Aid-Macroeconomic Policy Environment- Growth Nexus: Evidence from Selected Asian Countries

Transcription

1 Aid-Macroeconomic Policy Environment- Growth Nexus: Evidence from Selected Asian Countries Saima Liaqat* 1, Temesgen Kifle 2 and Mohammad Alauddin 3 Abstract This study empirically investigates aid effectiveness debate in light of Burnside and Dollar (2000) hypothesis that policy enviroment the recipient country is critically important for aid effectiveness.the data on ten developing countries of South and Southeast Asian region over a period of form the empirical basis of this investigation. In line with Burnside and Dollar (2000) a policy index is constructed which includes inflation, budget deficit/surplus and openness. Two stage least squares (2SLS) and Generalised Method of Moments (GMM) techniques were employed to test the model. Two major findings resulted from this study. First, aid had a negative impact on economic growth during the study period for the Asian region. Second, there was no evidence to suggest that aid policy interaction had any impact on economic growth implying that the good policy enviroment is not a condition for economic growth in context of aid effectiveness. This in effect refutes the Burnside - Dollar aid effectiveness hypothsis. Keywords: foreign aid, macroeconomic policy, economic growth, Asia JEL Classification:F350, O19, O38, P45, O11, O53 1 * Corresponding author: Visiting PhD Scholar at University of Queensland, Australia /PhD scholar at University of the Punjab Lahore, Pakistan. contact2saima@hotmail.com 2 Lecturer, School of Economics, University of Queensland, Australia. t.kifle@uq.edu.au 3 Associate Professor, School of Economics, University of Queensland, Australia. m.alauddin@uq.edu.au

2 1. Introduction A large body of literature addresses the foreign aid-nexus and broadly fall into two categories: 1) unconditional growth effect studies 2) conditional growth effect studies. The former implies that growth effect the recipient economies without requiring any conditions (Hansen and Tarp, 2001; Lensink and Morrsey, 2001; Moreira, 2005). On the other hand, the latter claims that growth effect is conditional on a good policy environment (Burnside and Dollar, 2000; Collier and Dollar, 2002; Collier and Hoeffler, 2004). The main purpose of this study is to empirically test the validity of these two strands of studies in the context of the Asian region. Both the Harrod-Domar and Solow growth models identified physical capital formation as the main driving force that can accelerates economic growth process. In other words, the output is dependent upon new investment and its productivity. Lack of savings was regarded as the main constraint to economic growth in the developing world. The low per capita income was the chief factor that limited the capacity to generate the savings required for investment purposes. Later, Chenery and Strout (1966, 1979) introduced the two-gap model by expanding the Harrod- Domar model. The two gap model made it possible to predict the amount of capital inflow required to maintain a specified rate of growth. According to the two-gap model, the low income countries investment is constrained either by shortage of exports income (trade gap) or shortage of domestic savings (savings gap). To fill these two gaps, inflow of foreign capital is required. The main purpose of aid is to supplement the domestic sources of finance, improve the amount of the capital stock and hence increase amount of investment. However, if the aid money is spent

3 on government consumption expenditures rather than public investment purpose, then result of this wasteful expenditure could be negative on growth. The paper tests whether 1. Aid is a significant determinant of GDP. 2. Policy variable and its interaction with foreign aid affect the economic growth one way or the other. 3. Government policies do affect the level of economic growth either positively or negatively (even in the absence of aid). The third hypothesis is a must for the second hypothesis. Because if the policy does not play any role in the absence of aid, then it is theoretically incorrect to believe that it plays any role when interacting with aid variable. The remainder of the paper is structured as follows. Following a brief review of the relevant literature in Section 2, the data and estimation methods are discussed in Section 3. Section 4 presents the empirical model and the main findings of this study. Finally, Section 5 presents concluding remarks and policy implications. 2. A Brief Review of Literature The aid effectiveness debate has moved away from traditional foreign exchange and savings gap theories to the policy and institutional gaps in recent years. Burnside and Dollar (2000) represented the first significant work that had generated debate on the aid-policy-growth relationship. In their study they conducted panel growth regression for 56 developing countries over six four years periods ( ), and empirically tested two main questions; first, whether aid affect growth positively in presence of good policies and secondly whether donors allocate more aid to countries with economic good policies. The results of their study showed that in presence of sound fiscal, monetary and trade policies aid had a positive and strong impact

4 on economic growth. Another finding was that the multilateral aid was allocated in favour of good policy but no such evidence was found in case of bilateral aid. Later, World Bank considered the Burnside and Dollar (2000) findings while assessing aid and stated that donors should direct aid to those countries where there were sound economic policies in the absence of which aid would be wasted given that the sole purpose was to foster economic growth. The methodology of the Burnside-Dollar study was claimed to be seriously flawed by many the critics. A large number of studies re-examined the robustness of policy view and assessed the relationship of aid and growth in this context. Dalgaard and Hansen (2001), Brumm (2003), Ram (2004), Feeny (2005), Karras (2006) found no evidence that good policy environment as a condition to economic growth which rendered the Burnside-Dollar claims invalid to a significant extent. Collier and Dollar (2002), Dalgaard et al., (2004), Denkabe (2004), Salisu et al., (2010) results showed that sound and stable macroeconomic policy environment was necessary for effective contribution of aid to economic growth. This study will use more recent data ( ) data and empirically estimate the aid-growth link in the context of macroeconomic policy for ten selected Asian economies. Such an analysis is required for recent time as many developing countries are still dependent on aid. It will help policy makers and donors in understanding how the aid money has been utilised in last three decades by the recipients and how it has contributed to economic growth in their respective environments. Drawing on recent modelling techniques (2SLS and GMM); this study investigates this relationship by considering the endogeneity of some growth determinants, incorporating policy variable, including aid policy interaction term and quadratic term in the model. 3. Data and Research Methodology The empirical model is estimated by considering 10 developing countries of Asian region covering the period 1984 to This is an unbalanced panel study. The main source of data is World Bank and OECD-DAC (The complete list of countries and variables is provided in appendices).in this part, Section 3.1 provides the summary statistics of variable used in this

5 study. Section 3.2 explains the procedure for the construction of policy index. Section 3.3 addresses the issue of endogeniety of aid while Section 3.4 gives the overview of estimation techniques used in this model 3.1. Descriptive Statistics Table 1. Summary statistics of variables used in this analysis. Number of Variables observations Mean Minimum Maximum Standard Deviation Growth (per capita) Aid Population growth openness Inflation Budget surplus/ deficit Domestic credit to private sector Infrastructure The origin of this study is based on the analysis by Burnside and Dollar ( ). According to them aid spurs growth but only in those countries where there is strong policy environment. Burnside and Dollar introduced a policy index and used the interaction term between this policy index and aid to investigate the impact of policy in growth regressions. In the same line the following model has been constructed to estimate the relationship between aid policy and economic growth. ΔYit = β0 +β1yio + β2 (Aid/GDP)it + β3(policy)it +β4(aid/gdp* Policy)it + β5 (Aid/GDP) 2 it +β6{(aid/gdp)* Policy}it +βz(z)it +λt +μit i = 1,.., 10; t = 1,., 32 (1) Where i denotes country and t time. The dependent variable (ΔYit) is the annual average growth of real GDP per capita.yi0 is initial year real GDP per capita, (Aid/GDP)it captures foreign aid as a percentage of GDP and Pit is a policy index affecting growth per capita. Zit is a vector of all exogenous variables that affect growth and allocation of

6 foreign aid, and λt and μit represent the constant that may change over time (or intercept dummy) and error term, respectively. The data are averaged over four years interval (e.g., , ,.., ).The main reason for making the data averaged is that it helps in smoothing yearly fluctuations that arise in the dependent and independent variables Construction of Policy index Before estimating the above growth model it is necessary to construct a suitable policy index that can represent fiscal, monetary and trade policies. This study has constructed a composite policy index to capture the effect of three policies, namely fiscal policy, monetary policy and trade policy. Later, along the same line with Burnside and Dollar (2000), weights are assigned to the coefficients of variables obtained in the growth regression. The Policy index can be expressed in the following equation: Policy index = f {β (budget surplus) +β (inflation rate) +β (openness)} Here, budget surplus/deficit was used to capture the effect of fiscal policy whereas inflation rate was used to represent the monetary policy. Openness variable was also used to capture the effect of trade policy. It is important to note that Burnside and Dollar (2000) used a Sach and Warner (1995) dummy variable to capture the effect of trade policy. In contrast, this study used trade openness index. For the construction of Trade Openness index the following formula is used Trade Openness Index = (Exports + Imports) / (Gross Domestic Product) The reason to avoid dummy variable for measuring trade openness and using a proper openness index is because a dummy variable is unable to classify the type of economy (either closed or open) and thus it gives a biased result (Rodrik and Rodriguez, 2000). The main advantage to use the policy index is that it weighs policies according to their correlation with the growth regression. For this purpose, we first run the OLS regression on growth equation by excluding the aid variable. The following is an estimated regression Policy index equation that accounts for monetary, fiscal and trade policies.

7 [ Policy = (budget surplus/deficit) (inflation) *** (trade openness) *** ] 4 (2) The policy index weighs the three policy variables according to their GDP per capita growth. The constant 4.42 can be interpreted as the predicted growth rate of the country, given budget surplus, inflation and trade openness, and it has the mean characteristic of all the variables that are included in the model (Burnside and Dollar, 2000). The coefficient of budget deficit/surplus is positive but insignificant. The coefficient of inflation variable is negative and significant at p< It shows that a one percentage point increase in inflation rate decreases per capita growth rate by around 0.2 percentage point. This could be because inflation depresses the economic growth in two ways; first, it creates uncertainty about the future profits which results in conservative investment strategies, and second, it decreases the international competitiveness of a country by making the exports expensive and making balance of payment situation worse. The term trade openness is used here as a proxy for trade liberalization/less trade restrictions. The coefficient of trade openness is negative and significant at p< The negative sign of openness variable indicates the existence of relatively larger share of imports than exports. As noted by Abbas (2014) developing countries are facing higher trade barriers and restriction for agriculture products and related production sectors which are negatively impact their exports performance and results in high trade deficit The coefficients obtained in the above regression are used to create the policy index. For this purpose the values of each policy variable is substituted for a particular year. This has generated the series of policy index. Later, this policy variable is used for the estimation of the main regression model (1). The expected sign for each independent variable presented in model (1) is as follows. Official development assistance is expected to have a positive impact on growth via increasing investment. The quadratic term of aid is expected to be negatively related to economic growth showing diminishing returns to aid. The population growth rate is expected to have a negative impact on growth rate of per capita income. Financial development is expected to be positively related to per capita growth by increasing the provision of services to people. Before estimating the relationship between aid and growth it is important to consider the issue of endogeneity in this context. 4 The Breusch-Godfrey test confirms the absence of autocorrelation in the model.

8 3.3. Endogeneity of Aid In the empirical literature, especially in the aid effectiveness literature, the endogeneity of aid remained a main issue. While estimating the relationship between aid and growth, it is important to consider the issue of endogeneity which is expected to appear, as indicated in some previous studies (Hansen and Tarp, 2002; Easterly et al., 2003; Clemens et al., 2012). Endogeniety problem arises when a regressor is correlated with the error term. This problem could arise in three cases: (1) if there is some measurement error, or (2) if some of the variable is omitted (which correlate with one of the regressors) or (3) if there is simultaneity bias/ reverse causality. If the aid is endogenous then we have the following equation: cov (aid; μ) 0 Simultaneity of aid within growth regression implies that foreign aid influences the economic growth of the recipient, but at the same time economic growth also influences aid (donors may prefer to allocate aid to the recipients performing higher growth levels or may be inclined to provide aid to poor economies with poor growth levels). In case of endogeneity problem the OLS estimators are biased. There are some other econometric approaches, such as instrumental variable regression and Generalised Method of Moments that are used to correct the potential endogeneity problem. This study has employed both these techniques to address the issue and find robust estimates. 3.4 Estimation of Model To address the issue of endogeneity, a two-stage least squares (2SLS) method is first used following the approach of Burnside and Dollar (2000), Hansen and Tarp (2001), Dalgaard et al.,(2004) and Angeles and Neanidis (2009). While some researchers including Boone (1996), Dalgaard, Hansen and Tarp (2004) used lagged aid as instrument variable, Rajan and Subramanian (2008) pointed out that using lagged variable as instrument could be problematic because if the average growth rate depends on previous year aid flows (for instance average growth rate depends on aid flows of 1997) then this instrument would violate the endogenity requirement and would not be a suitable instrument.

9 Later, Generalised Method of Moments (GMM) method is used to address the potential endogeneity of aid issue. These estimators got great popularity in aid effectiveness literature (Dalgaard et al., 2004; Rajan and Subramanian, 2008; Angeles and Neanidis, 2009) Two- Stage Least Squares (2SLS) In OLS method it is assumed that the error term of the dependent variable is independent of all explanatory variables. When the above assumption is violated and error term of dependent variable correlates with the any of the independent variable, the OLS estimators give biased results. If the error term of the dependent variable correlates with any regressor of the model, two-stage least squares technique can be employed to solve this issue. 2SLS technique is in fact an extension of OLS method. 2SLS operates in two stages: in the first stage, it replaces the problematic variable with the instrument variable. While during the second stage, the estimated values obtained from stage one are used (instead of using actual values of problematic explanatory/endogenous variable) to estimate the model Arellano -Bond (GMM) Model Arellano and Bond (1991) proposed a technique to tackle the issue of endogeniety through estimating the dynamic panel models. This technique is also known as Arellano-Bond GMM. If the results of first stage statistics of 2SLS show that the instruments are weak then the estimators of fixed effect IV are likely to be biased. Through this approach, the endogenous variable is converted to first differences and lagged levels of endogenous variable are used to instrument these differences. This estimator is also called the difference- GMM estimator. According to Hansen (1982), Arellano-Bond estimation method initially transforms all the regressors into difference form and then uses GMM for further estimation. The following are the assumption of this method. 1. When there are few time periods (T) but large individuals (N) or N T. 2. The left hand variable is dynamic and it depends on its past realizations. 3. The regressors are not strictly exogenous which means that they are correlated with past and current realization of error

10 4. There is heteroskedasticiy and autocorrelation within the individuals but there is no such thing across them. 5. If there is fixed individual effects which implies unobserved heterogeneity. 6. A linear functional relationship. The main advantage of this method of over other commonly used methods is that it efficiently uses the instruments generated from the endogenous regressors. Foreign aid variable is assumed to be endogenous in aid growth- policy nexus model in this study. The causality runs in both directions- from foreign aid to per capita growth and vice versa. So, the regressors could be correlated with the error term. To fix this problem, we tried to apply two-stage least squares (2SLS-Fixed Effect). For this purpose the lagged value of aid has been used as instrument. However, to test the robustness of the results obtained in 2SLS- Fixed Effects Arellano Bond (1991) difference GMM estimation method was employed as well. This method was first proposed by Holtz Eakin, Newey Rosen (1988). Lagged levels of endogenous regressors along with two additional instruments land area and population size of countries were added. The reason for this is simple because aid money is disbursed on the basis geographic size and population of the recipients Sargan Hansen J-test After estimating the growth model by Generalized Methods of Moments (GMM), Sargan- Hansen J-test is computed to test the validity of model. In other words Sargan-Hansen test checks if the instruments of the model are weak or not. If the moment conditions are close to zero the model fits the data. This is also called test for over-identifying restrictions. H0: m (θ0) = 0 (the model is valid or moment conditions match the data well) H1: m (θ0) 0 (the model is invalid or data does not come close to meet the restrictions) Under the null hypothesis, J statistic is asymptotically χ 2 (chi squared) with k-l degree of freedom J T. [1/T T t=1 g (Yt,θ )] T W T[1/T T t=1 g (Y t,θ )] χ 2 k-l under null hypothesis H0

11 In this statisticθ is the GMM estimator of the parameter θ0(dimension of vector g), k shows the number of moment conditions and l is the number of estimated parameters (dimension of vector θ). W Tis a matrix that should converge in probability to Ω -1 and must be exactly equal to Ω -1. Under the alternative hypothesis, J-statistic is asymptotically unbounded J p In order to conduct his test, the value of J statistic is calculated from data which is a nonnegative 2 value. Later the result would be compare with 0.95 quantile of theχ k l. 2 χ k l Reject H0 at for example 95 percent confidence level if J q χ and do not reject H0ifJ q k l Empirical results/ estimation of foreign aid model For the estimation of growth model, unbalanced data for ten major South Asia and Southeast Asia countries has been collected for the period Later, data is averaged over four-year (expect for initial per capita GDP). For initial per capita GDP first observation at the start of each decade is taken. Average real GDP per capita growth is the dependent variable. The log of initial GDP per capita, population growth, domestic credit to private sector (Financial development), and Infrastructure policy were taken as control variables. Aid, aid squared, policy and aid policy interaction terms are main variables of the interest. Aid squared variable describe if there is diminishing returns of aid or not. Negative value of aid squared terms shows diminishing returns of aid and positive terms explain vice versa.

12 Table 2: Panel growth regression outputs for 10 South and Southeast Asian countries Dependent variable: Growth rate of per capita GDP. Independent variable / Estimation method 2SLS-Fixed Effect (within) regression using BD Policy index GMM regression using BD Policy index Intercept (4.0257) Log initial per capita GDP (0.5413) Aid/GDP *** (1.9197) Policy index *** (0.1588) Aid/ GDP*Policy (0.508) Aid 2 (Aid-Squared) ** (1.0836) Aid 2 / GDP * Policy (0.1980) Pop growth ** (0.8999) Domestic credit to private Sector ** (Financial Dev.) (0.0146) Infrastructure policy *** (Fixed telephone lines) (0.1119) ( ) (0.4517) *** (1.1153) *** (0.1708) (0.1313) (0.9177) (0.0150) (0.0903) Log of Land Area (sqkm) ( ) Log of capital formation (1.4655) Degree of freedom 39 Sigma_e P-value (F-test) J Statistics ᵡ2 (P-Value) Number of observations 49 Note: The dependent variable is real GDP per capita. The values in the brackets are robust standard errors. J-statistics shows Hansen Sargan test for the join null hypothesis that the instruments are not correlated with the error term and are excluded from the estimated equation. Here *** p<.01, ** P <.05 and * p <.10.

13 Since one of the explanatory variable (aid) is endogenous in this model OLS or Fixed Effects methods could potentially lead to biased results. This study first used Two-stage least Squares (2SLS) and then first-differenced GMM method to check the robustness of results. First differenced Generalized Method of Moments was introduced by Arellano and Bond in 1991.One of the major advantage of this approach is that no external instrument variable is required to run the regression to deal with the endogeneity problem. This treats the regression equation in firstdifferences and then the lag levels of the explanatory variables are used as instruments. The results of growth regressions carried out using 2SLS (fixed effects) and GMM estimation are reported in the above table. The two stages least squares (2SLS) results shows that aid is significant but negatively related to real per capita GDP and one percentage point increase in aid depresses real per capita GDP by 5.5 percentage point. These results are similar to that of Brumm, (2003). The aid squared term of aid is also significant but positive which implies that there is increasing return to aid. But when this aid squared term interact with the policy it becomes insignificant. Population growth is positive and significant using both 2SLS and GMM estimation methods which indicates that population growth enlarges labor force and increases growth. In short run population growth (following transitional theory) reduces per capita GDP but in long run population growth improves living conditions through increase in labor force, human capital and consumer force (Crenshaw et al., 1997). Financial development is measured here in term of domestic credit to private sector. The coefficient of financial development is negative and significant which reveals the fact that financial development benefits economic growth up to a certain threshold level and beyond that threshold level more finance results in negatively affecting economic growth. In other words, more finance is not necessarily good for the economic growth but it s the optimal level that matters (Law and Singh, 2014). Following

14 Easterly and Levin (2003) and Loayza et al., (2005) telephone penetration (Telephone lines per 1000 person) was used to represent the infrastructure. Better Information and communication technology reduces the cost of interaction and expands market boundaries (Roeller and Waverman, 2001). In developing countries fixed line phones are replaced with mobile phones. The cross elasticity of fixed line phones and mobile phone is positive which means they are perfect substitute. The consumers have to pay the price of both which perhaps negatively affect their per capita income. The coefficient of aid*policy is insignificant in both cases which negates the conditional effect hypothesis when aid is interacting with policy. The GMM results are displayed in column 2 of the same table. The results concerning to key explanatory variable suggests that aid is negatively associated with real per capita growth rate as a percentage point increase in aid depressing real per capita growth by percentage point. In other words aid is not effective in improving growth of GDP per capita in this region. Although, the policies, alone significantly and positively affect the growth of GDP per capita. But when they are interacted with aid they are insignificant. 5. Conclusion and Policy Implications The motivation of this study was to provide the donors and recipients an insight on the effectiveness of aid. The effectiveness of aid has been questioned over the past few decades. The main purpose of this analysis is to test the claim made by Burnside and Dollar (2000) that aid is only effective in good policy environment. This study contributes to the literature of aid, policy and growth as it s the first study of this kind which empirically tests the Burnside- Dollar (2000) aid effectiveness hypothesis in the context of Asian region with updated data. Our finding suggests that in general monetary, fiscal and trade policies seem to be crucial for growth but

15 when interacted with aid they have no impact on GDP per capita. These findings are useful for the donor community and the policy makers. All this implies that strong and consistent public policies (fiscal, monetary, trade) by the governments of recipients are very much important for growth. Another important finding of the study shows that there is a negative relationship between aid and GDP per capita which implies that aid is not very effective in improving growth of GDP per capita in this region rather it has been used wastefully. The last but very significant finding is that aid is not conditioned to the good policy environment and this finding is inconsistent with Burnside and Dollar (2000) claim but highly consistent with Hansen and Tarp, 2001; Easterly et al., 2004; Yusuf, 2012 results. All These findings are surprising for the donors as well as eye opener for governments of recipients. Foreign aid was primarily given to the region to eradicate the poverty but does it really help? Does it help in improving the per capita GDP in the region? Unfortunately, NO. The policy implication is that donor should keep a check on aid and build a strong accountability mechanism for recipients. On the other hand, it is also the responsibility of the recipient s government to reconsider their internal policies and try to control corruption and wasteful use of financial resources. Inconsistent government policies, political instability and poor institutional structure should also be attended on priority basis otherwise the desired positive results from foreign assistance could not be attained. Furthermore, since aid flows are highly unstable, so the policy of self-reliance should be promoted. The recipients should rely upon some sustainable and stable sources of financing such as exports and foreign direct investment.

16 APPENDIX TABLE Sample of Countries Classified by Geographical region South Asia Bangladesh India Nepal Pakistan Sri Lanka Southeast Asia Indonesia Malaysia Philippines Thailand Vietnam Sample of Countries classified by income group Low income countries Nepal Lower Middle income Countries Bangladesh, India, Indonesia, Philippines,, Pakistan, Sri Lanka, Vietnam Upper middle income countries Malaysia, Thailand

17 Description of variables used in regression models VARIABLE GDP Per capita growth (annual %) Log initial GDP per capita Net ODA received(%gni) Population growth (annual %) Policy Index Inflation consumer prices (annual %) Budget Deficit/Surplus Openness Index Index of Ethno linguistic Fractionalization DESCRIPTION Annual percentage growth rate of GDP per capita based on constant local currency. Aggregates are based on constant 2005 U.S. dollars. GDP per capita is gross domestic product divided by midyear population. GDP at purchaser's prices is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for depreciation of fabricated assets or for depletion and degradation of natural resources. Log per capita PPP real GDP for the first year of each time period, constant Net official development assistance (ODA) consists of disbursements of loans made on concessional terms (net of repayments of principal) and grants by official agencies of the members of the Development Assistance Committee (DAC), by multilateral institutions, and by non-dac countries to promote economic development and welfare in countries and territories in the DAC list of ODA recipients. It includes loans with a grant element of at least 25 percent (calculated at a rate of discount of 10 percent). Annual population growth rate for year t is the exponential rate of growth of midyear population from year t-1 to t, expressed as a percentage. Population is based on the de facto definition of population, which counts all residents regardless of legal status or citizenship--except for refugees not permanently settled in the country of asylum, who are generally considered part of the population of the country of origin. The policy index includes three policy variables namely inflation, budget deficit/ surplus, openness to represent monetary, fiscal and trade policies respectively. Inflation as measured by the consumer price index reflects the annual percentage change in the cost to the average consumer of acquiring a basket of goods and services that may be fixed or changed at specified intervals, such as yearly. The Laspeyres formula is generally used. The government budget balance is the difference between government revenues and expenses. The budget is balanced when outlays equal to receipts, the country reports budget surplus when revenues are higher than expenses and deficit when expenses exceed the revenues. The Openness Index is an economic metric calculated as the ratio of country's total trade, the sum of exports plus imports, to the country's gross domestic product. Openness Index = (Exports + Imports)/(Gross Domestic Product) It measures the social conflict; constant value is taken for each country. It measures the probability that two randomly selected persons from a given country will not belong to same ethnoliguistic group. Data has been taken from Easterly and Levin Dataset

18 Growth Aid (per capita) Growth (per capita) Aid Population Correlation Matrix Population growth Openness inflation Budget deficit growth Openness Inflation Budget surplus/deficit Domestic credit to private sector Infrastructure Domestic credit to private sector Infrastructure Acknowledgement I would like to show my gratitude to Dr. Hafiz Khalil Ahmad [Assistance Professor, University of the Punjab, Lahore, Pakistan] for his support and significant advice.

19 References Abbas, S. (2014). Trade liberalization and its economic impact on developing and least developed countries. Journal of International Trade Law and Policy, 13(3), Angeles, L., & Neanidis, K. C. (2009). Aid effectiveness: the role of the local elite. Journal of Development Economics, 90(1), Arellano, M., & Bond, S. R. (1991). Some tests of specification for panel data: Monte carlo evidence and an application to employment equations. Review of Economic Studies, 58(2), Arellano, M., & Bover, O. (1995). Another look at the instrumental variable estimation of errorcomponents models. Journal of Econometrics, 68(1), Burnside, C., & Dollar, D. (2000). Aid, policies, and growth. American Economic Review, 90(4), Brumm, H. J. (2003). Aid, policies, and growth: Bauer was right. Cato Journal, 23(2), 167. Boone, P. (1996). Politics and the effectiveness of foreign aid. European economic review, 40(2), Crenshaw, E. M., Ameen, A. Z., & Christenson, M. (1997). Population dynamics and economic development: Age-specific population growth rates and economic growth in developing countries, 1965 to American Sociological Review, 62(6), Clemens, M. A., Radelet, S., Bhavnani, R. R., & Bazzi, S. (2012). Counting chickens when they hatch: Timing and the effects of aid on growth. Economic Journal, 122(561), Collier, P., & Hoeffler, A. (2004). Aid, policy and growth in post-conflict societies. European economic review, 48(5), Collier, P., & Dollar, D. (2002).Aid allocation and poverty reduction. European Economic Review, 46(8), Dalgaard, & Hansen. (2001). On aid, growth and good policies Journal of Development Studies, 37(6), Dalgaard, C. J., Hansen, H., & Tarp, F. (2004). On the empirics of foreign aid and growth*. Economic Journal, 114(496), F191-F216. Denkabe, P. (2004). Policy, aid, and growth: a threshold hypothesis. Journal of African Finance and Economic Development, 6, Easterly, W., Levine, R., & Roodman, D. (2003).New data, new doubts: A comment on Burnside and Dollar's" aid, policies, and growth"(2000) (No. w9846).national Bureau of Economic Research. Easterly, W., Levine, R., & Roodman, D. (2004). Aid, policies, and growth: Comment. The American Economic Review, 94(3), Feeny, S. (2005). The impact of foreign aid on economic growth in Papua New Guinea. Journal of Development Studies, 41(6),

20 Hansen, H., & Tarp, F. (2001). Aid and growth regressions. Journal of Development Economics, 64(2), Hansen, L. P. (1982). Large sample properties of generalized method of moment s estimators. Econometrica, 50(4), Karras, G. (2006). Foreign aid and long run economic growth: empirical evidence for a panel of developing countries. Journal of International Development, 18(1), Knack, S., & Keefer, P. (1995). Institutions and economic performance: cross country tests using alternative institutional measures. Economics & Politics, 7(3), Loayza, N., & Fajnzylber, P. (2005). Economic growth in Latin America and the Caribbean: stylized facts, explanations, and forecasts. World Bank Publications. Law, S. H., & Singh, N. (2014). Does too much finance harm economic growth? Journal of Banking & Finance, 41, Moreira, S. B. (2005). Evaluating the impact of foreign aid on economic growth: A cross-country study. Journal of Economic Development, 30(2), Minoiu, C., & Reddy, S. G. (2010). Development aid and economic growth: A positive long-run relation. The Quarterly Review of Economics and Finance, 50(1), Ram, R. (2004). Recipient country's policies and the effect of foreign aid on economic growth in developing countries: Additional evidence. Journal of International Development, 16(2), Rodríguez, F., & Rodrik, D. (2000). Trade policy and economic growth: A skeptic's guide to the cross-national evidence. NBER/Macroeconomics Annual, 15(1), Röller, L. H., & Waverman, L. (2001). Telecommunications infrastructure and economic development: A simultaneous approach. American Economic Review, Salisu, A. A., & Ogwumike, F. O. (2010). Aid, Macroeconomic Policy Environment and Growth: Evidence from Sub-Saharan Africa. Journal of Economics Theory, 4(2), Sachs, J. D., Warner, A., Åslund, A., & Fischer, S. (1995). Economic reform and the process of global integration. Brookings Papers on Economic Activity, 1995(1), Solow, R. M. (1956). A contribution to the theory of economic growth. Quarterly Journal of Economics, 70(1), Yusuf, T. (2012). Foreign Financial Aid, Government Policies and Economic Growth: Does the Policy Setting in Developing Countries Matter? Zagreb International Review of Economics and Business, 15(1), 1-22.