- Email: [email protected]
Trade openness and manufacturing growth in Malaysia
Available online at www.sciencedirect.com
Journal of Policy Modeling 31 (2009) 637–647
Trade openness and manufacturing growth in Malaysia V.G.R. Chandran a,b,∗ , Munusamy c,1 a b
Department of Economics, KM 12 Jalan Muar, University Technology MARA, 85009 Segamat, Johor, Malaysia Department of Economics, Faculty of Economics and Administration, University of Malaya, 50603 Kuala Lumpur, Malaysia c Department of Business, Nilai International University College, 71800 Putra Nilai, Negeri Sembilan, Malaysia Received 1 November 2008; received in revised form 1 May 2009; accepted 1 June 2009 Available online 8 July 2009
Abstract This study investigates the long-run relationship between trade openness and manufacturing growth and further assesses the causal relationship between these variables. Contrary to some scholars belief that at national level, openness does not contribute to growth in Malaysia, our sector specific analysis suggest otherwise. In this aspect, we believe that in any attempt to establish relationship between openness and growth, the analysis should be sector specific since it is more relevant as well as assures a meaningful insight for policy makers. The results suggest that in the long-run, trade openness is positively related to manufacturing growth in Malaysia. Furthermore, the results also suggest that openness should be viewed as the long term policy initiative for the sector to benefit. Therefore, the policy direction for Malaysian manufacturing sectors should focus on long term trade openness policies. Nevertheless, to ensure sustainability, emphasis should be placed on how (which manufacturing sub-sectors) or when openness is actually important. Importantly, policy makers and scholars should understand that leveraging the benefits of openness also depend on whether the liberalized sector has the comparative advantage. © 2009 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. JEL classification: F41; C22 Keywords: Trade openness; Autoregressive distributed lag; Manufacturing growth; Granger causality
∗ Corresponding author at: Department of Economics, KM 12 Jalan Muar, University Technology MARA, 85009 Segamat, Johor, Malaysia. Tel.: +60 17 6843705. E-mail addresses: [email protected] (V.G.R. Chandran), [email protected] ( Munusamy). 1 Tel.: +60 12 6142596.
0161-8938/$ – see front matter © 2009 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jpolmod.2009.06.002
638
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
1. Introduction Trade openness has emerged as the main argument among economists and policy makers in explaining the growth phenomena in developing countries (Dawson, 2006; Dutta & Ahmed, 2001; Edwards, 1992; Salehezadeh & Henneberry, 2002; Weinhold & Rauch, 1999). Besides, due to continuous interest on the issue, new methods were also proposed (Lloyd & MacLaren, 2002; Ruíz Estrada & Yap, 2006). The positive contribution of trade openness towards growth stemmed from the notion that liberalization increases specialization and division of labor thus improving productivity and export capability as well as economic performance. In addition, with greater efficiency as a result of trade openness, many of the developing countries followed suit with the export-led strategies. It is widely recognized that trade openness has a positive effect towards economic growth. It is found that countries with more trade openness relatively outperformed the countries, with less openness (Thirwall, 1994; World Bank, 1993). A study by Lloyd and MacLaren (2000) among the East Asian economies supported a similar opinion that the rapid growth was largely caused by East Asia’s economic openness. Other noteworthy studies supporting the openness and growth relationship include Urata and Yokota (1994), Osada (1994), Kajiwara (1994), Hwang (1998), Edwards (1998) and Jonsson and Subramanian (2001). In contrast, some scholars (Harrison, 1996; Rodríguez & Rodrik, 2001), however, have been more reserve in supporting the openness-led growth nexus. Although there are a considerable number of studies examining the relationship between trade liberalization and growth in developing countries (including Dollar, 1992; Edwards, 1992; Sachs & Warner, 1995; Sarkar, 2008; Yanikkaya, 2003), these studies are still far from complete. First, in the case of Malaysia, previous empirical studies relying on cross-country panel data analysis showed mixed results. Second, while many studies have provided empirical evidence on the impact of trade openness and economic growth at nation level, analysis within the framework on manufacturing industries are solely lacking. The lack of analysis at sectoral level may have contributed to the empirically mixed results reported in the previous literature. This study shows that giving emphasis to the wrong sectors or treating all countries to be homogenous in nature may lead to biasness. Contrary to the previous studies (e.g. Sarkar, 2008; Yanikkaya, 2003), it shows that in the case of Malaysia, openness is important for manufacturing sectors. The preferred country and sector specific analysis in this study captured and accounted for the complexity of economic environment and histories of the manufacturing industry in Malaysia. Besides, Malaysia appears to be a suitable case study given the fact that it is one of the highest growth open economies among the developing countries. Malaysian gross domestic product (GDP) grew at an average rate of 6.7% between 1971 and 1990, while during the period between 1990 and 2000 Malaysia had the highest growth rate averaging 8.1% per annum outperforming other ASEAN economies (Malaysia, 1971, 1990, 2001). Malaysia was one of the most active among the ASEAN countries in liberalizing its investment regime in the manufacturing sector during the 1980s and 1990s. This policy offered many incentives including pioneer status tax holidays, expanded investment tax allowances for expansion projects, tax deduction for export promotions, the establishment of Free Trade Zones and other types of incentives to attract and draw FDI. Malaysia, compared to other ASEAN economies has somehow progressed well with its outward orientated strategies. Furthermore, the tariff rate in Malaysia has declined considerably over the years (Urata, 1994). The manufacturing sector has been regarded as the main driver for export performance as well as for the impressive economic growth in Malaysia. In 2006, the strong growth in manufacturing output (7.1%) contributed around 31.3% to real GDP growth (Malaysia, 2009). Hence, interference drawn from this study provides general understanding and guidance for policy formulation.
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
639
Third, many studies have ignored the causation between the two variables, which is deemed important in econometric exercise and for policy implication. Generally, higher growing nations can also exhibit higher level of openness2 . Hence, we tested the causality between trade openness and manufacturing growth in a multivariate system by incorporating other basic determinants of manufacturing growth such as labor and capital. This was due to the argument that bivariate causality may lead to biasness due to the omission of variable phenomenon (Al-Yousif and Yousif, 1999). Furthermore, from the theoretical point of view the inclusion of other important determinants (such as labor and capital) would avoid potential mis-specification. Lastly, we adopted a more recent cointegration test called the bounds test developed by Pesaran, Shin, and Smith (2001) to establish if the variables are co-moving. Mah (2000) cautioned on the inference problems inherent with cointegrating analysis that has a small sample size3 . The Error Correction Model (Engle & Granger, 1987), Johansen (1988) and Johansen and Juselius (1990) methods are unreliable for studies that have small samples. Hence, the proposed method has a major advantage over the traditional cointegration proposed by Engle and Granger (1987), Johansen (1988) and Johansen and Juselius (1992). The bounds test is robust in dealing with a small number of observations and does not require the regressors to be in the same order of integration. Therefore, serious questions concerning the robustness of the cointegration tests could be limited compared to other studies that used Johansen’s method2 for small observations (less than 100). The primary objective of this study is to address the above gaps and to re-examine the impact of trade openness on the Malaysian manufacturing sector. The reminder of the study is organized as follows. The data and empirical model applied in this study is described in Section 2. Section 3 describes the methodology of the study. Section 4 presents the empirical findings while Section 5 discusses policy implications and the conclusion. 2. Data and model specification Annual data was used from 1970 to 2003, which was obtained from the ‘Malaysia Economic Statistics-Time Series’ published by the Department of Statistics, Malaysia. The series include manufacturing value added output, net fixed capital, number of labor and trade openness. Due to data limitation for the computation of capital stock, following Mahadevan (2002), we used the fixed assets as the proxy for capital. Manufacturing value added was deflated using the manufacturing producers price index while fixed assets was deflated using gross domestic fixed capital formation deflator with 2000 as the base year. The trade openness is calculated as a ratio of manufacturing import plus export to manufacturing output. Based on a theoretical framework (Harrison, 1996; Tsen, 2005; Wadud & Nair, 2003), the relationship between manufacturing value added, capital, labor and openness in the multivariate model can be specified as follows: LVAt = f (LCAt , LWt , LOt , t)
(1)
where LVAt , LCAt , LWt and LOt are logarithmic manufacturing value added, fixed capital, labor and trade openness, respectively. The variable t in (1) is the linear trend representing the Hicks neutral technical progress. We included a dummy variable to account for the Asian financial crisis in 1997/1998.
2
For instance see Reppas and Christopoulos (2005) for export-led growth and growth-led exports hypothesis. Johansen’s cointegration analysis requires large observation to be accurate. Scholars, e.g. Cheung and Lai (1993) suggested the adjustments to the critical values are necessary when applied to sample sizes of 100 or smaller. 3
640
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
3. Methodology 3.1. Unit root, cointegration and Granger causality The testing procedure involves three steps. The series of variables in our study need to induce stationary for the estimated results to be reliable and unbiased as to avoid spurious regression (Granger & Newbold, 1974; Phillips, 1986). Thus, we first test the stationary of the series by using the Phillip–Perron (PP) test. The next step in the test involves testing the cointegration in the model. The works of Engle and Granger (1987) and Toda and Phillips (1993) have shown that ignoring the existence of cointegration in the series could lead to serious model mis-specification. There are two ways of performing the cointegration test; the Johansen cointegration using maximum-likelihood method (Johansen, 1988) and the fairly new method known as “bounds testing approach” suggested by Pesaran et al. (2001). Owing to the limited size of observation, the bounds test is preferred over the Johansen method, which is mainly appropriate for a large sample size. Thus, in this study, we perform the cointegration test using the autoregressive distributive lag (ARDL) method proposed by Pesaran et al. (2001). This involves testing the following Unrestricted Error Correction Models (UECM): LVAt = a0M + δM t + D97M +
n i=1
+
n i=1
γiM LWt−i
n
βiM LVAt−i +
n
λim LCAt−i
i=1
ξiM LOt−i + θ1M LVAt−1 + θ2M LCA + θ3M LW
i=1
+ θ4M LOt−1 + εMt
(2)
In order to test the absence of a long-run relationship in (2), we conduct a Wald-type (F-test) coefficient restriction test, which entails testing the following null hypothesis. H0 : θ1M = θ2M = θ3M = θ4m = 0
(3)
This test is known as the bounds test (Pesaran et al., 2001) and the computed F-statistics under the null hypothesis (3) is donated by F(LVA|LO, LCA, LW). The asymptotic distributions of the test statistics are non-standard regardless of whether the variables are I(0) or I(1). For this purpose, Pesaran et al. (2001) computed two sets of asymptotic critical values where the first set assumes variables to be I(0) and the other as I(1), which are known as lower bounds (LCB) and upper bounds critical values (UCB), respectively. A decision on whether cointegration exists between the dependent variable and its regressors is then made as follows. If the computed F-statistics is greater then UCB then we reject H0 and conclude that the dependent variable and the regressors are cointegrated. On the other hand, if the computed F-statistics is lesser then the LCB, we fail to reject H0 , thus, signifying that there is no cointegration. However, if the computed F-statistics falls between the UCB and LCB, then the results are inconclusive. Although cointegration implies the presence of Granger causality, it does not necessarily identify the short-run direction of causality. For this purpose we use the Granger causality methodology to test the direction of causality. Hence, with the presence of cointegration, the Granger causality
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
641
relationships are written as Vector Error Correction Models (VECM) given below: LVAt = a0M + D97M +
k
βiM LVAt−i +
i=1
+
k
γiM LWt−i
i=1
k
k
λim LCAt−i
i=1
ξiM LOt−i + αiM ECTt−1 + εMt
(4)
i=1
In (4) to test whether lagged first differences of LO Granger cause LVA, we impose restrictions on all the lagged LO using the Wald or F-test. This is equivalent to testing the short-run causality of LO on LVA. Additionally, we can also test for exogeneity of the dependent variable. This implies that changes in the dependent variable are a function of the level of disequilibrium in the cointegrating relationship, captured by the error correction term (ECT), as well as by changes in other explanatory variables. The non-significance of the ECT is referred to as long-run noncausality, which is equivalent to saying that the variable is weakly exogenous with respect to the long-run parameters. This requires satisfying the null H0 : αiM = 0. Finally, the non-significance of all explanatory variables including the ECT term indicates the econometric strong-exogeneity of the dependent variable, that is, the absence of Granger causality. The strong-exogeneity test does not distinguish between the short-run and long-run causality, but it is a more restrictive test which indicates the overall causality in the system. 4. Empirical findings We examined the order of integration of the variables using the Phillips and Perron (1988) unit root test (PP test). The PP test was used because it allows for milder assumptions on the distribution of errors. Further, the test controls for higher order serial correlation in the series and is robust against heteroskedasticity. Table 1 reports the Phillip–Perron test. The evidence support that all variables are non-stationary in their level (except LO with time trend), but become stationary after taking the first difference. Hence we can conclude that all variables to be random walk indicating that all variables are integrated of order one, I(1). Table 2 reports the cointegration result. The lag length selection was determined using the Akaike Information Criterion (AIC). Due to the small sample size, we allow a maximum lag Table 1 Phillip–Perron tests. Log Levels
LVA LW LCA LO
Log differences
Without trend
With trend
Without trend
With trend
−0.8903 −1.248 −1.697 −0.499
−2.486 −2.494 −1.912 −3.325***
−6.594* −5.505* −3.786* −9.548*
−6.726* −5.502* −4.1914** −9.254*
Since LO is only marginally significant at 10% level we conclude that it is non-stationary at level. However, the ‘bounds test’ does not require the series to be integrated in the same order. * Denotes significant at 1%. ** Denotes significant at 5%. *** Denotes significant at 10%.
642
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
Table 2 Cointegration test: ARDL approach. F-statistics
Lag
6.350
2
95% critical value bounds I(0) 4.01
FMA (LVA|LW, LCA, LO)
I(1) 5.07
Critical values obtained from Pesaran et al. (2001), Table C1.v: Case V with unrestricted intercept and trend in the model. The underlying ARDL equation passes series of diagnostic test. We also performed cointegration test using LO, LW and LCA as the dependent variables, however, there is no evidence of any long-run relationship between the regressors and its determinants.
length of three. The optimal lag length is found to be two. The long-run relationship between the manufacturing value added output, capital, labor and openness was tested using the ‘bounds test’. Based on the ‘bounds test’ (in Table 2), the computed F-statistic is 6.35, and the 5% critical bounds, lower critical bound (LCB) and upper critical bound (UCB) are 3.79 and 4.85, respectively. Since the computed F-statistics is above the UCB at the 5% significance level, there exists a long-run relationship between manufacturing value added output and the regressors (capital, labor and openness). The LW, LCA and LO appear to be the long-run forcing variable for LVA based on the bounds test. It shows that in the long-run, Malaysian manufacturing sectors benefited through trade liberalization. The ECT captures the long-run impact (see Table 3). The ECT for (4) is −0.458 and was found to be statistically significant at the 5% significance level. This suggests that in the long-run LFA, LW and LO Granger cause LVA. This means that causality runs interactively through the ECT from LFA, LW and LO to LVA. The magnitude of the ECT term suggests that a deviation from the equilibrium level of LVA during the current period will be corrected by 46% in the next period. Additionally, this provides further support for our cointegration analysis earlier. It also provides support for the endogenity of the dependent variable (LVA) with respect to the long-run parameters. However, in the short-run, the direction of Granger causality may be different. The direction of causality in the short-run is given in Table 3. The results show that in the short-run, none of the lagged differences are significant except LW. This implies that openness does not Granger cause manufacturing growth in the short-run. Therefore, it leads support that countries can only leverage the benefits of openness when it is treated as a long term affair. Discontinuity of any liberalization policy and institutional setting or treating these policies and settings as short term initiative could not provide enough synergy to foster growth. Table 3 Granger causality test. Dependent variable
[F-statistics] short-run Granger causality
LVAt
LWt−1
[4.329]**
LCAt−1
[1.560]
(t-Statistics) weak-exogeneity test LOt−1
[1.407]
[F-statistics] strong-exogeneity test
ECTt−1 −0.4583*
[7.552]*
Selection of lag terms is based on AIC which was also subjected to series of diagnostic test. Plot of CUSUM and CUSUMSQ shows that the line stays within the 5% significant level, hence, indicating no evidence of any structural instability. * Denotes significant level at 1%. ** Denotes significant level at 5%.
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
643
Table 4 ARDL estimates of long-run elasticities. Variables
LVA (dependent variable)
Constant LCA LW LO t
−3.0463 0.246** 0.813* 0.417** −0.160**
The estimation passed all the diagnostic test namely, LM test of residual serial correlation, Ramsey’s RESET test for functional form, normality test and heteroscedasticity test. Lag selection was based on AIC. To examine the stability of the long-run coefficient with the short-run dynamics, we also followed Pesaran and Pesaran (1997) by applying the CUSUM and CUSUMSQ proposed by Brown, Durbin, and Evans (1975). It is found that the line stays within the 5% significant level, hence, indicating no evidence of any structural instability. Dummy variable is not shown. * Denotes significant at 1% level. ** Denotes significant at 5% level.
Having found a long-run relationship between manufacturing value added and trade openness when LVA serves as the dependent variable, we proceed to estimate the long-run elasticities. In other words, we investigate the impact of trade openness on manufacturing value added output. The long-run elasticities were estimated using the autoregressive distributed lag (ARDL) model. Based on the AIC, the ARDL [1,2,0,0] was found to be the optimal model. In the long-run trade openness has a significant positive impact on manufacturing value added in Malaysia (Table 4). A 1% increase in trade openness will result in 0.42% increase in real manufacturing value added. This result is contrary to Yanikkaya (2003) who find no significant relationship between openness and growth in Malaysia. Furthermore, Sarkar (2008) also finds no relationship between the two variables. The contradicting results might be due to national level analysis which lacks emphasis on the liberalized sector. Chang, Kaltani, and Loayza (2009) showed that institutional setting complements the effects of openness on growth. Ever since manufacturing sectors in Malaysia becomes the driver of growth, the sectors enjoyed fairly well established institutional setting (e.g. education, financial, labor market, and public infrastructure). This might have also complemented to the positive significant effects of openness. Therefore, as shown by our analysis, it is unlikely that Malaysia’s open policy could not have contributed to the growth of manufacturing sector. Additionally, in the long-run, labor and capital is significant in influencing the manufacturing value added. However, compared to capital investment, trade openness played far greater role in manufacturing sectors. More importantly, this has interesting implications for policy makers. It shows that even though investment is important without freer export and import regime the manufacturing growth could be at stake. With manufacturing sectors being more labor intensive (as shown by the greater impact of labor) and less technologically progressed (with negative t), trade openness (both exports and imports) are crucial. Additionally, due to lack of domestically produced intermediate inputs and capital goods, the liberalization of import regime played a major role. This is consistent with the argument that trade provides access to investment and intermediate goods for development process (Yanikkaya, 2003). 5. Policy implications and conclusions While, in general, issues of openness and growth are still debatable (Stiglitz, 2004), the results of this study provide country and sector specific insights and policy implications. Due to the
644
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
inappropriateness of cross-country analysis and policy responses derived from it (Pritchett, 1991), future studies may need to replaced or at least supplemented with country level analysis taking into consideration the sectoral differences. This study confirms that Malaysia is too small a market to allow manufacturing firms to enjoy an efficient scale of production as such opening to trade would expand their scale of production and help achieve growth. Most importantly, the ambiguity of whether openness promotes growth in manufacturing is clarified in the case of Malaysia. Therefore, policy makers in Malaysia need to accept the fact that openness has promoted manufacturing growth, as what the results of the study suggest. Export performance of the manufacturing sectors has been the key drivers for growth performance while importation of machinery is phenomenal in acquiring the embedded technologies and know-how and to some extent technology transfer from abroad. The significance of openness variable in this study reflects the same phenomena whereby with technological regress (as reflected by the negative contribution of technology variable in the study), the viable options are to tap the global markets and to learn from others via the open policy. Considering the above, the policy direction of Malaysia should emphasize on more liberal policies, with emphasis on how (which manufacturing sectors) and when openness is actually important. With limited domestic market, the export-led strategy is not an option, but a must for the sector to grow. Likewise, any shocks to the export markets could also weaken other sectors namely the service sector which contribute nearly 55% to the country’s gross domestic product. Additionally, policy makers should simultaneously also direct policy towards institutional arrangements that would help leverage the benefits of an open policy such as that which is observed in Korea, Taiwan, and Hong Kong. Additionally, it is suggestive that moving from import substitution policy to outward oriented policy especially for small nations is vital to promote growth. As such policy makers in such small countries need to consider and assess the implications of directing policies towards being a more open economy. The insignificant coefficient of short-run Granger causality might indicate that high openness growth does not necessarily generate economic growth in the short-run. The policy implication is that sustainable and prolonged openness policy is needed for countries to realize the benefits of openness. Therefore, small developing countries need to consider open policy as a long term plan of the country. Conversely, policy makers should understand that the benefits of openness also depend on whether the country has comparative advantages in sectors that are liberalized4 . Certainly, as shown by our analysis, Malaysian manufacturing sectors, at large, is liberalization and had the comparative advantages (Chandran, Deviga, & Karunagaran, 2004) where consequently it benefited the sector. This has also cast some doubt on the results of previous studies that does not show any significant relationship between openness and growth. As a whole, the policy implication is straightforward. First, policy makers need to rely on sectoral analysis to gain insights on the effects of openness on growth, and if possible using a more disaggregated sectoral data. Second, it is clear that policy makers should pursue outward looking strategies and to cushion the impact of any vulnerability to exports markets or imports, market diversification strategy (e.g. without depending on single market like the US) is needed. And lastly, in short term, countries can hardly gain from trade openness. 4 For instance, in Malaysia, electrical and electronics (E&E) exports contribute to nearly 63% to the total manufacturing export earnings in 2006 (Chandran, 2008). Manufacturing sectors in Malaysia are liberal in the sense that they have low export duties or in some cases virtually no export duties and increasing approval of export permits (see Devadason, 2006). Manufacturing sector in particular the E&E still maintains its export competitiveness in large number of products (Chandran et al., 2004).
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
645
An interesting finding of this study is that it concurs with some of the observation of the crosscountry findings in that the significance of trade openness in the overall samples of developing countries are solely attributed to the Asian nations. In this aspect, when considering Malaysian manufacturing sector, Malaysia stands out to be one of the beneficiaries. The finding of the cointegration test supports the existence of long-run relationship between trade openness and manufacturing growth. Likewise, the Granger causality confirms that causality runs from trade openness (and other determinants in the multivariate model) to manufacturing growth in the longrun. This strongly indicates that Malaysia’s manufacturing growth was partially the result of the government’s open policies. The policy implication and conclusion set in this study is based on the case of Malaysian manufacturing sector. However, before it can be generalized and served as a policy drive for other developing or middle income countries, further research is needed especially by taking into account the sector specific nature of the country. Unlike Malaysia, the results may not be applicable if manufacturing sectors in other countries are deterred by poor institutional setting or lacks the comparative advantages. Acknowledgement We are grateful to James Kunaratnam for his comments and editorial work. We also thank the anonymous referees and the Editor of this journal for their valuable comments. References Al-Yousif, & Yousif, K. (1999). On the role of exports in the economic growth of Malaysia: A multivariate analysis. International Economic Journal, 13(3), 67–75. Brown, R. L., Durbin, J., & Evans, J. M. (1975). Techniques for testing the constancy of regression relations over time. Journal of the Royal Statistics Society B, 37, 149–192. Chandran, V. G. R. (2008). Electronics cluster in Penang, Malaysia: The current challenges. Background paper of UNIDO’s cluster study. Vienna: United Nations Industrial Development Organization. Chandran, V. G. R., Deviga, V., & Karunagaran, M. (2004). Specialization and competitiveness of ASEAN and China trade: Is a treat or opportunity? Research report. Shah Alam: University Technology Mara. Chang, R., Kaltani, L., & Loayza, N. V. (2009). Openness can be good for growth: The role of policy complementarities. Journal of Development Economics, 90(1), 33–49. Cheung, Y. W., & Lai, K. S. (1993). Finite-sample sizes of Johansen’s likelihood ratio tests for cointegration. Oxford Bulletin of Economics and Statistics, 55(3), 313–328. Dawson, P. J. (2006). The export–income relationship and trade liberalisation in Bangladesh. Journal of Policy Modeling, 28, 889–896. Devadason, E. (2006). Trade protection and employment in manufacturing: The case of Malaysia. Malaysian Journal of Economic Studies, XXXXIII(1–2), 69–84. Dollar, D. (1992). Outward-oriented developing countries really do grow more rapidly: Evidence from 95 LDCs, 1976–85. Economic Development and Cultural Change, 40(3), 523–544. Dutta, D. & Ahmed, N. (2001). Trade liberalization and industrial growth in Pakistan: A Cointegration Analysis. Working Paper. Australia: University of Sydney. Edwards, S. (1992). Trade orientation, distortions and growth in developing countries. Journal of Development Economies, 39(1), 31–57. Edwards, S. (1998). Openness, productivity and growth: What do we really know? The Economic Journal, 108, 383–398. Engle, R. F., & Granger, C. W. J. (1987). Cointegration and error correction: Representation, estimation, testing. Econometrica, 55, 251–276. Granger, C. W. J., & Newbold. (1974). Spurious regressions in econometrics. Journal of Econometrics, 111–120. Harrison, A. (1996). Openness and growth: A time-series, cross-country analysis for developing countries. Journal of Development Economics, 48, 419–447.
646
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
Hwang, I. (1998). Long-Run determinant of Korean economic growth: Empirical evidence from manufacturing. Applied Economics, 30(3), 391–405. Johansen, S. (1988). Statistical analysis of cointegration vectors. Journal of Economics Dynamics and Control, 12, 231–254. Johansen, S., & Juselius, K. (1990). Maximum likelihood estimation and inference on cointegration with applications to demand for money. Oxford Bulletin of Economics and Statistics, 52, 169–210. Johansen, S., & Juselius, K. (1992). Testing structural hypothesis in a multivariate cointegration analysis of the PPP and the UIP for UK. Journal of Econometrics, 53, 211–244. Jonsson, G., & Subramanian, A. (2001). Dynamic gains from trade: Evidence from South Africa. IMF Staff Papers, 48(1), 197–224. Kajiwara, H. (1994). The effects of trade and foreign Investment liberalization policy on productivity in the Philippines. The Developing Economies, XXXII-4, 492–507. Lloyd, P. J., & MacLaren, D. (2002). Measures of trade openness using CGE analysis. Journal of Policy Modeling, 24, 67–81. Lloyd, P. J., & MacLaren, D. (2000). Openness and growth in East Asia after the Asian crisis. Journal of Asian Economies, 11(1), 89–105. Mah, J. S. (2000). An empirical examination of the disaggregate import demand of Korea—the case information technology products. Journal of Asian Economics, 11, 237–244. Mahadevan, R. (2002). A DEA approach to understanding the productivity growth of Malaysia’s manufacturing industries. Asia Pacific Journal of Management, 19, 587–600. Malaysia (1971). The First Outline Perspective Plan 1971–1990. Kuala Lumpur: Government Printer. Malaysia (1990). The Second Outline Perspective Plan 1991–2000. Kuala Lumpur: Government Printer. Malaysia (2001). Eight Malaysia Plan (2001–2005). Kuala Lumpur: Government Printer. Malaysia (2009). Economic Report 2007/2008. Kuala Lumpur: Ministry of Finance. Osada, H. (1994). Trade liberalization and FDI incentives in Indonesia: The Impact on industrial productivity. The Developing Economies, XXXII-4, 479–491. Pesaran, M. H., & Pesaran, B. (1997). Microfit 4.0: Interactive econometric analysis. Oxford: Oxford University Press. Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approach to the level relationship. Journal of Applied Econometrics, 16, 289–329. Phillips, P. C. B. (1986). Understanding spurious regressions in econometrics. Journal of Econometrics, 33, 311–340. Phillips, P. C. B., & Perron, P. (1988). Testing for a unit root in time series regression. Biometrika, 75(2), 335–346. Pritchett, L. (1991). Measuring outward orientation in developing countries. Can it be done? Trade Policy Working Papers. The World Bank (WPS 566). Reppas, P. A., & Christopoulos, D. K. (2005). The export-output growth nexus: Evidence from African and Asian countries. Journal of Policy Modeling, 27, 929–940. Rodríguez, F., & Rodrik, D. (2001). Trade policy and economic growth: A skeptics guide to the cross-national evidence. In B. Bernanke, & K. Rogoff (Eds.), NBER macroeconomics annual 2000 (pp. 261–325). Cambridge, MA: MIT Press. Ruíz Estrada, M. A., & Yap, S. F. (2006). The openness growth monitoring model (OGM-Model). Journal of Policy Modeling, 28, 235–246. Sachs, J. D., & Warner, A. (1995). Economic reform and the process of global integration. Brookings Papers on Economic Activity, 1, 1–118. Salehezadeh, Z., & Henneberry, S. R. (2002). The economic impacts of trade liberalization and factor mobility: The case of the Philippines. Journal of Policy Modeling, 24(5), 483–489. Sarkar, P. (2008). Trade openness and growth: Is there any link? Journal of Economic Issues, XLII(3), 763–785. Stiglitz, J. E. (2004). Globalization and growth in emerging markets. Journal of Policy Modeling, 26, 465–484. Thirwall, A. P. (1994). Growth and development. London: The Macmillan Press Ltd. Toda, H. Y., & Phillips, P. C. B. (1993). Vector autoregressions and causality. Econometrica, 61(6), 1367–1393. Tsen, W. H. (2005). Openness, financial development and economic growth in Malaysia. International Journal of Business and Society, 6(1), 93–121. Urata, S. (1994). Trade liberalization and productivity growth in Asia: Introduction and major findings. The Developing Economics, XXXXII(4), 363–372. Urata, S., & Yokota, K. (1994). Trade liberalization and productivity growth in Thailand. The Developing Economies, XXXII(4), 444–459.
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
647
Wadud, I. K. M., & Nair, M. (2003). Impact of liberalisation on the competitiveness of the Australian clothing industry: An empirical study. International Journal of Management, 20(3), 325–334. Weinhold, D., & Rauch, J. E. (1999). Openness, specialization, and productivity growth in less developed countries. Canadian Journal of Economics, 32(4), 1010–1027. World Bank. (1993). The East Asian miracle: Economic growth and public policy. New York: Oxford University Press. Yanikkaya, H. (2003). Trade openness and economic growth: A cross-country empirical investigation. Journal of Development Economics, 72, 57–89.
Journal of Policy Modeling 31 (2009) 637–647
Trade openness and manufacturing growth in Malaysia V.G.R. Chandran a,b,∗ , Munusamy c,1 a b
Department of Economics, KM 12 Jalan Muar, University Technology MARA, 85009 Segamat, Johor, Malaysia Department of Economics, Faculty of Economics and Administration, University of Malaya, 50603 Kuala Lumpur, Malaysia c Department of Business, Nilai International University College, 71800 Putra Nilai, Negeri Sembilan, Malaysia Received 1 November 2008; received in revised form 1 May 2009; accepted 1 June 2009 Available online 8 July 2009
Abstract This study investigates the long-run relationship between trade openness and manufacturing growth and further assesses the causal relationship between these variables. Contrary to some scholars belief that at national level, openness does not contribute to growth in Malaysia, our sector specific analysis suggest otherwise. In this aspect, we believe that in any attempt to establish relationship between openness and growth, the analysis should be sector specific since it is more relevant as well as assures a meaningful insight for policy makers. The results suggest that in the long-run, trade openness is positively related to manufacturing growth in Malaysia. Furthermore, the results also suggest that openness should be viewed as the long term policy initiative for the sector to benefit. Therefore, the policy direction for Malaysian manufacturing sectors should focus on long term trade openness policies. Nevertheless, to ensure sustainability, emphasis should be placed on how (which manufacturing sub-sectors) or when openness is actually important. Importantly, policy makers and scholars should understand that leveraging the benefits of openness also depend on whether the liberalized sector has the comparative advantage. © 2009 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. JEL classification: F41; C22 Keywords: Trade openness; Autoregressive distributed lag; Manufacturing growth; Granger causality
∗ Corresponding author at: Department of Economics, KM 12 Jalan Muar, University Technology MARA, 85009 Segamat, Johor, Malaysia. Tel.: +60 17 6843705. E-mail addresses: [email protected] (V.G.R. Chandran), [email protected] ( Munusamy). 1 Tel.: +60 12 6142596.
0161-8938/$ – see front matter © 2009 Society for Policy Modeling. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.jpolmod.2009.06.002
638
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
1. Introduction Trade openness has emerged as the main argument among economists and policy makers in explaining the growth phenomena in developing countries (Dawson, 2006; Dutta & Ahmed, 2001; Edwards, 1992; Salehezadeh & Henneberry, 2002; Weinhold & Rauch, 1999). Besides, due to continuous interest on the issue, new methods were also proposed (Lloyd & MacLaren, 2002; Ruíz Estrada & Yap, 2006). The positive contribution of trade openness towards growth stemmed from the notion that liberalization increases specialization and division of labor thus improving productivity and export capability as well as economic performance. In addition, with greater efficiency as a result of trade openness, many of the developing countries followed suit with the export-led strategies. It is widely recognized that trade openness has a positive effect towards economic growth. It is found that countries with more trade openness relatively outperformed the countries, with less openness (Thirwall, 1994; World Bank, 1993). A study by Lloyd and MacLaren (2000) among the East Asian economies supported a similar opinion that the rapid growth was largely caused by East Asia’s economic openness. Other noteworthy studies supporting the openness and growth relationship include Urata and Yokota (1994), Osada (1994), Kajiwara (1994), Hwang (1998), Edwards (1998) and Jonsson and Subramanian (2001). In contrast, some scholars (Harrison, 1996; Rodríguez & Rodrik, 2001), however, have been more reserve in supporting the openness-led growth nexus. Although there are a considerable number of studies examining the relationship between trade liberalization and growth in developing countries (including Dollar, 1992; Edwards, 1992; Sachs & Warner, 1995; Sarkar, 2008; Yanikkaya, 2003), these studies are still far from complete. First, in the case of Malaysia, previous empirical studies relying on cross-country panel data analysis showed mixed results. Second, while many studies have provided empirical evidence on the impact of trade openness and economic growth at nation level, analysis within the framework on manufacturing industries are solely lacking. The lack of analysis at sectoral level may have contributed to the empirically mixed results reported in the previous literature. This study shows that giving emphasis to the wrong sectors or treating all countries to be homogenous in nature may lead to biasness. Contrary to the previous studies (e.g. Sarkar, 2008; Yanikkaya, 2003), it shows that in the case of Malaysia, openness is important for manufacturing sectors. The preferred country and sector specific analysis in this study captured and accounted for the complexity of economic environment and histories of the manufacturing industry in Malaysia. Besides, Malaysia appears to be a suitable case study given the fact that it is one of the highest growth open economies among the developing countries. Malaysian gross domestic product (GDP) grew at an average rate of 6.7% between 1971 and 1990, while during the period between 1990 and 2000 Malaysia had the highest growth rate averaging 8.1% per annum outperforming other ASEAN economies (Malaysia, 1971, 1990, 2001). Malaysia was one of the most active among the ASEAN countries in liberalizing its investment regime in the manufacturing sector during the 1980s and 1990s. This policy offered many incentives including pioneer status tax holidays, expanded investment tax allowances for expansion projects, tax deduction for export promotions, the establishment of Free Trade Zones and other types of incentives to attract and draw FDI. Malaysia, compared to other ASEAN economies has somehow progressed well with its outward orientated strategies. Furthermore, the tariff rate in Malaysia has declined considerably over the years (Urata, 1994). The manufacturing sector has been regarded as the main driver for export performance as well as for the impressive economic growth in Malaysia. In 2006, the strong growth in manufacturing output (7.1%) contributed around 31.3% to real GDP growth (Malaysia, 2009). Hence, interference drawn from this study provides general understanding and guidance for policy formulation.
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
639
Third, many studies have ignored the causation between the two variables, which is deemed important in econometric exercise and for policy implication. Generally, higher growing nations can also exhibit higher level of openness2 . Hence, we tested the causality between trade openness and manufacturing growth in a multivariate system by incorporating other basic determinants of manufacturing growth such as labor and capital. This was due to the argument that bivariate causality may lead to biasness due to the omission of variable phenomenon (Al-Yousif and Yousif, 1999). Furthermore, from the theoretical point of view the inclusion of other important determinants (such as labor and capital) would avoid potential mis-specification. Lastly, we adopted a more recent cointegration test called the bounds test developed by Pesaran, Shin, and Smith (2001) to establish if the variables are co-moving. Mah (2000) cautioned on the inference problems inherent with cointegrating analysis that has a small sample size3 . The Error Correction Model (Engle & Granger, 1987), Johansen (1988) and Johansen and Juselius (1990) methods are unreliable for studies that have small samples. Hence, the proposed method has a major advantage over the traditional cointegration proposed by Engle and Granger (1987), Johansen (1988) and Johansen and Juselius (1992). The bounds test is robust in dealing with a small number of observations and does not require the regressors to be in the same order of integration. Therefore, serious questions concerning the robustness of the cointegration tests could be limited compared to other studies that used Johansen’s method2 for small observations (less than 100). The primary objective of this study is to address the above gaps and to re-examine the impact of trade openness on the Malaysian manufacturing sector. The reminder of the study is organized as follows. The data and empirical model applied in this study is described in Section 2. Section 3 describes the methodology of the study. Section 4 presents the empirical findings while Section 5 discusses policy implications and the conclusion. 2. Data and model specification Annual data was used from 1970 to 2003, which was obtained from the ‘Malaysia Economic Statistics-Time Series’ published by the Department of Statistics, Malaysia. The series include manufacturing value added output, net fixed capital, number of labor and trade openness. Due to data limitation for the computation of capital stock, following Mahadevan (2002), we used the fixed assets as the proxy for capital. Manufacturing value added was deflated using the manufacturing producers price index while fixed assets was deflated using gross domestic fixed capital formation deflator with 2000 as the base year. The trade openness is calculated as a ratio of manufacturing import plus export to manufacturing output. Based on a theoretical framework (Harrison, 1996; Tsen, 2005; Wadud & Nair, 2003), the relationship between manufacturing value added, capital, labor and openness in the multivariate model can be specified as follows: LVAt = f (LCAt , LWt , LOt , t)
(1)
where LVAt , LCAt , LWt and LOt are logarithmic manufacturing value added, fixed capital, labor and trade openness, respectively. The variable t in (1) is the linear trend representing the Hicks neutral technical progress. We included a dummy variable to account for the Asian financial crisis in 1997/1998.
2
For instance see Reppas and Christopoulos (2005) for export-led growth and growth-led exports hypothesis. Johansen’s cointegration analysis requires large observation to be accurate. Scholars, e.g. Cheung and Lai (1993) suggested the adjustments to the critical values are necessary when applied to sample sizes of 100 or smaller. 3
640
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
3. Methodology 3.1. Unit root, cointegration and Granger causality The testing procedure involves three steps. The series of variables in our study need to induce stationary for the estimated results to be reliable and unbiased as to avoid spurious regression (Granger & Newbold, 1974; Phillips, 1986). Thus, we first test the stationary of the series by using the Phillip–Perron (PP) test. The next step in the test involves testing the cointegration in the model. The works of Engle and Granger (1987) and Toda and Phillips (1993) have shown that ignoring the existence of cointegration in the series could lead to serious model mis-specification. There are two ways of performing the cointegration test; the Johansen cointegration using maximum-likelihood method (Johansen, 1988) and the fairly new method known as “bounds testing approach” suggested by Pesaran et al. (2001). Owing to the limited size of observation, the bounds test is preferred over the Johansen method, which is mainly appropriate for a large sample size. Thus, in this study, we perform the cointegration test using the autoregressive distributive lag (ARDL) method proposed by Pesaran et al. (2001). This involves testing the following Unrestricted Error Correction Models (UECM): LVAt = a0M + δM t + D97M +
n i=1
+
n i=1
γiM LWt−i
n
βiM LVAt−i +
n
λim LCAt−i
i=1
ξiM LOt−i + θ1M LVAt−1 + θ2M LCA + θ3M LW
i=1
+ θ4M LOt−1 + εMt
(2)
In order to test the absence of a long-run relationship in (2), we conduct a Wald-type (F-test) coefficient restriction test, which entails testing the following null hypothesis. H0 : θ1M = θ2M = θ3M = θ4m = 0
(3)
This test is known as the bounds test (Pesaran et al., 2001) and the computed F-statistics under the null hypothesis (3) is donated by F(LVA|LO, LCA, LW). The asymptotic distributions of the test statistics are non-standard regardless of whether the variables are I(0) or I(1). For this purpose, Pesaran et al. (2001) computed two sets of asymptotic critical values where the first set assumes variables to be I(0) and the other as I(1), which are known as lower bounds (LCB) and upper bounds critical values (UCB), respectively. A decision on whether cointegration exists between the dependent variable and its regressors is then made as follows. If the computed F-statistics is greater then UCB then we reject H0 and conclude that the dependent variable and the regressors are cointegrated. On the other hand, if the computed F-statistics is lesser then the LCB, we fail to reject H0 , thus, signifying that there is no cointegration. However, if the computed F-statistics falls between the UCB and LCB, then the results are inconclusive. Although cointegration implies the presence of Granger causality, it does not necessarily identify the short-run direction of causality. For this purpose we use the Granger causality methodology to test the direction of causality. Hence, with the presence of cointegration, the Granger causality
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
641
relationships are written as Vector Error Correction Models (VECM) given below: LVAt = a0M + D97M +
k
βiM LVAt−i +
i=1
+
k
γiM LWt−i
i=1
k
k
λim LCAt−i
i=1
ξiM LOt−i + αiM ECTt−1 + εMt
(4)
i=1
In (4) to test whether lagged first differences of LO Granger cause LVA, we impose restrictions on all the lagged LO using the Wald or F-test. This is equivalent to testing the short-run causality of LO on LVA. Additionally, we can also test for exogeneity of the dependent variable. This implies that changes in the dependent variable are a function of the level of disequilibrium in the cointegrating relationship, captured by the error correction term (ECT), as well as by changes in other explanatory variables. The non-significance of the ECT is referred to as long-run noncausality, which is equivalent to saying that the variable is weakly exogenous with respect to the long-run parameters. This requires satisfying the null H0 : αiM = 0. Finally, the non-significance of all explanatory variables including the ECT term indicates the econometric strong-exogeneity of the dependent variable, that is, the absence of Granger causality. The strong-exogeneity test does not distinguish between the short-run and long-run causality, but it is a more restrictive test which indicates the overall causality in the system. 4. Empirical findings We examined the order of integration of the variables using the Phillips and Perron (1988) unit root test (PP test). The PP test was used because it allows for milder assumptions on the distribution of errors. Further, the test controls for higher order serial correlation in the series and is robust against heteroskedasticity. Table 1 reports the Phillip–Perron test. The evidence support that all variables are non-stationary in their level (except LO with time trend), but become stationary after taking the first difference. Hence we can conclude that all variables to be random walk indicating that all variables are integrated of order one, I(1). Table 2 reports the cointegration result. The lag length selection was determined using the Akaike Information Criterion (AIC). Due to the small sample size, we allow a maximum lag Table 1 Phillip–Perron tests. Log Levels
LVA LW LCA LO
Log differences
Without trend
With trend
Without trend
With trend
−0.8903 −1.248 −1.697 −0.499
−2.486 −2.494 −1.912 −3.325***
−6.594* −5.505* −3.786* −9.548*
−6.726* −5.502* −4.1914** −9.254*
Since LO is only marginally significant at 10% level we conclude that it is non-stationary at level. However, the ‘bounds test’ does not require the series to be integrated in the same order. * Denotes significant at 1%. ** Denotes significant at 5%. *** Denotes significant at 10%.
642
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
Table 2 Cointegration test: ARDL approach. F-statistics
Lag
6.350
2
95% critical value bounds I(0) 4.01
FMA (LVA|LW, LCA, LO)
I(1) 5.07
Critical values obtained from Pesaran et al. (2001), Table C1.v: Case V with unrestricted intercept and trend in the model. The underlying ARDL equation passes series of diagnostic test. We also performed cointegration test using LO, LW and LCA as the dependent variables, however, there is no evidence of any long-run relationship between the regressors and its determinants.
length of three. The optimal lag length is found to be two. The long-run relationship between the manufacturing value added output, capital, labor and openness was tested using the ‘bounds test’. Based on the ‘bounds test’ (in Table 2), the computed F-statistic is 6.35, and the 5% critical bounds, lower critical bound (LCB) and upper critical bound (UCB) are 3.79 and 4.85, respectively. Since the computed F-statistics is above the UCB at the 5% significance level, there exists a long-run relationship between manufacturing value added output and the regressors (capital, labor and openness). The LW, LCA and LO appear to be the long-run forcing variable for LVA based on the bounds test. It shows that in the long-run, Malaysian manufacturing sectors benefited through trade liberalization. The ECT captures the long-run impact (see Table 3). The ECT for (4) is −0.458 and was found to be statistically significant at the 5% significance level. This suggests that in the long-run LFA, LW and LO Granger cause LVA. This means that causality runs interactively through the ECT from LFA, LW and LO to LVA. The magnitude of the ECT term suggests that a deviation from the equilibrium level of LVA during the current period will be corrected by 46% in the next period. Additionally, this provides further support for our cointegration analysis earlier. It also provides support for the endogenity of the dependent variable (LVA) with respect to the long-run parameters. However, in the short-run, the direction of Granger causality may be different. The direction of causality in the short-run is given in Table 3. The results show that in the short-run, none of the lagged differences are significant except LW. This implies that openness does not Granger cause manufacturing growth in the short-run. Therefore, it leads support that countries can only leverage the benefits of openness when it is treated as a long term affair. Discontinuity of any liberalization policy and institutional setting or treating these policies and settings as short term initiative could not provide enough synergy to foster growth. Table 3 Granger causality test. Dependent variable
[F-statistics] short-run Granger causality
LVAt
LWt−1
[4.329]**
LCAt−1
[1.560]
(t-Statistics) weak-exogeneity test LOt−1
[1.407]
[F-statistics] strong-exogeneity test
ECTt−1 −0.4583*
[7.552]*
Selection of lag terms is based on AIC which was also subjected to series of diagnostic test. Plot of CUSUM and CUSUMSQ shows that the line stays within the 5% significant level, hence, indicating no evidence of any structural instability. * Denotes significant level at 1%. ** Denotes significant level at 5%.
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
643
Table 4 ARDL estimates of long-run elasticities. Variables
LVA (dependent variable)
Constant LCA LW LO t
−3.0463 0.246** 0.813* 0.417** −0.160**
The estimation passed all the diagnostic test namely, LM test of residual serial correlation, Ramsey’s RESET test for functional form, normality test and heteroscedasticity test. Lag selection was based on AIC. To examine the stability of the long-run coefficient with the short-run dynamics, we also followed Pesaran and Pesaran (1997) by applying the CUSUM and CUSUMSQ proposed by Brown, Durbin, and Evans (1975). It is found that the line stays within the 5% significant level, hence, indicating no evidence of any structural instability. Dummy variable is not shown. * Denotes significant at 1% level. ** Denotes significant at 5% level.
Having found a long-run relationship between manufacturing value added and trade openness when LVA serves as the dependent variable, we proceed to estimate the long-run elasticities. In other words, we investigate the impact of trade openness on manufacturing value added output. The long-run elasticities were estimated using the autoregressive distributed lag (ARDL) model. Based on the AIC, the ARDL [1,2,0,0] was found to be the optimal model. In the long-run trade openness has a significant positive impact on manufacturing value added in Malaysia (Table 4). A 1% increase in trade openness will result in 0.42% increase in real manufacturing value added. This result is contrary to Yanikkaya (2003) who find no significant relationship between openness and growth in Malaysia. Furthermore, Sarkar (2008) also finds no relationship between the two variables. The contradicting results might be due to national level analysis which lacks emphasis on the liberalized sector. Chang, Kaltani, and Loayza (2009) showed that institutional setting complements the effects of openness on growth. Ever since manufacturing sectors in Malaysia becomes the driver of growth, the sectors enjoyed fairly well established institutional setting (e.g. education, financial, labor market, and public infrastructure). This might have also complemented to the positive significant effects of openness. Therefore, as shown by our analysis, it is unlikely that Malaysia’s open policy could not have contributed to the growth of manufacturing sector. Additionally, in the long-run, labor and capital is significant in influencing the manufacturing value added. However, compared to capital investment, trade openness played far greater role in manufacturing sectors. More importantly, this has interesting implications for policy makers. It shows that even though investment is important without freer export and import regime the manufacturing growth could be at stake. With manufacturing sectors being more labor intensive (as shown by the greater impact of labor) and less technologically progressed (with negative t), trade openness (both exports and imports) are crucial. Additionally, due to lack of domestically produced intermediate inputs and capital goods, the liberalization of import regime played a major role. This is consistent with the argument that trade provides access to investment and intermediate goods for development process (Yanikkaya, 2003). 5. Policy implications and conclusions While, in general, issues of openness and growth are still debatable (Stiglitz, 2004), the results of this study provide country and sector specific insights and policy implications. Due to the
644
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
inappropriateness of cross-country analysis and policy responses derived from it (Pritchett, 1991), future studies may need to replaced or at least supplemented with country level analysis taking into consideration the sectoral differences. This study confirms that Malaysia is too small a market to allow manufacturing firms to enjoy an efficient scale of production as such opening to trade would expand their scale of production and help achieve growth. Most importantly, the ambiguity of whether openness promotes growth in manufacturing is clarified in the case of Malaysia. Therefore, policy makers in Malaysia need to accept the fact that openness has promoted manufacturing growth, as what the results of the study suggest. Export performance of the manufacturing sectors has been the key drivers for growth performance while importation of machinery is phenomenal in acquiring the embedded technologies and know-how and to some extent technology transfer from abroad. The significance of openness variable in this study reflects the same phenomena whereby with technological regress (as reflected by the negative contribution of technology variable in the study), the viable options are to tap the global markets and to learn from others via the open policy. Considering the above, the policy direction of Malaysia should emphasize on more liberal policies, with emphasis on how (which manufacturing sectors) and when openness is actually important. With limited domestic market, the export-led strategy is not an option, but a must for the sector to grow. Likewise, any shocks to the export markets could also weaken other sectors namely the service sector which contribute nearly 55% to the country’s gross domestic product. Additionally, policy makers should simultaneously also direct policy towards institutional arrangements that would help leverage the benefits of an open policy such as that which is observed in Korea, Taiwan, and Hong Kong. Additionally, it is suggestive that moving from import substitution policy to outward oriented policy especially for small nations is vital to promote growth. As such policy makers in such small countries need to consider and assess the implications of directing policies towards being a more open economy. The insignificant coefficient of short-run Granger causality might indicate that high openness growth does not necessarily generate economic growth in the short-run. The policy implication is that sustainable and prolonged openness policy is needed for countries to realize the benefits of openness. Therefore, small developing countries need to consider open policy as a long term plan of the country. Conversely, policy makers should understand that the benefits of openness also depend on whether the country has comparative advantages in sectors that are liberalized4 . Certainly, as shown by our analysis, Malaysian manufacturing sectors, at large, is liberalization and had the comparative advantages (Chandran, Deviga, & Karunagaran, 2004) where consequently it benefited the sector. This has also cast some doubt on the results of previous studies that does not show any significant relationship between openness and growth. As a whole, the policy implication is straightforward. First, policy makers need to rely on sectoral analysis to gain insights on the effects of openness on growth, and if possible using a more disaggregated sectoral data. Second, it is clear that policy makers should pursue outward looking strategies and to cushion the impact of any vulnerability to exports markets or imports, market diversification strategy (e.g. without depending on single market like the US) is needed. And lastly, in short term, countries can hardly gain from trade openness. 4 For instance, in Malaysia, electrical and electronics (E&E) exports contribute to nearly 63% to the total manufacturing export earnings in 2006 (Chandran, 2008). Manufacturing sectors in Malaysia are liberal in the sense that they have low export duties or in some cases virtually no export duties and increasing approval of export permits (see Devadason, 2006). Manufacturing sector in particular the E&E still maintains its export competitiveness in large number of products (Chandran et al., 2004).
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
645
An interesting finding of this study is that it concurs with some of the observation of the crosscountry findings in that the significance of trade openness in the overall samples of developing countries are solely attributed to the Asian nations. In this aspect, when considering Malaysian manufacturing sector, Malaysia stands out to be one of the beneficiaries. The finding of the cointegration test supports the existence of long-run relationship between trade openness and manufacturing growth. Likewise, the Granger causality confirms that causality runs from trade openness (and other determinants in the multivariate model) to manufacturing growth in the longrun. This strongly indicates that Malaysia’s manufacturing growth was partially the result of the government’s open policies. The policy implication and conclusion set in this study is based on the case of Malaysian manufacturing sector. However, before it can be generalized and served as a policy drive for other developing or middle income countries, further research is needed especially by taking into account the sector specific nature of the country. Unlike Malaysia, the results may not be applicable if manufacturing sectors in other countries are deterred by poor institutional setting or lacks the comparative advantages. Acknowledgement We are grateful to James Kunaratnam for his comments and editorial work. We also thank the anonymous referees and the Editor of this journal for their valuable comments. References Al-Yousif, & Yousif, K. (1999). On the role of exports in the economic growth of Malaysia: A multivariate analysis. International Economic Journal, 13(3), 67–75. Brown, R. L., Durbin, J., & Evans, J. M. (1975). Techniques for testing the constancy of regression relations over time. Journal of the Royal Statistics Society B, 37, 149–192. Chandran, V. G. R. (2008). Electronics cluster in Penang, Malaysia: The current challenges. Background paper of UNIDO’s cluster study. Vienna: United Nations Industrial Development Organization. Chandran, V. G. R., Deviga, V., & Karunagaran, M. (2004). Specialization and competitiveness of ASEAN and China trade: Is a treat or opportunity? Research report. Shah Alam: University Technology Mara. Chang, R., Kaltani, L., & Loayza, N. V. (2009). Openness can be good for growth: The role of policy complementarities. Journal of Development Economics, 90(1), 33–49. Cheung, Y. W., & Lai, K. S. (1993). Finite-sample sizes of Johansen’s likelihood ratio tests for cointegration. Oxford Bulletin of Economics and Statistics, 55(3), 313–328. Dawson, P. J. (2006). The export–income relationship and trade liberalisation in Bangladesh. Journal of Policy Modeling, 28, 889–896. Devadason, E. (2006). Trade protection and employment in manufacturing: The case of Malaysia. Malaysian Journal of Economic Studies, XXXXIII(1–2), 69–84. Dollar, D. (1992). Outward-oriented developing countries really do grow more rapidly: Evidence from 95 LDCs, 1976–85. Economic Development and Cultural Change, 40(3), 523–544. Dutta, D. & Ahmed, N. (2001). Trade liberalization and industrial growth in Pakistan: A Cointegration Analysis. Working Paper. Australia: University of Sydney. Edwards, S. (1992). Trade orientation, distortions and growth in developing countries. Journal of Development Economies, 39(1), 31–57. Edwards, S. (1998). Openness, productivity and growth: What do we really know? The Economic Journal, 108, 383–398. Engle, R. F., & Granger, C. W. J. (1987). Cointegration and error correction: Representation, estimation, testing. Econometrica, 55, 251–276. Granger, C. W. J., & Newbold. (1974). Spurious regressions in econometrics. Journal of Econometrics, 111–120. Harrison, A. (1996). Openness and growth: A time-series, cross-country analysis for developing countries. Journal of Development Economics, 48, 419–447.
646
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
Hwang, I. (1998). Long-Run determinant of Korean economic growth: Empirical evidence from manufacturing. Applied Economics, 30(3), 391–405. Johansen, S. (1988). Statistical analysis of cointegration vectors. Journal of Economics Dynamics and Control, 12, 231–254. Johansen, S., & Juselius, K. (1990). Maximum likelihood estimation and inference on cointegration with applications to demand for money. Oxford Bulletin of Economics and Statistics, 52, 169–210. Johansen, S., & Juselius, K. (1992). Testing structural hypothesis in a multivariate cointegration analysis of the PPP and the UIP for UK. Journal of Econometrics, 53, 211–244. Jonsson, G., & Subramanian, A. (2001). Dynamic gains from trade: Evidence from South Africa. IMF Staff Papers, 48(1), 197–224. Kajiwara, H. (1994). The effects of trade and foreign Investment liberalization policy on productivity in the Philippines. The Developing Economies, XXXII-4, 492–507. Lloyd, P. J., & MacLaren, D. (2002). Measures of trade openness using CGE analysis. Journal of Policy Modeling, 24, 67–81. Lloyd, P. J., & MacLaren, D. (2000). Openness and growth in East Asia after the Asian crisis. Journal of Asian Economies, 11(1), 89–105. Mah, J. S. (2000). An empirical examination of the disaggregate import demand of Korea—the case information technology products. Journal of Asian Economics, 11, 237–244. Mahadevan, R. (2002). A DEA approach to understanding the productivity growth of Malaysia’s manufacturing industries. Asia Pacific Journal of Management, 19, 587–600. Malaysia (1971). The First Outline Perspective Plan 1971–1990. Kuala Lumpur: Government Printer. Malaysia (1990). The Second Outline Perspective Plan 1991–2000. Kuala Lumpur: Government Printer. Malaysia (2001). Eight Malaysia Plan (2001–2005). Kuala Lumpur: Government Printer. Malaysia (2009). Economic Report 2007/2008. Kuala Lumpur: Ministry of Finance. Osada, H. (1994). Trade liberalization and FDI incentives in Indonesia: The Impact on industrial productivity. The Developing Economies, XXXII-4, 479–491. Pesaran, M. H., & Pesaran, B. (1997). Microfit 4.0: Interactive econometric analysis. Oxford: Oxford University Press. Pesaran, M. H., Shin, Y., & Smith, R. J. (2001). Bounds testing approach to the level relationship. Journal of Applied Econometrics, 16, 289–329. Phillips, P. C. B. (1986). Understanding spurious regressions in econometrics. Journal of Econometrics, 33, 311–340. Phillips, P. C. B., & Perron, P. (1988). Testing for a unit root in time series regression. Biometrika, 75(2), 335–346. Pritchett, L. (1991). Measuring outward orientation in developing countries. Can it be done? Trade Policy Working Papers. The World Bank (WPS 566). Reppas, P. A., & Christopoulos, D. K. (2005). The export-output growth nexus: Evidence from African and Asian countries. Journal of Policy Modeling, 27, 929–940. Rodríguez, F., & Rodrik, D. (2001). Trade policy and economic growth: A skeptics guide to the cross-national evidence. In B. Bernanke, & K. Rogoff (Eds.), NBER macroeconomics annual 2000 (pp. 261–325). Cambridge, MA: MIT Press. Ruíz Estrada, M. A., & Yap, S. F. (2006). The openness growth monitoring model (OGM-Model). Journal of Policy Modeling, 28, 235–246. Sachs, J. D., & Warner, A. (1995). Economic reform and the process of global integration. Brookings Papers on Economic Activity, 1, 1–118. Salehezadeh, Z., & Henneberry, S. R. (2002). The economic impacts of trade liberalization and factor mobility: The case of the Philippines. Journal of Policy Modeling, 24(5), 483–489. Sarkar, P. (2008). Trade openness and growth: Is there any link? Journal of Economic Issues, XLII(3), 763–785. Stiglitz, J. E. (2004). Globalization and growth in emerging markets. Journal of Policy Modeling, 26, 465–484. Thirwall, A. P. (1994). Growth and development. London: The Macmillan Press Ltd. Toda, H. Y., & Phillips, P. C. B. (1993). Vector autoregressions and causality. Econometrica, 61(6), 1367–1393. Tsen, W. H. (2005). Openness, financial development and economic growth in Malaysia. International Journal of Business and Society, 6(1), 93–121. Urata, S. (1994). Trade liberalization and productivity growth in Asia: Introduction and major findings. The Developing Economics, XXXXII(4), 363–372. Urata, S., & Yokota, K. (1994). Trade liberalization and productivity growth in Thailand. The Developing Economies, XXXII(4), 444–459.
V.G.R. Chandran, Munusamy / Journal of Policy Modeling 31 (2009) 637–647
647
Wadud, I. K. M., & Nair, M. (2003). Impact of liberalisation on the competitiveness of the Australian clothing industry: An empirical study. International Journal of Management, 20(3), 325–334. Weinhold, D., & Rauch, J. E. (1999). Openness, specialization, and productivity growth in less developed countries. Canadian Journal of Economics, 32(4), 1010–1027. World Bank. (1993). The East Asian miracle: Economic growth and public policy. New York: Oxford University Press. Yanikkaya, H. (2003). Trade openness and economic growth: A cross-country empirical investigation. Journal of Development Economics, 72, 57–89.