Do public investment and FDI crowd in or crowd out private domestic investment in Malaysia？

Applied Economics,2009,41,913–919Do public investment and FDI crowd in or crowd out private domestic investment in Malaysia? James B.AngDepartment of Economics,Monash University,Vic3145,AustraliaE-mail:james.ang@.auMotivated by the concern of a persistent decline in total investment in Malaysia during the post-crisis era,this article examines the long-run relationship between private domestic investment(PDI),public investment and foreign direct investment(FDI)in ing multivariate cointegration techniques,the results indicate a fairly robust cointegrated relationship between these variables during the period1960to2003.Both public investment and FDI are found to be complementary to,rather than competing with,PDI.I.IntroductionThere has been a sharp decline in total investment in Malaysia following the onset of the Asian financial crisis in1997–98.This decline has emanated pre-dominantly from private investment while public investment is boosted as part of the crisis manage-ment programme.However,it is not clear whether such government pump-priming efforts are comple-mentary to or competing with capital formation in the private sector.Therefore,this pattern of devel-opment has become a major concern of researchers and policy makers,and this warrants an investigation into the relationship between private and public investment in Malaysia.In an influential study,Aschauer(1989)shows that an expansion in public investment leads to higher rate of return of private investment,thereby stimulat-ing private investment in the United States.Hence, a crowding-in effect of public investment is found.Following the seminal work of Aschauer(1989), a number of studies have investigated the relationship between private and public investment.While most studies find a positive impact of public investment, a negative effect is found in several cases.1Against this backdrop,this article investigates the debatable issue of whether public investment complements or displaces capital formation in the private sector in the light of the Malaysian experience,over the period 1960to2003.It is hoped that the results of this study will add to our understanding of the subject issue. Empirical studies on this issue have so far focused on analysing the broad relationship between private and public investment.However,for a large foreign direct investment(FDI)recipient country like Malay-sia,the separation of FDI from private investment is necessary to avoid aggregation bias.Hence,the analysis in this article considers PDI,public invest-ment and FDI,where private domestic investment is defined as private investment less FDI.1See,e.g.Blejer and Khan(1984)for a sample of24developing countries,Ghali(1998)for Tunisia and Ghura and Goodwin (2000)for Asia and Latin America,among others.2009Taylor&Francis913This article proceeds as follows.Section II provides an overview of the trends and patterns of investment in Malaysia.Model and data are described in Section III.Section IV sets out the econometric techniques employed in this study.The results are presented and analysed in Section V and the last section concludes.II.Trends and Patterns of InvestmentPolicy context and investment patternsThis section describes investment behaviour in Malaysia over the period1960to2003.After achieving independence in1957,the country inherited a relatively well-developed infrastructure and an efficient administrative system from the British government,which appeared to be conducive forcapital formation.Given that capital formation is an important driver for sustained economic develop-ment,the Malaysian government enacted the Pioneer Industries Ordinance of1958to stimulate private investment.Malaysia started1960with a moderate level of both private and public investment where the shares of private and public investment in GDP were8.2and1.9%,respectively.However,this total investment rate of10.1%was much lower than the average of lower and middle income countries which stood at21.3%in the same year.The total investment rate in Malaysia increased rapidly to about17%by the mid-1960s,prior to declining in the late1960s due to the higher economic uncertainty triggered by the racial riots of1969.The government attributed the poor performance of investment activities to inadequacies of the1958 Ordinance,and many of its weaknesses were corrected in the Investment Incentives Act of1968. The1970s saw a massive increase in capital forma-tion,reaching a record high level in the early1980s. Nevertheless,investment rates were significantly curtailed when the global economic recession hit the country in1985.Confronted with huge govern-ment budget deficits and current account deficits following the recession,as well as decreasing domestic and foreign investment,Malaysia began to grant attractive investment incentives to attract local and foreign investors through the Promotion of Investment Act1986.Figure1presents the time-series plots of the private,public and total investment rates.As is evident,the private investment rate showed an upward swing in1987,and increased at a rapid pace for a few years.Rapid capital formation, however,did not exert excess pressure on balance of payments,due to the impressive saving records of the country.Total investment was primarily funded by domestic saving,supplemented by foreign saving.2The ability of its financial system to successfully mobilize a lot of domestic saving has enabled the total investment rate to increase gradu-ally over the years and reach a peak in1997,before the onset of the Asian financial crisis.3 Notwithstanding these past strong records,the private investment rate declined sharply in the after-math of the crisis.This sharp contraction in private investment resulted in a downward trend in the total investment rate.Since then,capital formation has been mainly supported by public investment. Composition of investmentFigure2reflects the shares of FDI,PDI and public investment in total investment.Private investment (which includes both FDI and PDI)accounted for more than80%of the total in1960.This share declined dramatically over the next two decades before picking up again in the1980s.On the whole, private investment dominated total investmentin 196019651970197519801985199019952000PercentageYearFig.1.Trends of total,private and public investment rates (1960–2003)Notes:Private investment rate is gross private fixed capital formation as a percentage of GDP.Public investment rate is gross public fixed capital formation as a percentage of GDP.Total investment rate is the sum of private and public investment rates.2For an analysis of the dynamic relationship between domestic saving and investment rates in Malaysia,see Ang(2007a). 3Ang and McKibbin(2007)provide an analysis of the dynamic relationship between financial development and economic growth in Malaysia.Their results show that financial depth and economic development are positively related.Similar findings are also obtained by Ang(2007b).914J.B.AngMalaysia prior to the 1997–98crisis.However,as would be expected,the share of private investment declined sharply after the crisis.From the above analysis,it appears the behaviour of investment in Malaysia is predominantly influenced by private,rather than by public investment,except when the latter is counter-cyclical.Noticeably,the post-crisis investment contraction has been entirely driven by a decline in private investment.Inflows of FDI constitute an important component of total domestic investment given that Malaysia is one of the most successful developing countries in attracting a large amount of FDI.According to the inward FDI potential index (a measure of the attractiveness of the host economy to foreign investors)published by UNCTAD (2004),Malaysia was ranked first among the Asian developing countries in 2003.The importance of FDI to domestic economic development is therefore undeni-able.Hence,the separation of FDI from private investment is necessary.III.Model and DataIt is a widely accepted view that public investment PUB t may be complementary to,rather than compet-ing with,private investment in developing countries.Public investment may facilitate and stimulate private investment through the provision of infrastructural support (Sundararajan and Thakur,1980;Blejer and Khan,1984;Greene and Villanueva,1991).This can raise the productivity of capital,and expand the overall resource availability by increasing output.On the other hand,public investment may also crowd out private investment.This occurs when additional public investment requires raising future tax and domestic interest rates,or if the public sector producesinvestment goods that directly compete with private goods.In addition,the utilization of additional physical and financial resources,which would other-wise be available to the private sector,may also depress private investment (Blejer and Khan,1984;Aschauer,1989).Thus,it appears that economic theory provides no a priori argument for or against the effect of public investment on private investment.FDI can have both positive and negative effects on PDI.It can stimulate domestic investment by providing new investment opportunities for local firms through the provision of machinery and technology which cannot otherwise be produced domestically (Sun,1998).Local firms can emulate the new technology introduced by foreign firms,which may stimulate domestic investment (Noorzoy,1979).Besides,an increase in domestic investment is likely to be accompanied by an increase in FDI inflows when there are more joint venture activities between local firms and foreign firms (Jansen,1995).FDI may also contribute to higher domestic invest-ment through introducing new industries to the host country (Lipsey,2004).In this way,FDI and PDI are complementary to each other.On the other hand,FDI may crowd out domestic investment if foreign firms compete with local firms for the use of domestic scarce resources,such as skilled labour,financial resources,etc.(Jansen,1995).FDI may also disrupt backward linkages in domestic manufacturing through the substitution of imports for domestic goods.Furthermore,FDI may substi-tute for domestic investment if foreign firms have an edge in technological or managerial expertise,or tax benefits provided by the host country (Noorzoy,1979).These new technologies embodied in FDI may accelerate technological obsolescence of traditional technologies used in developing coun-tries and therefore crowd out domestic investment (Kim and Seo,2003;Lipsey,2004).These suggest that the effects of FDI on PDI may vary from country to country,depending on the types of FDI,the trade policies adopted in the host country,and the strengths of the domestic firms.Therefore,the impact of FDI inflows on PDI is ultimately an empirical matter.The above theoretical considerations lead to formulation of the following empirical specification of the long-run PDI equation:PDI t ¼f ðPUB t ,FDI t Þð1Þwhere PDI t is the steady-state private domestic investment,PUB t is public investment and FDI t refers to foreign direct investment.The expected signs on PUB t and FDI t are indeterminate.YearP e r c e n t a g eposition of total investment in Malaysia (1960–2003)Public investment and private domestic investment FDI915Three dummy variables are incorporated into the above specification to account for the racial riots in 1969,the global economic recession that hit Malaysia in1985–1986,and the1997–1998Asian financial crisis,defined as:D69¼1if t¼1969 0otherwise;D85À86¼1if t¼1985À86 0otherwise;D97À98¼1if t¼1997À98 0otherwisePrivate investment is measured by gross fixed private capital formation,public investment(PUB t)is measured by gross fixed public capital formation, and private domestic investment(PDI t)is measured by private investment minus foreign direct investment (FDI t).FDI t includes equity capital,reinvested earn-ings,and other short-and long-term capital.It represents the net inflows of investment to acquire a lasting interest in an enterprise operating in the Malaysian economy.Annual data covering the period1960to2003are used in the analysis.The data series are directly obtained or compiled from the domestic sources, including the Money and Banking in Malaysia (BNM,1994)of Bank Negara Malaysia and various issues of Monthly Statistical Bulletin of Bank Negara Malaysia.Logarithms are taken on all the investment series for the usual statistical reasons,and to express in real terms,these series are deflated by the gross capital formation deflator.However,it must be highlighted that the estimation for PDI may be subjected to some measurement issues,since the data sources for private investment are not compatible with those obtained for FDI.In Malaysia,like many other countries,data on private investment are obtained directly from the national account statistics.Data on FDI are taken from the balance of payment accounts,which are estimated by BNM using capital flow data from commercial bank records,and supplemented by information directly obtained from a company survey on reinvested earnings of foreign-invested firms(multi-national enterprise subsidiaries).Although the incompatibility of data sources may be a concern,it is unlikely to distort the estimation results significantly.IV.MethodologyThe objective of our empirical estimation is to examine how the variables are related in the long-run.We construct a trivariate vector autoregressive (VAR)model for the estimation purpose.The testing procedure involves three steps.We begin by perform-ing an integration analysis using the augmented Dickey–Fuller(ADF)test.The second step is to test for cointegration using the Johansen approach for the VARs constructed in levels.If cointegration is detected,the third step is to estimate the long-run relationship in a vector error correction framework. We begin our analysis by maintaining the assump-tion that the data generating process for the relation-ship between PDI t,PUB t and FDI t is a log-linear VAR model at levels.The use of a VAR model is appropriate in this case since the underlying variables may be endogenous.The VAR model is given as:x t¼ þX pj¼1j x tÀjþ"tð2Þwhere x t¼½PDI t PUB t FDI t 0, is a vector of constant terms where ¼½ PDI PUB FDI 0and f j is a matrix of VAR parameters for lag j.The vector of error terms"t¼½"PDI"PUB"FDI 0$INð0, Þ, where is the variance-covariance matrix of the residuals.The VAR model in Equation2can be transformed into a vector error correction model(VECM)after some mathematical manipulation,as given in Equation3.Áx t¼ þx tÀ1þX pÀ1j¼1jÁx tÀjþ"tð3ÞwhereÁ¼1ÀL, is the long-run multiplier matrix. The above econometric specification hinges upon the assumption that the disturbance term"t is serially uncorrelated.Hence,it is critical to ensure that optimal lag order p of the underlying VAR is chosen appropriately in order to strike a satisfactory balance between the two competing concerns,in which the lag order is high enough to ameliorate the residual serial correlation problems while low enough so that the estimation is not subjected to over-parameteriza-tion problems.This is particularly important given that a small sample is used in the estimation of this study.V.ResultsThe ADF test results are reported in the Table1show that all variables are nonstationary in their levels but become stationary after taking the first difference. Hence,we conclude that all series are I(1)at the5% level of significance.916J.B.AngGiven that PDI t,PUB t and FDI t share common integration properties,we can now proceed to testing for the presence of a long-run cointegrating relation-ship between the variables.Since the Johansen approach is sensitive to the lag length used,we conduct a series of nested likelihood ratio tests on first-differenced VARs to determine the optimal lag length(p)prior to performing cointegration tests.Given the sample size,we have considered a maximum lag length of4.The optimal lag length is found to be2.To provide some robustness checks,we have also reported results from lag1to lag 4. Cointegration tests are performed for each VAR models at levels.In Table2,both the results of trace test and maximum eigenvalue test unanimously point to the same conclusion that there is one cointegrated equation,irrespective of the choice of lag length.Table3presents the cointegrating vectors and speed of adjustment coefficients for each model. Multivariate normality test shows that the residuals are Gaussian for each model,except for Model D which assumes four grange Multiplier(LM) tests are performed to examine for evidence of serial correlation in the residuals.The results show that there is no evidence of serial correlation in each model up to the third order.By normalizing the coefficient of PDI t to1,we obtain the long-run elasticities of PDI t with respect to other variables. Both PUB t and FDI t enter the equation significantly when one or two lags are assumed.However,only FDI t is statistically significant when three or four lags are used.Hence,it appears that the results produced based on three or four lags are less robust.This is probably due to the small sample size used in the estimation.In the remaining discussion,we will focus on the results based on two lags(i.e.Model B),which is the preferred model chosen based on the likelihood ratio tests.The coefficients on ECT tÀ1are statistically sig-nificant at the5%level in all models and have the correct sign,providing further evidence for the use of a VECM framework.In Model B,PDI t adjusts at the speed of30%every year,or it takes about3.3years,to restore equilibrium when there is a shock on the steady-state PDI relationship.There is evidence that public investment crowds in PDI.Specifically,a one percentage point increase in public investment will result in a0.281percentage point increase in PDI.The finding of a crowding-in effect from government investment is in line with a majority of the empirical studies,including Aschauer (1989)and Erenburg(1993)for the United States, Greene and Villanueva(1991)for23developing countries,Shafik(1992)for Egypt,Odedokun(1997) for48developing countries,Laopodis(2001)for three newly industrialized countries,Athukorala and Sen(2002)for India and Narayan(2004)for Fiji, which have consistently found public investment to play a complementary role in private domestic capital formation.Hence,this finding is not unique to the Malaysian experience.Public investment expenditure has been used in Malaysia as a key policy tool to stimulate economic growth,particularly after the Asian financial crisis. The evidence obtained in this study seems to suggest public investment is complementary to private investment given that a rise in government investment is associated with an increase in private sector investment.Hence,the ongoing pump-priming efforts made by the Malaysian government to revive total investment appear to be effective for private capital formation.The sharp decline in private investment during the post-crisis period may be attributable to other factors.The coefficient on FDI t is statistically significant with a positive sign.In particular,a1%increase in FDI inflows will lead to a0.985%increase in PDI. Using the likelihood ratio tests,the restriction of one for the parameter on FDI t cannot be rejected at the 1%level for Model B,suggesting that the crowding-in effect of FDI appears to be one-to-one.The results that FDI has a complementary role to play in domestic capital formation are consistent with the findings of van Loo(1977)and Noorzoy(1979)Table1.Augmented Dickey–Fuller testsLevels First-difference PDI tÀ1.972À5.775*** PUB tÀ2.667À8.091*** FDI tÀ3.160À8.491***Notes:AIC is used to select the lag length.The maximum number of lags is set to be four.***Indicates1%level of significance.Table2.Johansen cointegration testsTrace statistic( trace)Maximum eigenvaluestatistic( max)Lag(p)r¼0r0r2r¼0r¼1r¼2 134.517** 5.5340.79428.983*** 4.7400.794 232.641** 5.0390.91227.602*** 4.1270.912 333.560**7.1890.82926.371*** 6.3600.829 428.585* 5.660 1.97322.925** 3.687 1.973 Notes:*,**and***indicate10,5and1%level of significance,respectively.Public investment and private domestic investment FDI917for the Canadian experience,Jansen(1995)for the Thai experience,Sun(1998)for the Chinese experi-ence,Bosworth and Collins(1999)for58developing countries over the period1978to1995and Agosin and Machado(2005)in panel regressions of five Asian countries(including Malaysia).By examining the sectoral composition of FDI in Malaysia,Min(2003)found that FDI in the manufacturing industry has increased the speed of the relocation of labour from the nontradable to the tradable sectors.This may have enhanced export capacity and therefore benefited capital formation in the private sector.4The crowding-in effect observed in Malaysia may imply that the domestic financial markets are not too tight.In cases where there are difficulties in raising funds in the domestic market,a crowding-out effect is likely to be observed (Jansen,1995).The dummy variables which capture the effects of the racial riots of1969,i.e.D69,global economic recession,i.e.D85–86and the Asian financial crisis,i.e. D97–98,are found to be statistically insignificant and therefore excluded in the estimation.VI.ConclusionsThis article examines the long-run cointegrating relationship between PDI,public investment and FDI in Malaysia for the period1960to2003.The estimation draws upon a cointegrated VAR model to account for the potential endogeneity problems.The results suggest that both public investment and FDI stimulate PDI in the long run. The impact of FDI on PDI is more pronounced than that of public investment.AcknowledgementsThe author would like to thank Russell Smyth and Yew-Kwang Ng for suggesting this topic of research. Constructive suggestions and comments from two anonymous referees of this journal are much appre-ciated.The usual disclaimer applies.ReferencesAgosin,M.R.and Machado,R.(2005)Foreign investment in developing countries:does it crowd in domestic investment?,Oxford Development Studies,33,149–62. Ang,J.B.(2007a)Are saving and investment cointegrated?The case of Malaysia(1965–2003),Applied Economics, 39,2167–74.Ang,J.B.(2007b)Financial Development and the FDI-Growth Nexus:The Malaysian Experience.Applied Economics,84,245–33.Ang,J.B.and McKibbin,W.J.(2007)Financial liberal-ization,financial sector development and growth: evidence from Malaysia,Journal of Development Economics,84,215–33.Aschauer, D. A.(1989)Does public capital crowd out private capital?,Journal of Monetary Economics,24, 171–88.Athukorala,P.-C.and Sen,K.(2002)Saving,Investment, and Growth in India,Oxford University Press, New York,Oxford.Blejer,M.I.and Khan,M.S.(1984)Government policy and private investment in developing countries, International Monetary Fund Staff Papers,31, 379–403.BNM(1994)Money and Banking in Malaysia,Bank Negara Malaysia,Kuala Lumpur.BNM Monthly Statistical Bulletin,various issues,Bank Negara Malaysia,Kuala Lumpur.Bosworth,B.P.and Collins,S.M.(1999)Capital flows to developing economies:implications for saving and investment,Brookings Papers on Economic Activity, 1999,143–69.Erenburg,S.J.(1993)The real effects of public investment on private investment,Applied Economics,25,831–7. Ghali,K.H.(1998)Public investment and private capital formation in a vector error-correction model of growth,Applied Economics,30,837–44.Ghura,D.and Goodwin,B.(2000)Determinants of private investment:a cross-regional empirical investigation, Applied Economics,32,1819–29.Table3.Cointegrated equationsModel A Model B Model C Model D(p¼1)(p¼2)(p¼3)(p¼4) Intercept 4.582 5.5327.233 6.468 PUB t0.406***0.281**0.0610.169 FDI t0.817***0.985*** 1.257*** 1.082*** ECT tÀ1À0.285***À0.300***À0.327**À0.546*** 2NORMAL8.2129.8679.82912.637**2SERIALð1Þ 6.81411.69011.915 1.5462SERIALð2Þ 4.45411.2357.86410.2832SERIALð3Þ 6.8817.983 6.392 5.738 Notes:The dependent variable is PDI t,ECT refers to error-correction term;p is the lag length, 2NORMAL refers to thejoint Jarque–Bera statistic of the test for normal residuals;2 SERIAL ð1Þ, 2SERIALð2Þand 2SERIALð3Þare the LM teststatistics for no first-,second-and third-order serialcorrelation,respectively;**and***indicate5%and10%level of significance,respectively.4To shed more light on the complementary relationship observed between FDI and private domestic investment,a detailed analysis of the industrial composition of FDI and inter-industry relations is desirable.However,this is beyond the scope of the present study.918J.B.AngGreene,J.and Villanueva,D.(1991)Private investment in developing countries:an empirical analysis, International Monetary Fund Staff Papers,38,33–58. Jansen,K.(1995)The macroeconomic effects of direct foreign investment:the case of Thailand,World Development,23,193–210.Kim, D. D.-K.and Seo,J.-S.(2003)Does FDI inflow crowd out domestic investment in Korea?,Journal of Economic Studies,30,605–22.Laopodis,N.T.(2001)Effects of government spending on private investment,Applied Economics,33,1563–77. Lipsey,R.E.(2004)Home-and host-country effects of foreign direct investment,in Challenges to Globalization:Analyzing the Economics(Eds) R. E.Baldwin and L. A.Winters,University of Chicago Press,Chicago,pp.333–82.Min,B.S.(2003)FDI and trade:links in the case of Malaysia, Journal of the Asia Pacific Economy,8,229–50. Narayan,P.K.(2004)Do public investments crowd out private investments?Fresh evidence from Fiji,Journal of Policy Modeling,26,747–53.Noorzoy,M.S.(1979)Flows of direct investment and their effects on investment in Canada,Economics Letters,2, 257–61.Odedokun,M.O.(1997)Relative effects of public versus private investment spending on economic efficiency and growth in developing countries,Applied Economics,29,1325–36.Shafik,N.(1992)Modeling private investment in Egypt, Journal of Development Economics,39,263–77. Sun,H.(1998)Macroeconomic impact of direct foreign investment in China:1979–96,World Economy,21, 675–94.Sundararajan,V.and Thakur,S.(1980)Public investment, crowding out,and growth:a dynamic model applied to India and Korea,International Monetary Fund Staff Papers,27,814–55.UNCTAD(2004)World Investment Report2004,United Nations,New York.van Loo,F.(1977)The effect of foreign direct investment on investment in Canada,Review of Economics and Statistics,59,474–81.Public investment and private domestic investment FDI919。