72 International Organization loss of policy discretion potentially harms their ability to maintain their position in office.A floating exchange rate,on the other hand,gives politicians the flexibility to use external adjustment not only to counter local shocks but also to employ macro- economic policy for electoral or partisan advantage.This political dilemma raises an interesting question:Under what conditions will politicians commit to a fixed exchange-rate regime? We argue that politicians'incentives over the exchange-rate regime reflect the configuration of domestic political institutions,particularly electoral and legislative institutions.In systems where the cost of electoral defeat is high and electoral timing is exogenous,politicians will be less willing to forgo their discretion over monetary policy with a fixed exchange rate.In systems where the costs of electoral defeat are low and electoral timing is endogenous,politicians are more likely to adopt a fixed exchange-rate regime.Consequently,differences in domestic political systems can help account for variations in the choice of exchange-rate arrangements. In the first section we review the conventional literature about exchange-rate ar- rangements.In the second section we develop our argument concerning the relation- ship between domestic political institutions and exchange-rate regime choice.In the third section we draw on the optimal exchange-rate and international political economy literatures to identify control variables,including systemic influences,domestic eco- nomic conditions,and other political factors.In the fourth section we evaluate the importance of domestic political institutions on a sample of twenty countries using a constrained multinomial logit model.We present our conclusions in the final section. Choosing an Exchange-rate Arrangement Two broad literatures address the choice of exchange-rate arrangement.First,the optimal exchange-rate literature considers the type of exchange-rate commitment that is "best"given the characteristics of a nation's economy.2 This literature focuses on country characteristics such as economic openness,country size,and labor mobil- ity.More recent contributions argue that the optimal exchange arrangement depends not only on the structure of the economy but also on the sensitivity of the economy to domestic and international macroeconomic shocks.3 One major problem with this literature is that it does not specify the origin of politicians'policy preferences.In fact the conclusions and policy prescriptions reached 2.For example,Bosco 1987:Dreyer 1978:Heller 1978:Holden.Holden,and Suss 1979;Savvides 1990;and Wickham 1985.A related literature concerns optimal currency areas.This literature considers whether regions should participate in a currency union based on factors such as common vulnerability to shocks.It is optimal for countries experiencing similar shocks to join a currency union,whereas the existence of dissimilar shocks makes a floating exchange arrangement the more prudent choice.The seminal contributions of Mundell and McKinnon focus,respectively,on the importance of external bal- ance and price stability.See Mundell 1961;and McKinnon 1963.More recent variants of this literature examine the source of the shocks (for example,Tavlas 1993;Frankel and Rose 1996;and Eichengreen 1992a)and the question of whether the EC constitutes an optimal currency area. 3.See Fischer 1977:and Savvides 1990.72 International Organization loss of policy discretion potentially harms their ability to maintain their position in office. A floating exchange rate, on the other hand, gives politicians the flexibility to use external adjustment not only to counter local shocks but also to employ macro￾economic policy for electoral or partisan advantage. This political dilemma raises an interesting question: Under what conditions will politicians commit to a fixed exchange-rate regime? We argue that politicians' incentives over the exchange-rate regime reflect the configuration of domestic political institutions, particularly electoral and legislative institutions. In systems where the cost of electoral defeat is high and electoral timing is exogenous, politicians will be less willing to forgo their discretion over monetary policy with a fixed exchange rate. In systems where the costs of electoral defeat are low and electoral timing is endogenous, politicians are more likely to adopt a fixed exchange-rate regime. Consequently, differences in domestic political systems can help account for variations in the choice of exchange-rate arrangements. In the first section we review the conventional literature about exchange-rate ar￾rangements. In the second section we develop our argument concerning the relation￾ship between domestic political institutions and exchange-rate regime choice. In the third section we draw on the optimal exchange-rate and international political economy literatures to identify control variables, including systemic influences, domestic eco￾nomic conditions, and other political factors. In the fourth section we evaluate the importance of domestic political institutions on a sample of twenty countries using a constrained multinomial logit model. We present our conclusions in the final section. Choosing an Exchange-rate Arrangement Two broad literatures address the choice of exchange-rate arrangement. First, the optimal exchange-rate literature considers the type of exchange-rate commitment that is "best" given the characteristics of a nation's e~onomy.~ This literature focuses on country characteristics such as economic openness, country size, and labor mobil￾ity. More recent contributions argue that the optimal exchange arrangement depends not only on the structure of the economy but also on the sensitivity of the economy to domestic and international macroeconomic shock^.^ One major problem with this literature is that it does not specify the origin of politicians' policy preferences. In fact the conclusions and policy prescriptions reached 2. For example, Bosco 1987; Dreyer 1978; Heller 1978; Holden, Holden, and Suss 1979; Savvides 1990; and Wickham 1985. A related literature concerns optimal currency areas. This literature considers whether regions should participate in a currency union based on factors such as common vulnerability to shocks. It is optimal for countries experiencing similar shocks to join a currency union, whereas the existence of dissimilar shocks makes a floating exchange arrangement the more prudent choice. The seminal contributions of Mundell and McKinnon focus, respectively, on the importance of external bal￾ance and price stability. See Mundell 1961; and McKinnon 1963. More recent variants of this literature examine the source of the shocks (for example, Tavlas 1993; Frankel and Rose 1996; and Eichengreen 1992a) and the question of whether the EC constitutes an optimal currency area. 3. See Fischer 1977: and Savvides 1990