Optimal sequencing of credible reforms with uncertain outcomes

We study a two-period representative agent economy in which economic liberalization is modeled as adding a positive random variable to the marginal product of capital. We show Liberalization always raises the expected utility of agents. Agents may respond to this by increasing consumption in the first period. Consequently, consumption in the second period is sometimes smaller than the in the first depending on the realization of the random variable. This 'tail' may cause the government to reject liberalization if declining GDP enters negatively in the government's objectives. We apply these results to the Chilean experience of the 1980s.