Impact of household credit on education and healthcare spending by the poor in Peri-Urban areas, Vietnam

ed 6 to 18 years old, number of persons aged 18 to 60 years old, dummy of older than 60 years old, ward dummies, marital status*head’s gender. In some cases to meet the balancing property in estimation of propensity scores we added interaction terms in specifications of the propensity score. ATTK: Average Treatment Effect on the Treated, Kernel matching. ATTR: Average Treatment Effect on the Treated, Radius matching. entrepreneurial ability and skills. Our estimation strategy, therefore, sought to reduce bias and enhance the reliability of the PSM estimates. Our findings are in line with those of Nguyen (2008) and Quach and Mullineux (2007), who have found that access to credit programmes has a positive impact on household welfare and poverty reduction in rural Vietnam. Spending on education and healthcare affects the formation of human capital. Given their limited physical and financial assets, human capital becomes the most important asset for the poor in peri-urban areas. Therefore, policies aiming at reducing poverty in these areas should consider enhancing subsidized credit to the poor and/or providing subsidized or free universal education and healthcare services. 07.indd 113 3/7/14 9:36:39 AM Journa l o f Southeas t As ian Economies 114 Vo l . 31 , No . 1 , Apr i l 2014 APPENDIXES Appendix 1 Pre-treatment Income per Capita of the Poor and Their Sources of Loans Mean income is VND3.57 million with standard deviation as VND0.844 million, and median of VND3.6 million (see Figure 3). Monthly income per capita is just about VND300,000 (or US$18), slightly higher than the national official poverty line (equivalent to US$ 16.4).8 FIGURE 3 Pre-treatment per Capita Income Note: Unit of horizontal axis is VND1,000. 1. 0e -0 4 0 2. 0e -0 4 3. 0e -0 4 D en si ty 4. 0e -0 4 5. 0e -0 4 2000 30001000 5000 60004000 INCOME 07.indd 114 3/7/14 9:36:39 AM Journa l o f Southeas t As ian Economies 115 Vo l . 31 , No . 1 , Apr i l 2014 TABLE 5 Sources of Loans Sources of loans No. of loans Percentage Mean of loans (VND1,000) Formal 336 55% 9,327 VBSP 37 6% 9,622 Commercial banks 26 4% 54,923 JCSF 29 5% 4,564 Social political organizations 62 10% 4,564 HEPRF 182 30% 5,176 Informal 272 45% 5,229 Moneylenders, ROSCAs, pawnbrokers, others 51 8% 9,218 Friends, relatives, neighbors 221 36% 4,308 Overall 608 100% 7,494 source: Calculation from authors’ survey; VBSP- Vietnam Bank for Social Policies; JCSP — Job Creation Support Fund; HEPRF — The Hunger Elimination and Poverty Reduction Funds; ROSCAs- Rotating savings and credit associations. Appendix 2 The Average Treatment Effect on Monthly Average Education Expenditure in VND1,000 Using Matching Estimators for the Entire Sample with Reduced Bandwidth (0.01) and Radius (0.05) Control variables in the propensity score estimation Treated/controls Kernel matching Radius matching Head’s gender, head’s age, head’s education, marital 86.58 95.63 status, school-aged child ratio, and ward dummies (S1) 304/107 (35.71)* (33.35)** S2=S1plus initial income in log, initial assets in logarithm 78.95 91.25 304/101 (36.53)* (33.46)** Head’s gender, head’s age, head’s education, marital 83.34 85.99 status, number of children from 6 to 18, and ward 304/107 (33.07)* (34.69)* dummies (S3) S4=S3 plus initial income in log, initial assets in 77.98* 85.07* logarithm 304/101 (36.53) (35.46) 07.indd 115 3/7/14 9:36:39 AM Journa l o f Southeas t As ian Economies 116 Vo l . 31 , No . 1 , Apr i l 2014 Appendix 3 The Average Treatment Effect on Monthly Average Healthcare Expenditure in VND1,000 Using Matching Estimators for the Entire Sample with Reduced Bandwidth (0.01) and Radius (0.05) Control variables in the propensity score estimation Treated/controls Kernel matching Radius matching Specification 1 (S1) 95.36 109.33 304/103 (49.21)* (51.07)* S2=S1 plus initial income in log, initial assets in 94.85 99.086 logarithm 304/97 (50.91)+ (53.83)+ Specification 3 (S3) 118.799 129.242 304/107 (45.74)** (44.56)** S4=S3plus initial income in logarithm, initial assets 103.15 114.571 in log 304/102 (49.87)* (49.24)* Appendix 4 The Average Treatment Effect on Monthly Average Education Expenditure in VND1,000 Using Matching Estimators for the Entire Sample with Reduced Bandwidth (0.01) and Radius (0.05) Control variables in the Informal credit vs. Formal credit vs. Formal vs. propensity score estimation Non-borrowers Non-borrowers Informal ATTK ATTR ATTK ATTR ATTR Specification 1 (S1) 1.64 22.07 151.72 163.46 101.26 (45.59) (37.07) (47.33)** (51.03)** (44.89)* Specification 2 (S2) 14.42 12.94 140.52 143.62 116.54 (38.72) (41.57) (49.89)** (53.94)** (53.33)* Specification 3 (S3) 39.05 27.92 142.33 158.14 103.75 (44.19) (39.01) (50.62)** (46.47)** (46.77)* Specification 4 (S4) 3.59 13.17 138.56 145.36 109.70 (42.27) (40.86) (48.50)** (49.80)** (49.06)* Notes: ATTK: Average Treatment Effect on the Treated, Kernel matching. ATTR: Average Treatment Effect on the Treated, Radius matching. 07.indd 116 3/7/14 9:36:39 AM Journa l o f Southeas t As ian Economies 117 Vo l . 31 , No . 1 , Apr i l 2014 Appendix 5 The Average Treatment Effect on Monthly Average Healthcare Expenditure in VND1,000 Using Matching Estimators for the Entire Sample with Reduced Bandwidth (0.01) and Radius (0.05) Control variables in the Informal credit vs. Formal credit vs. Formal vs. propensity score estimation Non-borrowers Non-borrowers Informal ATTK ATTR ATTK ATTR ATTR Specification 1 (S1) 76.74 82.27 195.91 203.89 182.58 (44.28)+ (42.20)+ (92.66)* (95.21)* (101.87)+ S2=S1plus initial income in log, 62.94 65.15 171.79 177.72 190.70 initial assets in log (46.35) (44.34) (102.53)+ (98.07)+ (93.12)+ Specification 3 (S3) 67.11 68.71 210.84 195.42 171.16 (42.33) (44.44) (93.08)* (100.89)* (105.08)+ S4=S3plus initial income in log, 81.99 90.09 205.71 195.77 179.03 initial assets in log (46.22)+ (48.36)+ (91.63)* (98.01)* (102.92)+ Notes: ATTK: Average Treatment Effect on the Treated, Kernel matching. ATTR: Average Treatment Effect on the Treated, Radius matching. NOTES The authors thank, without implicating, Andrea Menclova, Asadul Islam and the participants of the Sixth Australasia Development Economics Workshop for their helpful comments and suggestions. Any remaining errors are those of the authors. 1. HCMC has twenty-four districts. District 9 has the fifth lowest population density, with a population of 227,816 (in 2008). 2. The list was provided by the District Department of Labour, Invalids and Social Affairs. Poor households in the list excluded temporary migrants. 3. When we reduced the bandwidth and radius to 0.01 and 0.05 for kernel and radius matching respectively, the results are robust. See Appendixes 2 to 5. 4. Some studies suggest that the estimation should be in the range of 0.1 to 0.9, but there are 44 observations with scores greater than 0.9 (about 11 per cent of the sample); if these are dropped, the estimates will be misleading (Crump et al. 2009). 5. The sets of variables used for estimating scores to draw Figures 1 and 2 are different. Each set of variables should affect both credit participation and outcomes. The impact on education expenditure is featured in Figure 1 and the impact on healthcare expenditure is featured in Figure 2. That is why the two figures are slightly different. 6. Basically, formal credit amounts are larger than informal ones, hence formal credit is considered as consituting a higher level of treatment or a full treatment. We expect that if loans have an effect on outcomes (i.e., education and health care expenditure), borrowing greater amounts will have a stronger impact on the respective outcomes. The mean of accumulated loans per household is VND8,317 thousand (about US$500) and VND15,135 thousand (about US$920) for informal and formal credit, respectively. The average size per loan is VND5,229 thousand (about US$317) and VND9,327 thousand (about US$566) for informal and formal credit, respectively. 7. The impact on both expenditure and school enrolment and/or attendance should be investigated to provide a more complete picture of the effects on education. The impact on child schooling was examined in a separate paper in International Development Planning Review (forthcoming). 8. Average exchange rate in 2008, US$1=VND16,481. 07.indd 117 3/7/14 9:36:39 AM Journa l o f Southeas t As ian Economies 118 Vo l . 31 , No . 1 , Apr i l 2014 REFERENCES Armendariz, de Aghion B. and Johnathan Morduch. The economics of microfinance. Cambridge, MA and London: The MIT Press, 2010. Beegle, K., Dehejia, R. and Gatti, R. “Why should we care about child labor? The education, labor market, and health consequences of child labor”. NBER Working Paper Series, Working paper 10980 (2004). Bryson, A., R. Dorsett and S. Purdon. “The use of propensity score matching in the evaluation of active labor market policies”. Policy Studies Institute and National Centre for Social Research, University of Westminster, 2002. Caliendo, M. and S. Kopeinig. “Some practical guidance for the implementation of propensity score matching”. Journal of Economic Surveys 22, no. 1 (2008): 31–72. Coleman, B. E. “Microfinance in Northeast Thailand: Who benefits and how much?”. World Development 34, no. 9 (2006): 1612–38. Consultative Group to Assist the Poor (CGAP). ”Is microfinance an effective strategy to reach the Millennium Development Goals?”. CGAP FocusNote no. 24 (2003). Washington, D.C.: Littlefield, E., Morduch, J., & Hashemi, S. Retrieved from . Crump, R., V. Hotz, G. Imbens and O. Mitnik. “Dealing with limited overlap in estimation of average treatement effects”. Biometrika 96, no. 1 (2009): 187–99. Cull, R., A.D. Kunt and J. Morduch. “Microfinance meets the market”. Journal of Economic Perspectives 23, no. 1 (2009): 167–92. Dehejia, R.H. “Practical propensity score matching: a reply to Smith and Todd”. Journal of Econometrics 125 (2005): 355–64. Dehejia, R. and Gatti, R. “Child labor, the role of income variability and access to credit in a cross-section of countries”. The World Bank, Policy Research Working Paper no. 2767, 2002. Dehejia, R. and Wahba, S. “Casual effects in non-experimental studies: Re-evaluating the evaluation of training programs”. Journal of the American Statistical Association 94, no. 448 (1999): 1053–62. ———. “Propensity score matching methods for nonexperimental causal studies”. Review of Economics and Statistics 84, no. 1 (2002): 151–61. Dias, M.C., H. Ichimura and G. Berg. “The matching method for treatment evaluation with selective participation and ineligibles”. Centre for Microdata Methods and Practice. CEMMAP working paper CWP33/07 (2007). Edmonds, E. “Child labor and schooling responses to anticipated income in South Africa”. Journal of Development Economics 81, no. 2 (2006): 386–414. Glewwe, P., H. Jacoby and E. King. “Early childhood nutrition and academic achievement: a longitudinal analysis”. Journal of Public Economics 81 (2000): 345–68. GSO. “Population and housing census, 1 April 2009”. Complete Report and Major Findings, General Statistical Office. Hanoi, Vietnam (2010). Hazarika, G. and S. Sarangi. “Household access to microcredit and child labor in rural Malawi”. World Development 36, no. 5 (2008): 843–59. Heckman, J. and J. Smith. “The pre-program earnings dip and the determinants of participation in a social program: Implications for simple program evaluation strategies”. NBER Working Paper Series, Working paper 6983 (1999). Heckman, J., H. Ichimura and P. Todd. “Matching as an econometric evaluation estimator: Evidence from evaluating a job training program”. Review of Economic Studies 64, no. 4 (1997): 605–54. Imbens, G. “Nonparametric estimation of average treatment effects under exogeneity: A review”. Review of Economics and Statistics 86, no. 1 (2004): 4–29. Imbens, G. and J. Wooldridge. “Recent developments in the econometrics of program evaluation”. Journal of Economic Literature 47, no. 1 (2009): 5–86. Islam, A. and C. Choe. “Child labour and schooling responses to access to microcredit in rural Bangladesh”. MPRA Working Paper no. 16842 (2009). Retrieved from Munich Personal RePEc Archive < muenchen.de/16842/1/MPRA_paper_16842.pdf>. Jacoby, H. and E. Skoufias. “Risk, financial markets, and human capital in a developing country”. Review of Economic Studies 64, no. 3 (1997): 311–35. Khandker, S. “Microfinance and poverty: Evidence using panel data from Bangladesh”. World Bank Economic Review 19, no. 2 (2005): 263–86. Lee, M.J. Micro-econometrics for policy, program and treatment effects. New York: Oxford University Press, 2005. Maldonado, J. and C. Gonzalez-Vega. “Impact of microfinance on schooling: Evidence from poor rural households in Bolivia”. World Development 36, no. 11 (2008): 2440–55. 07.indd 118 3/7/14 9:36:40 AM Journa l o f Southeas t As ian Economies 119 Vo l . 31 , No . 1 , Apr i l 2014 Morduch, J. “Does microfinance really help the poor? New evidence from flagship programs in Bangladesh”, 1998 . Mosley, P. “The use of control groups in impact assessment for microfinance”. International Labor Office, Geneva, Working Paper no. 19, 1997. Nguyen, C. “Is a Governmental Microcredit Program for the Poor really Pro-poor? Evidence from Vietnam”. Developing Economies 46, no. 2 (2008): 151–87. Quach, H. and A. Mullineux. “The Impact of Access to Credit on Household Welfare in Rural Vietnam”. Research In Accounting In Emerging Economies 7 (2007): 279–307. Pitt, M. and Khandker, S. “The impact of group-based credit programs on poor households in bangladesh: does the gender of participants matter?”. Journal of Political Economy 106, no. 5 (1998): 958–92. Pitt, M., S. Khandker, O. Chowdhury and D. Millimet. “Credit programs for the poor and the health status of children in rural Bangladesh”. International Economic Review 44, no. 1 (2003): 87–118. Rosenbaum, P. and D. Rubin. “The central role of the propensity score in observational studies for causal effects”. Biometrika 70 (1983): 41–55. ———.“Constructing a control group using multivariate matched sampling methods that incorporate the propensity”. American Statistician 39 (1985): 33–38. Smith, J. and P. Todd.“Does matching overcome Lalonde’s critique of nonexperimental estimators?”. Journal of Econometrics 125, no. 1–2 (2005): 305–53. United Nations. World urbanization prospects: The revision 2007. New York: United Nations, 2007. VDR. “Vietnam Development Report 2010: Modern Institutions”. Joint Donor Report to the Vietnam Consultative Group Meeting, Hanoi, Vietnam, 3–4 December 2009. Tinh Doan is Research Analyst at the Ministry of Business, Innovation and Employment, and Fellow at University of Economics and Business, Vietnam National University, Hanoi. John Gibson is Professor in the Economics Department, University of Waikato, New Zealand. Mark Holmes is Professor in the Economics Department, University of Waikato, New Zealand. 07.indd 119 3/7/14 9:36:40 AM