Effectiveness of targeting through village programs

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 

If we compare the proportion of neglected villages with the proportion

of poor households throughout the country using the BPS headcount

poverty rate, it is apparent that the percentage of villages classifi ed under

the IDT program as ‘neglected’ (31 per cent) was signifi cantly higher than

the percentage of poor households in 1993 (13.7 per cent according to

the BPS data). Sumarto et al. (1997: 12–13) have argued that the striking

difference between these two fi gures was due to the quite different concepts

used to determine poor households and ‘neglected’ villages.22 Household

poverty is established using an ‘absolute’ measurement; a household is

considered poor if its average consumption and expenditure falls below a

given threshold (the offi cial poverty line). In contrast, the ‘neglected’ village

status was determined using a relative measure; a village was deemed to

be ‘neglected’ if its Podes-derived score is one standard deviation from the

provincial mean. However by the theory of probability, about 30 per cent of

any distribution will be one standard deviation from the mean. Hence, using

this procedure, in every province, there must always be some villages that

will be defi ned as neglected, regardless of their actual condition. Moreover,

two villages of broadly similar socio-economic condition, but located in

different provinces, could be treated differently because of their relative

position within their respective provinces (Alatas, 1999).

Another problem that resulted from the method used to determine

‘neglected’ village status was a marked geographic disproportion. Half of

all poor households in Indonesia (around 50 per cent) are in Java while

more than 70 per cent of all the IDT villages were located outside Java.

This occurred because villages outside Java are mostly smaller in size and

population than villages on Java (Sumarto et al., 1997; World Bank, 1996: 3).

As a result, in terms of the number of IDT villages, there was a signifi cant

disproportion in the number of ‘neglected’ villages on average per province

off Java than on Java.

Drawing on the results of a 1996 BPS pilot study of IDT in 384 villages

in six provinces, Sumarto et al. (1997) demonstrate that the effectiveness of

the IDT program in targeting poor households was indeed quite low, at least

in the fi rst year of the program. Citing the results of the targeting in two

provinces, Central Java and West Nusa Tenggara, as examples, their study

shows that there were still a large proportion of the poor excluded from

the program (see Table 3.8). In Central Java, about 30 per cent of villages

were classifi ed as ‘neglected’, and 54 per cent of poor households lived in

these villages. However, the remaining 46 per cent of poor households did

not live in the IDT recipient villages and therefore did not benefi t from the

program. In West Nusa Tenggara, 58 per cent of all households in IDT

villages were poor – only slightly more than the percentage in Central Java,

but the percentage of IDT village was almost double that in Central Java.

Nevertheless, about 46 per cent of poor households in West Nusa Tenggara

lived in non-IDT villages.

Since this study only considers the experience of two selected provinces,

caution should be exercised in drawing conclusions about the entire IDT

program. However, these two cases certainly illustrate the mis-targeting

problem that arose as a result of the selection process. The two provinces

selected as examples are also important for poverty analysis. Central Java

is one of the most heavily populated provinces with a high incidence of

poverty in rural areas, while West Nusa Tenggara is one of the poorest

provinces in the country. Unfortunately, we do not have similar data to

assess the 1994 and 1995 programs.

Table 3.8 Comparison between neglected villages and location of poor

population in IDT and non-IDT villages, in Central Java and

West Nusa Tenggara, 1993

Central Java West Nusa Tenggara

Number of Number Number of

Number of poor people of poor

Type of village villages % (million) % villages % households %

IDT villages 2524 30 2.47 54 330 56 65 274 58

Non-IDT villages 6006 70 2.13 46 262 44 46 850 42

Total 8530 100 4.6 100 592 100 112 124 100

Source: Sumarto et al. (1997: Tables 2.9 and 2.10).

During the second and third years of the program, the selection procedures

were revised to eliminate fl aws that had been identifi ed in the selection

criteria that made it diffi cult to give suffi cient consideration to the level

of household consumption (World Bank, 1996). In the initial selection

process, too much weight had been given to infrastructure defi ciencies that

do not always refl ect low levels of household consumption. Consequently,

in late 1994, the fi eld observation criterion was dropped, and a smaller list

of variables more closely related to economic welfare was used to obtain

village scores. This new method may have been better at targeting poverty,

since it was a more accurate refl ection of the social and economic condition

of the inhabitants rather than physical location. In addition, villages with

fewer than 50 families were eliminated from the program for the 1995 round.

As a result, almost a third of the villages that had been selected in 1994

changed for the 1995 program. However, due to the change in the selection

criteria, the 1995 and 1996 data could not be directly compared with the

1994 data.

Despite the problems identifi ed in targeting the poor, the IDT program

still had positive impacts for the recipients. Alatas (1999: 25–6) revealed

that the IDT program increased the per capita total expenditure of village

recipients, although not per capita expenditure on health care. The program

also had a positive impact on employment, especially on female workers

in rural areas, indicating a loosening of the traditional labor market

constraints for females in some rural areas. There was also a small positive

impact on the school attendance rate. In addition, Akita and Setto (2000)

have argued that the IDT program has been a relatively successful fi scal

decentralization, which channeled funds from the center to regions, and

has had some signifi cant impact in reducing regional disparities, especially

between the western and eastern regions of Indonesia.