Foreign direct investment and economic growth: the growth accounting perspective.

Wang, Miao ; Wong, M.C. Sunny

I. INTRODUCTION

During the past few decades, world foreign direct investment (FDI) flows have increased dramatically, with an annual average growth rate of over 20% in the 1980s and nearly 40% by the late 1990s (UNCTAD 2006). Such a significant expansion of foreign capital flows has captured the attention of both policy makers and researchers. Growth theories predict that physical capital accumulation and technology improvement lead to better economic growth performance (e.g., Aghion and Howitt 1992; Romer 1990; Solow 1956). This provides a promising prediction about the growth effect of FDI in the host country, since FDI is considered to transfer both physical capital and intangible assets such as better technology. As a result, policy makers and governments of many host countries have attempted to attract more inward FDI by giving special incentives, such as tax breaks, to multinational corporations (MNCs). According to the UNCTAD (2000), between 1991 and 1999, 974 FDI regulatory changes have been made in over 100 countries to attract inward FDI.

However, the existence of an absolute growth effect of FDI is often debated in the empirical literature (Carkovic and Levine 2005; Choe 2003). Recent researchers turn their attention toward whether FDI promotes economic growth given certain social and/or economic conditions. Two influential studies, Borensztein, De Gregorio and Lee (1998) (hereafter, BDL [1998]) and Alfaro, Chanda, Kalemli-Ozcan and Sayek (2004) (hereafter, ACKS [2004]), capture the positive effect of FDI on economic growth under two conditions: a sufficient level of human capital and a well-developed financial system, respectively. Based on data from 69 countries over the period of 1970-1989, BDL (1998) find that inward FDI promotes the host country's economic growth only when the host country obtains a threshold level of human capital, measured by average years of secondary schooling. ACKS (2004) argue that FDI alone does not necessarily contribute to economic growth in the host country. However, according to a sample of 71 countries between 1975 and 1995, the host country with a well-functioning financial system benefits significantly from inward FDI.

While these important previous studies have determined the conditions under which FDI will affect overall growth, there has not been investigation into which components of growth (i.e., productivity growth and capital stock growth) are affected by FDI. (1) Such evidence is important for understanding the mechanisms by which FDI affects growth and can better inform policy.

In this paper, we seek to shed light on the empirical literature of FDI growth in the perspective of growth accounting. (2) Based on the conjectures from BDL (1998) and ACKS (2004), which will be explained later, we study the effect of FDI on different growth components and how human capital and financial development affect the interaction between FDI and components of growth. To our knowledge, although some conjectures have been made in the FDI literature, there is no systematic research on this issue. We also provide a possible connection between the results in BDL (1998) and the results in ACKS (2004). Their conditions can be fundamentally different catalysts for FDI to promote economic growth. (3)

On the one hand, empirical results from BDL (1998) suggest the existence of complementarity between human capital and FDI on economic growth. The authors also find that FDI does not significantly simulate domestic investment. Hence, they conclude (page 118) that "... the main channel through which FDI contributes to economic growth is by stimulating technological progress, rather than by increasing total capital accumulation in the host economy." They conjecture that FDI drives technological progress only when there is a sufficient level of human capital in the host country. The idea is later echoed by Xu (2000), who tests the impact of FDI on technology transfer and finds that U.S. MNCs contribute to total factor productivity (TFP) growth in the host country when the country reaches a minimum human capital threshold.

On the other hand, well-developed financial markets are crucial to facilitating investments, by creating economies of scale for investors and allocating capital to its most productive use. ACKS (2004) find that well-developed financial markets can significantly enhance the growth effect of FDI. They attribute the finding to two possible channels. They argue that better development of financial markets can lead to higher technological spillover effects of FDI. They also conjecture that well-developed financial markets can efficiently create backward-linkages between foreign firms and domestic firms in upstream industries in the host country. As a result, the existence of foreign firms will increase the domestic investment in upstream industries. (4) ACKS (2004) state that "[e]ven though backward linkages may allow existing [domestic] firms, which already produce inputs in the industry, to achieve economies of scale that may not have existed earlier, it also can encourage the creation of new firms." (p. 92) This indicates that FDI can stimulate capital accumulation under a well-functioning financial system.

In the following paper, we empirically test two hypotheses: (1) the positive growth effect of FDI given an adequate level of human capital actually reflects the positive effect of FDI on productivity or TFP growth given a certain level of human capital; and (2) the positive growth effect of FDI given a well-developed financial system represents the result of positive effect of FDI on capital accumulation, given the adequate level of financial development.

Using data from BDL (1998), we find that inward FDI promotes TFP growth only when the host country achieves a threshold of human capital between 0.69 and 1.08 yr of secondary schooling. FDI promotes capital growth only when the host country reaches a level of financial depth at 0.308, measured by liquidity liabilities of financial intermediaries as a share of gross domestic product (GDP). In addition, we also observe a quadratic effect of financial depth--the maximum effect of FDI on capital growth occurs when financial depth in the host country is 0.684.

The paper is organized as follows: in Section II we present data and variables; econometric results are presented and analyzed in Section III; and we conclude the paper in Section IV.

II. DATA AND VARIABLES

Data on all control variables in our study come from BDL (1998). We use this data set so that our results are comparable to previous studies. Our sample includes 69 countries; all variables are averaged over two decades: 1970-1979 and 1980-1989; Our methodology is similar to BDL (1998), employing the seemingly unrelated regression (SUR) technique. (5) A list of country names can be found in the Appendix. We estimate the following model, which will be explained later. Table 1 presents descriptive statistics for the variables used in the empirical analysis.

TFP [growth.sub.it] = [[beta].sub.0t] + [[beta].sub.FDI] [FDI.sub.it] + [[beta].sub.FDI x H] [FDI X School).sub.it] + [[beta].sub.H] [School.sub.it] +[[beta]'.sub.z][Z.sub.it] + [[epsilon].sub.it] (1)

Capital [growth.sub.it] = [[??].sub.0,t] + [[??].sub.FDI] [FDI.sub.it] [[??].sub.FDI] x FinDepth x [(FDI X FinDepth).sub.it] + [[??].sub.FDI] x [FinDepth.sup.2] x [(FDI x [FinDepth.sup.2]).sub.it] + [[??]'.sub.z][Z.sub.it] + [V.sub.it]. (2)

The dependent variables, per capita TFP growth (TFP growth) and per capita real capital growth (Capital growth) are calculated according to the technique commonly used in the growth accounting literature. Per capita real capital growth is calculated based on data from PennWorld table 5.6 and the growth rate is averaged over the periods of 1970-1979 and 1980-1989. The TFP growth is calculated based on a Cobb-Douglas production function: TFP growth = per capita real GDP growth - [alpha] x Capital growth, where [alpha] represents the share of capital in the production function. Gollin (2002) estimates labor share (l-[alpha] across countries and finds that the labor share ranges between 0.47 and 0.72. Using the basic estimate of the labor share in Gollin (2002), we calculate the TFP growth rate. (6) We include the interaction between FDI and human capital (FDI x School), and the interaction between FDI and financial depth (FDI x FinDepth). These interaction terms help to catch any complementarity between FDI and these variables. To account for possible quadratic effect of financial depth (Rioja and Valev 2004), we also include the interaction between FDI and the square term of financial depth (FDI x [FinDepth.sup.2]). In addition, the set of other control variables Z includes the following, with expected signs consistent with the previous literature in parentheses: human capital (+), financial depth (+), log value of initial GDP (-), government size (-), black market premium (-), and inflation rate (-).

FDI (FDI) in the study is measured as FDI inflows divided by host country's GDP; human capital (School) is measured as the average years of secondary schooling in the male population over the age of 25; financial depth (FinDepth) is measured as currency plus demand deposits and other interest-bearing liabilities of banks and nonbank intermediaries as a share of GDP. It is generally equal to M2/GDP and is a measure of the overall liquidity in the financial sector. King and Levine (1993) show that this measure of financial development is closely related to long-run economic growth. The log value of initial GDP ln/[GDP]) is the log value of real GDP in 1970 and 1980; government size (Govt Expenditure) is measured as the government expenditure as a share of host country's GDP; black market premium on foreign exchange (Black Market Premium) is the relative difference between parallel exchange market and official exchange market; and inflation rate (inflation) is a measure of percentage change in the GDP deflator. (7)

Our key variables of interest are the interaction variables. For example, a positive coefficient on FDI x School in TFP growth regression ([[beta].sub.FDO] x H > 0) indicates the complementarity between FDI and human capital. In other words, FDI will contribute to productivity growth when the host country reaches a certain level of human capital.

III. EMPIRICAL RESULTS

A. SUR Results

Table 2 reports SUR regression results. (8) First, we concentrate on the results from TFP growth regressions. Human capital (School) in the TFP regressions has a positive coefficient, though not statistically significant in SUR7 and SUR8. Interestingly, the human capital measure becomes less significant with the inclusion of the financial depth variable. The coefficient on the stand-alone financial depth variable is positive and significant in TFP regressions in SUR3-8 (.017-.025). Such a positive effect of financial depth on TFP growth is consistent with the results from Beck, Levine and Loayza (2000) as they find that financial development contributes to both productivity growth and physical capital growth.

The coefficient on FDI in the TFP regressions is negative, and significant in SUR2-8. The interaction variable FDI x School has a positive and significant coefficient in all regressions. The results present a strong complementarity between FDI and human capital in terms of promoting productivity growth. In other words, inward FDI alone does not promote host country's productivity growth; FDI will have a positive effect on productivity only when a threshold level of human capital is reached. TFP regressions in SUR5-8 suggest that this threshold is 0.69-1.08 yr of secondary education. This threshold schooling is within the range of 0.52-1.13 yr estimated in BDL (1998). For example, based on SUR7 specification, the effect of FDI on TFP growth is [[beta].sub.FDI] + [[beta].sub.FDI x school] X School. As a result, the threshold level of schooling for FDI to have a positive effect on TFP growth is 1.781/ 2.424 = 0.73 yr. Intuitively, the host country needs to have the ability to understand the technology embodied in inward FDI so that it can benefit from the technology. Among our sample of 69 countries, 38 countries reached 0.73 yr of secondary schooling over the whole sample period or over the period of 1980-1989. Furthermore, it appears that the level of financial depth does not positively interact with FDI in TFP regressions in a robust manner. Although interaction between FDI and financial depth measures have significant coefficients in SUR3, they become insignificant in SUR8 when schooling interaction is included. The above results confirm our hypothesis that the major factor linking FDI and productivity growth in the host country is the human capital level, which also confirms the conjecture by BDL (1998).

We also observe certain common results from capital growth regressions. The coefficient on stand-alone financial depth is positive in all regressions, which is again consistent with the results from Beck, Levine and Loayza (2000). The magnitude of the significant coefficient on financial depth ranges between .009 and .012. However, the coefficient on schooling is not statistically significant at conventional levels in capital growth regressions. Furthermore, SUR8 results indicate that there is no complementarity between FDI and schooling in terms of affecting capital growth. It appears that financial development is more crucial than human capital in terms of promoting physical capital growth; and human capital level in the host country does not seem to affect how inward FDI impacts capital growth. Similar to the results in ACKS (2004), the coefficient on FDI is not statistically significant. While the coefficient on the interaction between FDI and financial depth alone in SUR2 and SUR4 is positive but not statistically significant, we find that the coefficient on FDI x FinDepth significantly positive and coefficient on FDI x [FinDepth.sup.2] significantly negative in SUR3 and SUR5-8. The results confirm our hypothesis that the host country needs to reach a certain level of financial development for FDI to have a positive effect on capital growth. Furthermore, the complementarity between financial depth and FDI is quadratic. As financial depth increases, the effect of FDI on capital growth increases; but it increases at a decreasing rate. For example, according to SUR7 specification, the effect of FDI on capital growth is calculated as [[beta].sub.FDI + [[beta].sub.FDI] x FinDepth x FinDepth + [[beta].sub.FDI] x [FinDepth.sup.2] x [FinDepth.sup.2]. The coefficient on FDI is -1.005; the coefficient on FDI x FinDepth is 4.206 and the coefficient on FDI x [FinDepth.sup.2] is -3.074. So, the effect of FDI on capital growth is -1.005+4.206 x FinDepth - 3.074 x [FinDepth.sup.2]. FDI will positively affect capital growth only when financial depth reaches the value of 0.308 in the host country and the maximum positive effect of FDI occurs when financial depth is 0.684. In other words, a country can benefit from inward FDI when the total liquid liabilities of the financial system is above 30.8% of GDP. (9) We also calculate the threshold level in ACKS (2004) based on their estimated coefficients and find that the threshold level of financial depth in their study is approximately 0.6. (10) Interestingly, our study suggests that FDI can start to have a positive effect on capital stock growth at a lower level of financial development. In our sample of 69 countries, 36 countries have obtained a financial depth threshold level of 0.308 over the whole sample period or over the second decade.

B. Control for Endogeneity

Potentially, cross-country study of FDI-growth can suffer from an endogeneity problem. In other words, it can be true that better capital growth rate or TFP growth rate also influences inward FDI in the host country. To control for the potential endogeneity problem, we adopt instrumental variables (IV) estimation. Ideally, instruments should be highly correlated with FDI, but not with the error term. In reality, perfect instruments are hard to obtain. Our instruments include log value of land area of the host country, regional dummies for East Asia and South Asia, political institution variables, and other explanatory variables in the regressions. (11) As in BDL (1998), we apply three-stage least square (3SLS) estimation to control for endogeneity. Results are reported in Table 3. The IV estimation results are not qualitatively different from the results from original SUR regressions. FDI itself does not necessarily promote either productivity growth or capital growth. The interaction between FDI and human capital has a positive and significant coefficient in TFP growth regression. Again, the coefficient on FDI x FinDepth is significantly positive, and the coefficient on FDI x [FinDepth.sup.2] is significantly negative in the capital growth regression.

IV. CONCLUSIONS

Many countries', especially developing countries', governments give MNCs favorable treatment such as a tax break. The purpose is to encourage more inward FDI. As mentioned above, the empirical results indicate that more attention is needed for the research on FDI and growth. We should clearly understand exactly how FDI affects physical capital growth and TFP growth. While there are a number of studies that look at the final effect of FDI on economic growth and many argue that FDI promotes economic growth given certain conditions in the host country, few papers compare the channels through which FDI impacts economic growth. Studying these channels--through physical capital growth and TFP growth can provide better policy guidance to policy makers. For example, if a host country needs to achieve a certain level of education for FDI inflows to have a positive effect on TFP, the government's first priority should be allocating more resources to education instead of a monetary incentive to MNCs.

Based on data from 69 countries over two decades, we find that FDI promotes capital growth when a certain level of financial depth is reached, and FDI contributes to productivity growth when a certain level of human capital is reached in the host country. Interestingly, the interaction between FDI and human capital is not significant in capital growth regressions; and the interactions between FDI and financial depth measures are not significant in TFP growth regressions. The results suggest that human capital is the major factor linking FDI and productivity growth and financial depth is the major factor linking FDI and capital growth.

ABBREVIATIONS

3SLS: Three-Stage Least Square

FDI: Foreign Direct Investment

GDP: Gross Domestic Product

IV: Instrumental Variables

MNCs: Multinational Corporations

SUR: Seemingly Unrelated Regressions

TFP: Total Factor Productivity

doi: 10.1111/j.1465-7295.2008.00133.x

APPENDIX
List of Countries

Algeria
Benin
Botswana
Cameroon
Central African Republic
Congo, Rep.
Gambia, The
Ghana
Kenya
Lesotho
Malawi
Mali
Mauritius
Mozambique
Niger
Rwanda
Senegal
Sierra Leone
Swaziland
Togo
Tunisia
Uganda
Congo, Dem. Rep. (Zaire)
Zambia
Zimbabwe
Barbados
Costa Rica
Dominican Republic
El Salvador
Guatemala
Haiti
Honduras
Jamaica
Mexico
Trinidad and Tobago
Argentina
Bolivia
Brazil
Chile
Colombia
Ecuador
Guyana
Paraguay
Peru
Uruguay
Venezuela
Bangladesh
Myanmar
Hong Kong, China
India
Indonesia
Iran, Islamic Rep.
Israel
Jordan
Korea, Rep.
Malaysia
Pakistan
Philippines
Singapore
Sri Lanka
Syrian Arab Republic
Taiwan, China
Thailand
Yemen, Rep.
Cyprus
Greece
Malta
Turkey
Papua New Guinea

REFERENCES

Abramovitz, M. "Resource and Output Trends in the United States since 1870." American Economic Review, 46, 1956, 5-23.

Aghion, P., and P. Howitt. "A Model of Growth Through Creative Destruction." Econometrica, 60, 1992, 323 51.

Alfaro, L., A. Chanda, S. Kalemli-Ozcan, and S. Sayek. "FDI and Economic Growth: The Role of Local Financial Markets." Journal of International Economics, 64(1), 2004, 89-112.

Barro, R. Determinants of Economic Growth: A Cross-Country Empirical Study. Cambridge, MA: MIT Press, 1997.

Barro, R., and J.-W. Lee. "International Comparisons of Educational Attainment." Journal of Monetary Economics 32, 1993, 363-94.

Beck, T., R. Levine, and N. Loayza. "Finance and the Sources of Growth." Journal of Financial Economics, 58(1-2), 2000, 261-300.

Borensztein, E., J. De Gregorio, and J.-W. Lee. "How does Foreign Direct Investment Affect Economic Growth?" Journal of International Economics, 45(1), 1998, 115-35.

Carkovic, M., and R. Levine. "Does Foreign Direct Investment Accelerate Economic Growth?" in Does Foreign Direct Investment Promote Development?, edited by T. Moran, E. Graham, and M. Blomstrom. Washington, DC: Institute for International Economics, 2005, 195-220.

Choe, J.-I. "Do Foreign Direct Investment and Gross Domestic Investment Promote Economic Growth?" Review of Development Economics, 7, 2003, 44-57.

Denison, E. F. Trends in American Economic Growth. 1929-1982. Washington: The Brookings Institution, 1985.

Frankel J. A., and D. Romer. "Does Trade Cause Growth?" American Economic Review, 89, 1999, 379-99.

Ghani, E., and V. Suri. "Productivity Growth, Capital Accumulation, and the Banking Sector." Working Paper, World Bank, 2000.

Gollin, D. "Getting Income Shares Right." Journal of Political Economy, 110, 2002, 458-74.

Jorgenson, D. W. "Productivity and Postwar U.S. Economic Growth." Journal of Economic Perspectives, 2(4), 1988, 23-41.

Jorgenson, D. W., and K. J. Stiroh. "Raising the Speed Limit: U.S. Economic Growth in the Information Age." Brookings Papers on Economic Activity, 1, 2000, 125-211.

King, R. G., and R. Levine "Finance and Growth: Schumpeter might be right." Quarterly Journal of Economics 108, 1993, 717-38.

Markusen, J. R., and A. J. Venables "Foreign Direct Investment as a Catalyst for Industrial Development." European Economic Review, 43, 1999, 335-56.

Neuhaus, M. The Impact of FDI on Economic Growth: An Analysis for the Transition Countries of Central and Eastern Europe. Heidelberg: Springer, 2006.

Oliner, S. D., and D. E. Sichel. "The Resurgence of Growth in the Late 1990s: Is Information Technology the Story?" Journal of Economic Perspectives, 14(4), 2000, 3-22.

Page, J. M. "The East Asian Miracle: An Introduction." Worm Development, 22, 1994, 615-25.

Rioja, F., and N. Valev. "Does One Size Fit All?: A Reexamination of the Finance and Growth Relationship." Journal of Development Eeonomics, 74, 2004, 429-47.

Romer, P. "Endogenous Technological Change." Journal of Political Economy, 98, 1990, 71-102. Solow, R. M. "A Contribution to the Theory of Economic Growth." Quarterly Journal of Economics, 70, 1956, 65-94.

--. "Technical Change and the Aggregate Production Function." Review of Economics and Statistics, 39, 1957, 312-20.

United Nations. World Investment Report 2000. UNCTAD, United Nations, New York, 2000.

World Investment Report 2006. UNCTAD, United Nations, New York, 2006.

Xu, B. "Multinational Enterprises, Technology Diffusion, and Host Country Productivity Growth." Journal of Development Economics, 62, 2000, 477-93.

Young, A. "The Tyranny of Numbers: Confronting the Statistical Reality of the East Asian Growth Experience." Quarterly Journal of Economics, 110, 1995, 641-80.

(1.) We will use the terms "capital stock growth," "capital accumulation," and "physical capital growth" interchangeably in this paper.

(2.) Pioneered by Abramovitz (1956) and Solow (1957), the components of growth, namely capital accumulation and productivity growth, have been widely studied in the growth accounting literature. For example, Young (1995) argues that the rapid growth in East Asian countries (Hong Kong, Singapore, South Korea, and Taiwan) is mainly due to better utilization of labor and capital, not productivity growth (Page 1994). See also Denison (1985), Jorgenson (1988), Jorgenson and Stiroh (2000) and Oliner and Sichel (2000) for further discussion.

(3.) Some recent studies also turn their attention to the impacts of FDI in the perspective of growth accounting. These studies discuss the effects of FDI on capital accumulation and technological growth and in turn the effect of FDI on economic growth. For example, Neuhaus (2006) shows theoretically FDI not only raises the level of physical capital (capital accumulation or capital widening), but also improves the quality of physical capital (technological change or capital deepening). On the other hand, Ghani and Suri (2000) examine the impacts of FDI and bank-lending on economic growth in Malaysia. The authors conclude that capital accumulation is the key driver of economic growth in Malaysia. Bank-lending is positively associated with capital accumulation, but not with TFP growth, while FDI is positively associated with TFP growth. Our paper builds upon previous important studies in the aspect that we discuss the channels through which FDI affects physical capital growth and TFP growth. In other words, we estimate empirically economic conditions necessary for FDI to positively affect physical capital growth and TFP growth.

(4.) See Markusen and Venables (1999) for detailed theoretical argument on backward linkages with the existence of foreign firms.

(5.) This SUR technique allows unobserved factors to affect TFP growth and capital growth at the same time. It is commonly used in the literature of economic growth (See also Barro and Lee 1993 and Barro 1997). Similar to BDL (1998), we estimate the regressions of TFP growth and capital growth simultaneously and constrain the coefficients in each regression to be equal across the two decades except the constant. We also run a pure cross-sectional regression by averaging variables across two periods. The cross-sectional regression provides qualitatively similar results to the SUR regression. Results from the cross-sectional regression are not provided, but available upon request.

6. Assume a Cobb-Douglas production function of Y = [AK.sup.[alpha] [L.sup.1-[alpha]], where A represents total factor productivity level; K and L represent capital and labor, respectively; [alpha] represents the share of capital input used in the production. We use the basic measure of labor share in table 2 of Gollin (2002, p. 470) for the estimation of capital growth and TFP growth. Furthermore, we employ the alternative adjustments of labor share in Gollin (2002) to estimate capital growth and TFP growth. The results do not change substantially, which can be obtained upon request.

(7.) See BDL (1998) for the detailed discussion on the measurement of all variables and data.

(8.) Regional dummy and institutional factors included in the regressions are also from BDL (1998). The institutional factors include number of wars during the sample period, political rights, and number of assassinations.

(9.) According to the quadratic expression, we find that the expression equals 0 when FinDepth are 0.308 and 1.05. Recall that the level of financial depth is defined as M2/ GDP, which should not exceed one theoretically. It implies that the growth effect of FDI is always positive when FinDepth is greater than 0.308. In our sample of 69 countries, Malta is the only country whose values of FinDepth are greater than 1.

(10.) ACKS (2004) do not explicitly calculate the threshold level of financial depth. Our calculation is based on their results of regression (3) in table 4 (p. 100).

(11.) To make our results more comparable to previous influential studies, we adopt the same instrumental variables used in BDL (1998). The land area of host countries is obtained from Frankel and Romer (1999).

MIAO WANG and M. C, SUNNY WONG *

* We would like to thank Bruce Blonigen, Brian Brush, David Clark, Joseph Daniels, Jim Granato, two anonymous referees, and the participants at the Midwest Economics Association Meeting 2007 for their helpful comments and suggestions. This research is supported by the Miles Research Grant from the College of Business Administration, Marquette University, and the Marquette University Regular Research Grant. All errors and omissions are our own.

Wang. Assistant Professor, Marquette University, Department of Economics, 606 N. 13th Street, Milwaukee, WI 53233. Phone 1-414-288-7310, Fax 1-414-288-5757, E-mail grace.wang@mu.edu.

Wong: Assistant Professor, University of San Francisco, Department of Economics, 2130 Fulton Street, San Francisco, CA 94117. Phone 1-415-422-6194, Fax 1-415-422-6983, E-mail mwong11@usfca.edu

TABLE 1
Descriptive Statistics

 Capital-Growth TFP-Growth FDI

 70-79 80-89 70-79 80-89 70-79 80-89

Mean 0.015 0.005 0.018 0.000 0.003 0.002
Max 0.049 0.035 0.064 0.065 0.026 0.039
Min -0.016 -0.022 -0.051 -0.052 -0.002 -0.004
Standard
 deviation 0.013 0.012 0.026 0.026 0.004 0.005

 Initial-Income School Government-Size

 70-79 80-89 70-79 80-89 70-79 80-89

Mean 7.335 7.596 0.647 0.905 0.174 0.183
Max 8.937 9.327 3.680 2.920 0.365 0.344
Min 5.958 6.107 0.010 0.060 0.060 0.057
Standard
 deviation 0.728 0.811 0.644 0.671 0.077 0.069

 Financial-Depth Inflation Black MarkPremium

 70-79 80-59 70-79 80-89 70-79 80-89

Mean 0.319 0.394 0.155 0.237 0.253 0.340
Max 1.586 1.327 0.834 1.682 1.673 2.709
Min 0.104 0.124 0.057 0.021 -0.004 -0.006
Standard
 deviation 0.220 0.256 0.141 0.339 0.367 0.481

TABLE 2
Seemingly Unrelated Regression (SUR) Results

Panel A SUR1 School Interaction Term only

 Capital Stock
Variables Growth TFP Growth

FDI -0.1497 (0.3164) -0.8424 (0.6261)
FDI x School 0.4604 * (0.2518) 1.5777 *** (0.5290)
School 0.0033 (0.0021) 0.01 *** (0.0040)
FDI x FinDepth
FDI x [FinDepth.sup.2]
FinDepth
Ln (GDP) -0.0022 (0.0018) -0.0061 * (0.0034)
No. of Obs 69 69
R-Squared .05 .18
 .12 .16

 SUR2 Financial Interaction
Panel A Linear Term only

 Capital Stock
Variables Growth TFP Growth

FDI -0.0845 (0.4604) -1.3768 * (0.8277)
FDI x School
School
FDI x FinDepth 0.2272 (0.5737) 2.0595 * (1.1945)
FDI x [FinDepth.sup.2]
FinDepth 0.0058 (0.0056) 0.0179 (0.0114)
Ln (GDP) -0.0007 (0.0016) -0.0026 (0.0029)
No. of Obs 66 66
R-Squared .02 .14
 .07 .06

Panel A SUR3 Financial Interaction Terms

 Capital Stock
Variables Growth TFP Growth

FDI -1.2112 * (0.7030) -3.03 ** (1.2711)
FDI x School
School
FDI x FinDepth 4.677 ** (2.1845) 8.6435 ** (3.9468)
FDI x [FinDepth.sup.2] -3.27 ** (1.5391) -4.8089 * (2.6764)
FinDepth 0.0098 * (0.0058) 0.0259 ** (0.0120)
Ln (GDP) -0.0014 (0.0016) -0.0035 (0.0029)
No. of Obs 66 66
R-Squared .09 .17
 .06 .06

 SUR4 Financial Interaction and
Panel A School Interaction

 Capital Stock
Variables Growth TFP Growth

FDI -0.09 (0.4595) -1.4084 ** (0.6708)
FDI x School 2.053 *** (0.7580)
School 0.0035 (0.0022) 0.0072 * (0.0039)
FDI x FinDepth 0.4435 (0.5746)
FDI x [FinDepth.sup.2]
FinDepth 0.0033 (0.0056) 0.0198 ** (0.0097)
Ln (GDP) -0.0022 (0.0019) -0.0067 ** (0.0033)
No. of Obs 66 66
R-Squared .01 .21
 .1 .13

 SUR5 Financial Interactions and
Panel B School Interaction

 Capital Stock
Variables Growth TFP Growth

FDI -1.1162 (0.6967) -1.5372 ** (0.6787)
FDI x School 2.2179 *** (0.7664)
School 0.0032 (0.0022) 0.0073 * (0.0039)
FDI x FinDepth 4.3951 ** (2.0957)
FDI x [FinDepth.sup.2] -2.8865 ** (1.4540)
FinDepth 0.0066 (0.0058) 0.0189 ** (0.0097)
Ln (GDP) -0.0025 (0.0019) -0.0068 ** (0.0033)
Govt Expenditure
Black Market Premium
Inflation
Regional and No No
 institutional factors?
Threshold value (no. of 32.21% (33) 0.6930 yr (39)
 countries reaching
 threshold)
No. of observations 66 66
R-squared .08 .21
 .09 .14

 SUR6 Financial Interactions and
Panel B School Interaction (w/ Control Variables)

 Capital Stock
Variables Growth TFP Growth

FDI -0.7653 (0.6027) -1.3409 ** (0.6587)
FDI x School 1.8975 ** (0.7619)
School 0.0019 (0.0019) 0.0069 * (0.0037)
FDI x FinDepth 3.7227 ** (1.8437)
FDI x [FinDepth.sup.2] -2.8012 ** (12862)
FinDepth 0.0106 * (0.0056) 0.0199 ** (0.0099)
Ln (GDP) -0.0045 *** (0.0017) -0.0082 *** (0.0033)
Govt Expenditure -0.0764 *** (0.0144) -0.0579 ** (0.0275)
Black Market Premium -0.0066 *** (0.0022) -0.0091 * (0.0049)
Inflation -0.006 * (0.0032) -0.0104 (0.0085)
Regional and Yes Yes
 institutional factors?
Threshold value (no. of 25.42% (43) 0.7066 yr (39)
 countries reaching
 threshold)
No. of observations 66 66
R-squared .38 .34
 .31 .11

 SUR7 Regional Dummy and
Panel B Institutional Factors

 Capital Stock
Variables Growth TFP Growth

FDI -1.0058 * (0.6013) -1.7818 *** (0.6462)
FDI x School 2.4249 *** (0.7452)
School 0.0007 (0.0019) 0.0036 (0.0037)
FDI x FinDepth 4.2064 ** (1.8446)
FDI x [FinDepth.sup.2] -3.0749 ** (1.2933)
FinDepth 0.0111 ** (0.0057) 0.0171 * (0.0098)
Ln (GDP) -0.0038 ** (0.0019) -0.0113 *** (0.0036)
Govt Expenditure -0.0684 *** (0.0151) -0.028 (0.0281)
Black Market Premium -0.006 *** (0.0022) -0.0053 (0.0047)
Inflation -0.0066 ** (0.0032) -0.0163 ** (0.0076)
Regional and Yes Yes
 institutional factors?
Threshold value (no. of 30.88% (36) 0.7347 yr (38)
 countries reaching
 threshold)
No. of observations 66 66
R-squared .4 .41
 .34 0.36

Panel B SUR8 All Regressors

 Capital Stock
Variables Growth TFP Growth

FDI -1.0988 * (0.6091) -2.4814 ** (1.1966)
FDI x School 0.0165 (0.5525) 2.2923 ** (1.1443)
School 0.0007 (0.0020) 0.0031 (0.0036)
FDI x FinDepth 4.5678 ** (2.1987) 3.6513 (4.5614)
FDI x [FinDepth.sup.2] -3.352 ** (1.4131) -2.9419 (2.7239)
FinDepth 0.0116 ** (0.0057) 0.0257 ** (0.0118)
Ln (GDP) -0.0041 ** (0.0019) -0.0119 *** (0.0036)
Govt Expenditure -0.0695 *** (0.0154) -0.0323 (0.0279)
Black Market Premium -0.0058 *** (0.0022) -0.0057 (0.0047)
Inflation -0.0066 ** (0.0032) -0.0143 * (0.0078)
Regional and Yes Yes
 institutional factors?
Threshold value (no. of 31.19% (35) 1.0824 yr (26)
 countries reaching
 threshold)
No. of observations 66 66
R-squared .40 .41
 0.33 0.36

Note. Standard errors in the parentheses. *** significant at 1%;
** significant at 5%; * significant at 10%.

TABLE 3
Instrumental Variables Regression (3SLS) Results

 SUR1 School Interaction
Panel A Term only

 Capital
Variables Stock Growth TFP Growth

FDI -0.2594 (0.4780) -1.6108 * (0.9541)
FDI x School 0.5308 (0.3383) 2.0765 *** (0.7061)
School 0.0031 (0.0022) 0.0087 ** (0.0042)
FDI x FinDepth
FDI x [FinDepth.sup.2]
FinDepth
Ln (GDP) -0.0021 (0.0018) -0.0058 * (0.0034)
No. of Observations 69 69
R-Squared .05 .18
 .12 .14

 SUR2 Financial Interaction
Panel A Linear Term only

 Capital
Variables Stock Growth TFP Growth

FDI -0.477 (0.6394) -1.5869 (1.1524)
FDI x School
School
FDI x FinDepth 0.6672 (0.7540) 2.3043 (1.5499)
FDI x [FinDepth.sup.2]
FinDepth 0.0049 (0.0057) 0.0173 (0.0120)
Ln (GDP) -0.0007 (0.0016) -0.0026 (0.0029)
No. of Observations 66 66
R-Squared .01 .14
 .08 .07

 SUR3 Financial
Panel A Interaction Terms only

 Capital
Variables Stock Growth TFP Growth

FDI -1.1198 (0.7558) -3.1361 ** (1.3556)
FDI x School
School
FDI x FinDepth 4.4284 ** (2.3164) 8.9255 ** (4.1599)
FDI x [FinDepth.sup.2] -3.1198 ** (1.6096) -4.9614 * (2.7843)
FinDepth 0.0097 * (0.0058) 0.0259 ** (0.0120)
Ln (GDP) -0.0013 (0.0016) -0.0036 (0.0029)
No. of Observations 66 66
R-Squared .09 .17
 .06 .06

 SUR4 Fin. Interaction and
Panel A School Interaction

 Capital
Variables Stock Growth TFP Growth

FDI -0.5907 (0.6046) -1.6229 ** (0.8081)
FDI x School 2.2385 *** (0.8559)
School 0.0(134 (0.0022) 0.007 * (0.0039)
FDI x FinDepth 1.0075 (0.7185)
FDI x [FinDepth.sup.2]
FinDepth 0.0022 (0.0056) 0.0201 ** (0.0097)
Ln (GDP) -0.0022 (0.0019) -(1.0067 ** (0.0032)
No. of Observations 66 66
R-Squared -.01 .22
 .11 .13

 SUR5 Financial Interactions and
Panel B School Interaction

 Capital
Variables Stock Growth TFP Growth

FDI -1.0519 (0.7390) -1.4716 ** (0.6896)
FDI x School 2.1586 *** (0.7740)
School 0.0032 (0.0022) 0.0073 * (0.0039)
FDI x FinDepth 4.2334 ** (2.1982)
FDI x [FinDepth.sup.2] -2.7942 * (1.5075)
FinDepth 0.0065 (0.0058) 0.0189 ** (0.0097)
Ln (GDP) -0.0025 (0.0019) -0.0068 ** (0.0033)
Govt Expenditure
Black Market
Premium
Inflation
Regional and No No
 Institutional
 Factors?
Threshold value 31.32% (35) 0.6817 yr (39)
 no. of countries
 reaching threshold)
No. of Observations 66 66
R-Squared .08 .21
 .09 .14

 SUR6 Financial Interactions and
 School Interaction
Panel B (w/ Control Variables)

 Capital
Variables Stock Growth TFP Growth

FDI -0.6811 (0.6399) -1.2799 ** (0.6690)
FDI x School 1.8389 *** (0.7694)
School 0.0019 (0.0019) 0.007 * (0.0037)
FDI x FinDepth 3.4967 * (1.9344)
FDI x [FinDepth.sup.2] -2.6673 ** (1.3334)
FinDepth 0.0105 * (0.0056) 0.02 ** (0.0099)
Ln (GDP) -0.0045 *** (0.0017) -0.0082 *** (0.0033)
Govt Expenditure -0.0766 *** (0.0145) -0.0585 ** (0.0275)
Black Market -0.0066 *** (0.0022) -0.0091 * (0.0049)
Premium
Inflation -0.006 * (0.0032) -0.0104 (0.0085)
Regional and Yes Yes
 Institutional
 Factors?
Threshold value 23.79% (46) 0.6960 yr (39)
 no. of countries
 reaching threshold)
No. of Observations 66 66
R-Squared .38 .34
 .31 .11

 SUR7 Regional Dummy and
Panel B Institutional Factors

 Capital
Variables Stock Growth TFP Growth

FDI -0.9554 (0.6408) -1.731 *** (0.6574)
FDI x School 2.3757 *** (0.7539)
School 0.0007 (0.0019) 0.0037 (0.0037)
FDI x FinDepth 4.0716 ** (1.9414)
FDI x [FinDepth.sup.2] -2.995 ** (1.3443)
FinDepth 0.011 ** (0.0057) 0.0171 * (0.0098)
Ln (GDP) -0.0039 ** (0.0019) -0.0113 *** (0.0036)
Govt Expenditure -0.0686 *** (0.0151) -0.0286 (0.0281)
Black Market -0.006 *** (0.0022) -0.0053 (0.0047)
Premium
Inflation -0.0066 ** (0.0032) -0.0163 ** (0.0076)
Regional and Yes Yes
 Institutional
 Factors?
Threshold value 30.15% (36) 0.7286 yr (38)
 no. of countries
 reaching threshold)
No. of Observations 66 66
R-Squared .4 .41
 .34 .36

Panel B SUR8 All Regressors

 Capital
Variables Stock Growth TFP Growth

FDI -1.0605 * (0.6507) -2.4086 * (1.2755)
FDI x School 0.0204 (0.5532) 2.3065 ** (1.1511)
School 0.0007 (0.0020) 0.0031 (0_0036)
FDI x FinDepth 4.4566 ** (2.3008) 3.419 (4.7781)
FDI x [FinDepth.sup.2] -3.2853 ** (1.4697) -2.807 (2.8368)
FinDepth 0.0116 ** (0.0057) 0.0256 ** (0.0119)
Ln (GDP) -0.0041 ** ((1.0019) -0.0119 *** (0.0036)
Govt Expenditure -0.0696 *** (0.0154) -0.0325 (0.0279)
Black Market -0.0058 *** (0.0022) -0.0057 (0.0047)
Premium
Inflation -0.0066 ** (0.0032) -0.0143 * (0.0078)
Regional and Yes Yes
 Institutional
 Factors?
Threshold value 30.78% (36) 1.0442 yr (27)
 no. of countries
 reaching threshold)
No. of Observations 66 66
R-Squared .4 .41
 .33 .36

Note: Standard errors in the parentheses. *** significant at 1%;
** significant at 5%; * significant at 10%.