Microeconomic non-experimental analysis
Using rigorous techniques of microeconomic non-experimental analysis to more fully characterize the effectiveness and distributive effects of public interventions in developing countries.
Research led under this program aims to provide a fuller characterization of the effectiveness and distributive effects of various phenomena and public interventions (i.e. “treatments”). These treatments include social protection schemes, labour market legislation, job training, education, migration, access to credit, agricultural development programs, etc.
It is important for policymakers and donors to know whether these policy interventions or programs will provide the expected benefits. This knowledge is also important for generating political support for the continuation, expansion or modification of the programs.
Policymakers are interested in answers to questions such as:
- What is the impact of a program or policy on beneficiaries?
- What would be the impact of expanding eligibility to the program?
- Do the poor gain the most from the intervention?
- Do some benefit who should not?
- Would it be possible to have the same impact at lower cost?
This research group encourages rigorous research to improve the design and/or implementation of such public interventions. Research supported through the PEP microeconomic group provides reliable evidence to assist policymakers facing challenges such as:
- How should national priorities for public and government spending be defined?
- Which dimensions of welfare are most important and should be prioritized?
- Which populations (region, age group, sex, education, etc.) should be targeted and how?
- How does poverty and inequality change over time?
To answer such questions, PEP research led under this program uses rigorous econometric techniques, which in the impact evaluation literature are normally classified as non-experimental approaches.
These approaches are used to resolve selection bias in assessing the treatment effect. They are:
- Regression discontinuity design
- Propensity score matching
- Instrumental variables
- Difference in difference
- Endogenous Switching Regression
- A combination of the above techniques
Parametric (e.g. quantile regressions) and non-parametric techniques are also often used to understand the heterogeneous effects of the treatment.
While the tools mentioned above are used to evaluate the impact of a given program or shock, it is sometimes useful to anticipate the potential effects of a specific treatment or shock on individual behaviour and welfare. For this, the PEP Microeconomic group proposes and develops micro-simulation techniques.
Microeconomic policy research advocates the use of existing nation-wide micro-based data. As such, new data collection is not usually carried out. The microeconomic policy research group strives for good comparative international research while retaining a focus on national results and impact. Over the years, an increasing number of projects supported within this group have made use of panel data.
The PEP microeconomic policy research group has yielded major contributions to the field. These include developing two world-renowned software tools for distributive analysis, used and valued worldwide throughout the international development research community:
- Distributional Analysis/Analyse distributive (DAD) – DAD is designed to facilitate the analysis and the comparison of social welfare, inequality, poverty and equity across distributions of standard living. Its features include the estimation of a large number of indices and curves that are useful for distributive comparisons as well as the provision of asymptotic standard errors to enable statistical inference.
- Distributional Analysis Stata Package (DASP) – The main purpose of DASP is to produce a comprehensive package of STATA modules to help analyse the distribution of living standards. In particular, it helps estimate the most popular statistics (e.g. indices, curves) used for poverty, inequality, social welfare, and equity analysis, estimate the differences in such statistics, perform the most popular distributive decomposition procedures, and check the ethical robustness of distributive comparisons.
RELATED LINKS, RESOURCES, PROJECTS & RESEARCHERS
More about past PEP microeconomic policy research
Multidimensional poverty analysis
PMMA program research on multidimensional poverty has made use of recently developed or neglected methodologies in economics, which can be highly effective in answering the concerns of policymakers that poverty and well-being cannot be considered solely on the basis of money-metric assessments. Such a concern has, inter alia, been strongly expressed in the development and monitoring of the Millennium Development Goals (MDGs). PEP’s original achievements in this area have been noted by many institutions and organizations around the world.
Public spending and its impact on poverty and equity
This has scored particularly well in the context of several national Poverty Reduction Strategy Paper (PRSP) exercises in Africa. In 2007, in order to contribute to improve the understanding of the impact of public spending on poverty and equity, the PEP PMMA research program has broadened its methodological approach to incorporate intra-household allocation of well-being and poverty dynamics
Growth and Poverty Dynamics
This theme includes:
- Comparing poverty across time and space, using statistically and normatively reliable methods;
- Analyzing the impact of demographic and socio-economic transformations (such as emigration, urbanization, environmental changes, changes in family sizes and composition, changes in human capital, changes in trade and openness) on poverty and inequality;
- Analyzing the impact of growth and redistribution on the evolution of poverty, and in particular, the heterogeneity of that impact;
- Analyzing the impact of growth in factor incomes on poverty alleviation;
- Analyzing the pro-poorness of growth and policy;
- Using panel data to look at individual and household poverty dynamics;
- Analyzing social protection, security, risk management, and vulnerability;
- Studying poverty duration, transient vs. permanent poverty.
This is an area in which current research activities are very strong both in the policy and the academic communities. Understanding the impact on poverty and inequality of national and sectoral growth, inequality, migration and/or demographic changes for instance is at the top of several international and national policy institutions' agenda. Individual poverty dynamics – including issues related to mobility, vulnerability and transiency of poverty – is also currently the focus of much academic research work, both from an ethical and a statistical point of view. Researchers everywhere have been meaning to evolve around the understanding of the experience of poverty across different categories of households, characterized not merely by household incomes, but also by different kinds of gender and age compositions. Comparative work across countries and time (made easier by a research structure such as that of PEP) is particularly valuable in this area, since the results are usually country specific.
Policy impact analysis
Many countries are designing and implementing social policies targeted to specific populations, e.g. social protection to poor people, job training programs to the youth and unemployed, agricultural development programs to farmers. It is important for policymakers and donors to know whether the programs have the expected benefits – and such knowledge is also important for generating political support for the continuation or expansion of the programs. This theme thus encourages the conduct of rigorous research to help improve the design and the implementation of such programs. This is often done in the context of welfare and job training programs in industrialized countries; which is also increasingly applied in developing countries.
Policymakers, for instance, are likely to be interested in answers to questions such as:
- What is the impact of a program on participants?
- What would be the impact from expanding eligibility to the program?
- Do the poor gain the most from the program?
- Do some benefit who should not?
- Would it be possible to have the same impact at lower cost?
Indeed, this is an area for which interest has grown strongly over the last decade and which has materialized, through PEP, in research projects on, e.g. the poverty and employability effects of workfare programs in Argentina, the effect of business development services for female microfinance clients in Peru, the poverty impact of the national micro-credit program in Vietnam, etc... The specific characterization and increasing importance given to such assessments has led to the creation of the PEP Policy Impact Evaluation Research Initiative (PIERI), now the Experimental Research group, funded by AusAID in 2007.
Intra-household allocation of well-being
This analysis is undertaken with a view to going from household-based poverty analysis to individual-based poverty analysis. Such analysis can matter significantly for looking at equity, gender issues, child welfare, social services access, among many other aspects. It can help assessing the true extent of inequality in the distribution of resources, as well as the extent of inequity in access to goods and services. This is fundamental to understanding the effect of policy and growth on welfare, poverty and exclusion. This is also an area for which PEP researchers have long expressed an interest.
In the course of the first phase of PEP, this interest materialized in part through the creation of a Gender Challenge Fund in collaboration with MIMAP’s Gender Network. Through this Fund, PEP was enabled to award additional grants for projects dealing with gender issues within both the macro and micro programs. More generally, the aim of such thematic focus was to encourage researchers to go beyond the usual household-based measurement of well-being to consider individual well-being, and how it varies across gender and age. Such analysis can matter significantly for looking at equity, gender issues, child welfare, social services access, etc. which is the focus of much of current international initiatives, from both an ethical and a statistical point of view.
The above-mentioned list of themes of PEP micro policy research is not exhaustive. The group has supported a number of projects on gender analysis and child well-being, as well as using micro-simulation techniques and experimental data analysis. The micro group's scientific support team has also led a number of special programs, such as the PEP-OPHI Human Development and Capabilities Initiative, and collaborated in initiatives supported under the macro/modeling program, to assist in the use of micro-simulation techniques.
In all cases, PEP microeconomic policy research work emphasizes the importance of reliability and robustness in drawing out ethical and statistical conclusions.