The World Conference on Environment and Development
) has defined sustainable development (SD) as development that meets present needs without comprising the ability of future generations to meet their own needs. This concept mainly consists of three dimensions, such as economics, social and environmental. The United Nations Conference on Trade and Development, UNCTAD
) has promoted foreign direct investment (FDI) as a potential driver for SD. SD goals have become an important mission, targeting many countries around the world, including Singapore. As one of the founders of the Association of South East Asian Nations (ASEAN) group, Singapore could act as a role model for other member countries in pursuing this ultimate goal. As a small island country, Singapore has done remarkably well in the decades of the twentieth century and has been one of the fastest growing economies in East Asia. Overall, the country has successfully achieved high economic growth through free trade and investment strategies implementation, which has led the country to become the top destination for investment, due to its favorable lending rates to foreign investors, implementation of a simple regulatory system, availability of tax incentives, a high-quality infrastructure, political stability, strong financial market and the absence of corruption. In the recent world investment report by UNCTAD
), Singapore is the fifth largest recipient of FDI in the world and the third largest of the East and Southeast Asian countries. Based on Figure 1
below, the country experienced a drastic fall in FDI inflows in 2007 as a result of the global economic crisis. The fall in 2008 was basically far greater than in late 1997, which happened during the Asian financial crisis. Interestingly, one year after the event, the FDI inflows into Singapore drastically increased and achieved its highest point in 2013, with a total value of US$
64,793,170,000. Overall, the success of this country in attracting a higher inflow of FDI is due to the effectiveness of its Economic Development Board (EDB) that maintains a network of overseas promotion offices and an international advisory council that includes global heads of leading multinational corporations (MNC).
Apart from achieving high economic growth, the country is also very concerned about the risks of negative externalities, such as pollution, as a result of industrialization in the country. To keep pollution under control, the Singapore government has implemented very strict regulatory measures. The third pillar for SD, known as income distribution, proxied by the GINI coefficient, was also observed in this study. Besides increasing the trend of CO2
emissions, there is also an increasing trend of income inequality in the country. Singapore’s income gap is one of the widest among developed countries. To narrow this gap, the government has made some effort to raise wages at the bottom and increase taxes on wealth at the top. However, relying only on the government side to maintain environmental quality as well as narrowing the income gap may not be enough in the long run. Thus, this research paper would like to test the potential role of FDI inflows that could bring favorable impacts on economic growth, income distribution and environmental quality, which represent the three pillars of SD. Given a strong record of FDI inflows to Singapore, this variable could be the best indicator for the realization of SD as suggested by UNCTAD. The findings of this research paper could help the country to prepare a better road map for achieving SD as this goal is also being targeted under ASEAN Vision 2025 as well as Sustainable Development Goal (SDG) 2030. The rest of the paper is structured as follows: Section 2
briefly explains the literature review; Section 3
focuses on the methodology; Section 4
displays data; Section 5
presents the empirical analysis and the last section concludes the paper with policy recommendations.
5. Empirical Findings
The first analysis performed was the descriptive statistics analysis that best described the basic information of the variables of the three econometric models. Table 1
displays the information such as mean, median, maximum, minimum, standard deviation, skewness, and kurtosis. For example, the highest mean of LNEN (12.08) detected is from Indonesia while the lowest mean (6.11) is recorded from the Philippines. This means that Indonesia used the highest amount of energy consumption due to its huge population as compared to the Philippines who used less energy consumption in the country. Meanwhile, the mean and median that implied the normal distribution of the data for every variable in each ASEAN-5 country were close enough to each, thus, provided more robust analysis. The minimum and the maximum value showed that the there was an overall increasing trend in the variables. Meanwhile, the standard deviation revealed that the average or typical distance scores varied from the mean. For example, in the case of Malaysia, the standard of typical distance that LNGDP values varied or spread from the mean was by about 0.482, whether its 0.482 above 8.126 or 0.482 below than 8.126.
The second analysis began with testing the stationarity of the data. This step was important to make sure that variables used in the study were not integrated of order more than I (1). The presence of I (2) and beyond could violate the requirements of using the ARDL estimation and critical value table proposed by Pesaran et al.
) and Narayan and Narayan
). Based on the results of Augmented Dickey-Fuller (ADF) and Philipp Perron (PP) displayed in Table 2
below, it can be seen that there is a mix stationarity either at I (0) or I (1) (at the level or at first difference) and thus justifies the use of ARDL cointegration test.
To test the existence of cointegration relationship among the variables, we performed ARDL bounds F
-test for cointegration that was displayed in Table 3
below. The maximum lag of 4 was set in each proposed model using Akaike Information criterion (AIC). The value of F
-statistics for each model (5.54, 3.25 and 3.97) was higher than the upper I (1) critical value table (for k
= 6 and 5) and significant at 1, 10 and 5 per cent level, respectively. Thus, this confirmed the existence of a long-run relationship between all variables in the three models
Next, before we proceed with long run and short run elasticities outcome, it is very important to make sure that our proposed models are reliable and free from any econometrics problems. The results from Table 4
presented the diagnostic tests for all three models. The diagnostics checking performed in these models has passed four major tests of serial correlation, functional form, normality, and heteroscedasticity, given that the probability value for each test is larger than a 10% significant level. This means that the stochastic error term is white noise, the specifications of the models are well specified, normally distributed with zero mean and constant variance, hence the models are robust.
To further enhance the reliability of our results, the three models were diagnosed with stability tests using the cumulative sum of recursive residuals (CUSUM) and the cumulative sum of the square of recursive residuals (CUSUMSQ). The results displayed in Figure 2
suggest that the coefficients (showed by the blue line) of the three models are stable and consistent as the results are still within the critical bound (represent the two red lines). This implies that the obtained results from this research paper can be used for policy inference.
After confirming the reliability of the proposed models, we now move to the core part of the analysis. Table 5
shows results for the long run elasticities for each model. Based on the output from the model of economic growth, it was found that FDI had a weaker positive impact on the Singapore economic growth at 10% significant level. Meanwhile, HC, TO and FD had stronger positive impact on the Singapore economy. An increase of FDI, HC, TO and FD by 1% could increase Singapore economic growth by 0.07%, 0.80%, 0.57% and 0.63%, respectively. The biggest impact on economic growth in Singapore is human capital development (HC). Realizing the importance of HC on generating growth, over the years, public expenditure on education has consistently been the second highest (after defense) in the government’s annual fiscal budget in Singapore. Next, the positive impact of FD on growth is in line with previous studies by Hermes and Lensink
) and Alfaro et al.
) who find that countries with better financial systems and financial market regulations can exploit FDI more efficiently and achieve a higher growth rate. The positive relationship between TO and growth in Singapore may come from three channels as suggested by Anderson and Babula
). First, trade provides access to foreign intermediate inputs and technologies that are not available in Singapore as this country is a small city-state with no natural resources. Second, trade also increases market sizes for new product varieties. Third, trade allows for diffusion of general knowledge across geographical boundaries that further facilitate the R&D process and subsequent innovation. Among past empirical studies that support the positive openness-growth nexus include Wacziarg and Welch
), Squalli and Wilson
) and Sakyi et al.
). Next, the positive impacts of FDI on growth validate FDI-led growth hypothesis for Singapore and this outcome could be driven by the positive impact of absorptive capacity variables (HC, FD and TO) on growth. The availability of high skilled workers, stability in the financial market and the implementation of trade liberalization, increase the confidence of foreign investors to invest more in this country. LAB and DI, on the other hand, are found to have a negative impact on growth. However, given that DI is not significant at any level this means that DI is not suited to explain growth in Singapore.
On the second model, we found that the rise of DI and FD can reduce the GINI coefficient, thus improving income distribution in the country. Precisely, a 1% increase in DI and FD will reduce the GINI coefficient by 0.99% and 0.55%. The negative relationship between DI and GINI is explained by the intuition that increases in domestic investment spending means more people getting jobs, which implies that more people are earning, thereby putting a downward pressure on income inequality. Meanwhile, recent literature reviews have taken a stance that financial development is an important factor in reducing income inequality (Ang 2010
; Clarke et al. 2006
). According to Galor and Zeira
), easy access to financial resources boosts investment activities that directly increase the income of poor segments of the population by generating employment opportunities. Besides, easy access to financial resources enables poor people to feed their children and educate them for a better future that results in income distribution improvement in (Canavire-Bacarreza and Rioja 2009
). FDI, on the other hand, has a positive relationship with GINI. This means that higher FDI inflows in Singapore will result to higher income disparity within the society when the multinational companies (MNC) is focusing on only recruiting high-skilled labor which in turn will result in an increase in income inequality (Lee and Vivarelli 2006
). The last two determinants GDP and TO were not significant at any level and thus, failed to have any relationship with income distribution in Singapore.
Based on the model of environmental quality, it shows that both GDP and EN have a positive relationship with CO2
emissions in Singapore. Ang
) and Halicioglu
) stated that higher economic development in the country might result in higher energy consumption and thus lead towards a higher release of CO2
emissions. The expansion of the Singapore economy that led towards environmental degradation in this study were similar to the research findings by Ali et al.
) in Singapore. Meanwhile, the positive relationship detected between GDP and CO2
emissions were in line with previous studies such as, Cole
). Next, negative signs for both FDI and TO on CO2
emissions mean that higher FDI inflows and increase in trade liberalization have successfully reduced environmental degradation in the country. To be more precise, a 1% increases in FDI and TO will decrease the release of CO2
emissions by 0.22% and 0.34%, respectively. FDI inflows received by Singapore might focus on modern and cleaner technologies in the production process to ensure best environmental practices that could reduce air pollution. Furthermore, the negative sign of TO shows that the country could specialize in clean and service intensive products for export and import pollution-intensive products from their trading partners. Lastly, FD does not influence CO2
emissions in Singapore, given that there is no significant level detected between these two variables.
shows the outcome for short run elasticities based on error correction model (ECM). Special attention is given to the expected results based on lag 0 on each model. Based on the model of economic growth, DI, FDI, and HC are found to have a positive relationship with growth in the short run. LAB, TO and FD do not have any relationship with growth as it is not significant relationship at any level. Next, based on the model of income distribution, it is revealed that GDP, DI and FDI have a significant relationship with income distribution, captured by Gini coefficient. Based on these three variables, GDP and DI exhibit negative sign while FDI exhibit positive sign. Lastly, based on the model of environmental quality, it is shown that FDI improve the environmental quality in the short run. To conclude, in the short run, it is found that FDI inflows have led to greater economic growth and environmental quality but at the same time increases the income disparity within the country. TO and FD did not influence any of the three models in the short run.
The long run elasticities on each model were supported by the negative and significant value of error correction term (ECT). ECT represents the speed of adjustment for each model and the negative value means that the variables will converge in the long run. The highest speed of adjustment is detected for a model of economic growth (−0.67), followed by a model of environmental quality (−0.39) and a model of income distribution (−0.35). Approximately, 67 per cent, 39 per cent and 35 per cent disequilibria from the previous year’s shock reconverged on the long run equilibrium in the current year. Overall, the R-square for all three models suggests that almost 94 per cent and above of the variables in equations for Singapore explain the dependent variable.
The Toda and Yamamoto
) granger non-causality test through vector autoregressive (VAR) model was conducted to investigate the direction of causality between determinants for model economic growth, model of income distribution and model of environmental quality. This method is valid regardless of whether a series is I (0), I (1) or I (2), non-cointegrated or cointegrated of any arbitrary order. The optimum lag for each model was detected using VAR lag order selection based on AIC. The optimum lag detected for a model of economic growth was 4. Meanwhile, the optimum lag for a model of income distribution as well as a model of environmental quality was 5. One extra lag (dmax = 1) was added to the optimal lag of the VAR model for implementing the Granger non-causality test using the Toda and Yamamoto approach. Next, to make sure that each model was dynamically stable, inverse roots of AR polynomial was performed. Based on Figure 3
below, it can be confirmed that all models are dynamically stable given that the inverted roots (dotted blue) for each model are all strictly inside the circle.
The results of the Toda-Yamamoto granger non-causality test is shown in Table 7
below while the figure illustration can be viewed in Figure 4
Based on the model of economic growth, we can confirm that four bidirectional causalities existed between (a) GDP and LAB, (b) LAB and FD, (c) LAB and TO and (d) LAB and HC. This relationship shows that Singapore depends highly on its labor to generate economic growth, deepening its financial sector, increases the production for export activities and advancement in the labor skills through human capital development. The unidirectional causality was detected between FD on GDP, TO and FDI. A strong financial sector in Singapore could assist the country to achieve a higher growth rate, increasing its value of trade as well as attracting a higher amount of FDI into the country. TO had a unidirectional causality with FDI and this supports Bhagwati’s hypothesis that the more open the country, the more attractive the country will be a hub of investment by foreign investors. Next, HC unidirectional granger caused TO and GDP of the country. With high skilled labor available in the country, it will produce more output and a higher volume of output will increase exports for international trade. FDI had a unidirectional relationship with LAB, which means that through foreign investment, more MNCs will open their business in the country, thus giving an opportunity to the locals to get jobs. Lastly, under the model of growth, we found that DI unidirectional granger caused FDI, FD, and LAB.
Next, based on the model of income distribution, we found ten strong bidirectional causalities at 1% significant level, detected between (a) GINI and FD, (b) GINI and FDI, (c) GINI and DI, (d) FD and FDI, (e) FD and DI, (f) FD and GDP, (g) FDI and GDP, and (h) DI and GDP. Meanwhile, unidirectional causality was found from GINI to GDP. This means that the economic growth of the country is a prerequisite for better income distribution in the country. Next, unidirectional causality was found from TO to FD. The rapid expansion of Singapore’s financial market leads to an introduction of new financial products and thus increases exports of the country. We also found that TO had a unidirectional causality to GDP, DI, and FDI. A unidirectional relationship running from FDI to DI was also shown in the model of income distribution.
Based on the model of environmental quality, only three bidirectional causalities were found running between (a) TO and FDI, (b) EN and FDI and (c) EN and TO. The unidirectional causality running from CO2
emissions to TO, EN and FDI reflected that the country’s environmental condition could influence openness to trade, level of energy consumption and level of FDI inflows. Lastly, the unidirectional causality can be seen from GDP to TO, EN and FDI. This means that per income capita growth of the country can be a prerequisite for the practice of trade liberalization, stimulation of higher energy used, as well as reasons for higher foreign investment. Similar to GDP, the unidirectional causality was also found running from FD to EN, TO and FDI. The illustration of granger causality can be seen in Figure 4