Throughout much of the postwar period, the level of central bank independence in most industrial democracies remained relatively stable. In the late 1980s and early 1990s, however, politicians in many industrial democracies reformed their central bank institutions, granting their central banks increased independence from direct political control.
I argue that this wave of central bank reform reflects long-term developments in the party systems of the industrial democracies. Changes in the policy preferences of constituents of the main governing parties and decreases in the ability of cabinet ministers to deliver promised economic outcomes increased the potential for intraparty conflicts over economic and monetary policy and in turn hurt the ability of the main governing parties to attain and retain office. Therefore, many political parties had to alter their electoral strategies and policy priorities in order to balance conflicting interests, rebuild social coalitions, and maintain electoral viability (Boix 1998; Garrett 1998a; Kitschelt 1994). Part of this strategic repositioning includes institutional reforms—reforms that are designed to reconcile diversified constituent demands while preserving the electoral success of the party. Central bank independence, I contend, represents one of those reforms.
In this chapter, I elaborate on this argument. The first section examines changes to the party systems of the industrial democracies since the 1970s, illustrating the increased potential for intraparty conflict over economic and monetary policy. This heightened potential for intraparty conflict, I argue, magnified the political benefits of adopting an independent central bank. The second section provides empirical support for this argument by examining the effect of economic internationalization and central bank independence on cabinet durability in 16 parliamentary systems between 1970 and 1992. The results show that an independent central bank protects cabinet durability, particularly under conditions of economic openness—precisely where one would expect more conflicts over monetary policy. In contrast, dependent central banks contribute to higher cabinet durability in relatively closed economies—where the potential for conflict between backbench legislators, coalition partners, and cabinet ministers is lower.
I contend that the political consequences of an independent central bank—fewer conflicts over monetary policy—can help explain variations in the timing and nature of central bank reform across the industrial democracies. Politicians in systems where economic internationalization has shortened cabinet durability have incentives to adopt an independent central bank as a way to prolong their tenure in office. Politicians in systems where cabinet durability is relatively unaffected by changes in economic openness have fewer incentives to adopt an independent central bank. Linking economic internationalization, cabinet durability, and central bank independence therefore can help explain variations in the timing and nature of central bank reform. In the third section, I evaluate these predictions.
The increased potential for intraparty conflict over economic and monetary policy in the 1990s is a product of developments in the party systems of the industrial democracies. Through the 1950s and 1960s, these party systems enjoyed tremendous stability (Lipset and Rokkan 1967). Individual voters, ensconced within distinct subcultures, tended to identify closely with a particular party, helping to ensure that the main parties had stable and consistent levels of electoral support. Additionally, party systems often mirrored traditional class cleavages; that is, parties appealed either to the working class or to the bourgeoisie (with some exceptions such as Christian Democratic parties). Each party had different economic priorities, but within each party, constituents possessed relatively similar preferences over economic and monetary policy.
Since the early 1970s, however, these party systems have been in flux. Voters identify less and less with their former party affiliations. These dealigned voters cast ballots on the basis of instrumental calculations and policy programs rather than cultural attachments (Barnes 1997; Dalton, Flanagan, and Beck 1984; Lewis-Beck 1988; Lewis-Beck, Norpoth, and Lafay 1991). Consequently, vote levels for many of the traditional governing parties became less predictable from election to election (Maguire 1983; Pedersen 1983). In some systems, new parties appeared to challenge the main governing parties on both the Left (Kitschelt 1989; Richardson and Rootes 1995; Shull 1999) and the Right (Betz 1994; Kitschelt 1997). As a result, the main governing parties have pursued new electoral strategies to cope with the changing electoral environment (Kitschelt 1994; Muller and Strom 1999; Przeworski and Sprague 1986). Some parties moderated their policy positions; others became more extreme. These strategic decisions, however, often entailed internal party debate and controversy, as parties fought about how best to reshape their electoral coalitions in this new political environment.
This party-system instability reflects two developments, both related to longer-term transformations in the economies of the industrial democracies: (1) changing constituent demands over economic and monetary policy and (2) the decreasing ability of governments to deliver promised policy outcomes. Both developments increased the potential for intraparty conflict over economic policy within the main governing parties and made it more difficult for them to attain and retain office.
Changes in constituent demands for economic and monetary policy are in part a consequence of three economic developments: greater economic wealth, increased economic openness, and the rise of the service sector. Each of these processes diversifies the policy incentives of politicians within the traditional governing parties.
First, the postwar period brought the industrial democracies unprecedented levels of economic development and stability. Citizens are wealthier, more educated, and have greater access to information. In many countries, a strong middle class has emerged—a middle class that does not always respond to traditional class-based political strategies (see, for example, Przeworski and Sprague 1986; Ramseyer and Rosenbluth 1993; Tarrow 1990). Economic prosperity has also allowed new issues to become prominent, including the “post-materialist” agenda, which focuses on the environment, women’s rights, and community participation (Esping-Andersen 1999; Inglehart 1997; Inglehart and Abramson 1994).
Second, increased economic openness changes constituent interests over economic and monetary policy (Alt and Gilligan 1994; Frieden 1991; Frieden and Rogowski 1996; Keohane and Milner 1996; Rogowski 1989). Economic internationalization alters the basis of economic interests from factors—primarily the worker—owner-class cleavage—to sectors, groups defined by asset-specific investments in a similar area of output. Actors within a sector have common interests that often cut across classes. For instance, both workers and owners involved in the export market may have similar policy preferences. In turn, their policy demands will differ from those of sectors that are insulated from international competition. These conflicts may center on the nature of wage bargaining (Franzese 2000; Golden, Lange, and Wallerstein 1999; Iversen 1996) or newly salient policy issues, such as exchange-rate policy (Frieden 1991).
Economic internationalization also affects the relative influence of economic sectors over economic and monetary policy (Frieden 1991). Owners of internationally mobile assets—in particular, capital—can threaten to withdraw those assets if they are dissatisfied with the government’s policies. By deciding whether to invest their assets domestically or move them abroad, owners of mobile assets influence the success of macroeconomic policy and in turn, politicians’ electoral fortunes. Therefore, politicians must respond to their demands if they wish to attract and retain these economic assets for their country, even if these sectors do not form part of the government’s electoral coalition (Frieden 1991; Keohane and Milner 1996; Maxfield 1997). This “exit” threat increases their policy influence relative to owners of immobile factors—often labor. Consequently, the increased political power of owners of mobile assets expands the number of constituent pressures placed on politicians within the governing parties. It contributes to the possibility of conflict between legislators who appeal to asset-specific interests (e.g., workers in the home district) and cabinet ministers who have to consider the overall performance of the economy.
Third, the provision of services, rather than manufacturing, accounts for an increasingly large portion of economic production (Iversen and Cusak 2000; Krugman 1994). This trend is a result of the global overcapacity in the production of manufactured goods and in some countries, government policy. Governments in some countries have compensated for the loss of job demand in other economic sectors by expanding the service-oriented public sector (Franzese 2000; Iversen 1999; Iversen and Wren 1998). The changing sectoral composition of the economy creates a number of potential conflicts for political parties. Politicians must face difficult decisions about how to manage decline in key manufacturing sectors. Further, the wage and policy demands of the service sector, which is largely insulated from international competition, may reduce the competitiveness of exposed manufacturing interests.
These economic developments have altered the socioeconomic coalitions underlying political parties in the industrial democracies over the past 20 years (see, for example, Pempel 1998). The economic policy demands made by voters have diversified. New issues have become salient. New cleavages have emerged, reflecting divisions between exposed and sheltered sectors, manufacturing and service sectors, rising and declining industries, the private and the public sector, and between owners of mobile and specific assets. Because a party’s policy reputation and voters’ partisan identification evolve relatively slowly, these altered economic interests do not always line up with existing political parties. Consequently, the diversification of constituent preferences has increased the potential for intraparty conflict over economic policy. Parties must determine which constituents to satisfy and which to ignore. They must decide how to balance the policy demands of traditional constituents with appeals to new interests. These decisions entail distributional conflicts, shaping both the fortunes of affected constituents and the careers of party politicians.
Since the 1970s, the ability of governments to deliver promised economic outcomes has also decreased substantially. As government budget deficits ballooned in the industrial democracies, fiscal policy became less useful as a policy instrument. At the same time, the link between monetary policy and economic outcomes became less predictable. Increased economic openness during this period intensified the problems of economic management. Consequently, cabinet ministers had less ability to satisfy party legislators and their constituents, increasing the probability that backbench legislators and coalition partners would withdraw their support from the cabinet.
The emergence of high public debt in many industrial democracies limited the flexibility of fiscal policy. During the 1970s and early 1980s, many governments ran high budget deficits, increasing their public debt burden substantially (de Haan and Sturm 1994, 1997; Franzese 2000; Hallerberg and von Hagen 1999; Roubini and Sachs 1989a, 1989b; von Hagen 1992). In Italy, Ireland, and Belgium, the ratio of public debt to GDP actually exceeded 100 percent. Consequently, many governments had to devote an increasing portion of their budgets to service interest payments, limiting the possibility of expanding successful programs or launching new spending initiatives. Higher interest rates during the 1980s exacerbated this problem. Political parties were therefore severely limited in their ability to use fiscal policy to manage and expand their social coalitions. Furthermore, governments could not easily use fiscal policy to smooth short-term fluctuations in macroeconomic performance. As a result, government ministers faced a temptation to rely on monetary policy to manage the economy.
At the same time, however, established monetary-policy choices no longer produced predictable outcomes. Prior to the 1970s, monetary policy reflected the idea of the Phillips Curve, which predicted a relatively stable trade-off between inflation and unemployment. Policymakers could simply expand the economy to a point at which the cost of higher inflation balanced the benefit of lower unemployment. In the 1970s, however, both theoretical critiques and empirical evidence undermined confidence in the Phillips Curve. As a consequence, it became less clear how monetary policy affected overall economic performance—and what policies policymakers should pursue (Clarida, Gali, and Gertler 1999; Sargent 1999).
From a theoretical perspective, Friedman (1968) noted that the existence of a long-run Phillips Curve requires that economic agents not anticipate the level of inflation when negotiating their wage contracts, an unrealistic assumption about individual behavior. If economic agents build inflation expectations into their wage demands, then a long-run trade-off between inflation and unemployment does not exist. The best outcome a policymaker could achieve would be to keep employment near the level it would be if firms and workers had expectations of nonaccelerating inflation.
In the short run, on the other hand, a trade-off between unemployment and inflation is possible. Monetary expansions can affect real economic variables—growth, employment—if economic agents are surprised by policy changes (Lucas 1972). The economic boom, however, will last only until economic agents have adjusted their inflation expectations. It is a point of debate in the literature how long this adjustment takes—and as a result, how much the economy benefits (Granato 2000; Hall and Franzese 1998; Iversen 1998; Krugman 1994).
Even as economists attacked the Phillips Curve theoretically, events were undermining it. During the 1970s, the industrial economies suffered from stagflation—a combination of high inflation and high unemployment—an outcome unexplained by the Phillips Curve. Ironically, therefore, as growing public debt forced policymakers to rely on monetary policy for macroeconomic management, the consequences of monetary policy became less predictable—and the policy guidelines for policymakers more opaque.
During this period, the industrial economies also became more exposed to the international economy. International trade continued to grow (Milner and Keohane 1996). Financial-market deregulation sharply increased capital mobility as well (Goodman and Pauly 1993; Kapstein 1994; Quinn and Inclan 1997; Simmons 1999). These increased levels of economic openness exacerbated the problems of policy management. Openness exaggerates the economic and social consequences of policy actions (Franzese 2000; Frieden 1991). Advances in technology shorten the time lag between policy choices and policy outcomes. Policymakers must therefore respond more quickly and more accurately to exogenous shocks in order to meet the demands of their constituents.
Additionally, the volume and pace of international markets can overwhelm many national policy tools (Andrews 1994). By the early 1990s, for example, transactions in international currency markets totaled well over $1 trillion each day, dwarfing the reserves of any individual country and limiting the ability of countries to affect the exchange rate through market intervention. As a result, national macroeconomic policies are often less effective in delivering policy outcomes (Goodman 1992; Goodman and Pauly 1993).
Fewer barriers to international economic activity also reduce national policy autonomy, making it more difficult for political leaders to manage the macroeconomy. Both monetary and fiscal policy can be affected by increased openness (Clark 1999; Clark and Hallerberg 2000; Oatley 1999; Simmons 1994). According to the Mundell-Fleming conditions, maintaining a fixed exchange rate in a world of capital mobility will limit the ability to use monetary policy for domestic-policy objectives. With floating exchange rates and capital mobility, however, fiscal policy may become less effective (Frieden 1991; Milner and Keohane 1996).
Decreased policy effectiveness hurt the ability of cabinet ministers to achieve promised economic outcomes—with predictable political consequences. Poor economic performance led to sharp electoral losses for incumbents. Parties that were in office during the stagflation of the 1970s were exceptionally hard hit, losing their reputation for capable economic management.
The combination of diversified constituent demands and decreased policy effectiveness increased the possibility of intraparty conflict over economic and monetary policy during the 1980s and 1990s—conflicts that threatened the ability of the main governing parties to win and retain office. Indeed, intraparty conflicts occurred throughout the industrial democracies and across the ideological spectrum.
In Left parties, traditional working-class constituents battled not only export interests but also middle-class postmaterialists and public-sector workers for control of party programs (Kitschelt 1994, 1999). In Germany, the Social Democratic–led government of the 1970s and early 1980s was divided between moderates and trade unionists, contributing to the collapse of the government in the early 1980s. In Britain, the Labour Party splintered as it adjusted its policies in the wake of Thatcher’s victory in 1979. The Labour Party originally pulled leftward, prompting some moderates to leave the party in 1981 to found the more centrist Social Democratic Party. After another electoral defeat in 1983, Labour Party leaders struggled to pull the party back toward the center—a long and tumultuous process that culminated with the election of Tony Blair in 1997. In France during the early 1980s, François Mitterrand’s decision to maintain the commitment to the EMS involved sweeping changes to the Socialists’ constituent base. By pursuing economic rigueur, Mitterrand rejected a focus on blue-collar constituents and reached out to middle-class professionals in the nontradables sector, creating conflict and dissension within the party (Frieden 1994; Loriaux 1991). In the United States, Democratic Party centrists and trade unionists battled over trade and environmental issues (Shoch 2000).
Right parties faced similar conflicts. The Italian Christian Democrats suffered internal battles over the best response to union militancy and the economic shocks of the 1970s, pursuing stop-go policies that alternately sought to appease the unions and then tighten macroeconomic policy. In Britain, policy conflicts between Tory Wets and radical free-market Thatcherites divided the Conservative Party during the 1980s and 1990s (Hall 1986). The Republican Party in the United States exhibited similar divisions.
Although these conflicts often centered on economic and monetary-policy issues, they represented more fundamental concerns about the long-term trajectory and viability of these parties. Changes in both the policy preferences of constituents and the government’s policy efficacy meant that parties were operating in a new political environment—an environment that not only challenged parties to maintain their traditional constituents but also provided them with opportunities to expand their electoral coalitions.
One strategy to enhance the party’s fortunes in this changed political environment involved the support of institutional reform. Institutional reform can provide outlets for the representation of new interests, demonstrate policy commitment to affected constituents, or enhance policy effectiveness. Institutional reforms can therefore potentially help parties balance the interests of party politicians with divergent policy incentives and reestablish an electoral coalition. Where the increased potential for intraparty conflict threatened the ability of parties to win and maintain office, party politicians had incentives to pursue institutional reform.
I argue that central bank reform reflects such calculations. A more independent central bank can help parties overcome internal party divisions over policy. First, it acts as a check on the cabinet’s policy discretion. The policy information furnished by a more independent central bank provides credible assurances about the cabinet’s policy behavior to backbench legislators, coalition partners, and constituents. A more independent central bank also enhances monetary-policy effectiveness. It signals to economic agents that monetary policy will be insulated from excessive partisan and electoral manipulation (Cukierman 1992; Hall and Franzese 1998; Maxfield 1997). Consequently, economic agents will adjust their behavior more quickly, improving policy efficacy. Making the central bank more independent does, of course, entail some political costs: the cabinet loses the ability to manipulate policy for short-term electoral or partisan gain. But where intraparty conflicts threaten the party’s ability to win and stay in office, the political benefits of an independent central bank outweigh these costs.
In particular, I argue that economic internationalization has increased the political value of an independent central bank. Internationalization diversifies constituent demands and decreases policy effectiveness, increasing the possibility of intraparty conflict. In systems where these conflicts threaten the parties’ ability to attain and retain office, politicians have more incentive to adopt an independent central bank. In contrast, where the configuration of domestic institutions prevented intraparty conflict from hurting the parties’ ability to remain in office, politicians had fewer incentives to adopt an independent central bank. In these systems, central bank reform will be delayed or put off entirely.
In this section, I examine cabinet durability in parliamentary systems to show that the political value of an independent central bank is higher under conditions of increased economic openness.1 Economic internationalization—increased capital mobility and exposure to international trade—has increased the possibility of policy conflict between backbench legislators, coalition partners, and cabinet ministers, making cabinets less durable. An independent central bank, however, can prevent some of these potential conflicts and as a consequence help politicians prolong their position in office. If independent central banks do prevent conflict over monetary policy, I expect cabinets in parliamentary systems with independent central banks to survive longer than cabinets in similar systems with dependent central banks.
The political-science literature argues that cabinet duration is a function of the cabinet’s attributes, such as majority status and number of parties in the coalition, and the bargaining environment, such as the fragmentation of the party system and party-system polarization (Alt and King 1994; King, Alt, Burns, and Laver 1990; Laver and Schofield 1990; Warwick 1994). Two other factors are also likely to influence expected cabinet duration: the diversity of policy incentives faced by politicians in the governing party(ies) and the ability of the government to determine policy outcomes. The divergence of policy incentives promotes conflict between and within the governing parties. If politicians within each party face different incentives over economic policy, it is more difficult not only to form a government but also to keep it together. Legislators will be more likely to withdraw their support from the cabinet if their policy demands are not met. Consequently, the more diverse the interests in the governing party(ies), the less stable the government.
The government’s ability to influence economic performance also influences cabinet duration. Government ministers need to deliver policy outcomes in order to maintain the support of backbench legislators and coalition partners. If government ministers do not have the capability to produce promised outcomes—if, for instance, the link between policy and outcomes is unclear or vulnerable to external shocks—dissatisfied backbenchers and coalitions partners may bring down the government. Therefore, the more the government can shape economic outcomes, the longer the expected cabinet duration.
I argue that the divergence of policy incentives within the governing party(ies) and the government’s policy effectiveness is a function of economic internationalization and the configuration of domestic political institutions.
As previously discussed, economic internationalization diversifies constituent demands over economic and monetary policy and reduces the ability of cabinet ministers to deliver promised economic outcomes. Both of these developments increase the likelihood that backbench legislators and coalition partners may withdraw their support from the government, leading to shorter cabinet duration. Therefore, systems with higher levels of exposure to the international economy will have shorter cabinet duration.
I measure economic internationalization along two dimensions (Milner and Keohane 1996). The first dimension reflects barriers to the movement of economic assets across borders. Increased economic internationalization implies fewer restrictions on the movement of capital and traded goods across borders, including no capital controls, full convertability of currencies, no restrictions on the current account, and no tariff or nontariffbarriers to trade. As a proxy for this variable, I use a measure of capital account openness compiled by Quinn (1997). This variable ranges from 1.5 (high restrictions) to 4 (no restrictions).2
The second dimension reflects the degree of exposure to the international economy. I measure this by the sum of imports and exports as a proportion of gross domestic product. This trade openness variable ranges from 18 percent to 148 percent of GDP. I expect both capital account openness and trade openness to have a negative effect on cabinet durability.
Domestic institutions mediate the effects of economic internationalization on cabinet durability (Garrett and Lange 1996). Three sets of institutions influence this relationship: coalition and party-system attributes, the level of unionization, and the institutional status of the central bank.
Coalition and Party-System Attributes
According to the political-science literature, cabinet durability is a function of coalition attributes and party-system attributes (Alt and King 1994; King, Alt, Burns, and Laver 1990; Laver and Schofield 1990; Warwick 1994). Coalition attributes reflect the majority status of the government and the number of parties in the government. According to the literature, single-party majority governments tend to be most durable, minimum-winning coalitions slightly less durable, and oversize and minority governments least durable.
I include dummy variables for government type in the analysis: single-party majority, minimum-winning coalition, oversize coalition, single-party minority, and coalition minority.
Party-system attributes include fragmentation of the political system and political polarization. Political scientists argue that the more fragmented and polarized the political system, the shorter the expected cabinet duration. I include a variable for party-system fractionalization, which measures the number of effective political parties in the system (Rae 1971). This variable should have a negative effect on cabinet durability. Polarization is measured by the electoral support for extremist parties. More support for extremist parties also implies shortened duration.
The level of unionization influences the relationship between economic internationalization and cabinet durability. At both high and low levels of unionization, cabinet durability will be high. Moderate levels of unionization, however, may hurt cabinet durability.
First, consider the effect of unionization on the diversity of policy incentives within the governing coalition. At one end, strong centralized unions decrease the diversity of policy demands faced by politicians in the governing parties. Large encompassing unions can coordinate the particularistic demands of workers or firm-level unions, even in situations where workers have different market interests (Garrett 1998a; Lange and Garrett 1985; Olson 1982). As a result, governing politicians will face more coherent policy demands in systems with-strong unions. The coherence of these policy demands will reduce the potential for policy conflict within the governing party(ies) and therefore contribute to higher cabinet durability.
At the other end of the spectrum, where unions are weak, politicians face working-class constituents with a wider variety of policy interests, but these workers lack the organizational power to press their claims. Politicians in the governing parties will not face strong pressure to respond to union demands. Union weakness will therefore reduce the potential for policy conflict and help increase cabinet durability.
In systems where unionization is moderate, however, workers will likely have both a greater variety of policy interests and significant organizational power. The diversity of these demands—and the ability of the unions to press their claims—provides politicians within the governing parties with conflicting incentives over economic policy, which in turn shortens cabinet durability.
The level of unionization will also have a curvilinear effect on the government’s policy effectiveness. High levels of unionization with centralized wage bargaining may enhance the government’s policy effectiveness. Strong unions can better coordinate a response to changes in government macroeconomic policy (Franzese 2000; Garrett 1998a; Hall and Franzese 1998; Iversen 1998, 1999; Lange and Garrett 1985; Scharpf 1987). This coordination deters militant wage behavior, which in turn can improve real economic performance. Very low levels of unionization, on the other hand, may also improve the government’s policy effectiveness by facilitating market adjustments to government policy. Weak unions do not have the organizational strength to prevent wages from adjusting quickly to market-clearing levels.
At moderate levels of unionization, however, unions cannot coordinate wage demands or facilitate wage adjustments to policy changes. As a result, the government may have to pursue overly tight monetary policy to signal its commitment to low inflation and induce wage moderation among workers (Cukierman 1992). This will hurt the government’s policy effectiveness—and its ability to hold a governing coalition together.
I include a variable measuring the level of unionization in the economy as the proportion of unionized workers in the work force.3 I also include the squared value of unionization.
The effect of unionization on cabinet durability will also depend on the level of trade openness. Unions in open economies may be less able to coordinate the policy demands and policy response of their members (Iversen 1996). Consequently, I include interaction terms between trade openness and the unionization variables.
Central Bank Independence
Independent central banks will help increase cabinet durability in the face of economic internationalization. First, an independent central bank helps alleviate potential conflicts over monetary policy between politicians by helping to check the cabinet’s discretion over monetary policy. Second, an independent central bank also strengthens the government’s policy effectiveness by providing credibility to the government’s policy choices (Cukierman 1992). An independent central bank therefore should increase cabinet durability.
I include a measure of central bank independence. Because systems with independent central banks should have longer cabinet duration than systems with dependent central banks, the coefficient should be positive.
The effect of central bank independence on cabinet durability will also vary according to the openness of the economy and the organization of economic interests. First, in systems with full capital mobility, an independent central bank will provide credibility with international markets. I include an interaction term by multiplying capital account openness by central bank independence. At higher levels of capital openness, I expect central bank independence to have a stronger positive effect on cabinet durability.
Second, independent central banks will also affect cabinet durability based on the organization of interests within the economy. In particular, independent central banks can better deter militant wage behavior where unions are organized (Franzese 2000; Hall and Franzese 1998; Iversen 1998). To capture this effect, I multiplied central bank independence with the unionization*trade openness interaction variables.
The partisanship of the government may affect the relationship between economic internationalization and cabinet durability. However, the literature does not offer a clear prediction. Recent work suggests that economic internationalization will likely shorten the expected duration of Left governments more than Right governments. The socioeconomic consequences of economic internationalization will more likely cause conflict within the Left’s traditional constituent base (Garrett 1995; Golden, Langer, and Wallerstein 1999; Milner and Keohane 1996). Economic internationalization creates divisions between workers in sectors exposed to international competition and those insulated from the international economy (Iversen 1996). Additionally, internationalization enhances the policy influence of financial interests, a constituent not in the traditional Left social coalition. These developments will likely create tension and policy conflict in Left parties and lead to less-stable governments.
On the other hand, Left governments have enjoyed long tenures in the many small, open economies of western Europe (Katzenstein 1985). These parties have used corporatist bargaining and extensive welfare states to insulate their constituents from international economic volatility and create stable political systems (Cameron 1978; Lange and Garrett 1985).
I include a measure of Left government strength based on Cameron (1984). I do not expect that partisanship will influence cabinet durability by itself. However, I do expect the consequences of economic internationalization to be more apparent in Left governments. Consequently, I multiplied this variable with three other variables: capital restrictions, trade openness, and central bank independence.
Table 7 contains the summary statistics for all the variables employed in the empirical analysis.
Sample
The sample includes 182 cabinets from 16 countries: Australia, Austria, Belgium, Britain, Canada, Denmark, Finland, France, Germany, Ireland, Italy, Japan, the Netherlands, New Zealand, Norway, and Sweden.4 The sample extends from approximately 1970 to 1992. For each country, the sample begins with the first cabinet installed after 1970 and ends with the last cabinet that dissolved in the early 1990s. The sample includes no censored data.
TABLE 7. Descriptive Statistics
|
|
Standard |
|
|
Cabinet durability |
0.55 |
0.32 |
0.041 |
1 |
Time to election |
39.46 |
13.70 |
3 |
59 |
Single-party majority |
0.23 |
0.42 |
0 |
1 |
Single-party minority |
0.21 |
0.41 |
0 |
1 |
Oversize coalition |
0.21 |
0.43 |
0 |
1 |
Minimum-winning coalition |
0.21 |
0.41 |
0 |
1 |
Fractionalization |
0.69 |
0.10 |
0.41 |
0.87 |
Polarization |
0.13 |
0.12 |
0 |
0.42 |
Trade openness |
63.64 |
29.41 |
18.01 |
148.68 |
Capital account openness |
2.84 |
0.64 |
1.5 |
4 |
Central bank independence |
0.33 |
0.14 |
0.17 |
0.69 |
Unionization |
0.48 |
0.17 |
0.10 |
0.84 |
Partisanship (Left) |
0.34 |
0.42 |
0 |
1.31 |
Inflation |
8.09 |
4.79 |
0.47 |
23.58 |
Unemployment |
5.51 |
3.28 |
0.17 |
16.7 |
N = 182
The number of cabinets varies substantially across the countries in the sample. Canada and Germany had the fewest governments (7 each), whereas Italy and Belgium had the most governments (20 and 18, respectively).
Dependent Variable
For the dependent variable, I could simply use the number of months that a cabinet is in office. Constructing the dependent variable in this manner, however, leads to complications that could bias the results. First, countries have different constitutionally mandated electoral terms—36, 48, or 60 months—that define the maximum number of months a government can hold office. If the dependent variable is simply the number of months in office, I would not be able to discern whether a cabinet that lasted 36 months completed its maximum term or if the cabinet ended prematurely. Second, measuring the dependent variable as the number of months in office obscures the durability of cabinets that do not form at the beginning of the electoral term. Consider two hypothetical cabinets in a system with a 48-month electoral term. The first cabinet forms immediately after an election and serves two years before collapsing. The second cabinet forms with only two years left before a mandated election and then serves out its maximum term. If the dependent variable were measured as months in office, both would be counted as 24 months, even though the latter served its maximum term.
To overcome these issues, I constructed a dependent variable that measures cabinet duration as the proportion of the maximum term available when the cabinet takes office. For example, a cabinet that lasts 36 months out of a possible 48 would score a 0.75. A cabinet that formed only 12 months before constitutionally mandated elections and survived for those 12 months is coded 1.
I include a control variable, time to election, measuring the maximum number of months that a government could potentially be in office.5 For example, in a country with a 60-month electoral clock, a government that formed immediately after an election would score a 60 because it could potentially be in office for 60 months. A government that formed with only one year left before constitutionally mandated elections would receive a 12 because it could be in office for only 12 months. The longer the potential time a government can be in office increases the possibility that a government will collapse during its term. Consequently, I expect this variable to have a negative coefficient.
Methodology
I employ simple OLS regression. But cross-country differences in cabinet durability, potential problems with serial correlation and high-influence observations, and a skewed dependent variable suggest that simple OLS will produce inefficient parameter estimates. Consequently, I also use three alternative estimators.
The first challenge to OLS is that instances of cabinet durability within a certain country may not be independent of one another. The literature on cabinet formation, as well as clear cross-national differences in cabinet duration, suggests that country-specific factors, such as similar political institutions or practices, will affect all cabinets in that country. As a consequence, there may be unequal variation in cabinet durability across countries. For example, cabinet duration in Italy may differ a great deal, whereas the length of cabinets in Germany is relatively similar. To address this issue, I also run the model using OLS regression with robust (Huber-White) standard errors. This method assumes that observations are independent across countries but not necessarily independent within countries.
A second concern arises from the possibility of serial correlation in the residuals; that is, the errors may be related to one another, either across countries or over time. For example, the duration of a cabinet in Italy at time t is likely to be related to the duration of prior cabinets in Italy because those observations share a similar political environment. To correct for this, I use OLS regression with panel-corrected standard errors and panel specific AR(1) as suggested by Beck and Katz (1995, 1996).
Third, OLS regression is vulnerable to outlying or influential observations. These observations can cause marked changes in parameter estimates. Both the skewed distribution of the dependent variable and distinct patterns of cabinet durability across countries suggest that outliers may be a problem. Consequently, I estimate the model using a variant of robust regression.6 Robust regression provides unbiased and efficient parameter estimates even when there are significant departures from normality.
I report the results from all four estimation strategies. Although OLS is the most conservative estimation technique and should have the largest standard errors, comparison of the standard errors across the four models provides an important robustness check.
Results
Table 8 contains the statistical results. Column entries are parameter estimates, and standard errors are in parentheses. Models I, II, and III are estimated using OLS. Models IV, V, and VI are estimated using OLS with robust standard errors, OLS regression with panel-corrected standard errors and panel specific AR(1), and robust regression, respectively.
Model I provides a baseline model of cabinet duration, based solely on attributes of the governing coalition and the party system. The model as a whole is statistically different from zero. Time to election, fractionalization, and polarization are statistically significant and, as predicted, negative. The negative coefficient on the time-to-election variable indicates that the longer potential term in office, the less likely the cabinet is to last the entire period. Consistent with the literature, the estimates for the fractionalization and polarization variables indicate that more heterogeneous party systems produce cabinets that survive for shorter times. Surprisingly, none of the government types is statistically different from zero.
Model II includes the central bank independence variable. The parameter estimates for the attributes of the governing coalition and the party system remain similar. The central bank variable is, as predicted, positive and significant. According to the results, systems with independent central banks have higher levels of cabinet durability. For a hypothetical government beginning a potential four-year term in office, increasing the level of central bank independence from 0.2 (about one-standard deviation below the mean) to 0.5 (about one standard deviation above the mean) increases expected durability by about four months.
Models III–VI add variables capturing economic internationalization, domestic institutions, and the interactive terms. Interestingly, inclusion of these terms makes the government-type variables statistically significant. (The omitted category is coalition-minority governments). Additionally, the magnitudes of the government-type parameter estimates are consistent with the literature. Single-party majority governments and minimum-winning coalitions are the most durable. Both single-party minority and coalition minority governments are least durable. Party-system fractionalization remains significant and in the predicted direction, but polarization becomes statistically insignificant.
TABLE 8. Models of Cabinet Durability
The standard errors across models III–VI are fairly consistent. As expected, the OLS standard errors (model III) tend to be slightly larger. With a few exceptions, none of the interaction terms is individually significant. T-statistics test whether a parameter estimate is statistically different from zero holding all other variables constant. With interactive terms, however, this is not really possible. Instead, I test for the joint significance of different sets of variables. These tests, included in table 8, indicate that most sets of variables are jointly significant. But again, the OLS estimates appear to be the least efficient of the four models.
Fig. 3. Effect of central bank independence on cabinet durability
The interactive terms, however, make interpretation of individual variables difficult. Consequently, I present the results graphically. Figure 3 shows the effects of central bank independence on cabinet durability. Holding all other variables constant, I generated predicted values of cabinet durability using the sample range of central bank independence (King, Tomz, and Wittenberg 2000). The x-axis represents the range of variation for the central bank independence variable. The y-axis represents the predicted values of cabinet duration.
I have argued that central bank independence is likely to improve cabinet durability where conflicts over monetary policy are likely—that is, in an open economy. In other words, there is an interactive effect between central bank independence and the level of economic openness. Thus, holding other variables at their means, I varied the level of economic openness. To calculate predicted cabinet durability in a closed economy, I restricted capital account openness to be at its lowest level in the sample (1.5) and trade openness to be one standard deviation below its mean (30 percent of GDP).
The line in figure 3 marked “Effect of CBI at low economic openness” illustrates how central bank independence (CBI) affects cabinet durability in a closed economy. The slope of this line is negative. In a closed economy, central bank independence shortens cabinet durability; that is, higher levels of central bank independence lead to shorter cabinet durability.
The line marked “Effect of CBI at high economic openness” indicates how central bank independence affects cabinet durability in an open economy. Here, I restricted capital account openness to be at its highest level in the sample (4) and trade openness to be one standard deviation above its mean (100 percent of GDP). The slope of this line is positive, indicating that central bank independence actually increases cabinet durability in an open economy; that is, an independent central bank helps prolong cabinet durability in situations where conflicts over economic policy are very likely.7 In other words, higher levels of economic openness increase the political value of an independent central bank.
Politicians will want to reform their central bank when it is in their interest to do so. Because they are interested in attaining and retaining office, if institutional reform can help the government survive longer, politicians are likely to pursue those reforms. Where central bank reform is unlikely to increase cabinet durability, politicians will have less incentive to reform the central bank. The results indicate the conditions under which an independent central bank provides higher levels of cabinet durability. Consequently, they can provide predictions concerning the conditions under which politicians are likely to pursue reform.
The implications for central bank reform shown in figure 3 are straightforward: increased economic openness provides an incentive for politicians to adopt an independent central bank. Under conditions of low trade and capital account openness, politicians will prefer a dependent central bank because that institutional arrangement provides them with the highest levels of expected cabinet duration. At higher levels of economic openness, however, politicians can increase their expected cabinet durability by adopting an independent central bank. Table 9 reports the level of economic openness in 1990 for a subsampie of systems with relatively dependent central banks—that is, systems in which central bank reform was likely. A comparison of the level of trade openness and capital account openness for reforming and nonreforming countries indicates that countries that reformed early in the 1990s tended to have higher levels of openness than late reformers or nonreformers.
TABLE 9. Economic Openness and Central Bank Reform
|
|
|
Capital |
|
|
| |||||
Italy |
1981; 1992 |
0.38 |
3.42 |
0.33 |
Center-Right |
New Zealand |
1990 |
0.54 |
3.50 |
0.42 |
Left |
Belgium |
1993 |
1.43 |
3.16 |
0.55 |
Center-Left |
France |
1993 |
0.45 |
3.08 |
0.10 |
Right |
Spain |
1994 |
0.35 |
3.00 |
0.21 |
Left |
Britain |
1997 |
0.51 |
4.00 |
0.38 |
Left |
Sweden |
1998 |
0.59 |
3.16 |
0.83 |
Left |
Ireland |
1998 |
1.12 |
2.83 |
0.35 |
Center-Right |
Japan |
1998 |
0.20 |
2.50 |
0.25 |
Right |
Australia |
|
0.34 |
3.00 |
0.59 |
|
Norway |
|
0.75 |
2.92 |
0.54 |
|
aCountries that ranked with less than 0.50 on the average scale of independence are included.
bExports plus imports as a proportion of GDP in 1990.
cAverage measure of capital account openness, 1985–1990. 1 = closed; 4 = open (Quinn 1997).
dProportion of unionized workers in the workforce in 1990 (Golden, Lange, and Wallerstein 1998).
eGovernment partisanship is for date of reform.
The choice of central bank institutions, however, depends not only on the openness of the economy but also on the level of unionization. Figure 4 illustrates the interactive effect of unionization and central bank independence on cabinet durability under different conditions of economic openness. The x-axis is the level of unionization and the y-axis is predicted cabinet durability.8 The four lines represent different levels of central bank independence (independent, dependent) and economic openness (open, closed).9
Consider first the two lines marked “closed economy.” As expected, unionization has a U-shape effect on cabinet durability; that is, low and high levels of unionization contribute to higher cabinet durability than moderate rates of unionization. But in a closed economy, a dependent central bank produces higher expected cabinet durability at all levels of unionization. This difference is statistically significant; for closed economies, central bank independence hurts cabinet durability regardless of unionization.
Fig. 4. Effect of unionization on predicted cabinet durability
At high levels of economic openness, however, the situation is reversed. Moderate levels of unionization provide the highest predicted cabinet durability, and high union density hurts cabinet durability. Further, in an open economy, an independent central bank produces higher predicted cabinet durability, although this effect diminishes at higher levels of unionization. At high levels of unionization in an open economy, the institutional status of the central bank appears to have no effect on expected cabinet durability.10 Highly organized union organization may already coordinate policy demands and enhance policy effectiveness, making an independent central bank unnecessary. Therefore, in open economies with high levels of unionization, politicians have less incentive to support an independent central bank than in systems with lower levels of unionization.
These results suggest that systems with high levels of unionization and economic openness will be slower to adopt an independent central bank. In fact, the countries that reformed their central banks in the early 1990s—France, Italy, New Zealand, and Spain—tend to have relatively low levels of unionization (see table 9). Countries with higher levels of unionization, such as Norway and Sweden, did not reform their central banks during this time period.
Fig. 5. Effect of central bank independence on predicted cabinet durability for different levels of partisanship and economic openness
Figure 5 shows the effect of central bank independence on expected cabinet durability for different levels of economic openness and partisanship.11 According to the results, economic openness actually helps the cabinet tenure of Left governments.12 This may reflect the success of Left parties in small open democracies (Katzenstein 1985; Lange and Garrett 1985). The results also indicate that central bank independence affects expected cabinet durability for both Left and Right governments in the same manner. In a closed economy, central bank independence has a negative relationship on expected cabinet durability for both Left and Right governments. For an open economy, central bank independence has a slightly positive affect on cabinet durability regardless of partisanship.
The results imply that government partisanship does not affect the probability of central bank reform. Under conditions of economic openness, both Left and Right governments benefit from an independent central bank. Indeed, central bank reform has occurred under both Left and Right governments (see table 9). In New Zealand, Spain, Britain and Sweden, Left governments instituted independent central banks. In Italy, a Center-Right government initiated central bank reform. In France, a Right government carried out the central bank reform, although the Socialists had also endorsed central bank reform in the 1993 election campaign.
In one sense, the findings in this chapter are not particularly controversial. As with much of the recent literature in international political economy, I make a connection between increased economic internationalization and central bank reform. What is different is the mechanism that links internationalization and central bank reform. Much of the literature contends that economic openness—particularly international capital mobility—presents an external constraint on policymakers, forcing them to choose certain types of policies and institutions, including an independent central bank.
Instead, I provide an explicitly political mechanism—cabinet durability—to link internationalization and reform. Increased economic openness alters the policy preferences of constituents in the electoral coalitions of the traditional governing parties, creating the potential for policy conflict among constituents and among party politicians. Economic internationalization also decreases the ability of the government to deliver macroeconomic outcomes. Both of these developments are likely to shorten cabinet durability and hurt the ability of the traditional parties to maintain their position in office.
Party leaders therefore have to consider policy innovations and institutional reforms to ensure that party legislators with different interests will still cooperate with one another to keep the party in office. Because an independent central bank can help prevent conflicts between party politicians with different incentives over monetary policy, central bank reform is one way for party politicians to overcome the potential conflicts created by economic internationalization and protect the party’s viability. Central bank reform represents a response by party politicians to the electoral dealignment brought about by increases in economic openness.
Linking economic internationalization, central bank independence, and cabinet durability helps account for the timing of central bank reform in the late 1980s and early 1990s. Despite international economic shocks—the end of the Bretton Woods monetary system and the oil shocks—and high levels of domestic inflation in the 1970s and early 1980s, politicians in most countries did not attempt to reform their central bank institutions. The stability of central bank institutions reflects the continuity of political incentives within the major parties. Although the economic shocks of the 1970s began to change constituent interests, electoral dealignment did not occur immediately. Voters continued to support the political party with which they identified. Party politicians continued to court traditional social coalitions for electoral support. Consequently, political leaders had little incentive to adopt an independent central bank.
By the mid-1980s, however, the political consequences of economic internationalization were apparent—the interests of the electoral coalitions underlying the traditional governing parties had changed enough to create sustained conflicts over economic and monetary policy. Where this electoral dealignment threatened the ability of the governing party(ies) to remain in office, party leaders moved to grant the central bank more independence. Where the social and economic consequences of internationalization did not threaten the ability of the governing party(ies) to hold office, politicians had less incentive to choose an independent central bank. The next chapter details the cases of Italy, where politicians moved early to grant the Bank of Italy more independence, and Britain, where central bank reform was delayed until the late 1990s, to illustrate the argument.
