Arnold et al., 2000, “Enhancing Policy and Institutional Support for Industrial Technology Development in Thailand,” Draft Report, World Bank.
Barro, R., and J-W. Lee, 1994, “Sources of Economic Growth,” Carnegie-Rochester Conference Series on Public Policy, Vol. 40, 1–46.
Ben-David, D., 1993, “Equalizing Exchange: Trade Liberalization and Income Convergence,” Quarterly Journal of Economics, Vol. 108(3), 653–680.
Coe, D. T., and R. Moghadam, 1993, “Capital and Trade as Engines of Growth in France,” International Monetary Fund Staff Papers, Vol. 40, 542–566.
Collins, S., and B. Bosworth, 1996, “Economic Growth in East Asia: Accumulation versus Assimilation,” Brookings Papers on Economic Activity, Vol. 2, 135–191.
Dollar, D., 1992, “Outward-Oriented Developing Economies Real Do Grow More Rapidly: Evidence from 95 LDCs, 1976-85,” Economic Development and Cultural Change, Vol. 40, 523–544.
Felipe, J., 1999, “Total Factor Productivity Growth in East Asia: A Critical Survey,” The Journal of Development Studies, Vol. 35, 1–41.
Jonsson, G., and A. Subramanian, 2001, “Dynamic Gains from Trade: Evidence from South Africa,” international Monetary Fund Staff Papers, forthcoming.
Oliner, S., and D. Sichel, 2000, “The Resurgence of Growth in the late 1990s: Is Information Technology the Story?” Federal Reserve Board Working Paper, February.
Rodriguez, D., and F. Rodrik, 1999, “Trade Policy and Economic Growth: A Skeptic's Guide to the Cross-National Evidence,” NBER Working Paper No. 7081, National Bureau of Economic Research.
Sachs, J., and A. Warner, 1995, “Economic Reform and the Process of Global Integration,” Brookings Papers on Economic Activity, Vol. 1, 1–118.
Siamwalla, A., 2000, “Anatomy of the Thai Economic Crisis,” in Thailand Beyond the Crisis, by P. Warr (editor), Routledge, London, forthcoming.
Tinakorn, P., and C. Sussangkarn, 1998, “Total Factor Productivity Growth in Thailand: 1980-95,” manuscript, Thailand Development Research Institute.
Young, A., 1995, “The Tyranny of Numbers: Confronting Statistical Realities of the East Asian Growth Experience”, Quarterly Journal of Economics, August, 641–680.
ANNEX: GROWTH ACCOUNTING
1. The simplest growth accounting exercise starts with a Cobb-Douglas production function:
where Y, K, and L denote output, capital, and labor, respectively, A refers to the technological or TFP level, and the parameters a and (1-α) are the technological factor shares. In growth terms, expression (1) translates to:$$$
where ΔY/Y, ΔK/K, and ΔL/L, denote growth in output, capital, and labor, respectively, while ΔA/A, is the technological progress or, put differently, TFP growth. In the empirical analysis, TFP growth is calculated residually given observations of, Y, K, L, and an estimation of the parameter α.
2. The marginal product of capital can be derived from expression (1) as dY/dK = αY/K. Thus, given an estimate of α, it is easy to empirically calculate the marginal product of capital at any point in time.
3. Different methods for estimating the capital share, α, have been proposed, including the national income approach and the regression approach. However, both approaches suffer from statistical shortcomings, and an alternative approach introduced by Sarel (1997) is used in the current paper. This approach uses detailed data on the compensation for the use of capital inputs for nine different economic activities for 26 countries (one-digit ISIC classification). The average of the sample is then defined as the typical capital share in the specific activity. The aggregate value of the capital share is derived by weighing the sector-specific capital shares by their respective share in GDP. The estimated capital shares in the different sectors are: (i) agriculture, 0.275, (ii) mining and quarrying, 0.601, (iii) manufacturing, 0.308, (iv) utilities, 0.538, (v) construction, 0.189, (vi) commerce, 0.232, (vii) transport and communication, 0.320, (viii) financial and business services, 0.604, and (ix) government and other services, 0.081.
4. When estimating the marginal productivity of ‘dwelling’ and ‘machinery and equipment’, respectively, the following production function is used
i.e., the capital stock is decomposed into dwelling, KD, and machinery, KM, respectively, where β denotes the factor share of KD. In the empirical analysis, β is assumed to 0.26, which is equal to the average share of dwelling in total capital between 1980–2000.
Prepared by Gunnar Jonsson.
Indeed, per capita output growth is exclusively driven by TFP growth in the simplest neoclassical growth model (assuming that population growth equals employment growth).
See the Appendix for a more complete description of the growth accounting framework and the methodology for estimating the factor shares.
Sarel (1997) shows that the estimated capital shares in a specific sector do not vary with the level of income.
The labor share was defined as one minus the capital share. All data is from the National Economic and Social Development Board (NESDB).
The marginal product of capital is calculated as the capital share divided by the capital output ratio (see Appendix for details).
Tinakorn and Sussangkarn (1998) also find empirical evidence of a degree of pre-crisis over-investment in Thailand. In particular, they show that there exists a non-linear relationship between productivity and capital growth, implying that the capital accumulation exhibited diminishing returns.
See, for example, Young (1994, 1995), Collins and Bosworth (1997), Sarel (1997), and Felipe (1999). Owing to differences in methodology, estimation technique, data, and sample period, the results in these studies differ with regard to the estimated rates of TFP growth. However, the general conclusion is that capital accumulation explained a substantial part of East Asia’s exceptional growth performance during the 1980s and early 1990s.
The differences between the numbers for the capital-output ratio and marginal product of capital in the figure and the table above are due to different data sources and estimates of Thailand’s capital stock. The Text Table is based on Sarel (1997), who derives capital stock data from investment flows using Penn World Tables, while the Text Figure uses actual capital stock data as published by the NESDB.
Alternatively, a cyclical element in the measure of TFP growth could appear if the labor market is inflexible, and the firms engage in some labor hoarding behavior.
The potential GDP series was extracted from the actual GDP series by using a Hodrick-Prescott (HP) filter (with the smoothing parameter set to 100). The resulting series indicate that potential GDP dropped from nearly 8 percent in the 1980s to about 3 percent in 2000. A potential GDP growth rate of about 4 percent during the second half of the 1990s is consistent with the observation that, during the same period, capacity utilization in the manufacturing sector fell by 20 percentage points while manufacturing production grew at a modest rate.
See IMF (1999), Thailand—Selected Issues, SM/99/304 (Chapters I and III).
See Appendix for a description of the assumptions underlying the calculations of the marginal product of different kinds of capital.
The main exception seems to be the Philippines, where the evolution of TFP growth is the opposite of the other countries.
It is possible that firms might increase the rate of capital scrapping in the coming years as a response to the previous over-investment. As a result, the depreciation rate could rise, implying that a given amount of gross fixed capital formation will generate a relatively smaller increase in the net capital stock.
The official unemployment rate was, on average, 2½ percent during 1991–95. However, this figure excludes the seasonally unemployed, who typically account for another 2½ percent of the labor force. The seasonally unemployment rate has remained fairly stable during the 1990s at around 2 percent per year, although it temporarily increased to 2.7 percent in 1998.
The participation rate was on average 74½ percent in 1991–95, compared with 69 percent in 1996–2000.
The definition of ‘underemployed’ is divided into the ‘severely underemployed’ defined as working less than 20 hours a week, and ‘moderately underemployed’ defined as working 20–34 hours per week. See Siamwalla (2000) for further discussions.
The share of the population below age 13 has fallen from 34 percent in 1980 to 29 percent in 1990 to 26 percent in 2000.
See, for example, Dollar (1992), Sachs and Warner (1995) and Ben-David (1993). Rodriguez and Rodrik (1999) have noted, however, that a more relevant question is whether liberal trade policy is good for growth, rather than whether openness—defined in terms of trade outcomes—is positively related to growth, and that the empirical evidence for the former proposition is less convincing.
It can be noted, however, that Thailand’s simple average tariff rate is higher than in several other Asian countries, although its effective rate is much lower, in part owing to a complex system of rebates and exemptions.
A similar positive relationship between openness and TFP growth have been found also for some other economies using time-series data; see, for example, Coe and Moghadam (1993) and Jonsson and Subramanian (2001).
However, the primary enrollment rate has fallen somewhat during the 1990s.
ICT-export is defined as computer and communication equipment and electronic components, while ICT-investment is defined as investment in telecommunication, computer hardware, and computer software.
A more recent study by the National Science and Technology Development Agency indicates that R&D expenditures had doubled by 1999 (i.e., it increased to 0.2 percent of GDP), which still, however, is much lower than in many other countries in Asia.