«Ravi Kanbur and Lyn Squire• September, 1999 • Ravi Kanbur is T.H. Lee Professor of World Affairs at Cornell University and Director of the World ...»
THE EVOLUTION OF THINKING ABOUT POVERTY:
EXPLORING THE INTERACTIONS
Ravi Kanbur and Lyn Squire•
Ravi Kanbur is T.H. Lee Professor of World Affairs at Cornell University and Director of the World
Development Report at The World Bank; Lyn Squire is Director of the Global Development Network, at The
World Bank in Washington, DC.
Noémi Giszpenc provided excellent research assistance. The authors would like to thank the participants at the Symposium on “Future of Development Economics in Perspective” held in Dubrovnik, 13-14 May, 1999 for their many helpful comments and suggestions.
THE EVOLUTION OF THINKING ABOUT POVERTY:
EXPLORING THE INTERACTIONS
THE EVOLUTION OF THINKING ABOUT POVERTY:
EXPLORING THE INTERACTIONSRavi Kanbur and Lyn Squire September, 1999
I. INTRODUCTIONEradicating, or at least reducing, poverty lies at the heart of development economics.
While development seeks to benefit all members of society, the poor command our special attention. Any reasonable definition of poverty implies that significant numbers of people are living in intolerable circumstances where starvation is a constant threat, sickness is a familiar companion, and oppression is a fact of life. In Alfred Marshall’s words, “The study of the causes of poverty is the study of the causes of degradation of a large part of mankind.”1 Improving the lives of the poor must be at the top of our agenda.
What does development economics have to say about reducing poverty? And how has thinking evolved overthe last quarter of a century? This paper explores these issues through the evidence and analysis available in the literature in general and through the World Bank’s World Development Reports on poverty.2 These reports, drawing on evidence from around the world, summarize contemporary thinking on the subject and are therefore useful instruments for assessing progress in our understanding of the problem and our ability to solve it.
The breadth of the topic requires some selectivity. We focus on two questions. How should poverty be defined and measured? And, what policies and strategies reduce poverty so defined? The questions are of course related – the definition of poverty drives the choice of policies. To organize the discussion, we take advantage of the fact that the definition of poverty has broadened over the last quarter of a century, and, as it has broadened, so the relevant set of policies has expanded. Beginning with a focus on command over market-purchased goods (income), the definition of poverty has expanded to embrace other dimensions of living standards such as longevity, literacy, and healthiness. And, as we learn more about and from the poor, the concept has developed further to reflect a concern with vulnerability and risk, and with powerlessness and lack of voice. Our review of the evolution of thinking about poverty leads us to two broad conclusions.
First, broadening the definition of poverty does not change significantly who is counted as poor – at least not as far as aggregate measures are concerned. While this is a simplification, and we can find evidence to the contrary in the literature, it reflects the fact that the many aspects of poverty – income, health, political rights, and so on – are often closely correlated. While aggregate measures may be largely unaffected, the broader definitions allow a better characterization of poverty and the terrible hardships burdening the poor, and therefore increase our understanding of poverty and the poor. This deeper understanding will often be critical to the design and implementation of specific programs and projects to help people escape poverty.
Marshall, A. “Principles of Economics”, p3, eighth edition, 1925.
World Development Reports, 1980 and 1990. A new report on poverty will be issued in 2000.
Second and related, broadening the definition of poverty changes significantly our thinking about strategies to reduce poverty. In part this is obvious. A broader definition naturally expands the set of policies that are relevant to the reduction of poverty. As more aspects of poverty are recognized, so more policies become relevant to fighting poverty – moving beyond income to include health, for example, introduces a new set of policy instruments. But, there is another, more subtle and important consequence. The various aspects of poverty interact in important ways, such that policies do more than simply add up – for example, improving health increases income-earning potential, increasing education leads to better health outcomes, providing safety nets allows the poor to take advantage of high-return, high-risk opportunities, and so on. Poverty-reducing strategies must recognize the interactions among policies – the impact of appropriately designed combinations will be greater than the sum of the individual parts. The interaction between the various dimensions of poverty is a theme that runs throughout this review.
These two issues – definition and strategy – are explored in more depth in the following sections as we trace the historical evolution of poverty through its various manifestations.
Section two looks at the definition of poverty as it emerged from the pioneering efforts of Rowntree at the turn of the century. His focus on income (or expenditure) led naturally to a strategy based on growth in national income. But growth in national income will only help the poor if they share in that growth. The key interaction then was the link between growth in national income and changes in inequality, and a fear that progress on one front (growth) would lead to setbacks on the other (inequality) with uncertain implications for the poor.
Section three explores the incorporation in the 1980s of other dimensions of poverty – longevity, literacy, healthiness. This brought new policies into play but also revealed two new interactions. One was within the new set of policies – healthier children perform better at school; better educated mothers have healthier families and so on. And the other was between progress in human development and increases in national income. From one perspective better health and education can be viewed as an investment in human capital that, like investment in physical capital, should yield a return in the form of increased income. From another perspective better health and education can be seen as improvements in the quality of life in their own right.
Indeed, in this view growth in national income only has value to the extent that is leads to longer lives, better health, and greater literacy. Either way, there is an interaction between the two sets of outcomes.
In section four we incorporate the findings from the analysis of panel data and from a range of participatory techniques that have recently come to the fore and which seek to elicit views about poverty from the poor themselves. This has led to today’s concern with risk and vulnerability, powerlessness, and lack of voice. Here too important interactions emerge.
Reducing exposure to risk offers an immediate benefit to the otherwise vulnerable but it also provides a platform to escape long-run poverty – lower exposure to risk frees the poor to engage in riskier, but more profitable, production and investment strategies including investing in their children’s education. Similarly, giving the poor voice reduces their sense of isolation, an immediate benefit. But once they have a greater say in the selection and design of programs to assist them, then they are also more committed to implementation. In direct opposition to the isolation the poor often endure, successful implementation of income-earning projects, health programs, and safety nets calls for inclusion and active participation in a wide range of circumstances.
We conclude in section five by restating the central proposition of this review. As the definition of poverty has expanded and new dimensions have been introduced, the degree of interaction among the elements has also increased because each element contributes to wellbeing in the broad but also contributes to the achievement of other elements. With this perspective as background, section five offers some views on the most important outstanding issues in need of further research.
II. INCOME AND CONSUMPTIONConventional Measures According to normal usage poverty is “The state of one who lacks a usual or socially acceptable amount of money or material possessions.”3 This definition contains two important ideas. First, the definition of poverty will be different at different times and in different societies – what is “socially acceptable” in, say, India may differ from that in the U.S.A. And second, the focus is on the ability to purchase goods and services (money) or on their ownership (material possessions). As we shall see, many attempts to measure poverty incorporate these two ideas.
Benjamin Seebohm Rowntree, an early measurer of poverty, arrived at a “socially acceptable” amount of money by estimating the budget required “to obtain the minimum necessaries for the maintenance of merely physical efficiency”4 appropriate to the specific circumstances of the City of York at the turn of the century. Based on the nutritional content of various foods and their local prices, Rowntree concluded that fifteen shillings would provide the minimum budget for food for a family of six for one week. Adding an allowance for shelter, clothing, fuel, and other sundries, he arrived at a poverty line of twenty-six shillings for a family of six that implied a poverty rate of almost 10 percent in York. While this same approach has been used in other countries and other times, the resulting poverty line is sensitive to local circumstances.
Thus, the Indian and U.S. poverty lines are based more or less on this approach, yet when both are expressed in 1985 PPP dollars, the latter is ten to twenty times as large as the former, depending on the size of the household. This difference reflects the tendency for poverty lines to change over time within countries as average incomes rise and views about the “minimum necessaries” evolve. A study of poverty lines – budgets for “minimum subsistence” – used in the U.S.A. in the period 1905 to 1960 found that they rose 0.75 percent in real terms for each 1.0 percent increase in the real disposable income per capita of the general population.5 Thus, compared with the absolute poverty threshold established by Mollie Orshansky for 1963, minimum subsistence budgets before World War I were, in constant dollars, between 43 and 54 percent as large; by 1923, a “minimum subsistence level” was 53 to 68 percent as large. An “emergency” budget during the Depression year of 1935 was 65 percent, and a low-income line for 1957 was 88 percent of Orshansky’s threshold.
Merriam-Webster’s Collegiate Dictionary (1995).
Rowntree, 1910, p.86.
Cited in Fisher, 1996.
That popular conceptions of the amount needed to “get along” rise with increases in overall incomes is neatly captured in a quotation from 1938: “A standard budget worked out in the [1890’s], for example, would have no place for electric appliances, automobiles, spinach, radios, and many other things which found a place on the 1938 comfort model. The budget of 1950 will undoubtedly make the present one look as antiquated as the hobble skirt.”6 As technology progresses and the general standard of living rises, three effects have an impact on poverty: new consumption items, initially viewed as luxuries, come to be seen as conveniences and then as necessities; changes in the way society is organized may make it more expensive for the poor to accomplish a given goal – say, when widespread automobile ownership leads to a deterioration in public transportation; and finally, general upgrading of social standards can make things more expensive for the poor, as when housing code requirements that all houses have indoor plumbing add to the cost of housing.
Once a poverty line has been established, it could be applied to data on incomes or on expenditure. Most analysts favor expenditure. In many cases, expenditure is far easier to measure. It also has a conceptual advantage. If incomes vary over time in fairly predictable ways (as they are likely to do in a rural economy, say), households can to some extent smooth their living standards from income variability. Anand and Harris (1994) address the choice of a welfare indicator using data from Sri Lanka. They conjecture, and find, that income is a noisy indicator of “permanent” income, while household total expenditure per capita is less noisy and thus preferred.7 Within this broad approach, many attempts have been made to improve estimates of poverty lines and overcome a host of conceptual and empirical difficulties.8 The value of this effort depends on the use to be made of poverty lines and hence the required level of precision.
We discuss and judge some of these efforts from the perspective of two possible uses of poverty lines – as a means of measuring poverty worldwide and monitoring changes over time, and as a means of designing specific actions oriented towards the poor.