Research plays a crucial role in both learn- ing from past policies and thinking criti- cally about future policies. Without research, the conceptual and empirical foundations for policy making would be weak, new knowledge or data to inform policy decisions would be scant, there would be little or no innovation and little would be known about what succeeds and what fails in the fight against poverty. Without critical inspection by researchers, failed policy orthodoxies would often persist, successful policies might be dropped for the wrong reasons, and new fads and fashions would get more traction than they deserve.
To realize the potential for research to inform development practice, research must address relevant questions and it must provide credible answers. In short: rigor must come hand-in-hand with relevance. That is the guiding principle for all research done in the World Bank’s Development Research Group (DECRG). Research Highlights provides an annual report on the group’s efforts in implementing that principle.
Although research activities are found throughout the Bank, DECRG is the Bank’s sole unit devoted to research. The challenges for our research span the gamut of development policy, including access to finance, human development, rural and urban infrastructure, redistribution and insurance, macroeconomic stabilization, trade reform, and environmental sustainability. Indeed, the Research Group is one of the few places in the Bank where all sectors come together this way.
The Research Group aims to maintain world-class expertise in the core sector work in which the World Bank is involved. Within each sector, we endeavor to address the most pressing research issues, while avoiding duplication and keeping an eye on the cross-sectoral issues relevant to overall impacts on the Bank’s key development goals.
Using limited resources (about 1 percent of the administrative budget of the Bank), the Research Group attempts to focus on the highest priority areas within each sector and to take account of how the whole program fits together. The key challenge is to prioritize well and to exploit knowledge overlaps and synergies across sectors.
DECRG’s work differs from that of academia because most academic researchers draw their ideas from the work of other scholars. To be useful World Bank research must draw on inputs from nonresearchers actively involved in policy design and implementation in developing countries. Bank researchers learn directly from practitioners. DECRG’s managers strongly encourage engagement with Bank operations; indeed, each member of the department’s research staff is asked to cover about one-third of his or her time directly supporting the Bank’s work outside the department. Much of this "cross-support" is to Bank operations, helping with analytic work and projects in developing countries.
This report does not attempt to represent all of DECRG’s research in 2007, but rather it selectively highlights some findings from research published that year, drawing from the larger body of work done by each research team. In the rest of this note I would like to draw out two important cross-cutting themes of DECRG’s work.
Sound data and methods
Development practitioners carry a set of tools to their project and policy dialogues with governments, including data, measures, and models. Research can play a valuable role in expanding and improving the tool kit routinely employed by policy makers and analysts. This includes the data collected and the methods used to analyze those data. The Bank has become a major producer of development data, and researchers have played a crucial role in this effort.
A number of the Bank’s most successful data initiatives started as research projects. The Living Standards Measurement Study (LSMS) is an important example. The LSMS started in the Bank as a research project around 1980, with the aim of greatly improving household survey data in developing countries. At a more macro level, successful efforts to build up international trade data and analytical methods illustrate how a body of knowledge that constitutes a "global public good" was established, due in part to Bank research. Other examples include the various cross-country databases that have been developed out of research projects aiming to assess country performance in economic, political, and social domains. The Bank’s global poverty monitoring effort started as a small research project for a background paper to the World Development Report 1990. The Bank’s Enterprise Surveys (formally known as Investment Climate Surveys) started in the research department, but were eventually mainstreamed into the relevant sectoral units in the Bank.
Theoretical work (typically done in academia) often points to new tools, some more useful than others. Choosing wisely is key—some tools are more promising for practitioners than others. Researchers in DECRG often demonstrate the usefulness of new analytic tools in real applications, often working with operational colleagues. Over the years, numerous software programs developed by DECRG researchers have been used to facilitate data analysis throughout the Bank and in its client countries.
These "global public goods" would never have emerged without initiatives by researchers and institutional support from the Bank’s administrative budget and donors. Without that support, we would not see as much progress as we would like toward improving evidence-based policy making
in the developing world.
This is a second cross-cutting theme of the Group’s work. Evaluative research rigorously assesses whether development policies are effective and gauges when they tend to be more effective. This is broader than impact evaluation and embraces both "micro" interventions in specific sectors and macro policies. It includes both ex ante and ex post evaluation.
The Bank is uniquely positioned to learn from evaluative research. The Bank’s practitioners search continually for operational solutions to pressing development problems; our researchers have the training and skills needed to inform that search and to help learn from both policy successes and failures.
Evaluative research must start with a clear understanding of the problem that a policy or project is addressing. Why is the intervention needed? What are the market or government failures it addresses? Researchers can often help in identifying the policy objectives (properly weighing gains across different subgroups of a population, and different generations) and the relevant resource, information, incentive, and political economy constraints.
This role of research in conceptualizing the case for intervention can be important when the capacity for policy making is weak and/or it is captured by lobby groups, advocating narrow sectoral interests. The existence of trade-offs between sectors (such as due to governmental budget constraints), spillover effects across sectors (costs and benefits to one sector from policies in another), and interaction effects (whereby attainments in one dimension influence the impacts of policies in another) beg for a broader perspective on the foundations of policy. DECRG’s cross-sectoral work can contribute in essential ways to the Bank’s overall development goals.
As is now widely appreciated, evaluative research must always assess impacts against explicit and relevant counterfactuals, such as the absence of the policy in question or some other policy option. This requires sound research designs, using good data and credible strategies for identifying causal impacts from those data—taking proper account of the likely sources of bias. DECRG’s approach to evaluative research strives to answer questions that are vital to development outcomes. The policy question should drive the research agenda, rather than the researcher’s preference for certain types of data or certain methods.
Starting with the question, not the method, often leads to data and methods outside the domain traditionally favored by the researcher’s own disciplinary background. For example, some DECRG research to understand persistent poverty and the impacts of antipoverty programs has drawn on theories and methods from a wide range of disciplinary perspectives in the social sciences. Good researchers, like good detectives, assemble and interpret diverse forms of evidence in testing empirical claims.
DECRG’s research often tests the assumptions made in operational work. Even field-hardened practitioners act on the basis of some implicit model of how the world works, which helps them rationalize what they do and explain how their development project is expected to have impact. Researchers can perform a valuable role in helping to make those models explicit and (where possible) assess their veracity. Research has long played a role in questioning and (if need be) rethinking the received orthodoxies of development policy. This fuels a deeper understanding of the real-world factors that impinge success and helps identify the most effective levers for policy makers.
Publishing is essential if evaluative research is to have an impact. Development policy-making draws on accumulated knowledge built up in large part from published research findings. In addition to the contribution to shared knowledge through proper documentation and dissemination, publishing has other benefits. It serves as a screening and disciplining device for researchers. Publishing in refereed professional journals also helps establish a researcher’s credibility. And publishing helps DECRG attract and retain the best researchers. Often the bright young researchers, with freshly minted PhDs who are attracted to the Bank’s research department become deeply engaged in operational work within the Bank.
Introduction to the 2007 edition of Highlights
In September 2007 I took over as research director from Alan Winters, who returned to the University of Sussex. On behalf of the whole department I thank Alan for his leadership and unswerving commitment to fulfilling the role of research in promoting the goals of economic development.
Following a 2006 reorganization of DECRG, our department now comprises six research teams: Finance and Private Sector Development, Human Development and Public Services, Macroeconomics and Growth, Poverty and Inequality, Sustainable Rural and Urban Development, and Trade and International Integration. Each team has its own research manager and core staff (listed in each section of this report). In all, DECRG has about 100 full-time equivalent researchers and about 80 of them are World Bank staff. The research teams are backed up by able administrative and information technology support teams.
Research is conducted both within and across the six teams, often in collaboration with researchers in other parts of the Bank and with colleagues in universities and research institutions throughout the world, including collaborators in most of the developing countries in which the department’s work is focused. In 2007, the department’s country-specific research spanned 72 developing countries (on top of cross-country comparative work).
The full report is available online at http://econ.worldbank.org/research/highlights2007. The online version provides a complete list of DECRG’s publications by team in 2007, which included 25 books, 175 journal articles, 90 book chapters, more than 180 working papers (that will be published in due course), and 12 new datasets. The online version also gives details on the Group’s outreach efforts, which included 12 featured articles, 12 research briefs, 14 conferences organized by DECRG or coorganized by staff with other institutions, and well over 500 presentations by research staff at seminars and conferences worldwide.
I very much hope that you enjoy reading this edition of Research Highlights. Please tell us what you think about the issues raised in these pages and bring up topics you think need more research. In the end, it is the active interaction with development thinkers and practitioners that will continue to assure that DECRG’s research remains relevant to our shared goals of achieving inclusive and sustainable economic development.
Martin Ravallion, Director