Effects of corruption on economic growth through transmission channels in developing coutries

UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM-NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS EFFECTS OF CORRUPTION ON ECONOMIC GROWTH THROUGH TRANSMISSION CHANNELS IN DEVELOPING COUNTRIES By NGUYEN NINH QUOC TRAN MASTER OF ARTS IN DEVELOPMENT ECONOMICS HO CHI MINH CITY, JULY 2013 A    UNIVERSITY OF ECONOMICS INSTITUTE OF SOCIAL STUDIES HO CHI MINH CITY THE HAGUE VIETNAM THE NETHERLANDS VIETNAM-NETHERLANDS PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS EFFECTS OF CORRUPTION ON ECONOMIC GROWTH THROUGH TRANSMISSION CHANNELS IN DEVELOPING COUNTRIES A thesis submitted in partial fulfilment of the requirements for the degree of MASTER OF ARTS IN DEVELOPMENT ECONOMICS By NGUYEN NINH QUOC TRAN Academic supervisor Dr NGUYEN TRONG HOAI HO CHI MINH CITY, JULY 2013 B    ABSTRACT This paper examines direct and indirect effects of corruption on economics growth in developing countries Indirect effects of corruption on economic growth will be examined by five transmission channels such as government size, capital investment, trade openness, human capital, and political instability With 35 developing countries in data set and OLS, 2SLS method, the result is that corruption has positive relationship with government expenditure, trade openness, human capital, and political instability That means more corruption will make more government expenditure, trade openness, human capital, and political instability However corruption only relates negatively with capital investment And through the effects of these transmission channels on economic growth, we can infer that totally corruption has positive relationship with economic growth The result in this paper is also an evidence of “grease the wheel” hypothesis in developing countries This is like a careful remind on the policy of fighting against corruption because in developing countries a characteristic of corruption is positively related with economic growth Keywords: Corruption, economic growth, government expenditure, capital investment, openness, human capital, political instability i    TABLE OF CONTENTS ABSTRACT i CHAPTER 1: INTRODUCTION 1.1- Research context 1.2- Research objective 1.3- Thesis structure CHAPTER 2: LITERATURE REVIEWS 2.1- CONCEPT OF corruption 2.2- Practical researches 2.2.1- Corruption greases the wheels of growth 10 2.2.2- Corruption is sand of the wheels of growth 10 CHAPTER 3: CONCEPTUAL FRAMEWORK 16 3.1- DIRECT EFFECT 16 3.2- INDIRECT EFFECTS 17 3.2.1- Government size 17 3.2.2- Capital investment 18 3.2.3- Trade openness 18 3.2.4- Human capital 19 3.2.5- Political instability 20 CHAPTER 4: METHODOLOGY AND DATA SOURCE 23 4.1- DIRECT EFFECT OF CORRUPTION AND TRANSMISSION CHANNELS ON GROWTH 23 4.2- INDIRECT EFFECTS OF CORRUPTION ON GROWTH THROUGH TRANSMISSION CHANNELS 24 ii    4.3- TOTAL EFFECT OF CORRUPTION ON GROWTH 25 4.4- DATA 25 CHAPTER 5: RESEARCH FINDINGS IN DESCRIPTIVE STATISTIC ANALYSIS 27 5.1- RELATIONSHIP BETWEEN TRANSMISSION CHANNELS AND GROWTH 27 5.2 RELATIONSHIP BETWEEN CORRUPTION AND TRANSMISSION CHANNELS AND EFFECT OF CORRUPTION ON GROWTH 32 CHAPTER 6: REGRESSION ANALYSIS 40 6.1- DIRECT EFFECT OF CORRUPTION ON GROWTH AND EFFECTS OF TRANSMISSION CHANNELS ON GROWTH 40 6.2- EFFECTS OF CORRUPTION ON TRANSMISSION CHANNELS AND GROWTH 45 6.2.1- Effect of corruption on government expenditure 45 6.2.2- Effect of corruption on capital investment 46 6.2.3- Effect of corruption on openness 48 6.2.4- Effect of corruption on human capital 49 6.2.5- Effect of corruption on political instability 50 6.3- TOTAL EFFECT OF CORRUPTION ON GROWTH 52 CHAPTER 7: CONCLUSION REMARKS 54 7.1- RESEARCH FINDINGS 54 7.2- THE DRAWBACK 55 7.3- POLICY IMPLICATION 56 REFERENCES 58 APPENDIX 61 iii    ABBREVIATION OLS: Ordinary Least Square 2SLS: Two Stages Least Square CI: Corruption Perceptions Index EEPC: Engineering Enrolment Per Capita iv    LIST OF TABLE Table 2.1: Summary of literature reviews…………………………………………… 15 Table 6.1: OLS of transmission channels and corruption on growth……………… 41 Table 6.2: 2SLS of gGX-CI……………………………………… ………………….45 Table 6.3: 2SLS of gFC-CI…………………………… ………… …………………47 Table 6.4: OLS of gFC-CI…………………………….………………………………47 Table 6.5: 2SLS of gTT-CI…………… …………………………………………… 48 Table 6.6: 2SLS of EEPC-CI…………….……………………………………………49 Table 6.7: 2SLS of PI-CI………….………………………………………………… 51 v    LIST OF FIGURES Figure 3.1: Effect of corruption on transmission channels and economic growth … 21 Figure 5.1: Scatter graph between g and gFC………… ……………………….…….28 Figure 5.2: Scatter graph between g and gGX…………………………… ………….29 Figure 5.3: Scatter graph between g and gTT……………… …………………….….30 Figure 5.4: Scatter graph between g and PI………… ……………………………….31 Figure 5.5: Scatter graph between g and EEPC……………………………………….31 Figure 5.6: Scatter graph between CI and gGX……………………………………….32 Figure 5.7: Scatter graph between CI and gFC…….………………………………….34 Figure 5.8: Scatter graph between CI and gTT….…………………………………….35 Figure 5.9: Scatter graph between CI and gTT (dropping out two outliner points) …35 Figure 5.10: Scatter graph between CI and PI….…………………………………… 36 Figure 5.11: Scatter graph between CI and EEPC… ……………………………… 37 Figure 5.12: Scatter graph between CI and g…………………………………………39 vi    CHAPTER 1: INTRODUCTION 1.1- RESEARCH CONTEXT Till now there is not the answer about the exact effects of corruption on economic growth The literatures have not come to one final agreement about the corruption on economic growth yet In early time, some authors argued that with an appropriate level of corruption, it can make benefit for economic growth In the case of rigid government administration, corruption can act as the “grease” to make smooth for economic operation Without corruption, people or enterprises may have to “queue in line” and waste a lot of time to have the final ideas from the government And the government also spends more labor force to deal with this situation Or like another argument on benefit of corruption, Acemoglu and Verdier (1998) showed that the cost for ensuring the government officials absolutely clean from corruption can be higher than the price of corruption Lui (1985) established “equilibrium queuing model of bribery” to examine the effect of bribery With this model, bribes can reduce the cost related with queuing Through that it improves the efficiency of public administration Moreover Beck and Maher (1986) also showed that bribery model would be “isomorphic” with bidding model in supplying goods and services to the government That means “in the absence of penalties for bribery, supplier firms would be indifferent between bribery and bidding institutions.” On the side of against the corruption, many empirical researches have come to the results that corruption was the “sand” for the wheels of economic operation In contrast of the idea of corruption is the “grease” for economic operation in rigid government administration, the corruption can make much more delays in administration to attract 1    more bribes The government officials not want to act quickly because they want to have more time and force the people or enterprises spend more bribes to them With this behavior the officials have tendency to give out their decisions basing on the price of corruption The enterprises with better quality but without corruption will have the disadvantaged decision from the authorized officials So a good project may come to an inability enterprise The benefit from this project can be destroyed by this enterprise So by corruption, effective allocation in economic has been made distortion This will have the negative effects on economic growth Moreover on the side of against the corruption, there are the ideas that corruption reduces the investment With corruption, the cost of project can become higher So this can drop out the good projects and reduce the investment Reduced investment will affect negatively to economic growth So until now we really look corruption like the impediment to the economic growth The report from World Bank in 1998 has seen corruption like the great obstacle to economic and social development However we does not still have enough the exactly theoretical framework to definitely confirm the impacts of corruption on economic growth We have the empirical researches to find out the effects of corruption in the specific cases with different methodologies Mauro (1995) showed that corruption reduced investment However investment is a main source of economic growth So corruption lowered economic growth through reducing investment As in Mauro (1995) corruption and growth (also investment) had negative relationship and significant in aspect of statistic and “in an economic sense” Tanzi and Davoodi (1997) conducted the research about “corruption, public investment and growth” With crosssection data of countries and regression method they found that more corruption made more public investment and lower government revenues Also higher corruption made “lower operation and maintenance expenditures” and “lower quality of public 2    First we also check the endogeneity of corruption variable The result shows that endogenous phenomenon exists at significant level of 10% (refer table A4 in Appendix) So 2SLS used here to eliminate the endogeneity is necessary The result in table 6.6 shows the negative relationship between corruption index and human capital investment And the coefficient of corruption index variable is significant at level of 10% When corruption index increases up unit, percent of number of enrolment in engineering over total population decreases approximately 0.2% It also means that more corruption will make more people invest in engineering area However how can we explain for that result? There is one idea that can make the result reasonable In fact under the society with much corruption, people with occupation such as law, politics, administrative management… (But not engineering) have more opportunity to corrupt However more opportunity to corrupt, to gain private benefit also means more difficulty to get these jobs because many people compete for these hot jobs And actually unless you were born in the high social level family or very rich family, you will be very difficult to get these jobs So the other choice is that you should invest in engineering department to easy get a job after you graduate In the table 6.1, human capital in engineering has positive relationship with growth However in this case the coefficient of EEPC is not significant So we can not make a link among corruption, human capital in engineering, and economic growth It is not possible to have a conclusion about the impact of corruption on growth through this channel 6.2.5- Effect of corruption on political instability 50    Table 6.7: 2SLS of PI-CI Dependent variable: Constant LY0 CI N Note: Standard errors are in parenthesis under coefficients PI -2.43 0.28 (0.19) -0.03 (0.24) 34 After the test of endogeneity of corruption variable, the endogenous phenomenon exists at significant level of 10% (refer table A5 in Appendix) So 2SLS used here to eliminate the endogeneity is necessary Here we have the negative relationship between corruption index and political instability index This means more corruption will make more political stability For detail when corruption index increases up unit, political instability index decreases 0.03 unit This result is different from the expectation at Conceptual Framework section that corruption relates positively with political instability This result is also different from the result in descriptive statistics section where corruption index also has positive relationship with political instability index Although here we have the negative relationship between corruption index and political instability index, however the coefficient of CI is not significant at all (and very small) So about statistic meaning, the effect of CI on PI is not so important in this case Because of this, in this case corruption nearly does not affect growth indirectly through transmission channel of political instability 51    6.3- TOTAL EFFECT OF CORRUPTION ON GROWTH Now we have had the effects of corruption on transmission channels and the effects of transmission channels on growth So the total indirect effect of corruption can be achieved Along with the direct effect of corruption on growth (in table 6.1), we can find out total effect of corruption on growth The direct effect of corruption on growth is from the result of table 6.1 The coefficient of corruption index variable is -0.714 This coefficient represents the direct effect of corruption on growth The indirect effects are from transmission channels So the indirect effect of corruption on growth = effect of corruption on transmission channel x effect of transmission channel on growth From that we have: Effect of corruption on growth through government size channel = -3.481 (table 6.2) x 0.136 (table 6.1) = -0.473 Effect of corruption on growth through capital investment channel is dropped out because the coefficient of CI in this case is not significant at all (table 6.4) Effect of corruption on growth through the openness channel = -6.671 (table 6.5) x 0.114 (table 6.1) = -0.76 Effect of corruption on growth through human capital channel is dropped out because the coefficient of CI in this case is not significant at all (table 6.4) Effect of corruption on growth through political instability channel is dropped out because the coefficient of CI in this case is small and not significant at all (table 6.7) So sum of indirect effects = -0.473 + (-0.76) = -1.233 52    So the total effect from the approach of transmission channels is: -1.233 + (-0.714) = -1.947 As the result of this chapter, the corruption index variable has negative relationship with growth That means corruption has positive relationship with growth In the context of sample data of developing countries and by the examining approach in this paper, the corruption is supporting for growth The countries with higher corruption have faster growth than less corruption countries 53    CHAPTER 7: CONCLUSION REMARKS 7.1- RESEARCH FINDINGS Basing on the data set of developing countries and using the methodology in this study, we come to some main conclusions like following: ‐ The corruption has positive relationship with government expenditure And through the positive relationship between government expenditure and economic growth, corruption relates positively with growth ‐ Corruption has negative relationship with capital investment More corruption will make capital investment decrease However capital investment must be one of the roots of economic growth So through this channel corruption relates negatively with economic growth However in this paper the coefficient that presents the effect of corruption on capital investment is not significant So in the context of this paper, capital investment does not show that it is an important transmission channel ‐ There is a positive relationship between corruption and the openness More corruption will create more openness With supporting effect of the openness on economic growth Corruption once again has positive relationship with economic growth ‐ There is a positive relationship between corruption and number of enrolment in engineering per capita (EEPC) which represents human capital More corruption will make more people invest in engineering educational area However effect of EEPC on economic growth is not significant So in the context of this paper, 54    human capital (EEPC) does not show that it is an important transmission channel ‐ And with the last channel, corruption relates negatively with political instability However in this case the effect of corruption on political instability is not significant and small So in the context of this paper, the political instability transmission channel is not meaningfull Summarizing all effects of corruption on economic growth through these channels, total effect of corruption on economic growth is positive Corruption has same direction relationship with economic growth 7.2- THE DRAWBACK Beside the achievements, this paper also has considerable drawback And this drawback is in methodology part According to the approach in methodology section, we use model (4.1) to estimate the direct effect of corruption on growth and the effects of five transmission channels on growth And the model (4.2) is used to find out the effect of corruption on every transmission channel So through model (4.1) and (4.2) we have a link to among corruption – transmission channel – economic growth This is how we estimation indirect effect of corruption on gowth through every transmission channel However also through this link, the drawback exists The model (4.2) expresses the relationship between corruption and transmission channels So this represents the correlation between corruption and every transmission channel And in model (4.1) corruption and transmission channels are all explanatory variables It is highly possible there is the phenomenon of multicollinearity exsiting in equation (4.1) And this phenomenon is going to affect estimation in (4.1) On the view of statistics, multicollinearity may not affect the coefficients of estimation However it will affect the assessment of statistic significance of the coefficients 55    because muticollinearity affects standart error estimation When multicollinearity exists, the OLS variances are often larger Larger variances also mean lager standard errors As a result, the t-stats values are going to become smaller in case of larger standard errors This will lead to wrong conclusion on statistic significant of coefficients For example in this paper, the coefficient of EEPC in regression of (4.1) is not significant at all This may be the result of multicollinearity phenomenon With insignificant coefficient of EEPC we cannot conclude the indirect effect of corruption on growth through this channel So the existing of multicollinearity phenomenon in the model (4.1) is the biggest drawback of this paper This drawback can be seen as an avoidace for further researches on this subject And another approach such as system of equations can be used to solve this issue in future 7.3- POLICY IMPLICATION Although corruption has positive relationship with economic growth, this should not be considered as a guideline for policy issues This should be considered like a characteristic, a feature of developing countries This can be seen as a property of development phase from developing countries to developed countries The developing countries should be careful in process of developing economic and fighting against corruption because in this phase it is possible that corruption goes in same direction with economic growth There should be more researches on the detailed relationship between corruption and economic growth in order to make sure that developing countries with fighting corruption are perfect and suitable If not fighting corruption can make intervention in developing economic Or if it is considered like a guideline, it should be used under any certain conditions What are these certain conditions? This will be the subject for studies beyond The 56    conclusion of this paper supports for the hypothesis “grease the wheel” of corruption The result of Meon and Weill (2008) was another evidence of “grease the wheel” hypothesis That paper concluded that “A possible policy implication of these results is that countries plagued with extremely inefficient institutional frameworks may benefit from allowing corruption to flourish This interpretation, however, is risky A country that would allow unfettered corruption may eventually find itself with an even worse global institutional framework, and thus be caught in a bad governance/low efficiency trap.” In that paper, corruption only supports for growth only in the condition of inefficient institutional frameworks Or in Kaouthar Gazdar (2012) showed that “corruption is positively associated with economic growth when the quality of governance is very low” Again in this paper low quality of governance is a condition on which corruption relates positively with growth The result in this paper is also an evidence of “grease the wheel” hypothesis in developing countries We should not use this result as a guideline and through that let corruption free in society because corruption is positively associated with growth Corruption can make many effects on economy and on society that we cannot predict its consequences Corruption can be positive relationship with growth only in some certain conditions Developing economic without corruption can be an ideal object we would like to have However the result in this paper is like a careful remind on the policy of fighting against corruption because in developing countries a characteristic of corruption is positively related with economic growth If we not care about the conditions existing parallel with corruption, fighting corruption can make difficulty for growth For example, in Meon and Weill (2008), fighting corruption should go with improve the quality of governance, government efficiency And there are some ways to improve the quality of governance such as modernizing institutions, strengthening the administration and enhancing human right, security and justice… 57    REFERENCES Alesina, A., Ozler, S., Roubini, N., Swagel, P (1992) Political Instability and Economic Growth National Bureau of Economic Research, Inc, NBER Working Papers: 4173 Acemoglu, D., Verdier, T (1998) Property Rights, Corruption and the Allocation of Talent: A General Equilibrium Approach Economic Journal, 108(450), pp 1381-1403 Aliyu, S U R., Elijah, A O (2008) Corruption and Economic Growth in Nigeria: 1986 -2007 MPRA Paper No 12504 Baro, R J (1999) Determinants of Economic Growth: A Cross-Country Empirical Study National Bureau of Economic Research, Inc, NBER Working Papers: 5698 Beck, P and Maher, M (1986) A Comparison of Bribery and Bidding in Thin Markets Economics Letters, 20(1), pp 1-5 Constantinos Alexiou (2009) Government Spending and Economic Growth: Econometric Evidence from the South Eastern Europe (SEE) Journal of Economic and Social Research 11(1) 2009, 1-16 Dzhumashev, R (2009) Is there a direct effect of corruption on growth? MPRA Paper No 18489 Gray, Cheryl, W., Kaufman, Daniel (1998) Corruption and development World Bank, PREM Notes ; no Public Sector Gundlach, Erich (1996) Openness and Economic Growth in Developing Countries Kiel Working Paper No 749 Available at SSRN: http://ssrn.com/abstract=1608 58    Hodge, A., Shankar, S., Rao, D S P., Duhs, A (2009) Exploring the Links between Corruption and Growth Australia: University of Queensland Discussion Paper No.392 Herrera, Santiago (2007) Public Expenditure and Growth World Bank, Policy Research Working Paper No.4372 Kaouthar Gazdar (2012) Does corruption “grease the wheels” of growth? Empirical Evidence from MENA countries Available at http://www.asectu.org/userfiles/Gazdar%20Kaouthar.pdf Kaufmann, D and Wei, S.-J (1999) Does "Grease Money" Speed Up the Wheels of Commerce?.National Bureauof Economic Research, Inc, NBER Working Papers: 7093 Lui, F T (1985) An Equilibrium Queuing Model of Bribery Journal of Political Economy, 93(4), pp 760-81 Lutz, M B., Ndikumana, L (2008) Corruption and Growth: Exploring the Investment Channel University of Massachusetts Economics Department Working Paper Series.Paper 33 Mankiw, N G., Romer, D., Weil, D N (1992) A Contribution to the Empirics of Economic Growth The Quarterly Journal of Economics, Vol 107, No 2, pp 407-437 Mauro, P (1995) Corruption and Growth.Quarterly Journal of Economics, 110(3), pp 681-712 Meon, P G., Sekkat, K (2003) Does Corruption Grease or Sand the Wheels of Growth? Public choice, 122, pp 69-97 59    Meon, P G., Weill, L (2008) Is corruption efficient grease? BOFIT discussion papers 20/2008 Mo, P H (2001) Corruption and Economic Growth Journal of Comparative Economics, 29(1), pp 66-79 Murphy, K M., Shleifer, A., Vishny, R W (1991) The Allocation of Talent: Implications for Growth Quarterly Journal of Economics, 106(2), pp 503-30 Murphy, K M., Shleifer, A and Vishny, R W (1991) The Allocation of Talent: Implications for Growth Quarterly Journal of Economics, 106(2), pp 503-30 Muscatelli, Anton, Darby, Julia and Li, Chol-Won (2000) Political Uncertainity, Public Expenditure and Growth CESifo Working Paper Series No 310 Available at SSRN: http://ssrn.com/abstract=263520 Pascual, M and Alvarez-García, S.(2006) Government Spending and Economic Growth in the European Union Countries: An Empirical Approach Available at SSRN: http://ssrn.com/abstract=914104 Pellegrini, L.,Gerlagh, R (2004) Corruption's Effect on Growth and Its Transmission Channels.Kyklos, 57(3), pp 429-56 Tanzi, V and Davoodi, H (1997) Corruption, Public Investment, and Growth International Monetary Fund, IMFWorking Papers: 97/139 Ugur, M., Dasgupta, N (2011) Corruption and economic growth: A meta-analysis of the evidence on low-income countries and beyond MPRA Paper No 31226 The content on website of Transparency International http://www.transparency.org The content on website of World Bank http://www.worldbank.org/ 60    APPENDIX Instrumental variables (2SLS) regression gGX Coef.            Std. Err.      z CI LY0 _cons ‐3.480672   2.543711    ‐1.37 3210442    1.940053     0.17 15.97564    12.30455     1.30 Number of obs 33 Wald chi2(2) 2.55 Prob > chi2 0.2792 R‐squared =        Root MSE 8.1386 P>z 0.171 0.869 0.194 [95% Conf Interval] ‐8.466254 1.504909 ‐3.481389 4.123478 ‐8.140835 40.09212 Tests of endogeneity Ho: variables are exogenous Durbin (score) chi2(1) = Wu‐Hausman F(1,29) =  2.36199 (p = 0.1243) 2.23571 (p = 0.1457) Table A1: 2SLS regression of CI and gGX and endogenous check Instrumental variables (2SLS) regression gFC Coef.             Std. Err.      z CI LY0 _cons 5723176      3.60273     0.16 ‐2.780484   2.747752    ‐1.01 30.78557     17.42728     1.77 61    Number of obs Wald chi2(2) Prob > chi2 R‐squared Root MSE P>z 0.874 0.312 0.077 [95% Conf 33 1.34 0.5112 0.0449 11.527 Interval] ‐6.488902 7.633538 ‐8.165978 2.605011 ‐3.371276 64.94242 Tests of endogeneity Ho: variables are exogenous Durbin (score) chi2(1) = Wu‐Hausman F(1,29) = 0.005418 (p 0.9413) 0.004762 (p 0.9455) Table A2: 2SLS regression of CI and gFC and endogenous check Instrumental variables (2SLS) regression gTT Coef.            Std. Err.      z CI LY0 _cons ‐6.670516   3.396108    ‐1.96 3.514511    2.782271     1.26 19.95362    17.74461     1.12 Number of obs 34 Wald chi2(2) 3.87 Prob > chi2 0.1444 R‐squared =        Root MSE 11.808 P>z 0.05 0.207 0.261 [95% Conf Interval] ‐13.32677 ‐0.0142666 ‐1.938641 8.967663 ‐14.82518 54.73243 Tests of endogeneity Ho: variables are exogenous Durbin (score) chi2(1) = Wu‐Hausman F(1,30) = 3.41073 (p = 0.0648) 3.34503 (p = 0.0774) Table A3: 2SLS regression of CI and gTT and endogenous check 62    Instrumental variables (2SLS) regression EEPC Coef.            Std. Err.      z CI LY0 _cons ‐.0019058   .0010095    ‐1.89 0037269    .000827        4.51 ‐.0177971   .0052746    ‐3.37 Number of obs Wald chi2(2) Prob > chi2 R‐squared Root MSE P>z 0.059 0.001 [95% Conf 34 21.3 0.251 0.00351 Interval] ‐0.0038843 0.0000728 0.0021059 0.0053479 ‐0.0281352 ‐0.007459 Tests of endogeneity Ho: variables are exogenous Durbin (score) chi2(1) = Wu‐Hausman F(1,30) = 2.90679 (p = 0.0882) 2.80459 (p = 0.1044) Table A4: 2SLS regression of CI and EEPC and endogenous check Instrumental variables (2SLS) regression PI Coef.            Std. Err.      z CI LY0 _cons ‐.0308337   .2356989    ‐0.13 2831546    .193097        1.47 ‐2.432909   1.231523    ‐1.98 Number of obs Wald chi2(2) Prob > chi2 R‐squared Root MSE P>z 0.896 0.143 0.048 [95% Conf 34 3.01 0.2221 0.0419 0.81954 Interval] ‐0.4927951 0.4311276 ‐0.0953086 0.6616178 ‐4.84665 ‐0.0191674 Tests of endogeneity Ho: variables are exogenous Durbin (score) chi2(1) = Wu‐Hausman F(1,30) = 3.08086 (p =0.0792) 2.98928 (p = 0.0941) Table A5: 2SLS regression of CI and PI and endogenous check 63    No 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 COUNTRY Argentina Azerbaijan Bangladesh Belarus Brazil Bulgaria Cape Verde Chile Colombia Ecuador El Salvador Ethiopia Ghana Guinea Jordan Kyrgyz Republic Lao PDR Latvia Lebanon Lithuania Macedonia, FYR Madagascar Mexico Mongolia Morocco Namibia Panama Romania Serbia Turkey Ukraine Uruguay Uzbekistan Venezuela, RB Vietnam List of countries in observation 64