Part Two
Problems and Policies: Domestic
Chapter 5
Poverty, Inequality, and Development
Key Concepts
This chapter takes up the question of growth versus income distribution and poverty. The chapter examines
eight questions:
What is the extent of relative inequality and poverty in developing countries?
What are the economic characteristics of the poor?
Who benefits from economic growth?
Are increased economic growth and more equitable income distributions compatible objectives?
Do the poor benefit from growth?
What is so bad about extreme inequality?
What policies will reduce absolute poverty?
5.1 and 5.2 deal with measurement issues relating to poverty and inequality. The concepts of size and
functional distribution of income are defined. Within this section, the following topics are covered:
A discussion of Lorenz curves and the Gini coefficient.
Measuring absolute poverty using the poverty gap (total and average poverty gap), the
Foster-Greer-Thorbecke index, and the newly introduced Multidimensional Poverty Index.
An important conclusion is that there is no clear relationship between the level of income per capita and
inequality.
5.3 seeks to explain:
Why inequality is a problem.
Three cases of dualistic development using Lorenz curves.
The relationship between inequality and per capita income or the Kuznets inverted-U hypothesis.
The relationship between growth and inequality.
The new edition includes updated data tables for various countries. Evidence suggests it is the character
of economic growth that determines how much living standards improve as growth increases.
5.4 takes up the topic of absolute poverty. The new edition adds and explains the concept of extreme
poverty and chronic poverty. New and more extensive data are presented on the degree of absolute poverty
in the world today, and the question of growth versus equity is debated in terms of which to pursue first.
Arguments are made for each case and include the following points:
Economic growth is not sufficient to eradicate poverty because it may benefit only the modern sector
with little or no trickle-down. The rich may not use their wealth to make productive local investments.
Investing in health and education improves worker productivity. Increasing the income of the poor
may be translated into demand-led growth.
Pro-growth arguments emphasize more savings by the higher income groups and the trickle-down theory.
In 5.5, the economic characteristics of the poor are described. It is argued that poverty is disproportionately
present among the rural population, women and minority/indigenous groups.
5.6 considers policy options for addressing poverty and inequality while still maintaining growth:
Remove factor price distortions. This topic is discussed further in Appendix 5.1.
Redistribution of asset ownership, such as land reform. This topic is discussed further in Chapter 9.
Progressive income and wealth taxes to reduce income inequality.
Direct transfer payments and public provision of goods and services to reduce the extent of poverty.
The price-incentive model, discussed in Appendix 5.1, suggests firms are profit maximizers who choose
the combination of inputs that will minimize cost. The many factor price distortions present in developing
countries send the wrong signal to firms in terms of choice over labor and capital. Graphical analysis using
isoquants and isocost lines is included. The appropriate technology concept is introduced and suggests that
developing countries that are relatively well endowed with labor should try to employ labor-intensive
technologies. The employment effect of removing factor price distortions depends on the elasticity of
substitution.
The Ahluwalia-Chenery Welfare Index is explained in detail in Appendix 5.2. GNP is shown to be a biased
indicator of development and welfare, and equal-weighted and poverty-weighted measures are
demonstrated to be better indicators of who is benefiting from the growth of production.
The chapter ends with a case study that demonstrates how the frameworks and multi-country statistical
studies of Chapters 2 and 5 can be applied to understand development and experiences in a comparative
perspective.
Lecture Suggestions
This is the beginning of Part 2 of the text that focuses on quality of life and equity issues. In particular the
authors stress again that growth can occur without a corresponding improvement in the living standards of
the entire population.
Income distribution data can be used to evaluate the effect of growth on welfare. The level of GNI per
capita is not a good indicator of the benefits of growth in terms of economic welfare. Topics to discuss in
this section include:
Size and functional distribution of income.
Lorenz curve and Gini coefficient.
Equal and poverty weighted measures of growth (Ahluwalia-Chenery Welfare Index, ACWI).
Reference to the Gini coefficient for different countries will show that there is no clear relationship
between income per capita level and inequality and that the type of growth pursued in terms of capital or
labor intensity can have a large impact on inequality. You might compare Brazil, Mexico, Taiwan, and
South Korea.
At this point, it would be a good idea to consider using labor market diagrams in figure 5.5 to demonstrate
how the impact of growth on the functional distribution of income hinges crucially on the relative
elasticities of labor demand and labor supply. For example, when the labor demand curve is relatively
inelastic a given increase in labor supply will lower real wage income’s share of increasing total income.
The downward pressure on real hourly wages overpowers the relatively small increase in hours of labor
demanded so real wage income in fact falls as total real income rises. In anticipation of the discussion of
asset ownership (pages 258-260) labor market diagrams again can be used to show inequality in asset
ownership, via capital market imperfections, can easily lead to growing and persistent inequality in the
size distribution of income. This can be done in anticipation of the latter discussion of asset ownership
(pages 258-269). Here you would want to draw sets of two labor market diagrams –one for unskilled labor
and the other for skilled with the labor demand curve much lower for unskilled labor and the supply curve
for unskilled labor much further to the right. Then after shifting the demand for skilled labor curve to the
right discuss what should happen as a result of rising real hourly wages in this segment of the market and
how capital market imperfections could be used to explain the persistence of the now larger gap between
the earnings of skilled and unskilled labor.
Stress that Lorenz curves and Gini coefficients measure relative income as opposed to absolute income.
Neither says anything about the extent of absolute poverty. When making policy decisions, it is important
to look at what is or is not happening to relative income and absolute poverty.
Examining the different cases of dualistic development can be an excellent teaching tool. Following the
Fields model, it is useful to set up numerical examples of stylized dual economies. One example of a dual
economy is to assume there are 10 people, five of whom are rich (each with $2) and five of whom are poor
(each with $1). The students can draw the Lorenz curve, and then illustrate how the curve shifts when
there is modern sector enrichment (those with $2 incomes get $3), traditional sector enrichment (those
with $1 get $1.25), and modern sector enlargement (one person with $1 now gets $2). The Fields’ book
(cited in footnote 17) as well as his Distribution and Development (cited in Further Reading) are excellent
sources of additional lecture material.
Tables in the text which illustrate poverty, growth and income distribution data for different countries,
combined with data on poverty from the 2010 World Development Indicators Report, can be used to
summarize some important propositions about growth, poverty, and income distribution, such as:
The Kuznets inverted-U hypothesis is a relatively weak relationship in that there are many countries
whose growth experience does not follow this pattern.
High inequality is neither necessary nor sufficient for high growth.
High growth is neither necessary nor sufficient for reductions in absolute poverty.
This helps emphasize that it is not just the growth process but also the character of growth that largely
determines the incidence of poverty and the extent of inequality. The character of growth can be identified by
the policy choices made by each country. These policy choices will be identified in later chapters.
When you cover the Ahluwalia-Chenery Welfare Index, it is advisable to allocate adequate lecture time
because for students to truly appreciate the importance of this index it is best for them to work through a
few examples themselves.
You might want to put together a worksheet where you have a fifteen-person economy and, for simplicity’s
sake income divided into terciles. Five have real per capita income of $1000 a year, five have real per
capita income of $2,000 while the remaining five have real per capita income of $3,000 a year. Ask first
what GNI would be in this example. Then ask what share of GNI goes to first the lowest tercile, then the
middle tercile and finally to the top tercile (16.67%, 33.33% and 50% respectively). Using these as
weights for equation A5.2.1 (see page 275) have them calculate G first for g3=50% and then for g1=50%.
Ask how these results confirm the statement made on page 276 that “the welfare weights attached to the
growth rates of different income groups are unequal, with a heavy social premium being placed on the
income growth of the highest quintile (here tercile) group”. This followed by attaching progressively
higher weights to the bottom tercile and carefully interpreting the results can be followed by a discussion
of the calculations found in Appendix 5.2 and the results found in table A5.2.1. It would be well worth the
time to find poverty-weighted ACWIs with more current numbers than those given in the table.
The growth versus economic welfare debate is related to the dilemma of how to provide basic human
needs without sacrificing growth. Policies to reduce income inequality can be discussed (increase
employment, improve education, redistribute assets, tax the wealthy, subsidize the poor) as can rural
development programs such as food for work programs.
For instructors who wish to emphasize the measurement of poverty, its global distribution and policies
aimed at alleviating its burden, the World Bank devotes the whole of its World Development Report to
this issue on occasions. The most recent Report devoted to this theme is the 2010 issue and contains a
lot of valuable information on this topic.
The price incentive model can be taught with or without graphs and equations. Basic supply and demand
arguments can convey the main idea of the model. Common factor price distortions can be discussed
(minimum wage, unions, multinational corporation hiring practices, low interest rates). Factor prices have
been distorted with good intentions, but poor results. The adoption of inappropriate technology can be
reviewed. The material in Appendix 5.1 is more technical and can be omitted without any loss of continuity.
This chapter lends itself well to numerical problems that clarify the main points. Some of these can be done
in class or for homework. Similar questions can be asked on exams. Here are three examples to draw from.
1. The following income distribution data are for Brazil.
Quintile Percent Share
Lowest 20% 3.0%
Second quintile 6.9%
Third quintile 11.8%
Fourth quintile 19.6%
Highest 20% 58.7%
Highest 10% 43.0%
(a) Carefully graph the Lorenz curve, labeling the axes.
Answer: The cumulative distribution by quintiles is 3.0, 9.9, 21.7, 41.3, and 100.
(b) Explain how to find the Gini coefficient, graphically.
(c) Brazil’s national income is about $300 billion. What is the approximate dollar income of the
bottom 20%? Bottom 40%?
Answer: $ 9 billion and $ 29.7 billion.
(d) Brazil’s population is approximately 150 million. Suppose that each household makes the average
income for its quintile. What is the level of poverty if the poverty line is $400 per capita?
Answer: 1st quintile: $300, below poverty line; 2nd quintile: $ 590 above poverty line; headcount
(e) Suppose one percent of national income were transferred from the richest 20% of households to
the poorest 20% of households. Show the effect on relative inequality.
Answer: Decreases by Lorenz criterion.
(f) Under the same transfer, what is the effect on poverty?
Answer: Poorest quintile receives $12B. H 30 but income shortfall drops to $ 0B. You may
Analogous questions can easily be constructed for other countries using the World Development
Report, 2004, Table 2.
2. Consider the following distribution of income in a 12-person economy, with the modern urban
wage 3, the traditional rural income 1, and the informal urban wage 2: (1,1,1,1,2,2,2,2,3,3,3,3).
The poverty line 1.25. Suppose rural incomes are raised to 1.5 through expanded agricultural exports.
What happens to relative inequality? Absolute poverty? Calculate the Ahluwalia-Chenery welfare index
for terciles (3 fractiles) under GNP weights and equal weights. Why is the equal-weighted index higher?
Answer: Inequality decreases. Poverty is eliminated. ACWI 0.083 with GNP weights. ACWI
3. Consider the following distribution of income in a 12-person economy, with the modern urban wage 10,
the traditional rural income 2, and the informal urban wage 4: (2,2,2,2,4,4,4,4,10,10,10,10).
(a) Suppose rural incomes are raised to 3, other things equal. Show the change using the Lorenz
curve. How would this type of growth be characterized? (It is suggested that you use the Fields
growth decomposition in answering. This shows traditional sector enrichment growth.)
(b) Calculate the Ahluwalia-Chenery welfare index for terciles (3 fractiles) under GNP weights and
equal weights. Which index is higher and why?
Answer: (GNP is 64, so the index with GNP weights is: (1/2) (1/8) (1/16). The index with equal
Discussion Topics
Discuss the relative importance of income inequality and absolute poverty as problems that need to be
solved. Present data from the developed countries for comparison. Discuss policy options and the
opportunity cost of devoting resources to such policies.
Sample Questions
Short Answer
1. About how many people are living in extreme poverty in the world? Which regions have the greatest
percentages of people living in extreme poverty?
Answer: There are various estimates, including the World Bank estimate of 1.4 billion people living
in extreme poverty in 2005. Percentage wise, 40.3% of those who reside in South Asia,
2. Explain the difference between size and functional measures of income distribution.
Answer: Size refers to family or personal income and function refers to factor shares.
3. What is a Lorenz curve? Draw one, labeling the axes.
Answer: Examined in the text.
4. Explain why a Lorenz curve can never lie to the left or above the 45 degree line.
Answer: It is constructed by ordering income recipients from poorest to richest. Note: if ordered
5. Using a Lorenz curve diagram, explain how to calculate the Gini coefficient.
Answer: Examined in the text.
6. Distinguish between economic growth by modern sector enlargement, modern sector enrichment and
traditional sector enrichment. For each of the three cases, show what happens to absolute poverty and
to relative inequality, using a precise measure of each.
Answer: Students may use either the headcount or the income shortfall measure for absolute
7. Generally speaking, higher income countries tend to have less income inequality than low-income
countries, however this does not always hold true. What could cause a low-income country to have
low-income inequality?
Answer: Low inequality can result due to very little income to be distributed in the first place.
8. Discuss the benefits of using a “poverty weighted” index of GNP growth as a measure of social
welfare as opposed to using the growth in GNP to measure change in social welfare.
Answer: Students should refer specifically to the Ahluwalia Chenery welfare index. Poverty weights
9. Increasing GNP is a necessary but not a sufficient condition for improving living standards in less
developed countries. True or false, explain.
Answer: True. Growth in income is necessary but it does not guarantee that all will benefit.
10. “The 1980s were a lost decade for the absolutely poor.” Evaluate this statement.
Answer: Answers will vary. You might expect some presentation of data.
11.What are the characteristics of the poor (be specific)?
Answer: A person is more likely to be absolutely poor if rural, agricultural, female or from a
female-headed family, landless, from a large family, from a family with a high ratio of
12. Why are about 100 million girls and women said to be “missing” in developing countries?
Answer: This is the number of females that would be proportional to the number of living males if
13. What are factor price distortions, and what are their major causes in developing countries?
Answer: Factors of production are paid prices that do not reflect the prices that would prevail under
a true competitive market, that is, their scarcity prices. Prices for capital and intermediate
14. Just what is it that makes a technology “appropriate” for a developing country?
Answer: A technology is appropriate if it fits with existing factor endowments.
15. Discuss the reasons why developing countries so often use a technology with an excessively high
capital labor ratio.
Answer: Factor price distortions and the unavailability of appropriate technology.
16. Discuss some policies or strategies that would give firms an incentive to substitute labor for capital
within modern production techniques.
Answer: Depends on lecture development.
17. The 2002 World Development Report provides the following information for Colombia and Thailand:
Per Capita
GDP ($)
Per Capita
GDP
(PPP $)
Population
below
National
Poverty Line
(%)
Population
below
$1 Day PPP
(%)
Poverty
Gap at
$1 Day
PPP (%)
Gini
Index
Colombia 2080 5890 17.7 11.0 3.2 57.1
Thailand 2010 6330 13.1 <2 <0.5 41.4
(a) Explain carefully what each of the entries in the final four columns of this table measures. What
concepts are being presented and what is their importance to economic development?
Answer: The student is asked to define and comment on the various ways of measuring poverty
(b) The table above shows that Colombia and Thailand are roughly at the same level of economic
development as measured by the level per capita national income. Their performance in terms of
poverty indicators is quite different. What factors may contribute to their differing performance?
Answer: The student is asked to comment on the distinction between per capita income growth and
18. Compare and contrast the similarities and the differences in the development experience of Ghana
and Cote d’Ivoire. How did the economic development policies followed by the two countries differ?
Answer: The case study on Ghana and Cote D’Ivoire discusses these issues in detail.
19. What is the Multidimensional Poverty Index and how is it a more improved measure of poverty?
Answer: The MPI identifies the poor at the household level and uses the dual cutoff method. First,
the cutoff levels within each of the dimensions and second, the cutoff of the number of
Multiple Choice
1. The absolute poverty line
(a) decreases as real income grows.
(b) shows the average income of the lowest income group.
(c) can be measured with the Lorenz curve.
(d) none of the above.
2. The Gini coefficient provides a measure of
(a) the level of poverty.
(b) the level of relative inequality.
(c) disguised unemployment.
(d) the rate of growth.
3. Kuznets’ inverted-U hypothesis
(a) implies that things must get worse before they get better.
(b) suggests that inequality will worsen and then improve as a country grows.
(c) suggests that inequality will improve and then worsen as a country grows.
(d) points out six characteristics of modern economic growth.
4. According to Kuznets, in the process of development inequality in an economy will normally
(a) first rise and then fall.
(b) first fall and then rise.
(c) remain about the same.
(d) show no definite pattern.
5. Poverty is better studied with size distribution measures than those based on factor distribution because
(a) labor income may be highly concentrated in well-paid modern sector workers.
(b) some poor farmers may receive a sizable share of income in rent.
(c) income from nonmarket activities such as foraging may be important.
(d) all of the above.
6. The number of people in the world who are absolutely poor is closest to
(a) a quarter-billion.
(b) a half-billion.
(c) one and a half billion.
(d) two billion.
(e) four billion.
7. With modern sector enrichment growth, inequality will
(a) first rise and then fall.
(b) first fall and then rise.
(c) remain about the same.
(d) none of the above.
8. With modern sector enlargement growth, inequality will
(a) first rise and then fall.
(b) first fall and then rise.
(c) remain about the same.
(d) all of the above.
9. Higher income countries tend to have lower levels of absolute poverty because
(a) more employment opportunities
(b) more public assistance
(c) greater entrepreneurship opportunities.
(d) all of the above.
10. One of the characteristics of the poor is that they are
(a) more likely to be employed in the modern industrial sector.
(b) more likely to come from small families.
(c) more likely to be well educated.
(d) more likely to live in a rural area.
11.Which of the following policies might increase labor intensity in industry?
(a) a decline in the cost of credit
(b) a decline in the minimum wage rate
(c) a decline in the elasticity of substitution
(d) all of the above.
12. Which of the following policies may decrease the level of capital intensity in industry?
(a) an increase in the cost of capital
(b) a decrease in the minimum wage
(c) an increase in the elasticity of substitution
(d) all of the above.
13. The Ahluwalia-Chenery welfare index
(a) is a method used to measure changes in absolute poverty.
(b) shows the value judgment implications of using the change in income per capita as a measure of
the change in development.
(c) is a method used to measure changes in inequality.
(d) is a method used to measure the growth rate of GDP.
14. About what percent of the world’s poorest people are female?
(a) 30
(b) 50
(c) 70
(d) 90
15. About how many girls and women are said to be “missing” in LDCs?
(a) 2 million
(b) 20 million
(c) 100 million
(d) 2 billion
16. Which of the following groups is(are) more likely to be poor?
(a) minorities
(b) indigenous people
(c) women
(d) all of the above.
17. Distribution of income according to percentiles, such as the highest 40% or lowest 20% is known as
the _______________ distribution of income.
(a) size
(b) functional
(c) GNP-weighted
(d) equal-weighted
18. What conclusion can be reached from the following data on income shares?
Percentage of Income Received
by
Lowest 40% Highest 20%
Bangladesh 17.3 45.3
Indonesia 14.4 49.4
(a) absolute poverty is more widespread in Bangladesh
(b) the size distribution of income is more unequal in Indonesia
(c) Bangladesh had adopted a strategy of redistribution with growth
(d) growth in Bangladesh is calculated using poverty weights rather than income weights
19. Developing countries who have adopted capital-intensive technologies tend to have
(a) relatively higher Gini coefficients.
(b) relatively lower Gini coefficients.
(c) Gini coefficients equal to one.
(d) Gini coefficients equal to zero.
20. Assuming that the Gini coefficient for Egypt is 0.403 and the Gini coefficient for Australia is 0.404, it
is possible to conclude that both Egypt and Australia have
(a) virtually the same number of households in absolute poverty.
(b) virtually the same percentage of households in absolute poverty.
(c) virtually the same level of the Human Development Index.
(d) none of the above.
21. Brazil’s growth rate during the 1960’s was 6.0% when poverty weights were used to evaluate growth,
compared with 8.2% when GNP weights were used to evaluate growth. One can conclude from these
numbers that
(a) average income growth was greater for poor households than for rich households.
(b) average income growth was greater for rich households than for poor households.
(c) more and more households were falling below the poverty line.
(d) the size distribution of income was getting worse.
22. The poverty gap is the
(a) absolute number of people below the international poverty line.
(b) percentage of the population below the international poverty line.
(c) consumption (measured in dollars) necessary to bring everyone below the poverty line up
to the line.
(d) percentage of a country’s total consumption necessary to bring everyone in the country below
the poverty line up to the line.
23. The functional distribution of income refers to the distribution of income between
(a) individuals or households.
(b) rural individuals or households.
(c) urban individuals or households.
(d) the factors of production (land, labor and capital).