POVERTY IN THAILAND

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 

In Thailand, as elsewhere, the measurement of poverty and the analysis

of its causes are controversial. Nevertheless, all major studies of poverty

incidence in Thailand agree on some basic points:

• Absolute poverty has declined dramatically in recent decades, with the

exception of a recession in the early 1980s and the period following

the Crisis of 1997–98.

• Poverty is concentrated in rural areas, especially in the northeastern

and northern regions of the country.

• Large families are more likely to be poor than smaller families.

• Farming families operating small areas of land are more likely to be

poor than those operating larger areas.

• Households headed by persons with low levels of education are more

likely to be poor than others.

The following discussion draws upon the offi cial poverty estimates produced

by the Thai government’s National Economic and Social Development

Board, which, like all other available poverty estimates, are based upon

the household incomes collected in the household Socio-Economic Survey

data. Despite their imperfections, these are the only data available covering

a long time period. These survey data have been collected since 1962. The

early data were based on small samples, but their reliability has improved

steadily, especially since 1975. Table 5.1 shows all of the available offi cial

data for the four decades from 1962 to 2002.

Table 5.1 Thailand: poverty incidence,a 1962 to 2002 (headcount measure,

per cent of total population)

Aggregate Rural Urban

1962 88.3 96.4 78.5

1969 63.1 69.6 53.7

1975 48.6 57.2 25.8

1981 35.5 43.1 15.5

1986 44.9 56.3 12.1

1988 32.6 40.3 12.6

1990 27.2 33.8 1.6

1992 23.2 29.7 6.6

1994 16.3 21.2 4.8

1996 11.4 14.9 3.0

1998 12.9 17.2 3.4

2000 14.2 21.5 3.1

2002 9.8 12.6 3.0

Poverty shareb 2000 100.0 92.6 7.4

Population sharec 2000 100.0 68.4 31.6

Notes:

a Poverty incidence means the number of poor within a reference population group expressed

as a proportion of the total population of that group. The headcount measure of aggregate

poverty incidence is the percentage of the total population whose incomes fall below a

poverty line held constant over time in real terms; rural poverty is the percentage of the

rural population whose incomes fall below a poverty line held constant over time in real

terms, and so forth.

b Poverty share means the number of poor within a reference population group expressed

as a proportion of the total number of poor within the whole population.

c Population share means the population of a reference group expressed as a proportion of

the total population.

Source: Data obtained from Development Evaluation Division, National Economic and

Social Development Board, Bangkok and Medhi (1993).

Table 5.1 focuses on the familiar headcount measure of poverty incidence:

the percentage of a particular population whose household incomes

per person fall below the poverty line.1 The table confi rms that most of

Thailand’s poor people reside in rural areas. The household survey data

are classifi ed according to residential location in the categories ‘municipal

areas’, ‘sanitary districts’ and ‘villages’. These correspond to inner urban

(historical urban boundaries), outer urban (newly established urban areas)

and rural areas, respectively. Poverty incidence is highest in the rural areas,

followed by outer urban, and lowest in the inner urban areas. When these

data are recalculated in terms of the share of each of these residential

areas in the total number of poor people and then the share of the total

population, as in the last two rows of the table, respectively, a striking point

emerges. In the year 2000 rural areas accounted for 93 per cent of the total

number of poor people but only 68 per cent of the total population.

Table 5.2 shows estimates of three measures of poverty incidence,

covering the years 1988 to 2002. The fi rst and fourth columns are based

upon the headcount measure of absolute poverty incidence. This measures

the proportion of the population (column 1) or the absolute number of

people (column 4) whose incomes fall below a poverty line established

by the National Economic and Social Development Board. Because

minimum needs vary with household size and location, the poverty lines

corresponding to households of particular sizes also vary according to

household characteristics. The second column (poverty gap ratio) measures

the average difference between the incomes of those below the poverty

line and the poverty line itself, expressed as a proportion of the poverty

line, while the third column is the average value of the square of this

difference (the squared poverty gap: see also Chapter 1). The three measures

capture different aspects of poverty, but because all three move in exactly

the same direction over time our discussion will focus on the fi rst, the

headcount measure.

The data reveal a very considerable decline in poverty incidence up to

1996, a moderate increase to 1998 and a further increase over the following

two years. Over the eight years from 1988 to 1996, measured poverty

incidence declined by an enormous 21.4 per cent of the population, an

average rate of decline in poverty incidence of 2.7 percentage points per

year. That is, each year, on average 2.7 per cent of the population moved

from incomes below the poverty line to incomes above it. Over the ensuing

two years ending in 1998 poverty incidence increased by 1.5 per cent of the

population. Alternatively, over the eight years ending in 1996 the absolute

number of persons in poverty declined by 11.1 million (from 17.9 million

to 6.8 million); over the following two years the number increased by 1.1

million (from 6.8 to 7.9 million). By this calculation, measured in terms

of absolute numbers of people in poverty, the Crisis reversed one tenth of

the poverty reduction that had occurred during the eight-year period of

economic boom immediately preceding it.

Table 5.2 Thailand: poverty incidence by different measures

Period Headcount Poverty gap Squared Number of

measure ratio poverty gap poor

(%) (in millions)

1988 32.6 10.4 4.6 17.9

1990 27.2 8.0 3.3 15.3

1992 23.2 6.8 2.8 13.5

1994 16.3 4.3 1.7 9.7

1996 11.4 2.8 1.1 6.8

1998 12.9 3.2 1.2 7.9

2000 14.2 4.1 1.7 8.9

2002 9.8 2.4 1.4 6.2

Source: As in Table 5.1.

During periods when incomes are steadily increasing, lags in reporting

changes in household incomes may not be important. But during periods

when past trends are suddenly reversed, as with the Crisis of 1997 and

beyond, these reporting lags can be very signifi cant. For this reason, for

assessment of the impact that the economic crisis had on poverty incidence,

the 1996 data are best compared with the 2000 data, not those of 1998. This

comparison roughly doubles the poverty impact of the Crisis – from 1.1

million additional poor to 2.1 million. The post-Crisis economic recovery

subsequently erased all of this increase in the absolute number of the poor,

so that by 2002 there were 6.2 million poor Thais, about half a million

fewer than in 1996.

From Table 5.3, it is apparent that one region, the northeast, accounted for

61 per cent of the poor in 2000, but only 34 per cent of the total population.

Every other region’s share of the total number of poor is smaller than its

share of the total population. Combining Tables 5.1 and 5.3, it is clear that

poverty is an issue for rural people, especially in the northeast. In 2000

rural people in the northeast accounted for 63 per cent of all poor people

in Thailand, but only 29 per cent of the total population.

Table 5.4 confi rms that poverty incidence is highest among households

with larger numbers of members. Combining the largest three household

size categories shown in the table (5 persons and over), this group represents

only 62 per cent of the total number of poor people compared with 44

per cent of the total population. From Table 5.5, it is apparent that age of

household head is also related to poverty incidence, but the relationship

is less dramatic. Households headed by persons over 60 years are more

likely to be poor than households headed by younger persons, accounting

for 42 per cent of all poor people in 1999 but only 27 per cent of the

population. Households headed by persons in their twenties are least likely

to be poor.

Table 5.3 Thailand: poverty incidence by region (headcount ratio, %)

Period North Northeast Central South Bangkok

and vicinity

1988 32.0 48.4 26.6 32.5 6.1

1990 23.2 43.1 22.3 27.6 3.5

1992 22.6 39.9 13.3 19.7 1.9

1994 13.2 28.6 9.2 17.3 0.9

1996 11.2 19.4 6.3 11.5 0.6

1998 9.0 23.2 7.7 14.8 0.6

2000 12.2 28.1 5.4 11.0 0.71

2002 9.8 18.9 4.3 8.7 1.4

Poverty share 2000 17.8 60.6 8.3 11.9 0.6

Population share 2000 18.8 34.2 23.3 13.3 10.4

Source: As in Table 5.1.

Table 5.4 Thailand: poverty incidence by household size (headcount ratio, %)

Period 1 person 2 persons 3 persons 4 persons 5 persons 6 persons 7 persons

1988 3.4 10.6 20.2 29.1 34.9 41.2 50.4

1990 3.7 9.2 16.1 23.0 28.3 34.3 43.2

1992 2.9 6.5 14.3 20.9 27.4 32.2 33.5

1994 1.0 3.2 8.6 16.4 19.4 23.7 27.9

1996 1.0 2.5 6.2 10.9 13.8 19.5 18.3

1998 1.2 3.0 6.6 10.7 17.1 20.2 22.0

2000 0.8 4.2 9.5 15.1 20.6 19.7 27.4

2002 0.2 3.4 11.1 24.1 21.4 20.4 19.1

Poverty

share 2000 0.11 2.34 10.72 24.96 24.85 14.76 22.26

Population

share 2000 2.24 8.91 18.07 26.47 19.31 11.99 13.01

Source: As in Table 5.1.

Table 5. 5 Thailand: poverty incidence by age of household head

(headcount ratio, %)

20–29 30–39 40–49 50–59 60–69 70 +

1988 26.4 33.0 35.1 32.4 32.4 33.8

1990 23.0 27.4 28.3 26.9 26.6 30.6

1992 17.3 23.2 24.3 23.3 22.3 27.5

1994 12.0 16.7 16.0 16.6 14.7 22.9

1996 7.7 12.0 11.7 10.5 11.6 14.0

1998 8.7 13.1 12.0 13.6 12.7 16.3

2000 7.8 16.7 16.6 16.3 14.8 17.9

2002 4.0 24.9 25.9 17.4 16.0 11.7

Poverty share 2000 0.29 19.52 21.94 16.05 42.19 *

Population share 2000 0.77 24.34 27.52 20.50 26.86 *

Note: * Data relate to ‘60 and over’ age category.

Source: As in Table 5.1.

Finally, Table 5.6 focuses on farm land-owning households. In 2000 these

households accounted for 26 per cent of the total population of Thailand,

but 56 per cent of all poor people. It is therefore not the case that poor people

are necessarily landless, or even with very small land holdings. The table

shows that among farming households, those with small land holdings (less

Table 5.6 Thailand: poverty among farm owners’ households by size of

holdings (headcount ratio, %)

Less than 5 rai 5 to 19 rai 20 rai or more

1988 67.7 56.2 32.9

1990 52.9 52.1 26.9

1992 41.2 46.3 31.2

1994 28.9 36.0 21.0

1996 37.2 29.9 12.1

1998 41.9 29.3 13.5

2000 45.4 43.6 20.8

2002 34.5 37.0 16.7

Poverty share 2000 10.76 62.48 26.76

Population share 2000 8.02 48.47 43.51

Source: As in Table 5.1.

than 5 rai) are the most likely to be poor, but poverty incidence in this group

is only marginally higher than for those with intermediate sized holdings

(5 to 19 rai).2 The reason is that those with very small land holdings obtain

signifi cant off-farm incomes, as well as on-farm incomes. As a proportion

of the total number of poor land holders, those with intermediate sized

land holdings are a more signifi cant group, accounting for more than one

third of all poor households.

More dramatic than any of these data, however, are recently released data

on the relationship between poverty incidence and education. According

to the National Economic and Social Development Board’s data, of the

total number of poor people in 2002, 94.7 per cent had received primary

or less education. A further 2.8 per cent had lower secondary education,

1.7 per cent had upper secondary education, 0.48 per cent had vocational

qualifi cations and 0.31 per cent had graduated from universities. Thailand’s

poor are overwhelmingly uneducated, rural and living in large families. But

they are not necessarily landless.

What caused the long-term decline in poverty? Long-term improvements

in education have undoubtedly been important, but despite the limitations

of the underlying household survey data, a reasonably clear statistical

picture emerges on the relationship between poverty reductions and the

rate of economic growth. The data are summarized in Table 5.7, which

divides the periods shown into high, medium and low growth categories.

During periods of rapid growth there was a dramatic decline in the incidence

of absolute poverty. As Table 5.1 shows, decline was not confi ned to the

capital, Bangkok, or its immediate environs, but occurred in rural areas as

well. Since 1988, the largest absolute decline in poverty incidence occurred

in the poorest region of the country, the northeast.

It is obvious that over the long term, sustained economic growth is a

necessary condition for large-scale poverty alleviation. No amount of

redistribution could turn a poor country into a rich one. Moderately

rapid growth from 1962 to 1981 coincided with steadily declining poverty

incidence. Reduced growth in Thailand caused by the world recession in

the early to mid-1980s coincided with worsening poverty incidence in the

years 1981 to 1986. Then, Thailand’s economic boom of the late 1980s and

early 1990s coincided with dramatically reduced poverty incidence. Finally,

the contraction following the Crisis of 1997–98 led to increased poverty

incidence. The recovery since the Crisis has been associated with signifi cant

poverty reductions.

In summary, the evidence from Thailand indicates that the rate of

aggregate growth is an important determinant of the rate at which absolute

poverty declines, even in the short run. However, the statistical relationship

is far from perfect. Reduction of poverty incidence must depend on more

than just the aggregate rate of growth.

Table 5.7 Thailand: GDP growth, poverty and inequality, 1962 to 2002

Year Annual Annual change in Annual change in

GDP growth poverty incidence Gini coeffi cient

(%) (headcount ratio)

Rapid growth periods

1986–88 9.75 –6.15 0.00

1988–90 10.27 –2.70 1.95

1992–94 7.01 –3.45 –0.45

Average 9.01 –4.10 0.50

Medium growth periods

1962–69 5.08 –3.60 0.20

1975–81 4.86 –2.18 0.23

1990–92 6.47 –2.00 0.60

1994–96 6.44 –2.45 –0.60

2000–02 5.58 –2.21 n.a.

Average 5.69 –2.49 0.11*

Slow growth periods

1969–75 4.15 –2.42 –0.15

1981–86 3.67 1.88 0.88

1996–98 –6.50 0.80 –0.20

1998–00 4.16 0.60 n.a.

Average 1.37 0.22 0.18*

Note: n.a. = not applicable.

Source: Poverty and GDP data from National Economic and Social Development Board,

Bangkok.