Total Factor Productivity Levels and Comparative Progress: The OECDs and The East Asian Countries

Total Factor Productivity Levels and Comparative Progress:

The OECDs and The East Asian Countries

Akaranant Kidsom, PhD**

Abstract

This study aims to estimate the relative Total Factor Productivity (TFP) levels and its comparative progress during 1970-2004 among the OECDs and the selected East Asian Countries-36 countries in total. Utilising data from PWT 6.2 and WDI, the TFP level accounting is performed. It is found that the OECD’s members generally still have a higher level of TFP. The tops five are Luxemburg, the USA, Ireland, Norway, and the UK respectively. Meanwhile, the bottoms five (from the bottom) are the Philippines, Indonesia, India, China, and Turkey.

Moreover, there is a strong evidence of TFP convergence among the selected countries. It suggests that the initial TFP level conversely relates to the speed of TFP growth (TFPG). Thus, the East Asian countries gain a comparatively higher rate of the estimated TFPG. Additionally, the study also shows that the countries’ openness to international trade could be a key to explain the differences in TFP levels among the nations. In contrast, there is no strong evidence supporting that foreign direct investment (FDI) relates to the TFPGs and thus TFP levels of the sampled countries.

Keywords: Total Factor Productivity (TFP), TFP Level, TFP Growth, Openness, Foreign

Direct Investment (FDI)

**

Mahanakorn Business School, Mahanakorn University of Technology, Bangkok, Thailand e-mail [email protected]

1. Introduction

Literatures have emphasized the importance and a positive effects of international trade and foreign direct investment on economic development. Frankel & Romer (1999) found that with-in and international trade raise income per worker through different channels. The first channel was that a proportion of trade share in GDP only had a significant positive effect on physical and human capital accumulations, but did not relate with the productivity levels of the nations. However, the other channel was that the proportion of total trade share in GDP only had a moderately statistically significant positive effect on productivity levels of the nations. Therefore, they concluded that although they found a strong positive relationship between trade and income per worker, it was not clear to determine a channel of trade contribution to income growth.

Moreover, Acemoglu & Zilibotti (2001) showed that the productivity and income per worker differences across the nations can be explained by the differences in ability to utilise the transferred technologies. They reasoned that numbers of technologies were invented and designed to fit the skills of the OECD countries’ workforces. Although the less developed countries (LDCs) can later access these technologies, it may not be fully utilised since workforces in these countries do not acquire adequate skills. A mismatch between the requirements of the new technology and the skills of LDC workers is a factor explaining low productivity in LDCs. That is international trade and FDI would help transferring technologies across the countries. However, the key element of productivity gain depends on workers’ skill to utilise the transferred technologies.

Utilising the data of 78 developing countries during 1960-85, BlomstrÖm et at (1992), likewise, found that FDI significantly influenced income growth only if the host country has ability to absorb technology from the multinational firms. A positive effect of FDI on economic growth was found for the case of the higher income developing countries only. For the lower income developing countries, other variables (such as secondary education, changes labour participation rates, and the initial distance behind the USA – i.e. income gaps between the US and the considered country) were the main factors explaining economic growth of the countries. They summarised that “the ‘least developed countries’ may learn little from the multinationals, because local firms are too far behind in their technological levels to be either imitators or suppliers to the multinationals”.

Alternatively, Landes (1990) reasoned that East Asian economic growth might be explained by the relatively homogeneous society with a high degree of historical and cultural uniqueness. With great unity and less ethnic conflict within the country, the development process would be made easier. Similarly, Easterly & Levine (1997) also concluded that ethnic diversity could explain at least 25 percent of the East Asia-Africa growth differences. High degree of ethnic fragmentation as in Sub-Saharan Africa significantly explained most of the growth obstructing factors (e.g. low schooling, political instability, underdeveloped financial systems, and distorted foreign exchange markets) found in the region.

In addition, Rodrick (2000) also showed that the differences in the quality of the country’s institutions (e.g. quality of the bureaucracy, rule of law, and risk of forced nationalisation of the business) can explain the growth performance of East Asian countries. The higher the quality of institutions, the better the growth performance. Meanwhile, Hall & Jones (1999) explained the productivity and output per worker differences by employing sociology framework in explaining an important role of social infrastructure on productivity differences across the sample countries. Parente & Prescott (2000), furthermore, suggested that monopoly rights play an important role in the economic development process. The poor countries are poor because inferior technologies have been used (and/or being used inefficiently). This is because there are too many monopoly rights existing in the market.

As literatures, especially the neoclassical and the endogenous growth model, state that the long-run economic development of a country solely depends on productivity or technical progress of the nation. This study aims to estimate the relative Total Factor Productivity (TFP) levels and its comparative progress during 1970-2004 among the OECDs and the selected East Asian Countries. There are 36 countries in total. Furthermore, the paper also investigates the relationship of TFP growth (TFPG) and country’s openness to international trade as well as the Foreign Direct Investment (FDI). The findings are presented below respectively.

2. Estimating Relative TFP Levels

There are numbers of ways to estimate relative TFP levels among the nations. The preferred methodology called `Translog Multilateral Index of Productivity' can be seen in Cave, Christensen, and Diewert (1982). The index is considered as a ‘superlative index’ as defined by Diewert (1976). However, subject to the number of considered countries (36 countries in total), for this study, a simple TFP level accounting, as seen in Hall & Jones (1999), is employed. Assuming a constant return to scale Cobb-Douglas production function and Hick’s neutral technological growth, the function can be written as:

it it it it A á K á L Y ln ln (1 ) ln ln = + ! + " ! where, it

Y denotes real output of a country i at period t

it

A denotes TFP levels of a country i at period t

it

K denotes capital input of a country i at period t

it

L denotes labour input of a country i at period t

á denotes capital’s share of the output and á

!

1 denotes labour’s share of the output

From Penn World Table (PWT) 6.2, real output (using PPP exchange rate) and a rough proxy of labour input (L), i.e. countries’ population, are provided. Regarding capital stock (K) estimation, the yearly investment in capital stock data can also be retrieved from the PWT 6.2. Then, a standard procedure calculating an initial capital stock and the subsequent data is employed1. The last variable needed to estimate is capital’s share. As literatures (Mankiw et al

1

The net physical capital stock in period t equals

_!

(

)

K % where It and δ denote

the gross investment in physical capital stock in period t and a constant annual depreciation rate (assuming 5% per year) respectively. Assuming the gross investment growth rate is approximately a constant rate, g, then it can be shown that the initial capital stock is:

! + = g I K 0

0 . The net physical capital stock at period t can be written as 1 1 1 ! ! ! + = t t t K ( ) I K "

(1992) and Kidsom (2004)) suggested, the capital’s share is assumed to be 1/3 of the output, thus the rest 2/3 belongs to the labour. Therefore, the TFP level can be calculated as the standard residual i.e. lnA_it =lnY_it "

_[

_]

For a comparative reason, the calculated TFP level of the US (for the year 2004) is normalised to one. Consequently, the estimated relative TFP levels of the considered countries are the relative levels comparing with that of the US (2004)2. A full list of considered countries and the respective abbreviations can be seen in the appendix. The charts below demonstrate the findings for the year 1970 and 2004.

Figure 1: Estimated Relative TFP Levels 1970 and 2004

Source: Calculation

Note: The data for India, Malaysia, and Thailand end at 2003.

The results show that the OECD’s members generally have a higher level of TFP. For the year 2004, the tops five are Luxemburg, the USA, Ireland, Norway, and the UK respectively. Meanwhile, the bottoms five (from the bottom) are the Philippines, Indonesia, India, China, and Turkey. Moreover, comparing the findings from 1970 to 2004, as seen from the figure above, it is clear that the East Asians (generally the poorer countries) have catching up in the term of TFP levels. Thus, the TFP levels of the considered countries have converged to each others. Hence, the less dispersion of the data, i.e. less value of the coefficient of variation (COV)3 of the estimated relative TFP levels, is observed. The further discussion on TFP convergence is presented next.

2

A full calculated result is available by request.

3

COV is the standard deviation of the data dividing by its mean.

AUS AUTBEL CAN_DNK FIN FRA GER GRCHUN ISL IRLITAJPN LUX MEX NLD NZL NOR POL PRTESP SWECHE TUR GBR USA KOR CHN INDIDN MYS PHL SGPTWN THA 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1 Relative TFP Levels 1970

Coefficient of Variation (COV) = 0.092

AUSAUTBELCANDNK FINFRAGER

GRCHUN ISLIRL ITA_JPN LUX MEX NLDNZL NOR POL PRTESP SWE_CHE TUR GBRUSA KOR CHN INDIDN MYS PHL SGPTWN THA 0.5 0.55 0.6 0.65 0.7 0.75 0.8 0.85 0.9 0.95 1 1.05 1.1 Relative TFP Levels 2004

3. Evidence Supporting TFP Convergence among the Nations

Figure below clearly shows that the TFP levels of the considered countries have been converging over the period of the study. The calculated COV of all the countries’ TFP level (called “COV” on the chart) dropped from the level of almost 0.1 to around 0.07 (around 30% decreasing rate). Meanwhile, considering only the East Asians (“COV East Asian”), over the period the calculated COV decreased around 8%. Lastly, for the OECDs (“COV OECD”), the COV declining rate was around 11% over the period of the study. That is, with-in a group of OECDs, the TFP convergence can be seen more clearly than that of the East Asians. However, TFP convergence between the two groups was the most visible.

Figure 2: Speed of TFP Convergence (measured by COV)

0 0.02 0.04 0.06 0.08 0.1 0.12 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 COV COV OECD COV East Asia

Speed of TFP Level Convergence

Source: Calculation

Furthermore, considering the speed of TFP growth, the average TFP growth rate (retrieving from the relative TFP levels)4 of the OECDs over the period of the study was around 0.26% per year (ranging from 0.12 % per year (New Zealand) to 0.44% per year

(Ireland)). At the same time, the average TFP growth rate of the East Asians was around 0.44% per year (ranging from 0.16 % per year (the Philippines) to 0.97% per year (China)). In

4

It should be noted that the main purpose of this paper is to compare the relative TFP levels of the considered countries rather than the TFP growth rate. Thus, the size of the estimated TFP growth in this paper should not be compared with the results from other studies aiming to estimate the “absolute TFP growth” especially those studies based on growth accounting.

summary, the estimated relative TFP growth rate of the East Asians is higher than that of the OECDs.

In addition, the study also found negative relationships between the initial relative TFP levels and its growth rates (TFPG). That is the lower the relative TFP level at the start

(beginning of the period of study-1970), the higher the TFPG. This is a confirmation of the findings above and the evidence supporting the Neo-Classical growth theory (the poorer country grow in a faster rate). A further investigation on a relationship between relative TFP levels, openness to international trade, and the FDI is presented next.

Figure 3: TFP Growth Rate (TFPG) and Initial TFP Level

0.00% 0.20% 0.40% 0.60% 0.80% 1.00% 1.20% 0.40 0.50 0.60 0.70 0.80 0.90 1.00

TFPG and Initial TFP Level

TFPG (% per year)

Initial TFP Level (1970)

Source: Calculation

4. Relationship between Relative TFP Levels, Openness, and FDI

This section explores the relationship between the TFP levels, openness to international trade and the FDI. As mentioned, the literatures suggest significant relationships between the factors stated above. This section reinvestigates the finding by utilising the results from the prior sections and the additional data of the degree of openness to international trade (from PWT 6.2) and the FDI data from the World Development Indicators (WDI, 2006), the World Bank.

Figure 4: Openness and Initial TFP Level

Openness and TFP Level

-50.00 100.00 150.00 200.00 250.00 300.00 350.00 400.00 450.00 500.00 0.40 0.50 0.60 0.70 0.80 0.90 1.00 1.10 (% of RGDPL) TFP Level: US 2004 = 1 Source: Calculation

The figure above shows a positive relationship between the countries’ openness to international trade (measured by a percentage of international trade to the real GDP) and the relatives TFP level. Thus, at this stage without any econometric treatment, it can be concluded that the findings support a hypothesis of positive relationship between international trade and the technical progress of the nations. Further investigations (such as co-integration analysis in a panel data and more econometric treatment for a seemingly heteroscedastic data as seen above) are needed.

Moreover, using WDI data (from 2000 to 2004)5, the average percentage of the FDI to the GDP of the considered countries (so called FDI/GDP) during the period can be calculated. For a comparative study, the numbers for the US 2004 is normalised to one, as seen in the previous section. Then the average relative FDI/GDP series can be compared with the relative TFP levels (2004). It is found that the calculated correlation coefficient of the two variables is a positive number but slightly low (r = 0.32). Consequently, there is no strong evidence to conclude that there is a positive effect of FDI on countries TFP levels.

However, for more investigation, the relative FDI/GDP data is divided in to two groups (i.e. the low FDI/GDP and the high FDI/GDP) by using the group median of 2.03 as a cutting

5

point. A t-test for equality means is employed (detailed results can be seen in the appendix). The test shows that, with a 95% significant level, the average relative TFP level for the group of high FDI/GDP is slightly higher than that of the low FDI/GDP group. Nevertheless, this is still not a conclusive finding. The more investigation is needed.

5. Conclusions and Further Investigations

This paper examines relative TFP levels and its relationship with country’s openness to international trade and FDI. The results show that the OECD’s members generally have a higher level of TFP. Meanwhile, the East Asians (i.e. poorer countries in general) have been catching up in a comparatively higher speed in term of the TFP growth. Thus, the relative TFP levels of all the considered countries have been converging over the period of the study. Although, the estimated relative TFP growth rate of the East Asians is higher than that of the OECDs, the TFP convergence with-in a group of OECDs can be seen more clearly than that of the East Asians. Nevertheless TFP convergence between the two groups was the most visible. Moreover, the study also found a negative relationship between the initial relative TFP levels and its growth rates. That is the lower the relative TFP level at the start (beginning of the period of study-1970), the higher the TFPG. This result is inline with the theory and prior studies. Furthermore, the evidence supporting a positive relationship between international trade and the technical progress of the nations also can be found. Nevertheless, there is no strong evidence to support the hypothesis of a positive effect of FDI on countries’ TFP levels. Even though it can be seen that the average relative TFP level of the group that acquires a high FDI proportion in the GDP is slightly higher than that of the lower proportion group.

Further investigations such as co-integration analysis in a panel data and more econometric treatment are needed to determine the size and the direction of the relationships between the variables considered.

References:

Acemoglu D, Zilibotti F 2001 Productivity Differences The Quarterly Journal of Economics

116(2): 563-606

BlomstrÖm M, Lipsey R E, and Zejan M 1992 “What Explains Developing Country Growth?” NBER Working Paper 4132

Cave D W, Christensen L R, and Diewert W E 1982 Multilateral Comparisons of Output, Input, and Productivity using Superative Index Number The Economic Journal 92: 73-86 Diewert W E 1976 Exact and Superlative Index Number Journal of Econometrics 4: 115-145 Easterly W, Levine R 1997 Africa’s Growth Tragedy: Policies and Ethnic Divisions The Quarterly Journal of Economics 112(4): 1203-1250

Frankel J A, Romer D 1999 Does Trade Cause Growth? The American Economic Review

89(3): 379-399

Hall R E, Jones C I 1999 Why Do Some Countries Produce So Much More Output per Worker than Others? The Quarterly Journal of Economics 114: 83-116.

Kidsom A 2003 "Reconsidering what caused the growth in East Asia ?", the 2nd HEWPEM Conference, University of Patras, Greece.

Kidsom A 2004 Total Factor Productivity and Economic Growth in East Asia Ph.D.

Dissertation, Department of Economics, School of Human Sciences, University of Surrey, UK. Landes D S 1990 Why are We So Rich and They So Poor? The American Economic Review

80(2): 1-13

Parente S L and Prescott E C 2000 Barriers to Riches The MIT Press, Cambridge

Rodrik D 2000 “Institutions for High-Quality Growth: What They Are and How to Acquire Them” NBER Working Paper 7540

Solow R M 1957 Technical Change and The Aggregate Production Function Review of Economics and Statistics XXXIX: 312-320

Solow R M 1962 Technical Progress Capital Formation, and Economic Growth American Economic Review 52: 312-320

Appendix I

List of Considered Countries

COUNTRY_NAME ISO_NAME COUNTRY_NAME ISO_NAME

Australia AUS China CHN

Austria AUT India IND

Belgium BEL Indonesia IDN

Canada CAN Korea KOR

Denmark DNK Malaysia MYS

Finland FIN Philippines PHL

France FRA Singapore SGP

Germany GER Taiwan TWN

Greece GRE Thailand THA

Hungary HUN Iceland ISL Ireland IRE Italy ITA Japan JPN Luxembourg LUX Mexico MEX Netherlands NLD New Zealand NZL Norway NOR Poland POL Portugal POR Spain ESP Sweden SWE Switzerland CHE Turkey TUR United Kingdom GBR United States USA

Appendix II

T-test for Equality of Means

Group Statistics 18 .9476 .04597 .01084 17 .9024 .07182 .01742 OPEN >= 2.03 < 2.03 TFPR N Mean Std. Deviation Std. Error Mean

Independent Samples Test

7.745 .009 2.231 33 .033 .04521 .02026 .00398 .08643 2.204 26.977 .036 .04521 .02051 .00312 .08730 Equal variances assumed Equal variances not assumed TFPR F Sig.

Levene's Test for Equality of Variances

t df Sig. (2-tailed) DifferenceMean DifferenceStd. Error Lower Upper

95% Confidence Interval of the

Difference t-test for Equality of Means