Assessing the Effectiveness of Proxy Means Testing in Targeting - A Case Study of Kenyan Cash Transfer for Orphans and Vulnerable Children

Amna BABAR

Master's thesis

Food Aid and Assistance Targetting in Conflict Areas

Abstract:

While universal provision of services and assistance has placed nation-building at heart in different countries for many years, very often than not, this provision is stratified to the most powerful, and marginalizes large populations of the needy and poor. This gives rise to the need of devising, implementing and evaluating mechanisms to ensure that the poor indeed benefit, namely from the mechanisms of 'targeting'. Answering the questions of 'where', 'when' and more importantly 'who' should be receiving assistance, targeting plays a pivotal role in determining the effectiveness of different assistance transfer modalities. This study reviews the various techniques of administrative and community based targeting at household and individual level and focuses on the econometric tool of Proxy Means Testing (PMT); with the primary objective of understanding the methodological discretions in the technique and their impact on its effectiveness. Using the updated dataset of Kenyan Integrated Household Budget Survey 2015/16, a PMT simulation is run for the country's cash transfer program for Orphans and Vulnerable Children (OVC) at different poverty thresholds by employing Ordinary Least Squares. The results indicate that targeting through proxy means testing performs markedly better than universal transfers and is significantly progressive in both, incidence and coverage. It is also encouraging to note that proxy means testing helps eliminate as much as 50% of inclusion errors otherwise implied by including all OVC households in the program. This, however, comes with a repeatedly emphasized trade-off with a large number of households incorrectly excluded from being selected as eligible for the transfer. It is therefore a policy matter and future research contributions on conducting an ex-poste analysis of actual beneficiaries of OVC-Cash Transfer can help modify the welfare proxies and increase the predictive power of proxy means tests as a targeting tool for the program.

Abstract:

While universal provision of services and assistance has placed nation-building at heart in different countries for many years, very often than not, this provision is stratified to the most powerful, and marginalizes large populations of the needy and poor. This gives rise to the need of devising, implementing and evaluating mechanisms to ensure that the poor indeed benefit, namely from the mechanisms of 'targeting'. Answering the questions of 'where', 'when' and more importantly 'who' should be receiving assistance, targeting plays a pivotal role in determining the effectiveness of different assistance transfer modalities. This study reviews the various techniques of administrative and community based targeting at household and individual level and focuses on the econometric tool of Proxy Means Testing (PMT); with the primary objective of understanding the methodological discretions in the technique and their impact on its effectiveness. Using the updated dataset of Kenyan Integrated Household Budget Survey 2015/16, a PMT simulation is run for the country's cash transfer program for Orphans and Vulnerable Children (OVC) at different poverty thresholds by employing Ordinary Least Squares. The results indicate that targeting through proxy means testing performs markedly better than universal transfers and is significantly progressive in both, incidence and coverage. It is also encouraging to note that proxy means testing helps eliminate as much as 50% of inclusion errors otherwise implied by including all OVC households in the program. This, however, comes with a repeatedly emphasized trade-off with a large number of households incorrectly excluded from being selected as eligible for the transfer. It is therefore a policy matter and future research contributions on conducting an ex-poste analysis of actual beneficiaries of OVC-Cash Transfer can help modify the welfare proxies and increase the predictive power of proxy means tests as a targeting tool for the program.

Keywords

targeting, proxy means test, incidence, coverage, inclusion error, exclusion error

Language used: English

Date on which the thesis was submitted / produced: 8. 6. 2020

Thesis defence

• Supervisor: prof. Maria Sassi

Palacký University Olomouc

Faculty of Science

Master programme / field:
Geography / International Development Studies