Investment Simulations

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 

We simulate the above model and determine which of a range of investment

options yields the greatest overall benefi t, the least leakage, and the least

cost. In simulating each option, the values of the other exogenous variables

are set at their mean values. In each case, both the direct (redistribution)

effect and the indirect (growth) effect are taken into account.

The base data set is the 1997 Family Income and Expenditure Survey. In

this exercise, a person is deemed poor if their standard of living, represented

by per capita consumption, falls below the absolute poverty line constructed

for that person’s province of residence. We use a social discount rate of 15

per cent to reduce streams of future benefi ts and costs to present terms.

Although high as a real discount rate, this is the rate set by the National

Economic and Development Authority on aid projects.

The cost of investment is assumed to be incurred in the first year.

Meanwhile, benefi ts are received starting year 2 (except those for education

whose stream of benefi ts begin in year 3) and remain constant over the

assumed economic life of the investment. The operating cost is assumed

to be fully recovered by charges to users. Assumptions on the cost of

maintenance vary by type of investment. Total benefi ts, as well as benefi ts

to the poor specifi cally, are calculated. The benefi t–cost ratios for different

expenditure packages are then computed and ranked. We discuss each

package briefl y.

Electricity

Presently, the proportion of households with no electricity is 30 per cent.

These are in the remote and poor barangays. Recently, the Department of

Energy has embarked on an electrifi cation project that seeks to ‘energize’

every barangay in the country. The source of energy need not be electricity.

In the far-fl ung barangays, especially those in the islands, the strategy is to

promote alternative energy sources, such as solar power, and small hydropowered

generators.7

In this experiment we assume that the present power-generating capacity

is suffi cient to provide electricity to every household. This is a reasonable

assumption given the current situation of excess capacity and the fact

that the poor households are likely to be small consumers of electricity.

The simulated project is such that in provinces where less than 90 per

cent of households have access to electricity, the intervention will result

in access for exactly 90 per cent of households. Meanwhile, in provinces

Table 6.5 Determinants of average living standards (per capita expenditure) by income quintile

1st 2nd 3rd 4th 5th

Explanatory Quintile Quintile Quintile Quintile Quintile

Variable (Poorest) (Richest)

Initial conditions

Schooling –0.010 0.080 0.107 0.075 –0.139

Local dynasty –0.104 *** –0.069 *** –0.055 ** –0.029 0.041

Political party 0.029 ** 0.013 0.022 * 0.022 * 0.030 **

Landlocked –0.067 *** –0.077 *** –0.070 *** –0.061 *** 0.041 **

Typhoon –0.064 *** –0.055 *** –0.046 *** –0.048 ** 0.059 ***

Irrigation 0.233 *** 0.157 *** 0.093 *** 0.008 –0.115 **

Farm size 0.010 –0.012 –0.011 0.010 0.072 ***

Time-varying variables

Per capita income 0.544 *** 0.621 *** 0.676 *** 0.798 *** 1.045 ***

Terms of trade 0.140 *** 0.149 *** 0.135 *** 0.119 *** –0.051

Roads –0.212 ** –0.264 *** –0.215 ** –0.051 0.478 ***

Electricity (100) 0.049 0.098 0.162 ** 0.143 ** –0.006

Agrarian reform 0.041 ** 0.033 ** 0.029 ** 0.026 * –0.009

Interactions

Schooling*Roads 0.110 ** 0.133 *** 0.102 *** 0.015 –0.251 ***

Schooling*Electricity (100) 0.007 0.019 0.009 0.002 0.002

Intercept 3.324 *** 2.760 *** 2.418 *** 1.625 *** 0.491

R-squared 0.758, 0.385 0.833, 0.498 0.864, 0.576 0.879, 0.610 0.854, 0.686

Notes:

Estimation is by three-stage least squares.

Instruments are actual values of schooling, roads, electricity, political-economy and geographic variables, terms of trade, and lagged values of the other

variables, including land Gini, tenancy and twice-lagged value of average income growth.

Data are for provincial panel covering the 1980s and 1990s.

The R-squared values apply to the level and the difference form of the estimated log (per capita expenditure) function.

***, ** and * denote signifi cance at the 1%, 5% and 10% level, respectively.

Source: Balisacan and Pernia (2003).

where the proportion of households with electricity is at least 90 per cent,

no intervention is provided. The simulated intervention results in 92 per

cent of households in total having access to electricity.

The development cost is estimated to be roughly Pesos 175 000 per linear

kilometer plus an average connection cost of Pesos 5000 per household.8

The total cost of the project is Pesos 561 million for public electricity utilities

plus Pesos 16 billion for households. The infrastructure is expected to last

for 15 years. An annual operating and maintenance cost of 5 per cent of

the total cost is assumed. Total incremental benefi t in terms of net income

change is Pesos 19.6 billion, but only Pesos 2.6 billion accrues to the poor.

This is equivalent to a leakage of 87 per cent. Poverty incidence is reduced

by 1.39 percentage points. The simulated project results in a benefi t–cost

ratio of 5.35.