Cynthia Arnson, Deputy Director, Latin American Program, Woodrow Wilson Center
Michael Edward Hess, Assistant Administrator, Bureau for Democracy, Conflict and Humanitarian Assistance, USAID
Mitchell A. Seligson, Centennial Professor of Political Science, Vanderbilt University
Aníbal Pérez-Liñán, Assistant Professor of Political Science, University of Pittsburgh
Steven E. Finkel, Daniel H. Wallace Professor of Political Science, University of Pittsburgh
Dinorah Azpuru, Assistant Professor of Political Science, Wichita State University
Michael Bratton, Professor of Political Science, Michigan State University
Michael Coppedge, Professor of Political Science, University of Notre Dame
Pamela Paxton, Associate Professor of Sociology, Ohio State University
Margaret Sarles, Division Chief, Strategic Planning and Research for Democracy, USAID

On October 27, 2005, the Latin American Program hosted a presentation of the results of a USAID-funded study of the effects of U.S. foreign assistance on democracy building worldwide. Cynthia Arnson noted that the latest wave of U.S. funding for democracy promotion began during the Reagan administration and has been expanded and institutionalized since then. Michael E. Hess stated that democracy building is an important aspect of the Bush administration's national security strategy, reflected in the increase in funding for democracy assistance since 2001.

Research team leader Mitchell Seligson said that the study focused on 165 countries from around the globe. Using the democracy indicators of Freedom House and Polity IV, the study found that since the end of the Cold War, there has been a worldwide trend towards democratization. Seligson noted that democracy assistance is still a very small proportion of the USAID portfolio, but contrasted the steady erosion of AID assistance in areas not related to democracy promotion prior to 2001 with a consistent increase in assistance for democracy and governance programs. For purposes of measurement, the study divided assistance for democratic governance into the four categories used by AID: civil society (including the media), rule of law (including human rights), elections (including parties), and governance (including decentralization and anti-corruption programs). The study found that foreign assistance is divided fairly evenly across regions, although rule of law assistance was highest in Latin America.

Aníbal Pérez-Liñán described the methodology used in the study, noting that the researchers looked at data from the years 1990-2003 and grouped some 280 variables into two categories. The first included variables such as economic growth and inflation that change from country to country and within countries over time. The second category consisted of constant factors within countries over time, such as the level of democratic governance at the start of the study, the degree of ethnic fractionalization, and income inequality. Pérez-Liñán noted that measuring advances or setbacks in the rule of law and governance categories were the hardest to quantify.

Steven E. Finkel summarized conclusions of the study. The model showed that U.S. aid spending in the area of democratic governance had a consistent, albeit modest, impact on both the Freedom House and Polity IV indicators. He added that the regional effects of aid were strong. Looking at both the immediate (within one or two years) and the lagged (within three years) effects of aid, the team found that aid for elections and civil society had a more contemporaneous impact, while the impact of rule of law spending lagged. Civil society spending demonstrated a strong correspondence between outlay and effect, noting that aid obligations appeared to exacerbate human rights problems, possibly as latent abuses became more well-known. The governance category, meanwhile, showed no immediate or long-term change, perhaps due to the difficulties with measurement. Finkel said that the study suggested that to maximize the impact of democracy assistance, overall outlays should be increased and more attention be devoted to the possibilities for synergy across sub-sectors.

Peer reviewers commended the team for assembling the most comprehensive and rigorous study to date on the effects of aid on democratic governance. Michael Bratton remarked, however, that the team did not match the powerful statistical model they used with an equally powerful conceptual or theoretical framework that would constitute a dynamic theory of democratic development. He also questioned the heavy reliance on Freedom House and Polity IV indicators, noting that while these are the two most comprehensive sources available, they may not fully capture the level of democratic governance. Bratton pointed out that U.S.-funded democracy initiatives may have benefited from good timing, as they came at the end of the Cold War. Finally, he called for more aid to sub-Saharan Africa, given the report's finding that the greatest advances in democratic governance occurred in that sub-region.

Michael Coppedge voiced concern that wealthy western democracies were excluded from the analysis. Had they been included, he argued, a more negative trend might have appeared. He questioned whether fourteen years was long enough to establish a reliable trend. Coppedge also expressed concern that the report did not adequately control for the aid spending of other countries. He suggested that more theorizing was needed to discern the difference between overall levels of spending and spending per capita. He also questioned whether different results would have emerged if the researchers had not adopted the same categories used by AID.

Pamela Paxton suggested that future research examine in greater depth the regional trends that the study uncovered. She added that the cumulative effects found in the study indicate that qualitative information would be useful to enhance the understanding of the impact of aid. She underscored the usefulness of the study for the work of USAID. The study, she said, would lead USAID to address questions of the consistency of their funding and facilitate long-term planning.

Comments during the discussion period focused on the quality of the environments in particular countries, and how grouping together such a large number of countries might have masked the particularities of individual cases and skewed overall results. Finkel replied that further examination is needed to apply general findings to particular cases, thereby enabling practitioners to determine what conditions must be in place to make aid most effective.

Prepared by Cynthia J. Arnson, Deputy Director, Latin American Program
and Elizabeth Bryan, Program Assistant