1
3The Wage Sector in Thailand and the Changes in Wage Inequality over Time
Introduction
This chapter provides an analysis of the changes in the degree of inequality in the wage sector in Thailand. We focus on wage earnings from the data collected by the successive labor force surveys. The unit of observation in these surveys is the individual. The analysis of changes in inequality among households — including both the wage and non-wage earners — is undertaken in a subsequent chapter.
The study of wage sector in Thailand is of crucial importance to the analysis of the globalization and its impact on welfare. The theoretical background for this analysis is provided in the previous chapter and here we intend to shed more light on the trends in the wage sector, the causes and the corresponding effect on welfare. The real wage rate between 1978 and 1985 remained the same for most of Thailand, but in some regions, such as the North, it dropped from 1.81 baht per hour to 1.66 baht while in Bangkok wages increased from 3.64 baht to 4.20 baht per hour. Minimum wage laws were first introduced in April 1973 after the legalization of unions in 1972. The laws initially covered only Bangkok. They were subsequently applied to the entire country, which was divided into three regions with three different scales for various types of activities; agriculture and government administration were exempted. By 1982 minimum wages in Bangkok had been raised by 100 percent while those in other regions had been raised by 50 to 70 percent. It has been argued that the sluggish real wage growth in 1970s till mid 1980s has some thing to do with the minimum wage not being enforced in most of Thailand.
It would be, however, incorrect to suggest that the trends in real wages in Thailand have been governed critically by labor legislation like minimum wages. We have seen that institutional wage determination outside the public sector has been weak in Thailand (see Chapter 2). Rather the stagnation of real wages in the seventies and through the first half of the eighties can be traced to the existence of a pool of ‘surplus labor’ in the Lewis sense in agriculture. Agriculture has been his mainstay of the Thai economy during its period of growth prior to the boom of the mid-eighties. But it has been noted that an unmusically depressed level of relative earnings in agriculture has characterized the Thai economy. Data for 1991 showed that Thailand, compared to neighboring Asian economies, had by far the lowest income per worker in agriculture (income per worker in agriculture relative to GDP per worker in the whole economy).[1] Sussangkarn has pointed out that the rate of transfer of labor out of agriculture lagged significantly behind the rate of decline in the share of GDP contributed by agriculture. This had led to a substantial deterioration in the relative income in agriculture in the decade 1975-1986. Possible reasons for this have been discussed (Sussangkarn 1994, pp 590-91 and 600-01)[2].
The wages did not start to rise generally until late in 1980. Thre are reasons to believe that the boom starting in the mid-eighties, with the focus in export growth shifting to labor-intensive manufacturing rather than agricultural products. Agriculture, historically the main sector off the Thai economy and the primary employer and major source of export, was left out of the investment boom. Agriculture, historically the main sector off the Thai economy and the primary employer and major source of export, was left out of the investment boom. The wage increase seems to have been caused by the exhaustion of the labor surplus in agriculture, which accompanied this boom, led by manufactured exports. The rate of decline in agricultural employment had been stepped up (Chapter 2).
The Lewis model and the Turning Point
Even if the relative income per worker in agriculture fell below the trend rate of growth in other sectors, one might be tempted to argue that the supply price of labor from agriculture would have increased over time, putting upward pressure on wages in the wage sector. This prediction, however, does not hold in the mechanism of wage increase visualized in the Lewis model. In the model, the supply price of labor to the wage sector is given not by the average product of labor in the self-employed farm sector, but by its marginal product. As long as surplus labor or ‘disguised unemployment’ prevails in the latter, the average product could increase without pulling up the marginal product. The impact of the exhaustion of the surplus labor being exhausted would presumably be felt most directly in the admittedly small wage sector in agriculture itself. Thus we would look for a significant increase in the trend of real wages in agriculture to signal the arrival of the ‘turning point’. This is in fact what we observe in Figure 3.1, most clearly after the mid-eighties. Further evidence is the higher rate of growth of wages in the North and Northeast regions of Thailand, which have historically contained the main reservoir of surplus labor (Kakwani).
Another relevant point pertaining to the Lewis model might be stressed here. Sometimes the transfer of labor from agriculture in the Lewis model is equated with increasing urbanization, with large rural-to-urban migration of labor. This is, however, not a realistic description of the transformation of labor markets in Thailand. The rate of urbanization had in fact been at a much slower rate in Thailand than even some of the other agrarian economies like Indonesia. Rather the transfer of labor is to wage and other non-agricultural activities within the rural economy. In many cases one would find some household members participating in the non-farm activities, even while the principal earner is engaged in self-employed farming.∙
Price Trends
One of the major interests in analyzing the trends in earnings in the wage sector is the light it throws on the changes in Thailand’s international competitiveness through its period of growth, crisis and recovery. The basis for this statement is that exports in large part originate in the wage sector, and hence wage trends are an integral part of the course of the index of the international competitiveness. Such an index is usually measured by the unit labor cost (ULC), expressed in the currency of the country’s major trading partners (for simplification let us call it dollars). The following formula applies:
UCL = W/V ∙ 1/e,
Where W denotes wage per worker, V is value added per worker, and e is the exchange rate (baht per dollar).
We can derive growth rates from this formula for the UCL and, expressing the growth rates of W and V as sums of the real growth rates plus then growth rates of their respective price indices (Pc the consumer price index and Pp the producer price index) we get the following relationship--all variables expressed as growth rates (Mazumdar, 1993):
UCL = f (wage relative to productivity; the ratio of consumer prices to producer prices; the exchange rate)
The ratio of consumer prices to producer prices is often called the domestic real exchange rate (DRER), because the former generally includes more non-tradable commodities.
The analysis of Thai competitiveness relative to Korea and other countries in Asia is pursued more fully in a later chapter. Here it is sufficient to mention that the growth period in Thailand saw an erosion of international competitiveness not only because of the accelerated rise in wages after the Lewis ‘turning point’, but also because of an adverse movement in the DRER. This, of course, is the story of the ‘Dutch disease’ when inflationary effects generated by movements in critical items in the external account are not compensated for by suitable correcting measures in monetary or exchange rate policies.
The Size of the Wage Sector
We begin this section by presenting some basic data on the size of the wage sector in Thailand and, in next section the distribution of employment and earnings in this sector. This material is expected to serve as background information to the detailed statistical analysis of inequality to follow in the later sections.
Table 3.1: Wage Earners in the employed labor force
1988 / 1990 / 1992 / 994 / 1996 / 1998 / 2000 / 2002Number / 7,593,887 / 8,387,425 / 9,660,547 / 10,668,495 / 11,721,114 / 11,056,958 / 12,350,380 / 12,522,930
%Workers / 27.97 / 30.21 / 32.2 / 35.99 / 39.44 / 38.13 / 41.62 / 42.92
Distribution
Female / 41.43 / 41.47 / 40.90 / 41.72 / 40.73 / 43.27 / 44.01 / 44.03
Public / 23.41 / 21.40 / 20.89 / 21.42 / 19.35 / 23.87 / 21.66 / 20.56
Farmers Hunters, etc / 25.10 / 19.53 / 17.33 / 16.57 / 12.32 / 13.51 / 16.14 / 12.20
Production Workers, Laborers / 30.00 / 36.18 / 38.18 / 38.11 / 41.97 / 35.52 / 34.14 / 36.79
Other / 44.90 / 44.29 / 44.49 / 45.32 / 45.71 / 50.97 / 49.72 / 51.01
Note: Full time Workers, (>32 hrs per week)
Source:Labor Force Survey files, various years
The more important points to notice in the data presented in Table 3.1 are the following:
- Wage earners in Thailand increased their share of the total workforce from 28% in 1988 to just under 40% in 1996, the year before the crisis. There was a slow-down in the post crisis years but the share still ended up being slightly higher above this level in 2002.
- Females are an important component — their share had a gentle negative trend until the crisis, ending up at 40.72% of all wage earners in 1966, less than one percentage point below the level of 1988, but it started increasing after the crisis, presumably reflecting the fact that males suffered from denial of wage employment more than females. Surprisingly this had little to do with the presumed female dominance of public wage employment: while close to 26% of female wage-earners where in public sector in 1988, this percentage decreased to 20% in 1996 and fluctuated around the same number till 2002.
- The public share of wage employment was declining before the crisis (going from 23 percent of all wage employment in 1988 to 19 percent in 1996), but the crisis saw a jump back to 24 percent in 1998, and it has fallen only slightly after that in the post crisis period (21% in 2002).
- Turning to the occupational distribution shown in the bottom panel, there was a marked shift from wage earning in agriculture as the proportion employed in this sector was halved between 1988 and 1996. Most of this shift was absorbed in production, while the tertiary occupations, as a whole, more or less maintained its share. In the post-crisis years the production activities reduced their share sharply, while the larger portion of the increase went to tertiary occupations.
Earnings of Wage Earners
Post-primary education was slow to develop in Thailand. Sussangkarn (1994) reported for the mid-eighties: “More than half of those with who finish the six years of compulsory education drop out of the formal education system… 75 percent of the workforce have only primary education. The gross enrollment ratio at the secondary level was only 30 percent. This is very low compared to the newly industrializing Asian countries.”
The following table provides a comparison of the education structure in Thailand with that of other developed-developing countries. It is clear that Thailand in 1970 had started off with a particularly low level of education — both at the primary and the secondary stage—compared to the other newly industrializing countries given in the table, and even by 1994 had not been able to catch up with them. In fact, the level of enrollment ratios at both dates closely resembled that of Indonesia. This in spit of the fact that Thailand in 1994 was much higher on the league table of per capita income, and had a rate of growth, in the decade preceding 1994, which was at least 40 per cent higher than that of Indonesia.[3] Thailand’s limited success in raising the enrollment ratio in secondary education is particularly surprising since it apparently spent more than double the share of its GNP in education compared o Indonesia.
Table 3.2: Trends in Education
Gross Enrollment Primary (%) / Gross Enrollment Secondary (%) / Public Education Exp. % of GNP / Secondary Pupils (% females)Country
/ 1965 / 1994 / 1970 / 1994 / 1980 / 1995 / 1970 / 1994Chile / 124 / 99 / 34 / 68 / 4.6 / 2.9 / 57 / 54
Indonesia / 72 / 114 / 12 / 48 / 1.7 / 1.3 / 11 / 44
Japan / 100 / 103 / 82 / 99 / 5.8 / 3.8 / 50 / 50
Korea / 101 / 101 / 35 / 98 / 3.7 / 3.7 / 39 / 47
Malaysia / 90 / 91 / 28 / 59 / 6.0 / 5.3 / 41 / 51
Mexico / 92 / 115 / 17 / 58 / 4.7 / 5.3 / -- / 48
Singapore / 105 / -- / 45 / -- / 2.8 / 3.0 / 51 / 50
Thailand / 78 / 91 / 14 / 48 / 3.4 / 4.2 / 41 / 50
Source: World Development Indicators
Table 3.3 brings together the cross-section picture of the educational profile of the Thai workforce for the years of our study, as well as the earnings accruing to different educational groups, together with a measure of the inequality of distribution by educational levels. Note that the statistics are for wage earners only. The table demonstrates the continued expansion of post-primary education in the years since 1988. But it appears that the proportion of those with primary education or less declined slowly over the 1988-96 growth period, but then dropped abruptly in the post-crisis years. The latter phenomenon is probably due to a shift of labor with low education from the wage to the non-wage sector, which the depression induced.
Table 3.2:Distribution of Employment and Earnings by Education levels
1988 / 1990 / 1992 / 1994 / 1996 / 1998 / 2000 / 2002% Of Workers
No Education / 3.33 / 3.35 / 2.72 / 2.65 / 3.44 / 2.13 / 2.44 / 2.02
< Pratom 4 / 2.46 / 2.46 / 2.25 / 1.98 / 1.79 / 1.54 / 1.58
Elementary / 57.31 / 57.3 / 56.86 / 55.48 / 54.49 / 46.75 / 46.26 / 44.66
Secondary / 13.68 / 14.53 / 15.35 / 16.88 / 18.48 / 22.17 / 23.12 / 25.00
Vocational / 6.71 / 6.3 / 5.99 / 6.14 / 4.99 / 5.58 / 4.73 / 5.38
Academic / 10.07 / 10.54 / 11.27 / 11.81 / 12.32 / 16.43 / 17.4 / 17.83
Teaching / 6.39 / 5.45 / 5.44 / 5.00 / 4.48 / 5.4 / 4.45 / 5.12
Years of Education (S.D.) / 7.41
(4.49) / 7.49 (4.42) / 7.63 (4.45) / 7.80
(4.45) / 7.77
(4.45) / 8.62
(4.59) / 8.66
(4.63) / 9.01
(4.45)
Share of Total Income
No Education / 1.414 / 1.52 / 1.14 / 1.20 / 1.40 / 0.83 / 1.06 / 0.76
< Pratom 4 / 1.24 / 1.25 / 1.11 / 1.09 / 0.93 / 0.74 / 0.77
Elementary / 36.92 / 38.14 / 34.94 / 34.41 / 35.01 / 28.22 / 28.01 / 26.33
Secondary / 16.56 / 16.41 / 16.74 / 17.61 / 18.81 / 20.31 / 20.41 / 20.75
Vocational / 10.49 / 9.09 / 9.18 / 9.20 / 7.14 / 7.15 / 6.32 / 6.51
Academic / 22.54 / 23.18 / 25.89 / 26.94 / 27.86 / 32.90 / 34.90 / 35.35
Teaching / 10.78 / 10.35 / 10.69 / 9.45 / 8.85 / 9.85 / 8.54 / 10.29
Gini Coefficient
No Education / 0.3505 / 0.2816 / 0.3011 / 0.3221 / 0.2444 / 0.2135 / 0.2340 / 0.2978
< Pratom 4 / 0.3767 / 0.2930 / 0.2904 / 0.3574 / 0.2736 / 0.2599 / 0.2483
Elementary / 0.4049 / 0.3629 / 0.3414 / 0.3252 / 0.2936 / 0.3030 / 0.2812 / 0.2897
Secondary / 0.4008 / 0.3827 / 0.3630 / 0.3406 / 0.3243 / 0.3445 / 0.3227 / 0.3271
Vocational / 0.3584 / 0.3397 / 0.3372 / 0.3411 / 0.3157 / 0.3193 / 0.3703 / 0.3419
Academic / 0.3937 / 0.3524 / 0.3363 / 0.3611 / 0.3502 / 0.3635 / 0.3554 / 0.4107
Teaching / 0.2286 / 0.2293 / 0.2089 / 0.2259 / 0.2293 / 0.2497 / 0.2525 / 0.2644
No of Sampled Workers / 12586 / 16669 / 19352 / 36177 / 36229 / 32966 / 33987 / 42913
Source:Labor Force Survey files, various years
Table 3.3 yields the following earnings relatives:
Table 3.4:Trends in Relative Earning of different educational attainments
1988 / 1994 / 1996 / 1998 / 2000 / 2002Elementary / 100 / 100 / 100 / 100 / 100 / 100
Secondary / 188 / 168 / 158 / 152 / 146 / 141
Academic / 346 / 367 / 352 / 331 / 332 / 337
Source:Labor Force Survey files, various years.
There is a clear hint of divergent trends in the education-related differentials at the lower end of the spectrum. The secondary-primary differential fell rapidly in the growth period until the crisis—which seems to have slowed down the downward trend. By contrast, the academic-elementary differential held its own or increased slightly in the growth period. It had a big dip immediately after the crisis, but increased slightly in the recovery. These divergent trends will be an important part of the story of trends in the inequality of wage earnings in our subsequent discussion.
Table 3.5 gives some idea of trends in real earnings per worker in different periods and occupational groups. Real wages grew strongly in the growth period starting in the mid-eighties, but started to slow down after 1994 even before the crisis. The crisis saw a significant decline and the recovery had not done much to this reversal of the trend by 2002.
Table 3.4: Annual Percentage rate of growth of Real Earnings
1988-1994 / 1988-1996 / 1988-2000 / 1988-2002 / 1996-2002All / 6.92 / 5.67 / 3.44 / 3.31 / 0.24
Male / 6.30 / 5.00 / 2.84 / 2.90 / 0.16
Female / 8.13 / 6.86 / 4.58 / 4.14 / 0.61
Private / 7.53 / 6.42 / 3.71 / 3.69 / 0.17
Public / 6.71 / 5.60 / 3.35 / 3.12 / -0.09
Farmers, hunters, etc / 8.31 / 6.89 / 4.10 / 2.57 / -2.92
Craftsmen, Production Workers, Laborers / 5.35 / 4.23 / 2.40 / 2.58 / 0.44
Source:Labor Force Survey files, various years.
An interesting part of the story of wage increase in the growth period is the relatively higher rate of growth in the lower rung of occupations. It matches the relatively higher growth of earnings of workers with low education noticed in table 3.3. The rate of increase of wages in the agricultural sector seems to have been much higher than for production workers. Also the private sector as a whole increased the earnings of its labor force at a higher rate than the public sector—which could be expected to employ a larger proportion of educated workers and also took the lead in wage increase in earlier periods.
Trends in Wage Inequality
There is a basic problem in any attempt to compare the distribution of earnings in the wage and the non-wage sector. Wage earnings accrue to individual earners, and are reported in the Labor Force Survey as such. But in the case of the non-wage sector, all earnings are pooled for the household, with different labor force participants contributing different amounts. The latter are imperfectly distinguished, because (a) the activities are often joint, and (b) the labor contributed by each worker is uncertain and of varying intensity. There is of course the additional problem that not all of the income of non-wage households is return to labor. Thus any comparison has to be based on total income accruing to households. This can only give an approximate measure of earning differences, because the earner-dependant ratio will vary between households. In sum, comparison of earnings between the wage and the non-wage sectors are not really possible with any degree of accuracy. Accordingly the measures distribution of earnings reported below are to be viewed only as giving an approximate over-all picture. We are on much more solid grounds in comparing household welfare. This will be attempted in a later from the data collected by the Household Economic Surveys (HES).
Nevertheless it has already been mentioned in the introduction that the analysis of trends in the wage sector is important not only because it is a rapidly growing sector, but also because it is crucial significance in the evolution of the export economy such as that of Thailand. Exports originated increasingly from the wage sector. The importance of the latter increased as Thailand’s exports shifted from agricultural commodities to labor-intensive manufacture and subsequently to higher technology items. As emphasized in the introduction to this chapter wage trends are one of the crucial elements in the evolution of international competitiveness over time. It should be noted that the trends in wage differentials by skill levels — which partly determine the changes in the inequality of distribution of wage earnings over time — are as important as the growth rate of average wages or earnings. This is because they are an important determinant of the ease with which a country can upgrade its composition of exports towards more skill-intensive commodities.
The following are the more important conclusions, which emerge from the data presented in Table 3.5. We notice first the trends in the distribution of individual wage earnings.