Abstract
There is a long-standing belief that accurate targeting of the poor can improve the effectiveness of development policy. While a number of factors, such as governance failure, inadequate targeting methods, political interference or lack of political will, might explain low targeting efficiency, this article focuses on improving indicator-based models that identify poor households more accurately. Using stepwise regressions along with out-of-sample validation tests, this article develops proxy means test models for rural and urban Malawi. The models developed have proved their validity and can therefore be used to target a wide range of development policies in the country. Likewise, the research here can be extended to other developing countries.
Abstract
Selon une croyance de longue date, un ciblage précis des pauvres peut améliorer l′efficacité des politiques de développement. Alors qu′un certain nombre de facteurs, tels que la mauvaise gouvernance, les méthodes de ciblage inadéquates, l′ingérence ou le manque de volonté politique, peuvent expliquer le faible taux de ciblage des programmes de développement, cet article cherche à améliorer les modèles conçus à partir d'indicateurs socio-économiques pouvant identifier plus précisement des ménages pauvres. En utilisant une procédure de régression séquentielle et de validation hors échantillon, ce travail a développé des modèles opérationnels pouvant servir à l'identification et au ciblage des ménages pauvres dans les zones rurales et urbaines du Malawi. Ces modèles se sont révélés bien valides dans des échantillons indépendants. Ils peuvent donc être utilisés pour cibler un large éventail de programmes de développement dans le pays. En outre, cette étude peut être étendue à d′autres pays en développement.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
See Chinsinga (2005) and Levy and Barahona (2002) for further details on community-based targeting.
See Coady et al (2002) and Grosh and Baker (1995) for further details on means tests.
We gratefully acknowledge the National Statistics Office of Malawi (NSO) for providing us with the data.
The list of indicators was reduced to 116 for the urban model as some of the variables were not relevant in urban areas.
Indeed, the initial sample is self-weighted; however, household replacements for various operational reasons during the survey period resulted in a non-self-weighted random sample, which justifies the use of weighted regressions.
For example, Grosh and Baker (1995) argue that, strictly speaking, ordinary least square is not appropriate for predicting poverty. Glewwe (1992) and Ravallion and Chao (1989) try to solve the problem of targeting using more complex poverty minimization algorithms. These methods are, however, difficult to implement and have limited applications compared to the methods used in this article.
The logarithm of expenditures was used instead of simple expenditures because the log function better approximates a normal distribution.
See SAS Institute (2003) for further details on the score procedure.
See Johannsen (2009) for further details on the percentile-corrected approach.
See the section ‘Targeting Ratios and Robustness Tests’ for further details on the BPAC.
For brevity reasons, only out-of-sample results are presented throughout the article. The results from the model calibrations are available upon request.
These rates differ slightly from official statistics because of errors in the weights of the IHS2 report.
When reducing the cut-off, this trade-off also applies to the WL method, which estimates the probability of being poor.
The variances were estimated at 0.70 for urban households versus 0.38 for rural households. For the four urban centers, they were set at 0.47, 0.95, 0.63 and 0.53 versus 0.34, 0.38, 0.26, 0.38, 0.38, 0.39, 0.38 and 0.35 for the eight agricultural development districts.
To allow for a stricter comparison of both estimation methods, we used the same indicator set to fit both regressions in separate simulations; however, the results do not differ from the observed performances.
The coverage of the poor, or poverty accuracy, is also known as sensitivity. On the other hand, the inclusion of non-poor, or inclusion error, is estimated as one minus specificity. It is the error of predicting non-poor as poor, expressed as a percentage of non-poor. It differs from leakage, which is expressed as a percentage of the poor (see Wodon, 1997 and Baulch, 2002 for further details on ROC curves).
We consider the band (70 per cent – 90 per cent) as a good range for a reasonably accurate proxy means test model.
A similar trend emerges when the models were calibrated to the international and extreme poverty lines.
Further findings revealed the same pattern under the urban model. These results are available upon request.
See Benson et al (2006) for a study on the perceptions of welfare by Malawian households.
References
Ahmed, A. and Bouis, H.E. (2002) Weighting what's practical: Proxy means tests for targeting food subsidies in Egypt. Food Policy 27: 519–540.
Ahmed, A., Rashid, S., Sharma, M. and Zohir, S. (2004) Food Aid Distribution in Bangladesh: Leakage and Operational Performance. Washington DC: The International Food Policy Research Institute. Discussion Paper No. 173.
Baulch, B. (2002) Poverty Monitoring and Targeting Using ROC Curves: Examples from Vietnam. University of Sussex: The Institute of Development Studies. Working Paper No. 161.
Benson, T. (2002) Malawi – An Atlas of Social Statistics. National Statistics Office and International Food Policy Research Institute, Washington DC.
Benson, T., Payongayong, E., Ahmed, A. and Sharma, M. (2006) Towards operational methods in measuring absolute income poverty. Paper presented at the 26th International Association of Agricultural Economists Conference; 16–18 August, Gold Coast, Australia.
Besley, T. and Kanbur, R. (1993) The principles of targeting. In: M. Lipton and J. Van Der Gaag (eds.) Including the Poor, Proceedings of a Symposium Organized by the World Bank and the International Food Policy Research Institute; December. Washington DC: The World Bank, pp. 67–90.
Braithwaite, J., Grootaert, C. and Milanovic, B. (2000) Poverty and Social Assistance in Transition Countries. New York: Saint Martin's Press.
Campbell, M.K. and Torgerson, D.J. (1999) Bootstrapping: Estimating confidence intervals for cost-effectiveness ratios. International Journal of Medicine 92 (3): 177–182.
Chen, S. and Ravallion, M. (2008) The Developing World Is Poorer Than We Thought, But No Less Successful in the Fight against Poverty. Washington DC: The World Bank. Policy Research Working Paper No. 4703.
Chinsinga, B. (2005) The clash of voices: Community-based targeting of safety-net interventions in Malawi. Social Policy and Administration 39 (3): 284–301.
Coady, D., Grosh, M. and Hodinott, J. (2002) The Targeting of Transfer in Developing Countries: Review of Experiences and Lessons. Washington DC: The World Bank.
Coady, D. and Parker, S. (2009) Targeting performance under self-selection and administrative targeting methods. Economic Development and Cultural Change 57 (3): 559–587.
Deaton, A. (1997) The Analysis of Household Surveys: A Microeconometric Approach to Development Policy. Washington DC: Johns Hopkins University Press.
Efron, B. (1987) Better bootstrap confidence intervals. Journal of the American Statistical Association 82 (397): 171–185.
Glewwe, P. (1992) Targeting assistance to the poor: Efficient allocation of transfers when household income is not observed. Journal of Development Economics 38 (2): 297–321.
Government of Malawi and World Bank. (2007) Malawi Poverty and Vulnerability Assessment: Investing in Our Future, Synthesis Report, Main Findings and Recommendations. A report prepared by a team of the Government of Malawi and the World Bank Officials, Zomba, Malawi.
Greene, W.H. (2003) Econometric Analysis, 5th edn. New Jersey: Pearson Education Inc. Prentice Hall.
Grootaert, C. and Braithwaite, J. (1998) Poverty Correlates and Indicator-based Targeting in Eastern Europe and the Former Soviet Union. Washington DC: The World Bank.
Grosh, M.E. and Baker, J.L. (1995) Proxy Means Tests for Targeting Social Programs: Simulations and Speculation. Washington DC: The World Bank. Working Paper No. 118.
Gujarati, D. (2006) Essentials of Econometrics, International edn. New York: McGraw Hill.
Hall, P. (1994) Methodology and theory for the bootstrap. In: R.F. Engle and D.L. McFadden (eds.) Handbook of Econometrics IV. Amsterdam: Elsevier Science B.V., pp. 2342–2379.
Hentschel, J., Lanjouw, J.O., Lanjouw, P. and Poggi, J. (2000) Combining census and survey data to trace the spatial dimensions of poverty: A case study of Ecuador. World Bank Economic Review 14 (1): 1471–1165.
Horowitz, J. (1999) The Bootstrap. The Department of Economics, Iowa: University of Iowa.
Houssou, N. and Zeller, M. (2010) Targeting the poor and smallholder farmers: Empirical evidence from Malawi. Quarterly Journal of International Agriculture 49 (4): 341–358.
IRIS. (2005) Note on assessment and improvement of tool accuracy. University of Maryland: The IRIS Center (Mimeograph).
Johannsen, J. (2009) Operational assessment of monetary poverty by proxy means tests: The example of Peru. In: F. Heidhues, J. von Braun and M. Zeller (eds.) Development Economics and Policy 65. Frankfurt, Germany: Peter Lang.
Levy, S. and Barahona, C. (2002) Targeted Inputs Programme (TIP): Main Report of the Evaluation Programme. Report submitted to the Ministry of Agriculture and Irrigation (Malawi) and the Department for International Development, UK.
National Statistics Office. (2005) Malawi Second Integrated Household Survey (IHS2): Basic Information Document. A report of the 2004–2005 IHS2 survey. Zomba, Malawi.
Poverty Monitoring System. (2000) The State of Malawi's Poor: The Incidence, Depth, and Severity of Poverty. Policy. Brief No. 2, Malawi.
Ravallion, M. (1996) Issues in measuring and modelling poverty. The Economic Journal 106 (438): 1328–1343.
Ravallion, M. and Chao, K. (1989) Targeted policies for poverty alleviation under imperfect information: Algorithms and applications. Journal of Policy Modelling 11 (2): 213–224.
SAS Institute. (2003) The Logistic Procedure: Effect Selection Methods. Cary, NC: SAS Institute.
Schreiner, M. (2006) A simple poverty scorecard for India. Microfinance Risk Management, Washington University in Saint Louis: The Center for Social Development.
United Nations. (2009) Millennium Development Goals Report. New York: The United Nations.
Valdivia, M. (2005) Is Identifying the Poor the Main Problem in Targeting Nutritional Program? Washington DC: The World Bank. Discussion Paper No. 7.
van Bastelaer, T. and Zeller, M. (2006) Achieving the microcredit summit and millennium development goals of reducing extreme poverty: What is the cutting edge on cost-effectively measuring movement across the $1/day threshold? In: S. Daley-Harris and A. Awimbo (eds.) More Pathways out of Poverty. Connecticut: Kumarian Press, pp. 1–41.
Weiss, J. (2004) Reaching the Poor with Poverty Projects: What is the Evidence on Social Returns? Tokyo: The Asian Development Bank Institute. Research Paper No. 61.
Wodon, Q. (1997) Targeting the poor using ROC curves. World Development 25 (12): 2083–2092.
Wooldridge, J.M. (2006) Introductory Econometrics: A Modern Approach, 3rd edn. Ohio: Thomson South-Western.
World Bank. (2007) Malawi Social Protection Status Report No. 40027-MW. Washington DC: The World Bank. Sustainable Development Network, Africa Region.
World Bank. (2008) Global Purchasing Power Parities and Real Expenditures 2005 International Comparison Program (ICP). Washington DC: The World Bank.
World Bank. (2009) Malawi – Country brief, www.worldbank.org, accessed 11 September 2009.
Zeller, M. and Alcaraz V., G. (2005) Developing and Testing Poverty Assessment Tools: Results from Accuracy Tests in Uganda. The IRIS Center, Maryland: University of Maryland College Park.
Zeller, M., Alcaraz V., G. and Johannsen, J. (2005) Developing and Testing Poverty Assessment Tools: Results from Accuracy Tests in Bangladesh. The IRIS Center, Maryland: University of Maryland College Park.
Zeller, M., Sharma, M., Henry, C. and Lapenu, C. (2006) An operational tool for assessing the poverty outreach performance of development policies and projects: Results of case studies in Africa, Asia, and Latin America. World Development 34 (3): 446–464.
Acknowledgements
The authors thank Hans-Peter Piepho at the University of Hohenheim and Todd Benson at the International Food Policy Research Institute (IFPRI) for helpful comments. We are grateful to our former colleagues on the poverty assessment team of the University of Goettingen, Gabriela Alcaraz V., Julia Johannsen and Stefan Schwarze; and Anthony Leegwater at the Center for Institutional Reform and Informal Sector (IRIS), University of Maryland. We also thank two anonymous referees. We acknowledge the German Academic Exchange Service (DAAD) for its financial support.
Author information
Authors and Affiliations
Appendices
Appendix A
Appendix B
Appendix C
Appendix D
Appendix E
Rights and permissions
About this article
Cite this article
Houssou, N., Zeller, M. Operational Models for Improving the Targeting Efficiency of Development Policies: A Systematic Comparison of Different Estimation Methods Using Out-of-sample Tests. Eur J Dev Res 24, 465–490 (2012). https://doi.org/10.1057/ejdr.2011.34
Published:
Issue Date:
DOI: https://doi.org/10.1057/ejdr.2011.34