Budgetary Response of Recipient Governments to International Aid Transfers

Transcription

1 Budgetary Response of Recipient Governments to International Aid Transfers By B. Ouattara The University of Wales (Swansea) UK Abstract This paper investigates the effects of international aid transfers on the public sector behaviour in the recipient economies. It contributes to the fiscal response literature on three main fronts. Firstly, the paper overcomes the problem of using econometric estimates to approximate public sector fiscal targets, by incorporating the economic relationships that determine the targets into the reduced form equations. Secondly, the equations derived from the solution of the model can be tested using relatively sophisticated econometric techniques including cointegration and panel data techniques. Thirdly, the paper departs from the existing literature by using the panel data estimation techniques to empirically test the model. The key findings of the paper, based on a relatively large sample of aid recipient countries over the period, can be summarised as follows. 1) Aid affects positively and significantly public investment; 2) The results also show that aid has a positive and significant on government consumption; 3) with respect to revenue, contrary to general belief we find no evidence that aid reduces government revenue collection effort; 4) aid affects negatively and significantly public sector borrowing; and 5) we also found evidence that debt servicing is major constraint to public investment in developing countries. Our results are robust across different estimation techniques. JEL classification: C5, F3, H2, H5, H6 Keywords: Public sector fiscal behaviour; aid; Panel data Corresponding address: Economics Department, James Callaghan Building, Singleton park, Swansea SA2 8PP (UK). Tel: 44-(0) Fax: 44-(0) b.ouattara@swansea.ac.uk. 1

2 1. INTRODUCTION The debate over the effectiveness of development aid has been revived in recent years, though much of the debate is still centred on the aid-growth nexus. One strand of the wider aid effectiveness literature, the fiscal response literature has, however, adopted a different route by focussing on the way aid transfers affect public sector behaviour in developing countries. The main argument put forward by advocates of this approach is that the primary recipient of aid transfers is the government and, therefore, exploring the way in which the latter uses these funds should be the first step in investigating the macroeconomic effectiveness of aid. An extensive literature has developed on the impact of aid flows on the fiscal behaviour of recipient countries since the seminal contribution of Heller (1975) in the American Economic Review. Studies that have addressed this issue include Mosley et al. (1987), Gang and khan (1991), Khan and Hoshino (1992) and Otim (1996), Gupta (1997), Franco-Rodriguez et al. (1998), Gang and Khan (1999), McGillivray and Ahmed (1999), McGillivray (2000), Franco-Rodriguez (2000), Mavrotas (2002), Mavrotas and Ouattara (2006), and McGillivray and Ouattara (2005). In addition to providing us with a better understanding of the macroeconomic effectiveness of aid, fiscal response studies could also play a crucial role in our understanding of the aid-poverty relationship. This channel is particularly an important one today given the new emphasis on reducing global poverty by the year Loosely speaking, it means that an understanding of the fiscal effects of development aid could offer some insight on how to better achieve the millennium development goals (MDGs). Models of fiscal response are generally based on welfare utility maximisation models, which are generally tested in a context of country specific studies using the non-linear three stage least square (NLTSLS) technique. The model in this paper, along the lines of previous work, analyses the budgetary response of recipient governments in developing countries to international aid transfers. This study differs in a number of important respects from previous fiscal response studies. Firstly, it develops a new model that overcomes the problem of having to use econometric estimates to approximate government fiscal targets. This is done by incorporating the economic relationships that determine these targets into solution of the model. Secondly, the equations derived from 2

3 the solution of the model can be tested using relatively sophisticated econometric techniques such as cointegration analysis (in the context of country specific studies) and cross-section and panel studies. Finally, this study applies the model using panel data technique for more than 80 countries, over the period ; as such this is the first application of a fiscal response model in the context of a relatively large sample. The remainder of the paper is organised in the following way. Section 2 outlines the settings of the model and derives its solution. Section 3 discusses the methodology and the data. Section 4 presents the results and their interpretation. Concluding remarks are left to the final section. 2. THE MODEL Like in previous fiscal response models, it is assumed that decision-makers in the public sector behave as a single individual with a well-behaved, homothetic preference map and with the utility function: U = f ( Ig, G, R, A, B ) [1] where Ig stands for public investment, G for government consumption, 1 R for government revenue (tax and non-tax), A for net foreign aid disbursements and B for the flow of public borrowing from other sources (domestic and foreign). It is then assumed that the public authorities minimise the following quadratic loss function: α 2 α 2 α 2 α 2 1 * 2 * 3 * 4 * U = α0 - ( Ig -Ig ) - ( G-G ) - ( R -R ) - ( B-B ) [2] Although G includes developmental expenditures (health, education, etc..) and non-developmental expenditures (wages, salaries, etc.) it is introduced here in aggregate form to make the model tractable. 2 Bihn and McGillivray (1993) provide detail discussion about the specification of the utility function. 3

4 where the starred variables are exogenously determined targets and α i > 0, i =1,...,4 represent the weight attached to each element of the utility function. The rationale for specifying the loss function in this manner is that the public authorities have some idea about the target values of the decision variables and they try to minimize the deviations of the actual values from their targets, subject to budgetary constraints, which also include foreign aid. 3 Following White (1994) it is then assumed that the public sector policy-makers face the following budget constraint: Ig + G = R + A + B [3] The above budget constraint assumes that expenditure (public investment + government consumption) must equal total government s receipts. In other words, it is assumed that the government runs a balanced-budget. 4 Some studies have introduced a second budget constraint to capture the distribution of government s receipts among expenditure types. However, the present paper is only interested in capturing the fiscal impact of aid flows (and not its allocation) and, therefore, specifying the budget constraint as in [3] is sensible. Moreover, given that we are only interested in the final reduced-form equations, introducing a second budget constraint would not affect our results. The Lagrangean is then applied to the maximisation problem, as below: 3 Some (see Franco-Rodriguez et al. (1998)) have endogenised aid on the basis that it is a government choice variable. However, given that the target for aid is generally set as aid commitments this implies that the impact of aid obtained in the reduced from equations will be that of the commitment values and not the disbursement values. This will tend to over-estimate the impact of aid, as the amounts committed are generally higher than those disbursed. 4 In some years the government might run a deficit and in other a surplus. Modelling this will be difficult so for simplicity we assume a balanced-budget. 4

5 α1 α2 α α L = α Ig Ig G G R R B B * * 3 * 4 * ( ) ( ) ( ) ( ) λ ( Ig + G -R - A -B ) [4] where λ is the Lagrange Multiplier. Taking the derivative of L with respect to the choice variables and λ, and solving the first order conditions through leads to the following semi-reduced form equations: 5 * * * * Ig = δ1ig + δ2 (A+ B + R G ) [5] * * * * G = δ3g + δ4 (A+ B + R Ig ) [6] * * * * R= δ5r δ6 (A + B Ig G ) [7] * * * * B= δ7b δ8 (A + R Ig G ) [8] where the δ ( i = 12,,... 8 ) are combinations of α ( i = 1,... 4 ). 6 j i The next step in the specification of the model is deriving the equations explaining the target variables. One problem faced by, virtually, all fiscal response studies is that the target variables included in the model are difficult to obtain from government sources. When they are available they only exist for very few years. To overcome this problem, target values have generally been estimated using Ordinary Least Square (OLS), Autoregressive or cointegration techniques. Given, that there is no consensus on the methodology used to derive these targets we adopt a different approach. Rather than deriving the target values, through an estimation technique, we incorporate the economic relationship that determines them (the targets) into the solution of the semi-reduced form equations, to obtain the full-reduced form equations. Although, any specification adopted 5 The software Mathematica 5 was used in solving the maximisation problem. 6 These coefficients can be obtained from the author upon request. 5

6 is bound to be ad hoc, in setting these targets we build on previous work on the main economic determinants of each fiscal variable. It is assumed that the target for public investment is approximated by the following economic relationship: * Ig = γ 0 + γ1y + γ2d+ γ3 A [9] where Y stands for GDP per capita. We expect that government s decision to investment will be affected by the income. For example, higher income would tend to generate extra resources, for the government, which could be used to finance public investment. D, debt service, is included to capture the fact that debt repayment and expenditure on investment are substitute as far the government is concerned. Loosely speaking, this means that higher debt expenditure would tend to reduce public investment expenditure. And A, aid disbursements, is included on the assumption that donors expect the recipient to use part of aid to finance public investment. The target for government consumption is assumed to be close to the following economic relationship: * G = η0 + η1y + η2d+ η3 TOT [10] where Y and D are defined as above. TOT represents the terms of trade. The inclusion of these variables is justified on the following grounds. Income is likely to have an effect on government consumption expenditure. The extent to which government can spend on consumption will also be affected by the amount of its resources it devotes to debt servicing. Finally, it is expected that changes in the terms of trade would affect government consumption. In one hand, improvement in the terms of trade could lead to an increase in consumption via an increase in government resources; declines in the terms 6

7 of trade, on the other hand, should lower revenue and thus reduce government consumption, ceteris paribus. The target for government revenue (tax and non-tax) is approximated by the following economic relationship: * R = µ + µ Y + µ Trade [11] where Y is defined as above, Trade is the sum of exports plus M is imports. Y captures revenue from income tax, indirect tax (VAT) on private consumption and non-tax revenues such as fines. The inclusion of trade is based on the fact that many developing countries get revenue from exports as well as imports. Finally the target for borrowing is approximated by the following relationship: * B = ε 0 + ε1 A [12] where A is aid disbursements. It is assumed that the public sector borrowing decision will be depends on how much aid funds is expected. Borrowing is generally more expensive than aid, which if disbursed in concessional form. Putting Equations [9]-[12] into Equations [5]-[8] gives the following full reduced-form equations: Ig = ω + ωy + ω D + ω A + ω Trade + ω TOT [13] G = ψ + ψ Y + ψ D+ ψ A+ ψ Trade + ψ TOT [14] R= π + π Y + π D+ π A+ π Trade+ π TOT [15] B= ρ + ρy + ρ D+ ρ A+ ρ Trade+ ρ TOT [16] 7

8 where the ωs, ψs, π s and ρs could be traced back to the attached to each element of utility function [2]. α s representing the weight 3. Methodology and Data Issues Methodology Each of the equations [13]-[16] are tested for a sample of aid recipient countries between 1970 and We consider four different estimation methods to ensure robustness of our results across different estimation techniques. The first estimation consists of using ordinary least squares OLS) with the pooled data. The drawback of the pooled OLS estimator is that it is likely to generate highly biased coefficients by ignoring both country specific effects and possible endogeneity of the right hand side variables. The second and third estimation methods consist of, respectively, applying fixed effects (FE) and random effects (RE) estimations techniques. These two techniques can handle systematic tendency of individual specific components to be higher for some units than for others (individual effects) and possible higher in some time periods than others (time effects). Another advantage of these two techniques is that they adjust for heteroskedasticity. However, it is important to note that even though FE and RE estimators appear to be preferable to the pooled OLS estimator, they still require some assumptions to be satisfied, such as the so-called strict exogeneity assumption. To cope with the potential endogeneity problem we use the instrumental variable approach based on the general moment method (GMM) estimator first proposed by Arellano and Bond (1991). Our preferred GMM technique is the system-gmm (SYS-GMM) which Blundell, Bond and Windmeijer (2000), Bond, Hoeffler and Temple (2001) and Hoeffler (2002) have shown superior to the standard GMM. The validity of the instruments used in the estimation process can be tested using standard Hansen/Sargan tests of overidentifying restrictions. The Hansen/Sargan test is asymptotically distributed as χ 2 and tests the null hypothesis 8

9 of validity of the (overidentifying) instruments. In addition to the Hansen/Sargan test, it is also important to check for the absence of serial correlation in the error term, as consistency of the estimates depends on it. First-order, AR(1), and second-order, AR(2), serial correlation tests are used for this purpose. While first order serial correlation is expected by construction, failure to reject the null hypothesis of absence of second order serial correlation leads to the conclusion that the original error term is serially uncorrelated. The test statistics are asymptotically distributed as standard normal variables. One problem often encountered in cross-section or panel studies relates to outliers. Their inclusion or exclusion can to a large extent alter the results of regression analysis. If useful generalisations are to be drawn, it becomes imperative to ensure that the results reflect the whole sample rather than being driven by a few outlying observations. To deal with this problem, different approaches have been used in the empirical literature. One of these approaches is to re-run the regressions without the outliers. The new trend in the literature is to use robust regression, which minimises the influence of outlying observations on the estimated equation rather than omitting them altogether. The data The data used here come from three main sources and covers the period of Data on aid (defined as net ODA) is obtained from the OECD-DAC online database. For estimation purpose this data was expressed in percentage of each country s GDP, obtained from the World Development Indicators (2000). Data on government consumption (G), government revenue (R), debt service (D), Trade (exports + imports) (all expressed as a percentage of GDP). Real GDP per Capita (Y) data comes from the World Bank Global Development Network database. Data on borrowing (B) is obtained from Global Development Finance (2000). Data on public investment (Ig) and terms of trade (TOT) are taken from World Bank Global Development Network Database. For the estimation the logs of the variables were taken. Figure 1 (in Appendix) show a plot of the 9

10 trends of aid and government fiscal variables over the period. Tables (1)-(4) (in Appendix) present summary statistics of the data for the individual equations. For the estimation we assembled a panel dataset of countries (81 to 99 countries). 7 The data are averaged over each of the five 5-year and one 6-year intervals composing the period of The number of countries in each equation was determined, mainly, by the availability of data on the dependent variable as well as the variable of interest aid. Countries that do not have enough data were automatically dropped from the regression. 4. Empirical Results Results related to equations Ig, G, R, and B are reported, respectively in Tables (5)-(8) (with t-statistics in brackets). Starting with the public investment equation, results in Table (5) show that foreign aid, our variable of interest, has a positive impact on public investment, in all four estimation techniques. The estimated coefficient of aid is found to be statistically significant at the conventional levels. This finding suggests that, at least part of foreign aid is used to finance public investment. Another interesting finding is that debt service exerts a negative significant impact on public investment. This implies that debt servicing is a constraint to public investment in developing countries, a situation known as debt overhang. Judging from the magnitude of the respective estimates for aid and debt service it can be concluded that, indeed, debt reduces more public investment than aid increases it. In terms of policy options this could mean that, if the macroeconomic objective of both donors and recipients is to stimulate public investment in the latter, a debt reduction might prove to be more effective than additional aid. The results derived from the government consumption equation (see Table (6)) indicate that aid flows tend to be associated with increases in government consumption. The estimation coefficient of aid with respect to government consumption is statistically significant. On average, it seems that the magnitude of the impact of aid on government consumption is slightly higher than its impact on public investment. This would tend to imply that aid is more pro-consumption than investment, and therefore inefficiently used 7 Each regression generates a different sample due to data availability for the endogenous variables. 10

11 (from a development perspective). However, such conclusion would be shortsighted, as part of government consumption is concerned with expenditure on health and education, and thus the formation of human capital which has a potential positive impact on economic growth. With respect to debt service, the results suggest that this variable does not have any significant effect on government consumption expenditure. This, together with the finding on investment, suggests that recipient governments would rather cut public investment than public consumption to finance debt repayments, ceteris paribus. The impact of aid on the government revenue collection effort, assessed in the regression reported in Table (7), is positive but statistically insignificant, implying that aid bears only a weak relationship to revenue effort. Put differently, the evidence suggests that, contrary to general belief, government in developing countries do not reduce their revenue collection efforts when aid is made available to them. In relation to the new poverty agenda this implies that increasing aid to meet the MDGs does not appear to jeopardise recipients efforts to mobilise domestic resources. Higher debt servicing tends to induce an increase in government revenue. One explanation could be that, in addition to reducing public investment, recipient government would also tend to increase revenue to finance the debt burden. Finally, the results in Table (8), which reports estimation of the borrowing equation, show that the effect of aid on public sector borrowing is negative and statistically significant, thus confirming the fact that public sector authorities in developing countries would substitute borrowing for aid. The evidence also suggests that increases in debt servicing tend to be associated with higher borrowing. One explanation could be that if the recipient government were to use its resources to finance its debt it would have to borrow to finance other form of expenditure, as there is a threshold for reducing public investment or increase tax revenues. 11

12 5. Conclusion One area of the wider aid effectiveness debate, the fiscal response literature, has attempted to look at the how governments in developing countries behave vis-à-vis of international aid transfers. The present paper uses the fiscal response framework as point of departure in modelling the impact of aid flows on public sector fiscal aggregates. It provides a new model with a solution that can be tested using more advanced econometric techniques such us cointegration, in the context of country specific studies, or panel data techniques, in cross-countries studies. The paper then tests the the model using panel data approach, for a group of developing countries over the period The findings in this paper, in terms of the effectiveness of development aid, are rather encouraging. We found a positive and significant relationship between aid flows and public investment. Although we did find that aid tends to increase government consumption we argued that this should not be a case for concern among donors, as this type of expenditure also includes education and health spending, which are two core elements of the Millennium Consensus. In terms of the analysis on the revenue side of the public sector, contrary to the conventional wisdom, aid does not reduce incentives for the mobilisation of public resources. We also found that recipient government would reduce other form of borrowing if aid receipts increase. Finally, the results appear to support the view that debt servicing has a negative effect on public investment. Two main policy recommendations can be derived from this study. Firstly, it is clear from the result that additional aid flows to meet the MDGs would be essential. Aid tends to increase government consumption, which includes expenditure on health and education. However, increasing aid is only the first step of the process. In addition, recipient governments have to actually commit the new resources towards the financing of health and education expenditure. Secondly, given that debt burden tends to constraint public investment implies that more aid coupled with a reduction in the debt burden of the aid receiving countries could help them reduce poverty effectively and at the same time enhance their growth prospect. 12

13 This paper has also some research implications. Firstly, in the light of the evidence presented here it is important for the aid-growth studies to consider the question of government fiscal behaviour vis-à-vis of aid flows before studying their broad macroeconomic impacts, as the relationship between aid and growth might not be a straightforward one as many aid-growth studies seem to assume. Secondly, it might also be worth analysing whether the fiscal behaviour of the recipient government depends on the modality (project versus programme) of aid. Thirdly, although panel studies are sophisticated and allow us to have more insight into economic relationships than previous techniques such as cross-section, country specific studies are still needed. Cointegration and error-correction estimation techniques can be used for this purpose. We hope this paper will provide the stimulus for such research. 13

14 References Arellano, M., and O. Bover, 1995, Another Look at the Instrumental Variable Estimation of Error Components Models, Journal of econometrics 68, Binh, T.N., and M. McGillivray, 1993, Foreign Aid Tax and Public Investment: A Comment, Journal of Development Economics 41, Blundell, Richard, and S. Bond, 1998, Initial Concditions and Moment Restrictions in Dynamic Panel Models, Journal of Econometrics 87, Blundell, Richard, and Stephen Bond, F. Windmeijer, Estimation in dynamic panel data models : improving on the performance of the standard GMM estimator (Institute for Fiscal Studies, London). Bond, S., A. Hoeffler., and J. Temple, 2001, GMM Estimation of Empirical Growth Models, Centre for Economic Policy Research Discussion Paper 01/525. Franco-Rodriguez, S., 2000, Recents Advances in Fiscal Response Models with an Application to Costa Rica, Journal of International Development 12, Franco-Rodriguez, S., M. Mcgillivray and O. Morrissey, 1998, Aid and and the Public Sector in Pakistan: Evidence with Endogenous Aid, World Development 26, Gang, I and H. Khan, 1991, Foreign Aid, Taxes and Public Investment, Journal Of Development Economics 34, Gang, I.N., and H. Khan, 1999, Foreign Aid and Fiscal Behaviour in a Bounded Rationality Model: different Policy regimes, Empirical Economics 24, Gupta Kanhaya, L., 1997, Public Fiscal Behaviour and Foreign Aid: Some Model Solutions, Economic Modelling 14, Heller, P., 1975, A Model of Public Fiscal Behavior in Developing Countries: Aid, Investment and Taxation, American Economic Review 65, Hendry, D.F., and J.A. Doornik, Empirical Economteric Modelling Using PcGive, Volume I (Timberlake Consultants LTD, London). Hoeffler, A., 2002, The Augmented Solow Model and the African Growth Debate, Oxford Bulletin of Economics and Statistics 64, Khan, H.A., and E. Hoshino, 1992, Impact of Foreign Aid on Fiscal Behaviour of LDC Governments, World Development 20,

15 Mavrotas, G., 2002, Foreign Aid and Fiscal Response: Does Aid Disaggregation Matter?, Weltwirtschaftliches Archiv 138, Mavrotas, G., and B. Ouattara, 2006, Aid Disaggregation, Endogenous Aid and the Public Sector in Aid-Recipeint Economies: Evidence from Côte d'ivoire. Review of Development Economics. (forthcoming). McGillivray, M., 2000, Aid and Public Sector Behaviour in Developing Countries, Review of Development Economics 4, McGillivray, M., and A. Ahmed, 1999, Aid, Adjustment and Public Sector Fiscal Behaviour in Developing Countries, Journal of Asia-Pacific Economy 4, McGillivray, M., and O. Morrisey, 2000, Aid Fungibility in Assessing Aid: re herring or true concern?, Journal of International Development 12, McGillivray, M., and B. Ouattara, 2005, Aid Debt Burden and Government Fiscal Behaviour in Cote d'ivôire, Journal of African Economies 14, Mosley, P., J. Hudson and S. Horrell, 1987, Aid, the Public Sector and the Market in Less Developed Countries, Economic Journal 97, OECD, DAC Online Satistics (OECD, Paris). Otim, S., 1996, Foreign Aid and Government Fiscal Behaviour in Low-Income South Asian Countries, Applied Economics 28. White, H., 1994, Foreign Aid, Taxes and Public Investment: a Further Comment, Journal of Development Economics 45, World Bank, World Development Indicators CD-ROM (World Bank, Washington D.C.). 15

16 APPENDIX Figure 1: Aid and Public Sector Fiscal Variables Trends aid public investment revenue debt service government consumption borrowing

17 Table (1) Summary Statistics for Public Investment Equation Variable Mean Std. Dev. Min Max Public Investment (%GDP) Real GDPPC Debt Service (%GDP) Aid (%GDP) Trade (%GDP) Terms of Trade Table (2) Summary Statistics for Government Consumption Equation Variable Mean Std. Dev. Min Max Government Consumption (%GDP) Real GDPPC Debt Service (%GDP) Aid (%GDP) Trade (%GDP) Terms of Trade Table (3) Summary Statistics for Revenue Equation Variable Mean Std. Dev. Min Max Revenue (%GDP) Real GDPPC Debt Service (%GDP) Aid (%GDP) Trade (%GDP) Terms of Trade Table (4) Summary Statistics for Borrowing Equation Variable Mean Std. Dev. Min Max Borrowing (%GDP) Real GDPPC Debt Service (%GDP) Aid (%GDP) Trade (%GDP) Terms of Trade

18 Table (5) Public Investment Equation Variables OLS FE RE GMM-SYS RGDPPC D A Trade TOT Constant.000 (.865) -0156** (-2.320).158*** (2.970).057*** (3.800).042*** (5.280) (-.091).000 (.370) -.116* (-1.710).094** (2.370).081*** (4.730).037*** (5.320) (-.530).000 (.720) -.146** (-2.480).138*** (4.790).068*** (6.280).392*** (6.390) (-.200) (-.290) -.260** (-2.470).135** (2.270).056*** (3.190).075*** (4.700) (-1.220) No of countries 81 R AR1 (p-value) AR2 (p-value) Hansen test (p-value) Notes: t-statistics in brackets. Finite sample corrected standard errors and robust standard errors options were used for these estimations. *, **, and *** represent significance at the 10, 5, 1 percent level, respectively. Table (6) Government Consumption Equation Variables OLS FE RE GMM-SYS RGDPPC.000 (.070) D.067 (1.170) A.169*** (3.310) Trade.080*** (10.030) TOT.031*** (4.830) Constant *** (4.210) No of countries 99 R AR1 (p-value) AR2 (p-value) Hansen test (p-value).000 (.390) -.083* (-1.290).093** (2.740) (-1.620).060*** (3.930) *** (10.250) (.090) -.099* (-1.650).126*** (4.120).039*** (3.920).028*** (4.240) 8.987*** (7.620) (.050).067 (1.080).169** (2.550).080*** (7.570).031*** (4.220) 4.280*** (3.370) Notes: t-statistics in brackets. Finite sample corrected standard errors and robust standard errors options were used for these estimations. *, **, and *** represent significance at the 10, 5, 1 percent level, respectively. 18

19 Table (7) Government Revenue Equation Variables OLS FE RE GMM-SYS RGDPPC D A Trade TOT Constant.001** (2.440).309** (2.530) (-1.550).139*** (6.590).029** (2.070) 4.347** (2.320).001** (2.750).191** (2.190) (-.220).069*** (3.450).069 (1.610) *** (6.540).001*** (3.810).217** (2.720) (-1.200).103*** (7.070).015*** (5.320) 8.584*** (5.470).001** (2.730).267** (2.570) (-.840).140*** (6.240).027** (2.130) 3.461* (1.780) No of countries 91 R AR1 (p-value) AR2 (p-value) Hansen test (p-value) Notes: t-statistics in brackets. Finite sample corrected standard errors and robust standard errors options were used for these estimations. *, **, and *** represent significance at the 10, 5, 1 percent level, respectively. Table (8) Borrowing Equation Variables OLS FE RE GMM-SYS RGDPPC D A Trade TOT Constant.000 (.078).161*** (3.960) -.052*** (-2.970).002 (.467).003*** (3.550) ** (-2.110).000 (.873).102*** (3.130) -.043** (-2.580).002 (.220).138*** (4.130) (-.280).000 (.507).181*** (7.440) -.048** (-4.000).029 (.392).016*** (5.760) -.990** (-2.240) (-.870).302*** (5.400) -.060** (-2.710).005 (.980).0125* (1.800) (-.980) No of countries 99 R AR1 (p-value) AR2 (p-value) Hansen test (p-value) Notes: t-statistics in brackets. Finite sample corrected standard errors and robust standard errors options were used for these estimations. *, **, and *** represent significance at the 10, 5, 1 percent level, respectively. 19