Measuring Poverty

К оглавлению
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 
17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 
34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 
51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 
68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 
85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 
119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 
136 137 138 139 140 141 142 

The standard approach is to establish a poverty line, normally refl ecting a

minimum necessary standard of living (or that adequate for a minimum

calorie intake), and to identify who falls below this line. Establishing the

poverty line can be complex, and the Appendix to Chapter 4 discusses the

experience in PRC, where there has been considerable discussion about the

level and trend in the poverty line. Once such a line is available there are

alternative ways of quantifying the degree of poverty. Of these the simplest

and most widely cited is the ‘headcount index’, which gives the proportion

of the population below the poverty line.

For our country cases Figures 1.5 to 1.9 show the offi cial estimates of the

poverty headcount (proportion of the population below the offi cial national

poverty line) for each country.3 As they are based on different poverty lines

the estimates are not directly comparable across countries. They show that

poverty remains very high in India and the Philippines and by offi cial fi gures

is now very low in PRC (although as we have noted the accuracy of the

offi cial poverty line used in PRC is widely disputed) (Riskin et al., 2001). In

all countries however there is a downward trend in poverty estimates and it

is the role of targeting programs in this process that we examine.

However, more sophisticated indicators are also available and are drawn

on in the country studies. These aim to assess the depth of poverty (that is

0

5

10

15

20

25

30

1990 1992 1994 1996 1998 2000 2002

Thailand

Figure 1.5 Poverty headcount (%): Thailand

0

10

20

30

40

50

60

1977–78 1983 1986–87 1989–90 1992 1993–94 1999–

2000

India

Figure 1.6 Poverty headcount (%): India

0

5

10

15

20

25

30

1996 1998 1999 2000 2001

Indonesia

Figure 1.7 Poverty headcount (%): Indonesia

0

5

10

15

20

25

30

35

40

1991 1994 1997 2000

Philippines

Figure 1.8 Poverty headcount (%): Philippines

how far on average the poor are from the poverty line) and the severity of

poverty (that is the distribution of income or consumption within the group

of the poor). The depth of poverty is captured by the ‘poverty gap’ measure,

which is the difference between the income (or consumption) of a poor

individual, and the income (or consumption) poverty line as a proportion

of the poverty line, aggregated for all of those in poverty and then divided

by the total population. Hence a poverty gap of 0.2 should be interpreted

to mean that averaged over the whole population the living standard of the

poor is 20 per cent below the poverty line. Hence, assuming away targeting

problems, the cost of removing poverty totally will be 20 per cent of the

poverty line multiplied by the total population.4

A variant of the poverty gap that refl ects distribution within the poor is

the squared poverty gap, which is calculated in the same way, except for the

important difference that the gap between the income of a poor individual

and the poverty line as a proportion of the line is squared, so that the larger

gaps are given a greater relative weight in the indicator.5 Hence for a given

average income of the poor, a worse distribution within the poor will result

in a higher value of this indicator, capturing a greater severity of poverty.

This indicator also has the convenient property that it is decomposable,

so that it can be calculated for different subgroups in the population, and

total poverty can be derived by weighting this poverty indicator for each

subgroup by their population share.

0

2

4

6

8

10

12

14

16

18

1986 1989 1992 1996 1998 2001 2002

PRC

Figure 1.9 Poverty headcount (%): PRC

As they offer different types of information, often all three indicators

are calculated for individual countries and their trends over time tracked.

However, the basic headcount indicator and the squared poverty gap can

give quite different perspectives because of the latter’s incorporation of a

distributional dimension.

Errors of targeting can in principle arise for several reasons; inaccurate

specifi cation of who are in fact poor; poorly designed programs that do not

reach the target group even if it is known accurately; and poor governance

in the implementation of schemes so that benefi ts leak to the non-poor.

Since targeting in its broad and narrow sense has been widely used over the

past two decades there is now a relatively long record of experience that can

be surveyed to attempt to establish generalizations about the effectiveness

or otherwise of particular measures. Experiences in our fi ve case-study

countries suggest that errors have been signifi cant and that in some cases

these programs have had only a minor impact on poverty reduction.

The rest of this chapter is structured as follows. The second section

looks at the scale of resources devoted to poverty-targeting measures in

the different countries and how this has changed over time. The third section

examines the criteria used in the different cases to identify who are ‘the

poor’. The fourth section looks at the type 1 and 2 errors associated with

different forms of targeting, looking in particular at the effectiveness of

the most common measures – location targeting, self-targeting, and broad

targeting. These errors raise questions about governance and the capacity

of states to mount effective targeting policies. The fi nal section looks at

poverty reduction in the countries covered and the role of targeted measures

in the process.