Columbia International Affairs Online: Working Papers

CIAO DATE: 10/2009

The Limits of Institutional Engineering: Lessons from Iraq

David Waldner

May 2009

United States Institute of Peace

Abstract

* Post-conflict, post-totalitarian societies like Iraq possess many economic, political, social, and cultural characteristics that are not conducive to democratic governance. A central pillar of democracy promotion is that judicious institutional engineering--crafting new institutions and other elements outlining the democratic rules of the game--can overcome these obstacles and engender stable democracies.
* Two theories explain the success or failure of institutional engineering: institutions as causes, or institutions as indicators. The first theory states that constitutions, electoral systems, and other components of democratic governance are the direct causes of democratic stability or collapse.
* The second theory states that institutions are indicators, not causes. Like barometers, they allow us to forecast impending storms or calm weather, but they do not cause these phenomena.
* Many methodological weaknesses are evident in the studies that supposedly support the "institutions as causes" theory. One type results from nonrandom selection of units. A second type results from nonrandom assignment to treatment. Left uncorrected, these weaknesses generate inaccurate and uncertain estimates of causal effects and may invalidate many studies of institutional design.
* Different types of methodological corrections for these biases exist, some statistical, others qualitative. All of them demand that we first model the selection process before estimating the effects of institutions; we must identify the origins of institutions to evaluate their effects correctly.
* By using a method called process tracing to scrutinize institutional engineering in Iraq, it becomes clear why intensified violence followed the drafting and ratification of the Iraqi constitution. It is not surprising that institutional engineering did not forestall violence; therefore, we can conclude that the Iraqi experience does not support theories of institutional design.
* Although the "institutions as causes" approach claims that "getting the institutions right" can create a stable democracy in otherwise unfavorable conditions, it is imperative to view the studies that support it with skepticism. Nonetheless, these weaknesses do not completely dismiss the theory of institutional engineering. One study of one country can never produce a definitive conclusion. Instead, policymakers should pay closer attention to the methodological underpinnings of all findings based on case studies and other forms of observational research.