![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
ABSTRACT
The aim of this article is to analyze the effect of the income distribution between labor and capital on the growth performance of Thailand from a post Keynesian view. It rests on the theoretical model of Bhaduri and Marglin (Citation1990) to see if an increase in the labor income share has a sufficient positive effect on consumption to offset a negative effect on investment and export demand. In order to investigate the question empirically we adopt and develop the approach of Stockhammer, Onaran, and Ederer (Citation2009). Several measures of the labor income share are calculated to take into account the fact that wage labor represents only half of the total labor force and check the robustness of our results. We also introduce a new treatment of external trade to better integrate the price competitiveness of Thailand. The econometric investigation shows that the growth regime is profit-led over the period 1970–2011, which shows that rebalancing the Thai economy will be difficult and requires an overall change of strategy going beyond a simple prolabor policy.
Notes
1In current prices Atlas method (for further details, see World Bank, World Development Indicators).
2Over the period 1970–85, the ratio of private investment to GDP was on average 20.3 percent and 27.3 percent for total gross fixed capital formation (GFCF). During the boom period (1986–96) it was 29.5 percent and 36.8 percent, respectively. After the Asian crisis (1997–98) and in 1999–2013, it fell to 15.5 percent and 21.4 percent, respectively (authors’ calculations based on data from the National Economic and Social Development Board (NESDB) of Thailand.
3Over the same period (average for 2010–13), the share of industry (which includes mining, manufacturing, and utilities) in GDP was 42 percent in China, considered the factory of the world, 35 percent in Indonesia, the largest ASEAN economy, and 33.3 percent for all Southeast Asia (United Nations, National Accounts Main Aggregates Database, GDP and its breakdown at constant 2005 prices in U.S. dollars).
4National Statistical Office of Thailand, Labor Force Survey, various years—1969 is the first year for which detailed data are available. For data consistency we used the third quarter of each year, which was the only one available in the early years.
5Authors’ estimations based on data from the Labor Survey of the National Statistical Office (NSO) of Thailand, various issues.
6Authors’ estimates based on NSO data.
7A methodological note detailing data issues, estimates of the adjusted labor shares, and econometric results is available from the authors on request.
8The empirical calculations will be based on this version of the private excess demand.
9Our empirical findings show that this is the case for only one measure of the labor share that we calculated. See the section on the estimation of the model equations and results.
10Thirlwall (2002, chap. 4) employs a similar exports function. He uses a domestic prices/foreign prices ratio instead of relative export prices. Ederer (2008) employs the same exports function as ours, but in his model, export prices are a function of domestic and imports prices, whereas in our model export prices along with domestic prices depend on (weighted) export prices of competitor countries instead of import prices.
11Razmi (2005) also uses a weighted sum of world imports instead of world income in estimating the exports equation in order to test the balance–of-payments-constrained growth model for India.
12Authors’ estimations based on data for employment by industry taken from the National Statistical Office of Thailand. We use data from the third quarter of each year for consistency over the period and to avoid seasonal fluctuations.
13Labor productivity is calculated by the authors as value added in the relevant sector, in constant U.S. dollars for 2005 (World Development Indicators, 2014) divided by data for employment by industry published by the National Statistical Office of Thailand, third quarter.
14In their study, LDCs are Argentina, China, India, Korea, Mexico, South Africa, and Turkey and for these countries the period covered is Citation1970–2007.
15The results of the ADF and EG are not presented due to space limitations but are available on request.
16Three outcomes are possible due to the nature of the test. The series might be cointegrated, all stationary, or not conclusive at all based on the test.
17Although normality is not a must for the residuals we report the probability values of the Doornik–Hansen test. Gelman and Hill (Citation2007) underline that “the regression assumption that is generally least important is that the errors are normally distributed. In fact, for the purpose of estimating the regression line (as compared to predicting individual data points), the assumption of normality is barely important at all.”.
18Since we hypothesize a structural break beginning in 1980, we applied the Chow test to that year. The same test is also conducted for the midsample date, 1990, but no signal of a break was detected.
19We report the highest VIF of the regressors of a given regression. Values greater than 10 point to multicollinearity.
20These estimations are comparable with those of Onaran and Galanis (Citation2014) who get elasticities in the range of 0.316 for Turkey to 0.845 for South Korea with a labor share calculated as in ALS1.
21This variable takes the value 1 for the years 1997 and 1998, and 0 otherwise. We also added the year 1998 because the impact of the crisis persisted.
22Since we do not have enough data for the import content of the exports of Thailand, we cannot address this issue in our current work.
23This variable takes the value 1 for the years 1973 and 1974, and 0 otherwise. We included the year 1974 since the effect of the oil crisis remained.
24For Turkey, ePULC is 0.35, and for China, −0.77 for the developing countries in the estimations of Onaran and Galanis (Citation2014), while ePPm is not significant for India, Mexico, and China, and for the others ranges between 0.12 (South Africa) and −0.36 (Argentina).
25We excluded D1973–1974 from the estimations with ULC2 because it was not significant.
26We also tried a third specification by including the dummy variable for the Asian crisis, but this variable was not significant. We opt for the second estimation in our calculations.
27The classical Keynesian equation ΔlnC =constant +eCY ΔlnY gives an income elasticity of 0.88. Although this coefficient cannot be directly used in our model, it gives an idea about the magnitude of the coefficient. Our estimates lie between 0.72 and 0.81.
28See the Appendix on the derivation of multipliers.
29Here it is implicitly assumed that domestic and imported goods are perfectly substitutable.
30In constant 2005 dollars (authors’ calculations based on World Bank, World Development).
Additional information
Notes on contributors
Bruno Jetin
Bruno Jetin, Universiti of Brunei Darussalam and Université Paris 13.
Ozan Ekin Kurt
Ozan Ekin Kurt, Université Paris 13.