Models of economic activity with frictions in coordinating trading have been shown to be capable of generating multiple steady states. (Peter Diamond, 1982, is the pioneering work; see my 1987a paper for a general discussion.) Less work has been done on out-of-steady-state dynamics in such models, which would enable us to examine what sort of fluctuations these models may generate. The absence of dynamics leaves open the question of which steady state the economy will reach, as well as whether the comovements of key variables resemble what is observed over the cycle. In my earlier paper (1987b), I presented a model combining search and aggregate demand approaches to unemployment to show how spillovers between product and labor markets could yield multiple equilibria. Here, a highly simplified dynamic model based on this work is presented, in which (self-fulfilling) sales expectations determine to which steady state the economy converges. These expectations are summarized by the asset values of firms which are producing output relative to those that are not. Two types of dynamic paths leading to stationary solutions can arise. The first is a saddle path. In addition, for certain parameter values, stable limit cycles emerge. Interestingly, over this cycle, asset values of firms (which one could interpret as stock market values) lead economic activity. I. Model Setup

Adi Brender Allan DrazenDecember 2008When democracy is new it is often fragile or unconsolidated, meaning that importantpolitical groups lack full commitment to the democratic process, so that its survival is notassured. What economic policies can a government use to try to prevent a reversion toautocracy?One answer to this question begins with the argument that the threat to democracycomes from anti-democratic elites Œthe army, groups such as the wealthy who bene–ttedmost under the old regime, the fioligarchsflŒwho are seen as basically anti-democratic andwho have the power to overthrow the new democratic regime. The fimassesflare seen asunambiguously pro-democratic. Under this view, policy to consolidate democracy shouldfocus on placating the anti-democratic elites, or, colloquially but not inaccurately, to fibuythem o⁄.flThis approach has received a superb treatment in Daron Acemoglu and JamesA. Robinson (2005).Targeted economic policy to consolidate democracy can be viewed more generally. Sur-vival of democracy (and policy to enhance that) can be thought of in terms of two fun-damental questions. The –rst is: How does the support of di⁄erent actors a⁄ect theprobability that democracy survives? The second is then: How do these important actorschoose whether or not to support the continuation of democracy or its overthrow? Thedesign of policy to address democratic fragility depends on the answers to these two basicquestions. The policy prescription in the previous paragraph, for example, follows from theanswers: it is the elites alone who are determinate; and, their support depends on givingthem enough that they are content not to overthrow democracy.Our research on the fragility of new democracies focusses on di⁄erent, but complemen-tary issues. While we agree that anti-democratic elites pose a serious threat to fragile1

We test whether good economic conditions and expansionary fiscal policy help incumbents get reelected in a large panel of democracies. We find no evidence that deficits help reelection in any group of countries independent of income level, level of democracy, or government or electoral system. In developed countries and old democracies, deficits in election years or over the term of office reduce reelection probabilities. Higher growth rates over the term raise reelection probabilities only in developing countries and new democracies. Low inflation is rewarded by voters only in developed countries. These effects are both statistically significant and quite substantial quantitatively.