exogenous shocks (e.g., natural disasters, terms of trade shocks, reversals in capital flows, etc.) and inappropriate policies (loose fiscal or monetary policy stance). Macroeconomic instability hurts more the poor, relatively more vulnerable to, for example, high inflation rates and recessions. According to Ames et al. (2001), and by the same line of reasoning, any poverty reduction strategy should be financed in a sustainable and noninflationary manner, in order to maintain macro stability. Hence, policymakers should define a set of attainable macroeconomic targets (i.e., inflation, external debt, growth and net international reserves) to sustain macroeconomic stability, and pursue macroeconomic policies (monetary, exchange rate, and fiscal) accordingly. Macroeconomic stability is essential to economic growth, and also for this reason macroeconomic stability should be promoted. Ames et al. (2001) point to an important consequence of low or negative output growth in a country: the ‘hysteresis’ phenomenon. This phenomenon operates typically through shocks to the human capital of the poor: e.g., poor families’ children tend to abandon school during crises, which will influence negatively poverty in long run. Macroeconomic Fundamentals of Poverty and Deprivation 16 Macroeconomic stabilization is, among others, characterized by the maintenance of low inflation goals that, by itself, appears to be essential for poverty reduction. Inflation can have a direct impact on poverty. In fact, poor people allocate a large share of their income to subsistence and, so, changes on the prices of goods and services that the poor consume, or changes on the government expenditures, significantly matter to them (Agénor, 2005; Ames et al., 2001). If the goods that are consumed in large amounts are kept under control by the government, inflation may have little impact on the poor; otherwise, it will affect negatively and significantly the poor. Reduction in subsidies of goods and services will have similar effects. The behavior of overall inflation also matters because poor people are more vulnerable to inflation than higher-income groups. Poor people income is normally defined in nominal terms, not benefiting from indexation mechanisms. Moreover, they lack access to assets such as land or art objects that are not subject to inflation depletion. Hence, lowering the level of inflation can benefit the poor. Nevertheless, some authors, namely Azis (2008), claim that these effects of inflation cannot and should not be generalized. Clarifying the effects of inflation on the poverty line or the effects of output reduction on the income of poor households is essential to choose the right policies. In fact, disinflation can also be critical to all society, including the poor, if it is accompanied by a contraction of the aggregate demand and employment. This will increase labor supply which may lead to downward pressures on wages, increasing poverty. Also, a reduction of the inflation level through tight macroeconomic policies increase real interest rates and reduce growth rates through the effect of the former on the level and efficiency of investment. Changes in aggregate demand correspond to another macroeconomic transmission channel that may have impact on poverty through changes in employment and wages (Agénor, 2005); e.g., fiscal shocks like wage cuts in the public sector may directly raise the poverty rate, particularly if it happens during periods when economic activity is subdued or in the absence of a proper safety net, since the public sector employees have normally low wages. Reduction in government transfers, cuts in current spending on goods and services or capital spending may also increase poverty by reducing the demand for labor and the aggregate demand. Macroeconomic Fundamentals of Poverty and Deprivation 17 Macroeconomic policies that change aggregate demand by affecting private spending are also possible (Agénor, 2005); e.g., fiscal adjustments such as increases in tax rates on wages or profits lowers the expected profit and net rate of return on capital, which may reduce private expenditure on consumption and investment, lowering the aggregate demand. Another way of lowering private expenditures is based on restrictive credit through tight monetary policy. Conversely, cuts in public expenditures can also increase private expenditures if they reduce the cost or increase the availability of bank credit to the private sector, increasing aggregate demand. Additionally, fiscal adjustments that reduce government expenditures also reduce the pressures for monetization of the deficit, which may pull inflation down. Real exchange rate appears also to be a crucial macroeconomic variable in affecting poverty. In order to understand how a depreciation of the real exchange rate can affect the poor, we need, first, to know where the poor are predominantly allocated in terms of economic activity and, second, if the poor tend to consume more of imported goods relative to non-tradable goods (Agénor, 2005). A real depreciation increases the prices of imported goods and fosters a reallocation of resources towards (agricultural) export sectors, raising the income of the corresponding workers (farmer and rural households). Inequalities in poverty may arise, because rural poverty can be decreasing while urban poverty is most probably increasing; this happens because while a reallocation of resources towards the agricultural sector is being made, the demand for labor in the urban areas can decrease, and also because the poor from urban areas tend to consume more imported goods, that are more expensive after the real depreciation. Moreover, the increase in the prices of imported goods (machinery and equipment), if not accompanied by a cut in tariffs, may reduce the demand for skilled workers. If we assume that skilled and unskilled workers are substitutes, the demand for unskilled workers will increase, raising employment and income for the poorer (as the poor are usually less skilled). If cuts in tariffs are implemented, the prices for imported goods may actually fall, raising the demand for skilled workers, and the opposite situation may occur. If the economy depends on crucial imported intermediate inputs (in particular, commodities), demand for labor may decrease, unemployment may rise, and poverty may increase. Hence, the external competitiveness of a country can have a direct impact on the poor (Agénor, 2005). Macroeconomic Fundamentals of Poverty and Deprivation 18 It is also important to refer that business cycles have asymmetric impacts on poverty. Recessions and crises tend to increase poverty rates significantly, whereas expansions tend to have a more limited effect. Hence, the ability of the institutional framework to smooth these cycles is essential. Recessions reduce the demand for labor and tend to put downturn pressures on wages, raising unemployment in the formal sector (Agénor, 2005). In developing countries, with rather imperfect credit markets and where no state benefits for the unemployed are available, individuals cannot afford to stay for long time unemployed, so they will move to the informal and the rural sectors. This will tend to put downturn pressures on wages in these two sectors as well. Also, in a recession, firms tend to fire first the unskilled workers while keeping the skilled ones. When the crisis ends, firms have incentives to recover the productivity losses. Given the high complementarity between skilled workers and physical capital, firms may be tempted to increased fixed investment instead of hiring unskilled workers. Hence, any pro-poor macroeconomic policy should aim at smoothing economic fluctuations, particularly, downturns (Epaulard, 2003). Furthermore, among other authors, Ames et al. (2001) claim also that countries should support structural reforms in order to improve and strengthen flexibility in markets’ adjustments. Hence, quality of institutions seems also to be determinant for achieving lower stabilization costs. Since the quality of institutions appears as a crucial determinant either for economic growth or for macro stabilization, in the next section we bring the institutional framework into discussion. 3.3. Institutional framework and poverty “Over the last decades, national governments across the developing world have implemented economic structural adjustment programs (ESAP)” (Marquette, 1997: 1141). ESAP programs have, within their principal objectives, decreasing state interventionism and improve regulation towards non-interventionism, privatization and deregulation. Some structural adjustments may have impact at the same time on growth and on stabilization, (e.g. reforms in the fiscal structure such as on budget and treasury management, public administration, and governance) will increase efficiency and transparency, benefiting the poor via the increase on efficiency per se and through the better use of public resources (Ames et al., 2001). “Poverty reduction - in the world or Macroeconomic Fundamentals of Poverty and Deprivation 19 in a particular region or country - depends primarily on the quality of economic policy. Where we find in the developing world good environments for the households and firms to save and invest, we generally observe poverty reduction”, (Collier and Dollar, 2001: 1800). So, changes in the institutional framework can have an important impact on both growth and stabilization, and thus, may be equally essential to affect poverty. “An economy with a robust system to control corruption, an effective government, and with a stable political system will create the necessary conditions to promote economic growth, minimize income distribution conflicts, and reduce poverty in developing countries” (Tebaldi and Mohan, 2010: 16). Tebaldi and Mohan (2010) argue that improving the quality of institutions is an essential step to fight poverty. The other mechanisms (government transfers, aid programs, etc.), will have only a limited and a short effect on reducing poverty if improvements on the quality of institutions are left out of the strategy. Their paper suggests that “(…) policies aimed at reducing poverty should first consider improving institutions in developing countries as a pre-requisite for economic development and poverty eradication” (Tebaldi and Mohan, 2010: 17). Is also defended that “(…) corruption, ineffective governments, and political instability will not only hurt income levels through market inefficiencies, but also escalate poverty incidence via increased income inequality” (Tebaldi and Mohan, 2010: 16). As already mentioned in the previous section, macroeconomic volatility can arise due to domestic policy misconduct resulting from failures in the institutional design of policy authorities regarding objectives and procedures (Ames et al., 2001). This biased policy framework can affect poor in various ways. As already referred, volatility tends to distort price signals and the expected rate of return to the investors, which may delay decisions and lower both private investment and growth rates. It can also lead to higher risk premium or credit rationing, and this will affect directly the capability of obtaining loans by individuals and small and labor-intensive firms, which may result in lower private investment and lower growth rates. Lower macroeconomic volatility signals higher policy credibility, which brings the benefits mentioned above. To this regard, policy credibility is an essential characteristic to promote. “If a policy lacks credibility, the private sector does not believe that the authorities are truly committed to their policy targets, and hence does not fully factor the authorities’ targets into its inflation expectations, for instance when setting wage bargains” (Ames et al., 2001: 20). The Macroeconomic Fundamentals of Poverty and Deprivation 26 Table 4.1: Structure of the IMD_D Weights (%) Dimensions Indicators Measurement units Rank ordering 22.5 Income Deprivation (proportion of people that live on income deprivation) At risk of poverty (at risk of poverty threshold) Percentage of total population Ascendant Inability to face unexpected financial expenses Percentage of total population Ascendant 22.5 Employment Deprivation (involuntary exclusion from work) Long-term unemployment Long-term unemployment in percentage of active population Ascendant Population in jobless households Percentage of people aged 18-59 Ascendant 13.5 Health Deprivation and Disability (physical and mental health) Hospital beds Beds per 100,000 inhabitants Descendant Suicide rate Percentage of suicide on the standardized death rate by 100 000 inhabitants Ascendant 13,5 Education, Skills and Training Deprivation Persons with upper secondary or tertiary education attainment Percentage of population aged15 -64 years Descendant Early leavers from education and training Percentage of total population Ascendant 9.3 Barriers to Housing and Service (accessibility of housing) Housing overburden Percentage of total population Ascendant Severe house deprivation Percentage of total population Ascendant Overcrowding rate Percentage of total population Ascendant 9.3 Crime (rate of recorded crimes) Crimes reported by the Police Total (all recorded offences)* Ascendant 9.3 Living Environment Deprivation (quality both in and out the house) Inability to keep home adequately warm Percentage of total population Ascendant Inability to afford a meal with meat, chicken, fish (or vegetarian equivalent) every second day Percentage of total population Ascendant Greenhouse gas emissions Total emissions* Ascendant Source: Eurostat (several years) (http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/themes, accessed in May, 2011). Note 1: Within each dimension, indicators have equal weights. Note 2: All measurement units marked with * were divided by the total population of each country. Note 3: Rank ordering scores from 1, referring to the least deprived country, to a maximum value (no larger than the number of countries in the sample), referring to the most deprived country (if two countries have the same value on one indicator, they will score the same on the ranking). Macroeconomic Fundamentals of Poverty and Deprivation 27 After deciding which indicators to use, we calculate the ranking order for each of them. The next step is, following the methodology in McLennan et al. (2011), the construction of the domains. Since all indicators have the same weight in each respective domain, we just have to average across the indicators’ ranking orders within each domain. After computing each domain ranking, McLennan et al. (2011) suggest that a transformation to an exponential distribution should be implemented according to the following formula:                   23 100 11ln23 erE (4.1) Where E is the transformed domain score, and r is calculated by dividing the domain ranking order for the maximum value in the ranking, R (r varies from 1/R for the least deprived country, to R/R for the most deprived country)1. The rank transformation through equation (4.1) enables to comprise the rank scores between 1 and 100, being that the countries scoring more than 50 are among the 10% most deprived for a given domain. Tables in Annex B show the transformed ranking for each domain. Table 4.2 shows the countries that were, on average, among the 10% most deprived in each dimension. Table 4.2: Most deprived countries by dimension Dimensions Countries among the 10% most deprived (average 2005-2008) Income deprivation Latvia; Lithuania Employment deprivation Poland, Hungary Health deprivation and disability Slovenia, Estonia Education, Skills and Training deprivation Portugal, Spain Barriers to Housing and Service Poland, Latvia Crime Denmark, Germany Living Environment Deprivation Poland, Czech Republic Source: Own calculations (see tables in Annex B). 1 In our study, the maximum value in the ranking (R) is 18 (total number of countries). If two countries, or more, have the same domain ranking, the maximum score (R) will not be 18, but a lower one. Macroeconomic Fundamentals of Poverty and Deprivation 28 We conclude from the table that Poland is the only country appearing three times in different dimensions, followed by Latvia, appearing two times. The remaining countries reported in the table appear only once. The crime dimension causes some suspicion, but it can be because most of the crimes, even the less serious ones, are reported in these countries (Denmark and Germany). As for the rest of the dimensions, excluding Education, Skills and Training deprivation, notice that the most deprived countries were the latest to join the European Union (1 of May of 2004) and have also been the ones most recently admitted in the Euro Area (Slovenia, 2007, and Estonia, 2011), or are still non-participants (Poland, Hungary, Czech Republic, Latvia and Lithuania). We can, albeit timidly, conjecture that the co-investment from European funds and the monetary and fiscal discipline are non-negligible for the performance in terms of deprivation. As for the Education, Skills and Training deprivation dimension, Portugal is, in all years, the most deprived country, followed by Spain. The next step was to calculate the IMD_D through a weighted average across the transformed domain scores for each country/year. Results can be seen in. Figure 4.1 depicts the evolution of the multidimensional deprivation index for the EU countries in our sample from 2005 to 2008. Figure 4.1: IMD_D (2005-2008), EU countries Source: Own calculations (see Annex C). Macroeconomic Fundamentals of Poverty and Deprivation 29 The IMD_D averages deprivation across seven domains. From Figure 4.1 it is possible to observe that Poland is the most deprived country for the first three years under study, 2005-2007, being that position occupied by Hungary in the last year, 2008. Poland and Latvia are always between the three most deprived countries during the time span considered. Hungary starts from the fifth position in 2005, behind Lithuania and Greece, to end up the most deprived country in 2008. This degradation results from a worse performance in all dimensions; however, if we look closely to the Employment deprivation dimension (Table B.2 in Annex B) it is possible to see clearly that Poland and Latvia are getting better, while Hungary is getting worse. Austria is, consistently, the least deprived countries. As for Portugal, which starts on the seventh position, maintains in 2006-2007 the sixth position, and ends on the fifth place, is possible to understand that is growing worse in almost all dimensions, particularly in the employment deprivation and in the crime dimensions (see tables in Annex B). If we calculate the average composite index for the 18 countries, it is possible to see that this value stays around 21-23 for all years. For Portugal, we confirm that the country is near 24 in 2005; decreases in 2006 following the EU18 average trend, but increases during the last two years (see Table 4.3). Hence, the IMD_D shows that deprivation is growing worse in Portugal. Table 4.3: IMD_D – Portugal vs EU18 average Country 2005 2006 2007 2008 EU18 average 22.59 21.07 22.51 21.96 Portugal 24.47 23.20 25.57 26.60 Source: Own calculations (see Table C.1 in Annex C). Motivated by the results presented in Alkire and Santos (2010: 30), in Table 4.4 we compare the monetary poverty (MP), measured by the indicator “At risk of poverty rates” (poverty line defined as 60% of the median of the equivalent disposable income, usually used to measure monetary poverty) with the IMD_D. If the position of a country in the ranking of the IMD_D is worse than the position in the ranking of the “At risk of poverty rates” indicator, then the IMD_D will appear in the table (shadowed areas) as the latter puts the country in a worse position. If, on the contrary, the position of a country in the ranking of the IMD_D is better or equal than the position in the ranking of the “At risk of poverty rates” indicator, the MP will appear because MP puts the country in a worse position. Macroeconomic Fundamentals of Poverty and Deprivation 30 Table 4.4: MP vs IMD_D Source: Own calculations. For the countries where IMD_D appears we can say that in those cases monetary poverty does not fully capture the effective extent of deprivation. We can observe that for the countries performing worse in the IMD_D - Poland, Latvia, or Hungary - monetary poverty fails, on most of the cases, to fully capture the effective degree of deprivation. On the contrary, for countries ranking lower in the IMD_D, like Austria, Luxembourg, or Denmark, monetary poverty overestimates effective deprivation. Again following Alkire and Santos (2010: 30), Figure 4.2 plots the average (2005-2008) monetary poverty (MP), measured by the indicator “At risk of poverty rates” and the average (2005-2008) of IMD_D. GEO/TIME 2005 2006 2007 2008 Czech Republic IMD_D IMD_D IMD_D IMD_D Denmark MP MP MP MP Germany (including former GDR from 1991) IMD_D IMD_D IMD_D IMD_D Estonia MP MP MP MP Greece IMD_D MP MP MP Spain MP MP MP MP France MP IMD_D MP IMD_D Cyprus MP MP MP MP Latvia IMD_D IMD_D MP MP Lithuania IMD_D IMD_D MP MP Luxembourg MP MP MP MP Hungary IMD_D IMD_D IMD_D IMD_D Austria MP MP MP MP Poland IMD_D IMD_D IMD_D IMD_D Portugal MP IMD_D MP IMD_D Slovenia IMD_D IMD_D IMD_D IMD_D Slovakia IMD_D IMD_D IMD_D IMD_D Finland IMD_D MP MP MP Macroeconomic Fundamentals of Poverty and Deprivation 31 Figure 4.2: IMD_D versus MP Source: Own calculations. We can observe that, broadly, Figure 4.2 offers the same conclusions of Table 4.4. In 11 out of the 18 EU countries under analysis, the ranking using MP dimension alone overestimates the degree of deprivation when compared with the IMD_D position. In particular, larger differences between MP and IMD_D appear in those countries performing better in terms of lower degree of multi-dimension deprivation (countries in the second half of the figure). Apparently, and in general, differences in MP are smaller across those countries when compared with those observed using a more comprehensive measure. Moreover, only in 2 countries – Portugal and Germany – relative deprivation can be assessed either on the basis of MP alone or with the IMD_D. In the next chapter we will use the IMD_D as the relevant variable for measuring poverty/deprivation phenomena and pursue an econometric study in order to analyze what are the main potential explanatory macroeconomic variables behind deprivation in our set of EU countries. 0 2 4 6 8 10 12 14 16 18 20 IMD_D MP Macroeconomic Fundamentals of Poverty and Deprivation 32 5. Deprivation and macro mechanisms: a panel data analysis 5.1. A brief review on methodology After the construction of the IMD_D, and following the line of argumentation in the previous chapters, namely in Chapter 3, we propose now to study the role of macroeconomic variables in explaining deprivation, as measured by the IMD_D index, using a sample of developed countries. Since we have IMD_D data for 18 countries during four years, we can use panel data methodology, which we next briefly review. There are, generally, three types of data available for empirical analysis: cross section, time series and panel data. Using cross section data, following our example, we could analyze empirically the 18 countries but only for one year each time. In time series data, we could have, following again our example, one empirical analysis for each country for the period of four years. In a panel data, the same cross section data is surveyed over time: “(…), panel data have space as well as time dimensions” (Gujarati, 2004: 636). Gujarati (2004) defends that there are many advantages of using panel data. In short, panel data, when compared with cross section or time series, enrich substantially the empirical analysis. In our analysis we consider an econometric model of the following type: ititit uDIMD  βX 1 _  , i = 1,…, 18, t = 1,…,4, (5.1) Where:  IMD_Dit is the dependent variable for country i at time t;   1 is the common intercept;  β is a vector of coefficients associated with the independent explanatory (macroeconomic) variables;  it X is a vector of independent explanatory (macroeconomic) variables for country i at time t;  uit is the random term for country i at time t;  i represents the ith cross-section unit (country); t represents time (t = 2005, 2006, 2007, 2008). Macroeconomic Fundamentals of Poverty and Deprivation 33 Hence, there are a maximum of N=18 observations (countries) and a maximum of T=4 time periods (years).2 If the number of time series observations is the same for all the cross-section observations, the panel is balanced, if not, it is called an unbalanced panel. In our case, we have a balanced panel. When using panel data we must choose between fixed effects model (FEM) or random effects model (REM) (Gujarati, 2004). “The simplest, and possible naive, approach is to disregard the space and time dimensions of the pooled data and just estimate the usual OLS regression” (Gujarati,2004: 641), which puts us in the position of choosing between fixed or random effects, the biggest challenge when using panel data. The fixed effects approach accounts for the possibility of different intercepts, changing across countries or/and years. Instead, random effects approach accounts for a random effect for each cross-section or/and years unit. As Gujarati (2004: 650) emphasizes: “The challenge facing a researcher is: Which model is better, FEM or [REM]?” REM can be heteroscedastic and autocorrelated, if we assume that the error component εi is correlated with one or more regressors. And, as Gujarati (2004: 650) highlights: “If the individual error component εi and one or more regressors are correlated, then the REM estimators are biased, whereas those obtained from FEM are unbiased”. Even so, when N is large and T is small, like in our case, significant differences on the resulting estimates from the two methods are expected. If we sustain that our cross-section units are not random drawings from a larger sample, FEM is a more suitable method (in fact, our sample includes 18 out of the 27 members of the European Union). Hence, we decided to use the fixed effects model. In line with equation (5.1), the FEM may have different configurations: ititiit uDIMD  βX 1 _  , i = 1,…, 18, t = 1,…,4 (5.2) itittit uDIMD  βX 1 _  , i = 1,…, 18, t = 1,…,4 (5.3) itititit uDIMD  βX 1 _  , i = 1,…, 18, t = 1,…,4 (5.4) Considering that the intercept changes across countries, but that the slope coefficients do not, FEM may be implemented by applying dummy variables to the intercept. Hence, we may re-write equation (5.2) as (and similarly for equations (5.3) and (5.4)): 2 See Chapter 4 for country and time lengths. Macroeconomic Fundamentals of Poverty and Deprivation 34 ititiit uDIMD  βXD11 _α  , i = 1,…, 18; t = 1,…,4, (5.5) Where:   1 is the fixed effect for one of the countries;  1i D is a vector of dummy variables, each of which corresponding to each of the remainder i-1 countries;   is the constant associated to each dummy variable, that should be added (+) or subtracted (-) to  1. In order to estimate the FEM we use the software Eviews that provides built-in tools for testing FEM against REM, and also for testing the joint significance of the fixed effects, cross-section or/and time series. 5.2. Model specification and results As mentioned before, the dependent variable in our model is the IMD_D. In line with the arguments and mechanisms presented throughout Chapter 3, the literature points to some relevant macroeconomic variables that should be used as independent variables in the model: public investment, GDP growth rate, inflation, unemployment rate, government budget, and quality of institutions, among others. However, among those invoked by the literature, we chose to exclude, on the following grounds, some of the variables from the model specification:  Inflation and current tax burden, referred as impinging negatively with poverty by some authors (Ames et al., 2001; Agénor, 2005), were not significant at an individual level and, when added to the model, worsened results on the significance for the other independent variables in use. Since inflation is now rather low and stable among the (developed) EU countries and tax burden is also rather similar across the sample, these variables appear to be more important in explaining poverty in developing countries than in determining deprivation in developed ones;  Short-term unemployment rates, with negative impacts on monetary poverty (DeFina, 2002; Agénor, 2005), were strongly significant at an individual level, but when treated together with other explanatory variables the model produced Macroeconomic Fundamentals of Poverty and Deprivation 35 significant changes on the significance results. Apparently, short-run unemployment rates are strongly correlated with indicators already used in the construction of the IMD_D (namely with those embedded in the Employment Deprivation dimension), as well as with other explanatory variables, namely governance index and government budget;  Finally, variables such as investment in education, consensually acknowledged as essential to fight poverty (e.g. Petrakis and Stamatakis, 2002; Tarabini, 2010) were also disregarded due to lack of related data across time and/or countries. After several trial and error experiences, we have established the explanatory macroeconomic variables of overall deprivation: overall public investment, GDP growth rate, Gini coefficient, government budget and a governance index. All the indicators were taken from the Eurostat database (http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/themes, accessed in July, 2011), except for the governance index, taken from the World Bank (http://info.worldbank.org/governance/wgi/sc_country.asp, accessed in July, 2011). Briefly,  Public investment refers to general government gross fixed capital formation, as percentage of GDP;  GDP growth rate is measured by the growth rate of GDP in volume, as percentage change on previous year;  Gini coefficient is a measure of disposable income inequality in each country; the coefficient varies between 0, for full equality, and 100, for maximum inequality;  Government budget is defined as total revenues less total expenditures of general government, as percentage of the GDP;  Governance index is used to capture the quality of the institutions in each country; it refers to the simple average of six indicators - the Worldwide Governance Indicators (WGI) - available at the World Bank site (http://info.worldbank.org/governance/wgi/sc_country.asp, accessed in July, 2011), covering 200 countries and territories and including six dimensions of governance starting in 1996: “Voice and Accountability, Political Stability and Absence of Violence/Terrorism, Government Effectiveness, Regulatory Quality, Macroeconomic Fundamentals of Poverty and Deprivation 42 References Agénor, P. R. (2005), ‘The macroeconomics of poverty reduction’, The Manchester School, Vol. 73, No. 4, pp. 369 - 434. Agénor, P. R., Nihal, B. and Karim, A. (2008), ‘The roads out of poverty? Assessing the links between aid, public investment, growth, and poverty reduction’, Journal of Development Economics, Vol. 86, pp. 277 - 295. Aidt, T., Dutta, J. and Sena, V. (2008), “Governance regimes, corruption and growth: Theory and evidence”, Journal of Comparative Economics, Vol. 36, pp. 195220. Akoum, I. F. (2008), ‘Globalization, growth, and poverty: the missing link’, International Journal of Social Economics, Vol. 35, No. 4, pp. 226 - 238. Alkire, S. and Santos, M.E. (2010), “Acute multidimensional poverty: a new index for developing countries”, United Nations Development Programme, Human Development Reports, United Nations Alves, N. (2010), ‘A Relação entre Educação e Pobreza em Portugal’, in Teixeira, A. C., Silva, S. T. and Teixeira, P. (2010) (1st edition), O que Sabemos sobre a Pobreza em Portugal?, Porto: VidaEconómica - Editorial, S.A. Ames, B., Brown, W., Devarajan, S. and Izquierdo, A. (2001), “Macroeconomics policies and poverty reduction”, Paperback, International Monetary Fund and World Bank. Atkinson, A.B. (2003), ‘Multidimensional deprivation: contrasting social welfare and counting approaches’, Journal of Economic Inequality, No. 1, pp. 51 - 65. Azis, I. (2008), “Macroeconomic policy and poverty”, ADB Institute, ADB Institute Discussion Paper No. 111. Colliers, P. and Dollar, D. (2001), “Can the world cut poverty in half? How policy reform and effective aid can meet international goals”, World development, Vol. 29, No. 11, pp. 1787-1802. DeFina, R.H. (2002), “The impact of macroeconomic performance on alternative poverty measures”, Social Science Research, Vol. 31, pp.29–48. Dobson, S. and Ramlogan-Dobson C. (2009), “Is there a trade-off between income inequality and corruption? Evidence from Latin America”, Economic Letters, Vol. 107, pp.102-104. Macroeconomic Fundamentals of Poverty and Deprivation 43 Epaulard, A. (2003), “Macroeconomic performance and poverty reduction”, International Monetary Fund, IMF Working Paper No. 03/72. Gujarati, D.N. (2004), Basic Econometrics (Fourth edition), McGraw−Hill, USA. Gundlach, E. and Paldam, M. (2009), “The transition of corruption: From poverty to honesty”, Economics Letters, Vol. 103, pp.146-148. Kaufmann, D., Kraay, A. and Mastruzzi, M. (2010), “The worldwide governance indicators: Methodology and analytical issues”, Global economy and development at Brookings, September 2010. Kraay, A. and Raddatz, C. (2007), “Poverty traps, aid, and growth”, Journal of Development Economics, Vol. 82, pp. 315 – 347. Lomasky, L. and Swan, K. (2009), ‘Wealth and poverty in the liberal tradition’, The Independent Review, Vol. 13, No. 4, pp. 493 - 510. Mabughi, N. and Selim, T. (2006), 'Poverty as social deprivation: a survey', Review of Social Economy, Vol. 64, No. 2, pp. 181 - 204. Marquette, C.M. (1997), “Current poverty, structural adjustment, and drought in Zimbabwe”, World Development, Vol. 25, No. 7, pp. 1141-1149. McLennan, D., Barnes, H., Noble, M., Davies, J., Garratt, E. and Dibben, C. (2011), “The English indices of deprivation 2010”, Department for communities and local government, March 2011. Misturelli, F. and Heffernan, C. (2008), 'What is poverty? A diachronic exploration of the discourse on poverty from the 1970s to the 2000s', The European Journal of Development Research, Vol. 20, No. 4, pp. 666 -684. OECD (2008), Growing Unequal? Income distribution and poverty in OECD countries, OECD Publishing, Paris: OCDE. Petrakis, P.E. and Stamatakis, D. (2002), “Growth and education levels: a comparative analysis”, Economics of Education Review, Vol. 21, pp. 513–521. Rodrigues, C.F. and Andrade, I. (2010), “Monetary poverty, material deprivation and consistent poverty in Portugal”, School of economics and management, Technical University of Lisbon, Department of economics, Working Paper No. 25. Silva, M. (2010), ‘Pobreza, Direitos Humanos e Democratização da Economia’, in Teixeira, A. C., Silva, S. T. and Teixeira, P. (2010) (1st edition), O que Sabemos sobre a Pobreza em Portugal?, Porto: VidaEconómica – Editorial, S.A. Macroeconomic Fundamentals of Poverty and Deprivation 44 Tarabini, A. (2010), “Education and Poverty in the global development agenda: Emergence, evolution and consolidation”, International Journal of Education Development, Vol. 30, pp. 204–212. Tebaldi, E. and Mohan, R. (2010), “Institutions and Poverty”, Journal of Development Studies, Vol. 46, No. 6, pp. 1047-1066. World Bank (2000), World Development Report 2000 ⁄2001: Attacking Poverty, New York and Washington: Oxford University Press, World Bank. WEBSITES http://epp.eurostat.ec.europa.eu/portal/page/portal/eurostat/home/ http://info.worldbank.org/governance/wgi/index.asp Macroeconomic Fundamentals of Poverty and Deprivation 45 Annex A: Systematic review on the macroeconomic determinants of poverty/deprivation Table A.1: Macroeconomic determinants and mechanisms of poverty/deprivation – a summary of the literature Economic Growth Macroeconomic Stabilization Institutional Framework Transmission mechanism Growth (promoted by foreign aid, investment in human and/or physical capital, etc.) increases average material living conditions; the potential positive impact of growth on poverty depends on the evolution of inequality. Macroeconomic stability is, per se, essential for any strategy of poverty reduction and also affects economic growth. For example: poor people are more vulnerable to inflation than higher-income groups; depending on how poor are allocated across activity sectors and on the composition of their consumption basket, real exchange rate affects reallocation of resources and, thus, poverty; changes in the aggregate demand may have direct impact on poverty or, indirectly, through changes in employment and wages; trade liberalization can smooth the impact of the crises in poverty, etc. Quality of institutions contributes for achieving both lower stabilization costs and enhancing economic growth. Explanatory variables Paper Research Aims Scope Real (per capita) GDP growth rate Inequality indicators Foreign aid Human capital Public capital Inflation Exchange rate Aggregate demand Trade liberalization Legal framework, economic and political organization and rules Main results Definition of poverty (indicator) Collier and Dollar (2001) Impact of external aid on fighting poverty. Developing countries. X X The positive impact of aid on poverty depends, not only on the quantitative amount of aid, but also (crucially) on the quality of the institutions and policies. The paper shows that if aid was used in the right way, the results would be twice more effective than they are now. Monetary poverty (poverty lines) Petrakis and Stamatakis (2002) The effect of human capital on growth across different levels of developmen t. General to all levels of developmen t countries. X X X Education strongly influences economic growth. Higher development levels demand for higher levels of education because it will generate higher levels of labor productivity. Education and development are complementary processes. Monetary poverty (economic growth) Macroeconomic Fundamentals of Poverty and Deprivation 46 Paper Research Aims Scope Real (per capita) GDP growth rate Inequality indicators Foreign aid Human capital Public capital Inflation Exchange rate Aggregate demand Trade liberalization Legal framework, economic and political organization and rules Main results Definition of poverty (indicator) Epaulard (2003) The role of economic growth and distributiona l patterns on poverty reduction. Developing countries. X X X X Growth is essential for poverty reduction strategies but if complemented with improvements in inequality lead to higher rates of poverty reduction. Moreover, corruption, inflation and trade liberalization also influence the link between growth and poverty. Monetary poverty (poverty lines) Agénor (2005) Overview of recent literature on the macroecono mics of poverty. Study of channels through which macroecono mic policy can affect the poor. SubSaharan Africa. Poor classes in the rural, informal, and formal sectors. X X X X X X Micro and measurement aspects are not enough to study poverty. A better study of the labor market in the developing countries is needed to overcome some distortions, and to avoid incorrect inference in assessing how a given policy measure affects the poor. The recent attempts to develop a model to be used for poverty analysis have failed. Monetary poverty (poverty lines) Kraay and Raddatz (2007) Empirical relevance of poverty traps caused by low savings/pro ductivity. Developing countries, in particular, African countries; general poor. X X Contrary to literature, the association of poverty traps with low savings and productivity appears not to be empirically relevant. The paper identifies a sharp increase in savings when capital stocks are very low, then a flat section followed by another increase in savings at very high capital stock levels. As for productivity, only constant and moderate increasing returns are found. Monetary poverty (poverty lines) Macroeconomic Fundamentals of Poverty and Deprivation 47 Paper Research Aims Scope Real (per capita) GDP growth rate Inequality indicators Foreign aid Human capital Public capital Inflation Exchange rate Aggregate demand Trade liberalization Legal framework, economic and political organization and rules Main results Definition of poverty (indicator) Agénor et al. (2008) Model that captures the links between aid, public capital (health, core infrastructur e and education), growth, and poverty reduction. Ethiopia; general poor. X X X X X Aid is crucial to sustain adequate levels of government spending and public investment. The results show that is important to improve aid while the management of public resources is reformed. This maximizes growth and reduces poverty. Monetary poverty (poverty lines) Azis, I. (2008) The role of growth and macroecono mic stability on poverty reduction. Thailand and Indonesia; general poor. X X X X X The mechanisms by which macroeconomic policies influence poverty rates cannot be generalized. Growth and macroeconomic stability are complements for poverty reduction. Monetary poverty (poverty lines) Marquette (1997) Assessment of an economical structural adjustment program (ESAP) regarding current and future poor. General poor in countries where the government has strong market control. X X ESAP programs are essential for economies strongly dependent on government control. ESAP, even if raising current poverty, will bring sustainable growth, and will be essential to reduce poverty in the future. Multidimensional poverty (measures of physiologic al and social deprivation) Ames et al. (2001) Study of the relations between macroecono mic stability, growth, and poverty reduction. General poor; general to all countries. X X X X X X X X X Growth is the most important factor influencing poverty. However, before fighting poverty, policy should first focus on macroeconomic stability. Growth impacts on poverty depend essentially on distributional patterns and sector composition. Multidimensional poverty (measures of physiologic al and social deprivation) Macroeconomic Fundamentals of Poverty and Deprivation 48 Paper Research Aims Scope Real (per capita) GDP growth rate Inequality indicators Foreign aid Human capital Public capital Inflation Exchange rate Aggregate demand Trade liberalization Legal framework, economic and political organization and rules Main results Definition of poverty (indicator) DeFina (2002) Methods to measure poverty, able to capture how aggregate economic conditions may have different effects on poverty reduction. General poor, but some other classes are also studied. United States. The way aggregate labor market changes affect poverty depends on how poverty is measured. The precise definition of poverty is critical to choose whether and how changes should be conducted. Multidimensional poverty (measures of physiologic al and social deprivation) Gundlach and Paldam (2009) Assessment of the longrun causality direction between income and corruption. General poor; general to all countries. X Corruption vanishes for higher levels of development. As a country develops there is a changeover from poverty to honesty. (implicit) Multidimensional poverty Dobson and RamloganDobson (2009) The relation between corruption and inequality. General poor; South American countries where the informal sector represents 25% to 35% of output. X X While conventional literature points supports the idea that if corruption decreases, income inequality will also decrease, the study shows that this is not actually true for countries where there is a large informal sector. Multidimensional poverty (measures of physiologic al and social deprivation) Tarabini (2010) The role of education and poverty in the current global developmen t agenda. General to developing countries; general poor. X X X Economic growth is insufficient for poverty reduction. Education is essential to fight poverty and, thus, a strong investment in education should be a priority in national development strategies. Multidimensional poverty (measures of physiologic al and social deprivation) Macroeconomic Fundamentals of Poverty and Deprivation 49 Paper Research Aims Scope Real (per capita) GDP growth rate Inequality indicators Foreign aid Human capital Public capital Inflation Exchange rate Aggregate demand Trade liberalization Legal framework, economic and political organization and rules Main results Definition of poverty (indicator) Tebaldi and Mohan (2010) The relevance of the quality of institutions to fight poverty. General poor; countries with bad quality institutions. X X X Poverty is directly linked with bad quality institutions. The other mechanisms (aid, etc.), will influence poverty only if good quality of institutions prevails: an effective government, a stable political system, and control on corruption. Multidimensional poverty (measures of physiologic al and social deprivations ) Macroeconomic Fundamentals of Poverty and Deprivation 50 Annex B: Transformed ranking per deprivation dimension Table B.1: Income deprivation Table B.2: Employment deprivation GEO/TIME 2005 2006 2007 2008 Czech Republic 10.0 5.9 7.2 5.9 Denmark 1.7 1.8 2.2 1.8 Germany (including former GDR from 1991) 3.5 11.0 22.8 17.4 Estonia 19.1 11.0 17.8 11.0 Greece 34.4 26.4 37.9 26.4 Spain 28.1 17.4 29.1 21.5 France 15.6 11.0 13.7 11.0 Cyprus 28.1 21.5 37.9 32.8 Latvia 43.0 100.0 100.0 100.0 Lithuania 100.0 55.7 52.4 55.7 Luxembourg 7.6 3.8 7.2 3.8 Hungary 23.2 32.8 37.9 32.8 Austria 5.5 3.8 4.5 5.9 Poland 57.1 41.5 52.4 41.5 Portugal 12.6 8.3 10.2 14.0 Slovenia 15.6 11.0 10.2 17.4 Slovakia 23.2 14.0 13.7 8.3 Finland 3.5 8.3 10.2 14.0 GEO/TIME 2005 2006 2007 2008 Czech Republic 17.4 17.4 17.8 11.6 Denmark 11.0 8.3 7.2 5.1 Germany (including former GDR from 1991) 55.7 41.5 52.4 50.4 Estonia 26.4 8.3 7.2 8.1 Greece 41.5 26.4 29.1 27.0 Spain 5.9 3.8 4.5 20.6 France 26.4 32.8 37.9 35.9 Cyprus 1.8 1.8 2.2 2.4 Latvia 21.5 8.3 7.2 11.6 Lithuania 14.0 11.0 7.2 20.6 Luxembourg 3.8 5.9 4.5 15.6 Hungary 32.8 32.8 52.4 100.0 Austria 11.0 11.0 7.2 8.1 Poland 100.0 100.0 100.0 35.9 Portugal 8.3 8.3 13.7 15.6 Slovenia 11.0 14.0 10.2 11.6 Slovakia 55.7 55.7 52.4 35.9 Finland 17.4 21.5 22.8 15.6 Macroeconomic Fundamentals of Poverty and Deprivation 51 Table B.3: Health deprivation and disability Table B.4: Education, skills and training deprivation GEO/TIME 2005 2006 2007 2008 Czech Republic 5.1 3.8 4.5 5.9 Denmark 27.0 32.8 37.9 32.8 Germany (including former GDR from 1991) 2.4 1.8 2.2 1.8 Estonia 50.4 41.5 52.4 41.5 Greece 8.1 8.3 10.2 8.3 Spain 27.0 26.4 29.1 21.5 France 15.6 14.0 13.7 17.4 Cyprus 11.6 11.0 13.7 14.0 Latvia 20.6 17.4 17.8 17.4 Lithuania 35.9 55.7 52.4 41.5 Luxembourg 8.1 21.5 37.9 11.0 Hungary 20.6 14.0 29.1 26.4 Austria 5.1 3.8 4.5 3.8 Poland 27.0 26.4 22.8 17.4 Portugal 27.0 26.4 29.1 26.4 Slovenia 100.0 100.0 100.0 100.0 Slovakia 11.6 5.9 7.2 14.0 Finland 27.0 26.4 37.9 55.7 GEO/TIME 2005 2006 2007 2008 Czech Republic 2.0 1.6 2.0 1.6 Denmark 15.6 9.2 19.7 20.6 Germany (including former GDR from 1991) 24.7 17.2 12.2 11.6 Estonia 12.2 9.2 12.2 14.2 Greece 39.8 44.5 39.8 44.5 Spain 54.1 58.5 54.1 58.5 France 24.7 29.6 31.0 29.6 Cyprus 39.8 35.9 24.7 29.6 Latvia 24.7 20.6 19.7 24.7 Lithuania 9.2 5.1 6.5 7.0 Luxembourg 31.0 35.9 31.0 35.9 Hungary 24.7 24.7 15.6 17.2 Austria 12.2 11.6 9.2 9.2 Poland 6.5 3.2 4.1 3.2 Portugal 100.0 100.0 100.0 100.0 Slovenia 6.5 7.0 4.1 7.0 Slovakia 4.1 3.2 4.1 5.1 Finland 19.7 14.2 9.2 9.2