Public investment, private investment, foreign direct investment and economic growth in vietnam further evidence at the provincial level

Pham Dinh Long & Le Nguyen Quynh Phuong | 445 Public investment, private investment, foreign direct investment and economic growth in Vietnam: Further evidence at the provincial level PHAM DINH LONG HCMC Open University – long.pham@ou.edu.vn LE NGUYEN QUYNH PHUONG University of Economics and Finance Abstract This paper identifies the linkages between public investment, private investment, foreign direct investment and economic growth using Vietnam provincial data in the period 2007–2014 Applying the PVAR model in considering the relations between these variables, we provide more empirical evidence on the impacts of different types of investments on the local economic growth of 63 provinces in Vietnam Keywords: PVAR model; public investment; private investment; FDI; economic growth; Vietnam provinces Introduction On the trend of modern economy, globalization is one of the main tendencies Opening up and integrating national and regional economies become a mandatory condition of development The issue of durable growth, high speed, and structural shift have become development management for all modern economies and Vietnam is not out of the trend There have been many previous studies focused on understanding the relationship between investment types and economic growth including domestic investment (including the public investment, private investment), foreign direct investment, and most of all in the countries Domestic investment (DI) has always been seen as a major contributor to the nation's economic growth especially in term of 446 | ICUEH2017 creating jobs for the economy Foreign direct investment (FDI) plays a key role in boosting the economy, contributing significantly to the economic growth of countries in general Developing countries in the world, as well as Vietnam, are facing issues of attraction, management, and use of investment sources that have not been effective in recent years Vietnam has access to state budget funds, encouraged in sector private investment as well as foreign direct investment from multinational companies, which is expected to make a positive change in promoting economic growth However, in the local context, it is not the exploitation of these resources that is the same, depending on the characteristics of each region, the policies and ways of using the sources of investment of managers are different All of these factors contribute to changing the impact of investment on local economic growth Studies on the impact of overall investment on economic growth have been done mostly within the country, but empirical evidence suggests that this effect is still an unsatisfactorily answered question Such as there is very little research going into understanding the simultaneous causal relationship between different types of investment and economic growth, most research focuses on studying the relationship of two of the factors Studying the causal relationships between these factors for a specific analysis within the provinces of a country is very important for policy makers to design preferential investment policies Therefore, the paper will analyze the current status and quantification of the causal relationship between public investment, private investment, foreign direct investment and economic growth in the provinces/cities Vietnam in the period 2007 - 2014 using the PVAR model From the results of the analysis, the study provides policy recommendations for each of these investments as well as a combination of these types of investments in a way that can sustain and promote local economic growth in Vietnam The outline of this paper is the following Section summarizes the survey of the literature Section contains an overview of the econometric methodology and a brief discussion of the data Section discusses the empirical results Finally, Section gives some conclusions and policy implications Pham Dinh Long & Le Nguyen Quynh Phuong | 447 Literature Review 2.1 Theoretical Studies In the history of developing countries and in general theories, any country must use the most internal resources to develop the economy By using the effective domestic capital to enhance its role and achieve the country's important goals Domestic capital is a decisive source, and foreign capital is an important additional source of initial development This is reflected in the role of additional funds for investment as the internal accumulation of the economy is low The role of domestic capital in any country is undeniable, but with globalization in the new era, foreign capital plays an important role in promoting economic development Of any country, especially underdeveloped, developing countries, of which FDI is the dominant (apart from official development assistance (ODA)) In the early 1980s, researches on the determinants of economic growth attracted new interest among economists In particular, the relationship between FDI and economic growth is a topic studied in developed economies, both theoretically and empirically The theory of endogenous growth (Romer (1986), Lucas (1988), Barro (1991), Mankiw, Romer and Weil (1992), Barro and Sala-i-Martin (2004) confirms that FDI promotes long-term economic growth by technological diffusion This theory forms the basis of later studies to identify the major factors explaining the differences in economic growth among nations On the other hand, the Eclectic theory of FDI, developed by Dunning (1988), provides another way of analyzing the relationship between FDI and economic growth Based on the analysis of competitive advantage, this theory shows that the attraction of FDI depends very much on the factors and characteristics of the host country Although in the short term the relationship between investment and economic growth tends to be low, but in the long run, investment rates are found to be closely linked to economic growth One country, if domestic savings are not enough for domestic financial investment, will depend on foreign capital In short, domestic capital has always been a major source of economic growth However, it can not be denied that the role of foreign capital, especially FDI, is also a potential source of economic growth for any country, its role in economic growth It 448 | ICUEH2017 should be explored in depth through empirical studies In addition, the relationship between these two sources of capital is also affirmed as complementary and complementary relationships, in each period in which they interact at different levels, This effect is also not the same for each country with its own economic characteristics 2.2 Empirical Studies Sumei Tang, E.A Selvanathan and S Selvanathan (2008) used the VAR and ECM regression model for quarterly time series data from the first quarter of 1988 to the third quarter of 2003 to examine short- and long-term relationships between FDI and DI, and economic growth in China In the course of the experiment, he focused on the Granger causal relationship between the variables and the function of the impulse response function The results show that instead of overriding DI, FDI has an additional relationship for DI FDI not only helps in overcoming the shortage of capital, it also stimulates economic growth through the addition of DI capital in China China's DI capital and economic growth are positively correlated; Large economic growth encourages large domestic investment and vice versa China's DI and economic growth have not had a significant impact on FDI inflows over the long term The causal relationship between economic growth and DI is two-way, but there is only one causal dimension from FDI to DI and FDI to economic growth China's DI has a greater impact on economic growth than the impact of FDI on stimulating economic growth and promoting domestic savings should be prioritized over FDI attraction in the design and implementation of the war Investment strategy and investment policies in China Samuel Adams (2009) used panel data for 42 sub-Saharan African countries between 1990 and 2003 to analyze the impact of FDI, DI on economic growth By the OLS regression method, the study shows that FDI in the latter period has a positive impact on economic growth, but the current FDI does not have a positive impact on economic growth This result is explained by the low level of development in sub-Saharan Africa, less developed economies that favor FDI, but through OLS and FEM regression, the study has also been shown that DI has a positive and significant correlation for economic growth In addition, the study results also show a trend that FDI is inversely proportional to DI Hoi Hoi Lean and Bee Wah Tan (2011) with annual time series from 1970 to 2009 study the relationship between FDI, DI and economic growth in Malaysia You use the Pham Dinh Long & Le Nguyen Quynh Phuong | 449 VAR and VECM models to study the relationship of economic variables in the short and long run First, they used the VAR model, verified the co-alignment relationship, and then continued using the VECM model to study the relationship of economic variables in the long run The empirical results of this study indicate that FDI, DI and long-term economic growth have a long-term relationship FDI has a positive impact on economic growth, whereas DI has a negative impact on economic growth In the long term In particular, the results of the study also show a causal relationship between DI and FDI, between economic growth and FDI in the short term In addition, the increase of FDI will have a positive impact on DI, resulting in additional effects from FDI on DI High economic growth will attract FDI flows in Malaysia DI is also an important factor in attracting foreign investment in the short term Expanding FDI inflows may be linked to an astonishing increase in DI, and both FDI and DI businesses can work together to further the growth of Malaysia Rehmat Ullah Awan, Khalid Javed and Falak Sher (2012) study the relationship between FDI, economic growth, exports, imports, DI in turn in South Asia including Bangladesh, India, Pakistan and Sri Lanka, Using the annual time series data from 1973 to 2010 The paper uses the VAR model and focuses primarily on causal relationships between FDI variables, economic growth, exports, imports DI, represented by fixed capital formation and additional commercial openness (with trade openings = (exports + imports) / GDP) The results of the study show that FDI has an impact on economic growth but this effect is weaker than the impact of exports on economic growth The paper also shows that other results such as economic growth affect imports in different periods but there is no reverse relationship of imports to economic growth Similarly, economic growth affects trade openness but there is no reversal of trade openness to economic growth There is also a two-way relationship between FDI and trade openness Tran Nguyen Ngoc Anh Thu (2014) focused on the impact of public investment on economic growth in Vietnam using the ARDL technique for 1988-2012 The results of the study show that there is no evidence of this effect in the short term but public investment has the effect of promoting long-term growth, although this is the lowest compared to private and capital investment Private direct foreign In addition, the study by Nguyen Hong Ha (2015) on the impact of FDI on economic growth in Tra Vinh province for data for the period 1999 to 2013 using VAR techniques and relationship 450 | ICUEH2017 testing Cause and effect, the study results show that the attraction of foreign direct investment has an impact on economic growth in Tra Vinh province and vice versa Tran Pham Khanh Toan (2014) captured on the impact of public expenditure on economic growth in Southeast Asian countries and lessons learned for Vietnam The study used the panel data for 1995-2012 and used the regression analysis technique The results of the study show the impact of public expenditure on the economic growth of Southeast Asian countries, reflected in the impact of the total public expenditure variable, public expenditure on health, expenditure Public security and defense have a positive impact on economic growth, and the public spending on education is negative In addition, variables such as labor force, private investment, FDI have a positive impact on economic growth, meanwhile, the openness of the economy, inflation is the opposite Based on the findings of the study, the author also provided some lessons for Vietnam to improve the impact of public expenditure on economic growth In addition to the theoretical observations surrounding the relationship between sources of investment capital for economic growth, there has been quite a number of empirical studies in the past regarding this relationship with other economic factors Studies use a variety of data types (time data, table data) with a variety of estimation methods and thus produce different results for each country, group, or region The results show that the relationship between investment capital and economic growth remains unclear On the basis of inheriting previous research, as well as extensive research, the implementation of a research project on the same subject in a Vietnamese locale has an important role to play To help policy makers have a pragmatic and general view of their local situation in order to make recommendations and implement effectively the competitiveness, sustainable economic development of the locality, close contribution and overall development of the whole economy in the country Method and Data This method used in this research is Panel Vector Autoregression (PVAR) By this way, the Panel VAR (PVAR) technique combines the traditional VAR approach treating all variables of the system as endogenous with estimation techniques for panel data, which allows for unobserved heterogeneity while the use of VAR models in time series Pham Dinh Long & Le Nguyen Quynh Phuong | 451 analysis is a common standard PVAR is one of appropriate estimation techniques in the context of dynamic panel data models This analysis focuses on test for Granger-causal relationships, the resulting ImpulseResponse Functions (IRFs), which estimates the response of particular variables in the system to innovations in another variable in the system, while holding all other shocks at zero To analyse the IRF, this method needs an estimate of their confidence intervals; since the matrix of IRF is constructed from the estimated VAR coefficients, their standard errors need to be taken into account Then, this method also presents the Forecast Error Variance Decompositions (FEVDs), which shows the percent of the variation in one variable that is explained by the shock to another variable, accumulated over time The variance decompositions show the magnitude of the total effect The equation for PVAR in this research is follow: LnGDPit = βo + β1LnGIit + β2LnPIit + β3LFDIit + β4LnLABORit +β5TOit + vit (1) Where is: + i: 63 provinces of Vietnam; + t: years from 2007 to 2014 Data used in this research are provincial data Dependent variable + GDP: real value of per captita GDP Independent variables (Explanatory variables) + GI: Public Investment + DI: Private Investment + FDI: Foreign Direct Investment Control variables + LABOR: Labor + TO: Trade Openness (counted by: (Export +Import)/GDP) 452 | ICUEH2017 Table 3.1 The calculation and the expected signs of the variables in the model Dependent variable GDP growth rate per capita Real GDP represented for economic growth GDP Independent variables (Explanatory variables) GI PI Investment from government sector + Investment from private sector + Public Investment Public Investment rate in GDP Private Investment Investment rate of nonstate sector in GDP FDI Foreign Direct Investment + Proportion of investment capital from foreign direct investment sector to GDP Employees aged 15 years and above in the Workforce + Annual rates of labor force growth Control variables LABOR Labor Force TO Total value of export and import of goods in the area + Volume of exports and imports of goods and services Total Import-Export turnover to GDP This paper uses secondary data from 63 provinces and cities in Vietnam for the period of 08 years from 2007 to 2014, so there will be a total of 504 observations Data from the following sources: - Gross domestic product (GDP) of provinces and cities in the period 2007-2014 is taken according to the actual value, measured in VND billion at the fixed price in 2010 taken from the Vietnam economic - the social statistical yearbook of 63 provinces and cities, provided by the General Statistics Office of Vietnam Pham Dinh Long & Le Nguyen Quynh Phuong | 453 - Private investment (PI), foreign direct investment (FDI) sector in provinces and cities in the period 2007-2014 are taken at current prices in VND billion from the statistical yearbook, provided by the General Statistics Office of Vietnam - The labor force over 15 years old in the economy (in thousands of units) of provinces and cities, provided by the General Statistics Office of Vietnam - The export and import value of goods in the area (unit: the US $ thousand) taken from the statistical yearbook of the General Statistics Office Then, trade openness data is calculated from this source The research data of the topic is taken from the reliable sources available in the provincial statistical yearbook and the General Statistics Office of Vietnam But for provincial GDP figures, some estimates tend to be higher than GDP figures for the whole country This is a restriction on the GDP figures of provinces and cities in this topic However, this topic does not take into account the overall GDP of the whole country but only to assess the impact of public investment on the GDP of provinces and cities Therefore, it is temporarily acceptable according to the GDP data available from the provincial statistical yearbook for analysis Whereas, the time series databases are often not only very large values but also different ones, most variables used in this model are got Logarithm from their values except trade openness Moreover, the getting logarithm also helps to eliminate the linear trend for data Result 4.1 Unit root Test First, like the VAR model, when estimating the PVAR model, it is necessary to ensure stationarity of all time series variables A commonly used test valid in large samples is the augmented Dickey–Fuller Test (ADF Test) through The Levin, Lin and Chu - LLC (2002), the test result shows that all LnGDP, LnGI, LnPI, LnFDI , LnLabor and Openness are I(0), so these time series are stationary 4.2 Optimal Lag An important aspect of empirical researches based on the PVAR model is the choice of lag order To find out optimal lag for this estimation, we base on J statistic and 454 | ICUEH2017 corresponding p-value, and moment model selection criteria (MMSC) developed by Andrews and Lu (2001) (MBIC, MAIC, MQIC criterion) These criteria are based on a high log-likelihood value, the smaller the value of an information criterion the better the result Since optimal lag length is chosen according to these criteria, the first is chosen Table 4.1 PVAR Lag Order Selection Lag CD J p-value MBIC MAIC MQIC 1 114.2077 3228587 -451.9009 -101.7923 -243.6297 63.95394 739278 -313.4518 -80.04606 -174.6044 23.64065 9436641 -165.0622 -48.35935 -95.6385 4.3 PVAR model with GMM estimation Table 4.2 Main result of a 6-variable VAR model Response to (95% confidence interval) Response of LnGDP(t-1) LnGI (t-1) LnPI (t-1) LnFDI (t-1) LnLabor (t-1) TO (t-1) LnGDP (t) -.1668642 (0.205) 1042577 (0.150) 1364178 (0.003) -.0716075 (0.090) 3.338253 (0.000) 0603157 (0.028) LnGI (t) 4616195 (0.065) 094512 (0.512) 9706284 (0.000) -.3197219 (0.003) -5.440385 (0.000) 0014753 (0.982) LnPI (t) -.8522748 (0.002) -.0606252 (0.661) 1.194189 (0.000) -.1774328 (0.026) 2.188309 (0.055) -.052237 (0.363) LnFDI (t) -.0623101 (0.808) 1877488 (0.160) 3540222 (0.000) 5093528 (0.000) -1.228988 (0.190) 1153056 (0.020) LnLabor (t) 052977 (0.001) -.0213812 (0.015) 0459218 (0.000) -.0167932 (0.000) 4737137 (0.000) -.0103271 (0.003) TO (t) 4541414 (0.252) 4023829 (0.036) -.2428815 (0.003) -.0808577 (0.397) 4471119 (0.759) 8904815 (0.000) Pham Dinh Long & Le Nguyen Quynh Phuong | 455 The results of the PVAR regression indicate that the big values of R2, Log Likelihood, small values of FPE and AICI Criterion, HQIC (Hannah Quinn Criterion), Small SBIC (Schwarz Criterion) are accordant in the model In general, the results of the model estimation show that in the short run, the relationships between these variables are unclear Hence, using the Granger causality test to better analyze the relationship between variables to find out what variables affect another 4.4 PVAR Stability Graph 4.1 PVAR stability According to the result, all the eigenvalues line inside the unit circle PVAR satisfies stability condition 456 | ICUEH2017 4.5 PVAR Granger Table 4.3 PVAR Granger Causality/Block Exogeneity Wald Tests Dependent variable: LnGDP Conclusion Excluded Chi-sq df Prob LnGI 2.071 0.150 LnPI 8.769 0.003 *** PI Ž GDP LnFDI 2.875 0.090 * FDI Ž GDP Dependent variable: LnGI Excluded Chi-sq df Prob LnGDP 3.409 0.065 * LnPI 70.137 0.000 *** PI Ž GI LnFDI 8.980 0.003 *** FDI Ž GI PI Ž GI FDI Ž GDP FDI PI GDP Ž PI FDI Ž GI Dependent variable: LnPI Excluded Chi-sq df Prob LnGDP 9.784 0.002 *** LnGI 0.192 0.661 LnFDI 4.942 0.026 ** PI GDP FDI Ž PI Dependent variable: LnFDI Excluded Chi-sq df Prob LnGDP 0.059 0.808 LnGI 1.973 0.160 LnPI 14.034 0.000 *** PI Ž FDI Notes: Robust Standard Errors in parentheses ***p