capra_05

"jump" from a bad economic equilibrium into a better one, permanently increasing their output and income. Experiments can be used to study complex processes like the effect of institutions on economic growth. The control that experiments afford allows structural parameters to be changed, policies to be added and subtracted, and economic outcomes to be precisely measured. In this paper, we study a simple experimental economy in which agents produce output in each period, and can allocate the output between consumption and investment (the experiment builds on the design of Lei and Noussair, 2002, 2003). Capital productivity is higher if total investment is above a threshold. Because of the threshold externality, there are two equilibria—a suboptimal “poverty trap” and an optimal “rich country” equilibrium—which differ by a factor of approximately three in the agent income they create. In baseline sessions, in which agents make independent decisions in a decentralized economy, the economies typically sink into the poverty trap and the optimal equilibrium is never reached. However, the ability to communicate before investing, or to vote on binding “industrial policy” proposals, improves average earnings. Combining both of these simple institutions enables all of the economies to escape the poverty trap. This experimental environment constitutes a platform onto which many more complex features can be added.

Title:	The Impact of Simple Institutions in Experimental Economies with Poverty Traps
Number:	05-08
Author:	C. Mónica Capra, Tomomi Tanaka, Colin Camerer, Lauren Munyan, Veronica Sovero, Lisa Wang, and Charles Noussair
Issue Date:	February 2005
Abstract:	The existence of multiple equilibria is one explanation for why some countries are rich while others are poor. This explanation also allows the possibility that changes in political and economic institutions might help poor countries "jump" from a bad economic equilibrium into a better one, permanently increasing their output and income. Experiments can be used to study complex processes like the effect of institutions on economic growth. The control that experiments afford allows structural parameters to be changed, policies to be added and subtracted, and economic outcomes to be precisely measured. In this paper, we study a simple experimental economy in which agents produce output in each period, and can allocate the output between consumption and investment (the experiment builds on the design of Lei and Noussair, 2002, 2003). Capital productivity is higher if total investment is above a threshold. Because of the threshold externality, there are two equilibria—a suboptimal “poverty trap” and an optimal “rich country” equilibrium—which differ by a factor of approximately three in the agent income they create. In baseline sessions, in which agents make independent decisions in a decentralized economy, the economies typically sink into the poverty trap and the optimal equilibrium is never reached. However, the ability to communicate before investing, or to vote on binding “industrial policy” proposals, improves average earnings. Combining both of these simple institutions enables all of the economies to escape the poverty trap. This experimental environment constitutes a platform onto which many more complex features can be added.