The design of central bank institutions represents an explicitly political choice. Therefore, an explanation of the variation of central bank institutions must consider politicians’ interests, particularly their desire to attain and hold office. It must take into account how different electoral, legislative, and government institutions affect politicians’ incentives over central bank structure. Finally, it must explain why politicians remain committed to a particular central bank institution, even if preferences over policy outcomes change, and describe the conditions under which politicians may alter the institution.
In this chapter, I develop an explanation of the choice of central bank institutions, focusing on the potential for intraparty conflict over monetary policy and the principal-agent relationship between backbench legislators, coalition partners, and the cabinet. At a fundamental level, party legislators and coalition partners delegate policy responsibility to the cabinet, allowing legislators and coalition partners to remain ignorant of issue areas while cabinet ministers develop expertise in their portfolio. Cabinet ministers, however, can exploit their informational advantages over backbench legislators and coalition partners to implement their own preferred policies—policies that may not benefit the party as a whole. The informational asymmetries of the policy process therefore can create potential conflicts between party legislators, coalition partners, and cabinet ministers. I argue that an independent central bank can help alleviate some of these conflicts by supplying credible information about monetary policy.
The nature of this intraparty principal-agent relationship conditions the choice of central bank institutions. In situations where conflict over monetary policy between party legislators and cabinet ministers is likely, politicians will choose an independent central bank. If conflict is unlikely, politicians will opt for a dependent central bank. Consequently, variations in the potential for intraparty conflict can help explain differences in central bank institutions across systems and over time.
This chapter develops the logic of the argument. The first section discusses how the informational asymmetries of the monetary-policy process affect the incentives of legislators, coalition partners, and cabinet ministers over the choice of central bank institutions. The second and third sections explore the institutional conditions under which information from a central bank bureaucrat can solve the potential conflicts created by the informational asymmetries of the policy process. The fourth section lays out the observable implications of the argument.
I begin with the assumption that politicians and parties are fundamentally motivated by a desire to retain office. Before any concern with either policy or “pork,” politicians and parties must hold office. Politicians’ preference for attaining (and retaining) office provides them with policy incentives. Because voters generally use economic outcomes, rather than policy outputs, to evaluate candidates and parties, politicians and parties must pursue policy outcomes that satisfy the preferences of their constituents.
But in a complex issue area such as monetary policy, politicians may have difficulty choosing policies that will produce desired outcomes (Krehbiel 1991). First, changes in monetary policy do not always have expected consequences. Monetary policy has indirect effects throughout the economy, often with uneven and variable lag times (Beck 1987; Friedman 1968). These outcomes are vulnerable to exogenous shocks and international influences. Second, changes in monetary policy may have different effects across economic sectors. Higher interest rates, for instance, may benefit banking and financial interests but hurt firms in the construction industry. Third, monetary policy takes time to affect the economy. Politicians must consider this lag to deliver outcomes in a timely manner. Finally, the relationship between policy and outcomes may itself change as new technology evolves or as individuals modify their behavior in response to current regulations. The advent of electronic banking, for example, significantly altered the conduct of monetary policy, forcing policymakers to reconsider how they estimate the money supply. Successful monetary policy-making therefore requires that politicians invest resources to develop the expertise necessary to predict the consequences of a variety of policy proposals.
Although individual legislators want policy outcomes to reflect the preferences of their constituents, each legislator must budget his or her limited time and resources toward achieving reelection. These individual electoral incentives create collective dilemmas for the legislative majority in the production and implementation of legislation (Kiewiet and McCubbins 1991).
Legislators face a dilemma in the development of policy expertise (Gilligan and Krehbiel 1989, 1990; Krehbiel 1991). Information about the relationship between policy choices and economic outcomes represents a collective good for legislators; it enhances the quality of legislation and ensures that policy leads to outcomes that accurately reflect the desires of the enacting coalition. Each legislator, however, has high opportunity costs in the development of the expertise necessary to legislate and to oversee policy implementation effectively. They can better spend their time and effort campaigning or focusing on issue areas with specific electoral benefits. As a result of these electoral pressures, no individual legislator may have an incentive to gather information about an issue area, limiting the quality of both initial legislation and legislative oversight of policy implementation.
Although legislators desire policies that enhance their reelectoral fortunes, they may not possess the ability to organize themselves to achieve those policies. To help solve these collective dilemmas, legislators delegate policy responsibility to cabinet ministers.
In forming a coalition government, parties make both policy compromises and policy logrolls—trading policies across issue areas. They also divide the cabinet portfolios (Laver and Schofield 1990; Laver and Shepsle 1994). In doing so, parties in a coalition delegate management of monetary policy to one party. Although formally constrained by the coalition agreement, the party controlling that dimension, the portfolio party, has some discretion in developing new policies and in responding to situations unforeseen in the coalition agreement.1
The nonportfolio party, which does not control monetary policy (although it may control other issue areas), possesses interests similar to those of the backbenchers in the portfolio party.2 Because the government’s economic-policy performance affects the nonportfolio party’s electoral support (even if it does not hold the monetary-policy portfolio), the nonportfolio party wants outcomes to reflect the interests of its constituents as closely as possible. At the same time, however, the nonportfolio party does not want to make costly investments in developing expertise in portfolios it does not control or in monitoring the portfolio party’s compliance with the coalition agreement. Instead, it prefers to devote resources to its own portfolios and electoral campaigns.
To overcome these dilemmas, backbench legislators and coalition partners delegate monetary-policy authority to the cabinet (i.e., the president, prime minister, or cabinet ministers). In parliamentary systems, in which the cabinet “emerges from” and “is responsible to” the legislature, the legislative majority explicitly delegates this responsibility to the cabinet (Lijphart 1984). But even in presidential systems, in which the executive is not responsible to the legislature, legislators still depend on the government. Cabinet ministers acquire policy expertise, allowing them to develop legislative proposals and oversee policy implementation. By gathering information about the relationship between monetary policy and macroeconomic performance, cabinet ministers may help guarantee that policy outcomes reflect the desires of their party’s and their coalition partners’ constituents.
As an incentive for ministers to develop expertise, legislators and parties grant ministers institutional power over monetary policy, including agenda control and discretionary authority. First, with agenda control powers, the minister can limit the policy proposals considered by the cabinet and by the legislature as a whole (Huber 1996; Laver and Shepsle 1994). The minister may therefore manipulate policy by acting as a gatekeeper.
Second, cabinet ministers also possess some discretion over monetary policy. They may choose policies without explicit approval as long as those choices do not cause backbench legislators or coalition partners to withdraw their support. In a parliamentary system, the cabinet needs the support of party legislators and—in multiparty systems—coalition partners to remain in office. In a presidential system, the government does not need the support of party legislators to remain in office, but it does require their support to achieve its policy goals.
The range of the cabinet’s discretion reflects the configuration of policy preferences among the cabinet’s legislative and coalition supporters, as well as the expected policy outcome that would result if the legislature vetoed the cabinet (Aldrich 1995; Austen-Smith and Banks 1990; Cox and McCubbins 1993; Laver and Schofield 1990; Laver and Shepsle 1996). If the policy preferences of party (or coalition) legislators lie relatively close to one another—if a consensus over outcomes exists within the government’s party or the coalition—the cabinet minister has less discretion. That is, the set of acceptable policies from which the minister can choose policy is relatively small. As the policy preferences of government supporters become more diverse, the minister’s discretion generally increases. In some situations, the cabinet can manipulate the agenda to build support for almost any policy.
The expected outcome of a legislative veto of cabinet policy also influences the level of cabinet discretion. In some systems, a veto may result in a simple policy change, bringing it closer to the ideal points of backbench legislators and coalition partners. In other systems, a veto may force a change in the partisan identity of the cabinet (Huber 1996). If the expected outcome of a veto lies close to the preferences of the party legislators and coalition partners, then the cabinet has less discretion. As the cabinet’s legislative and coalition supporters find the alternative less and less acceptable, however, the cabinet enjoys greater discretion.
In return for institutional authority over their portfolio, ministers develop the expertise necessary to forecast relatively accurately the relationship between policy outputs and outcomes. Although cabinet ministers must maintain the support of backbench legislators and coalition partners, they may be tempted to exploit their expertise to manipulate monetary policy for their own electoral and policy goals—goals that might differ from those of their legislative and coalition supporters.
In fact, the cabinet possesses a twofold informational advantage over backbench legislators and coalition partners in monetary policy. First, cabinet ministers are in a position to know more about the relationship between policy outputs and outcomes. The complex relationship between monetary policy and economic performance, including the lag between policy choices and results, allows the cabinet to avoid responsibility for bad outcomes they may have caused and to claim credit for good outcomes they did not create. For instance, the cabinet may propose a loosening of monetary policy, lowering interest rates. Legislators and coalition partners do not know whether economic conditions justify lower interest rates—for example, if inflation is under control and continued high rates would stifle potential economic expansion—or if the cabinet has lowered rates to create a short-term economic boom, with negative long-term consequences.
Second, cabinet ministers may conceal their actual monetary-policy choices from party legislators and coalition partners. No single policy indicator captures the direction of monetary policy, allowing the cabinet to obfuscate its intentions. For example, during the early 1980s, British policymakers announced several monetary targets and then emphasized the monetary aggregate that came closest to the announced target (Temperton 1991).
Furthermore, the cabinet has incentives to maintain secrecy surrounding its monetary-policy plans (Cukierman 1992). Secrecy allows the cabinet to generate politically expedient temporary booms in real economic conditions through surprise monetary expansions. Providing information to its legislative and coalition supporters about plans for monetary policy could actually negate the ability of the cabinet to manipulate short-term economic outcomes.
The delegation of authority to the cabinet solves the policy-information dilemma, but it creates the potential for new conflicts between legislators, coalition partners, and cabinet ministers—conflicts that could threaten the ability of the party to remain in office. First, the cabinet’s monetary-policy choices influence the electoral fortunes of party legislators and coalition partners. Although macroeconomic outcomes may not be a decisive factor for an individual legislator or for a nonportfolio party, empirical evidence across systems indicates that inflation negatively affects the incumbent government’s electoral fortunes (Lewis-Beck 1988; Powell and Whitten 1993). Moreover, the cabinet’s policy choices fundamentally shape the party’s social coalition, benefiting some constituents while hurting others. Over time, these policy decisions determine the party’s electoral strategy and condition the party’s viability. Backbench legislators and coalition partners must therefore decide whether to trust cabinet ministers to pursue policies that will help their electoral fortunes. If the cabinet’s incentives over monetary policy differ from those of its legislative supporters or coalition partners, backbenchers and coalition partners will likely decide that they cannot trust the cabinet.
Second, these information asymmetries can threaten the party’s position in office. Despite government ministers’ expertise, policy outputs may have unintended consequences, producing outcomes unacceptable to the cabinet’s legislative and coalition supporters. Without the technical expertise and oversight capability to evaluate monetary policy, party legislators and coalition partners may blame the cabinet for outcomes that the cabinet could not control, withdrawing their support even after ministers had attempted to pursue policies in the party’s (or coalition’s) overall interest. If the cabinet’s supporters can credibly threaten to punish the cabinet for negative policy outcomes—for instance, if a policy consensus exists in favor of low inflation—then the cabinet has an incentive to find ways to reassure party legislators and coalition partners about its policy behavior.
Accurate information about the cabinet’s monetary-policy choices and the consequences of those choices can help alleviate these potential conflicts.3 With this information, backbench legislators and coalition partners can implicitly threaten to punish the cabinet if it does not pursue policies that benefit them. At the same time, cabinet ministers can use this information to maintain the support of their legislative and coalition principals—and keep the party in office for a longer time.
The cabinet’s principals can monitor the cabinet’s policy behavior in three ways: by gathering information themselves, by observing the cabinet’s compliance with policy rules, and by relying on endorsements from outside actors. First, backbench legislators and coalition partners can monitor the cabinet’s policy choices themselves. Legislators can institute strong committee systems, allowing MPs to develop expertise in particular issue areas (Krehbiel 1991). In a coalition government, nonportfolio parties may demand deputy-level cabinet positions, which permit them to participate in the ministry’s policy-making process. These institutions, however, require costly investments in time and resources—resources that may be better employed in electoral campaigns or directed toward particularly salient issues for the party. The responsibility of cabinet oversight may also create new collective dilemmas for legislators. Individual legislators may lack incentives to gather information in a particular issue area, hurting the ability of party legislators to monitor cabinet activity. Finally, the cabinet’s control of the legislative agenda in some systems (and the existence of party discipline) may make it difficult for legislators to propose bills that challenge the cabinet’s position, limiting the incentive for legislators to gather information.
Easily observable policy rules represent a second way to monitor the cabinet’s policy choices. Policy rules provide ex-ante information about the cabinet’s responses to different circumstances.4 These policy rules not only make outputs more predictable, but they also allow backbench legislators and coalition partners to evaluate the cabinet’s policy performance against clear standards. Rules are an attractive option to monitor the cabinet when compliance usually leads to satisfactory outcomes. Strict adherence to policy rules, however, may not give the cabinet enough flexibility to respond satisfactorily to unforeseen or ambiguous circumstances. Additionally, the relationship between the policy rule and outcomes may change over time, so that compliance produces undesirable consequences.
Despite these limitations, European Community member states successfully employed a policy rule during the 1980s: exchange rate stability in the EMS. The EMS established clear policy guidelines for governments based on changes in the value of the country’s currency. These rules allowed each government’s legislative and coalition supporters, financial markets, and the public to easily monitor their cabinet’s policy actions (Bernhard 1998).
A third method to monitor the cabinet is to rely on information from actors affected by or involved in the policy process. These actors often possess policy expertise. Their endorsement or criticism of policy may allow the cabinet’s principals to infer the consequences of the cabinet’s policy choices (Calvert 1985; Lupia 1992). Additionally, these actors can help party legislators interpret cabinet behavior in the event of unanticipated situations.
Interest groups, for example, often provide information about the consequences of government policy in their issue area. They serve as “fire alarms” for party legislators, notifying them if policy choices adversely affect their constituents (Banks and Weingast 1992; Lupia and McCubbins 1993; McCubbins and Schwartz 1983). If the interest group has a known policy position or possesses strong ties with a political party—for instance, as unions and social democratic parties do in many European countries—then party MPs can rely on the interest group’s evaluation of the government’s policy choices.5
Backbench legislators and coalition partners may also use central bank bureaucrats to help them monitor the cabinet. Central bankers possess information about the cabinet’s policy choices and their economic effects. The cabinet relies on the central bank to implement its policy choices, ensuring that bureaucrats will be informed of policy changes. Moreover, as experts in the policy area, central bankers can forecast and evaluate the consequences of the cabinet’s policy.
But how can politicians ensure that central bankers will provide the information that they need to prevent potential intraparty conflicts? Under what conditions will backbench legislators and coalition partners find the central bank’s policy evaluations credible and informative? To answer these questions, I develop a game-theoretic model of the monetary-policy process to determine the conditions under which legislators and coalition partners can infer consequences of the cabinet’s policy choices. The results of the model have implications for the choice of central bank institutions and for the behavior of cabinet ministers, central bankers, and legislators in the policy process.
In this section, I informally sketch the logic of my model of the monetary-policy process. The appendix contains the formal model. The next section discusses, again informally, the results of the model.
The game has three players: the government (i.e., the president, the prime minister, the minister of finance), the central bank bureaucrat, and the legislator. The legislator represents the cabinet minister’s principals—either party legislators, coalition partners, or both. It may be useful to think of the legislator as a set of party MPs (or a coalition partner) who can act as a veto point to block legislation.
The model is one of asymmetric information. The information asymmetry stems from the ignorance of backbench legislators and coalition partners about the relationship between policy outputs and economic outcomes (Gilligan and Krehbiel 1989, 1990; Krehbiel 1991). The legislator knows the location of the status quo but does not know the outcomes associated with a policy change. A policy change could result in an outcome that backbench legislators and coalition partners prefer to the status quo, P1, or in an outcome that is worse for them than the status quo, P2 (see fig. 1).
Cabinet ministers and central bankers, on the other hand, have expertise in monetary policy. They have access to a variety of economic indicators, forecasts, and analyses, allowing them to predict the consequence of a policy change more accurately than backbench legislators or coalition partners. As a result, cabinet ministers and central bankers have full knowledge of both the location of the status quo and the consequences of a policy change.
The sequence of play is as follows. Nature first determines whether a policy change will lead to an outcome that benefits the legislator (with probability d) or hurts the legislator (with probablility 1–d). The government makes a proposal to either enact a new policy (M) or to retain the status quo (Q). The central bank then offers an explicity endorsement (E) or rejection (C) of the government’s proposal.6 In effect, the central bank says to the legislator, “Based on what I know about the consequences of a policy change, you should or should not approve the policy:’ In practice, central bankers are not usually so blunt (although they can be). More often, central bankers provide a forecast and analysis of policy outcomes, signaling whether good or bad news is on the economic horizon. They may also provide some policy prescriptions based on their analysis. From these pronouncements, backbench legislators, coalition partners, and the public can determine the bank’s opinion on the government’s choice. Finally, after observing the government and the bureaucrat, the legislator votes to accept or reject a policy change. Figure 2 presents a game tree for the sequence of moves.
Fig. 1. Potential outcomes of a policy change
The three players receive payoffs from the policy outcomes: the closer the outcome to the player’s preferences, the higher the payoff. The preferences of the legislator are the baseline, ranking the outcomes as follows: a policy change, P1, that would hurt them < status quo (sq) < a policy change, P2, that would help them. I evaluated the game based on different preference configurations for the government and bureaucrat, where each had preferences over the policy outcomes that (1) were the same as those of the legislature, (2) were opposed to those of the legislature, or (3) were in partial agreement with those of the legislature (e.g., sq < P2 < P1). This created a possible total of 36 different preference configurations (see appendix for details).
In addition to receiving payoffs from the policy outcomes, the government and the bureaucrat receive a payoff from the legislator’s action. If the legislator vetoes a government proposal, the government receives a penalty larger than any possible gain in utility from achieving his or her preferred policy outcome. A veto represents a challenge to the policy on the legislative floor or, in a multiparty government, when a coalition partner chooses to withdraw from the government. No government likes to suffer the loss in political reputation resulting from a legislative veto. In a parliamentary system, a legislative veto might also precipitate a vote of no confidence. In a coalition government, the withdrawal of the nonportfolio party could cause the government to collapse.
Fig. 2. Game tree and outcomes
Similarly, if the legislator votes against the central bank’s advice, the central bank receives a penalty larger than any possible gain from achieving the bureaucrat’s preferred policy outcome. Bureaucrats do not like to have their decisions vetoed (Ferejohn and Shipan 1990). A veto could hurt the central bank’s reputation or demonstrate its irrelevance to the policy process. A legislative veto could also represent an attempt to alter the bank’s institutional structure.
I evaluated two possible scenarios based on the central bank’s institutional status. In the first, the government has no ex-post power over the bank; that is, the government does not have the authority to dismiss central bankers or veto the bank’s policy decisions. In the second scenario, the government has ex-post power over the central bank: the authority to dismiss central bank directors, cut the bank’s budget, or veto the bank’s decisions. After the legislature votes on the policy, the government imposes a cost on the central bank if the bank’s comment contradicts the government’s proposal.7 This penalty is larger than any possible payoff that the bureaucrat could receive from achieving the preferred policy outcome.
The structure of the game has strong empirical analogues. The legal statutes of the Dutch central bank, for example, contain formal procedures to resolve conflicts between the government and the central bank, with the parliament as the final arbiter (The Economist, 10 February 1990; Eizenga 1987). If the government and the bank disagree over the proper course of monetary policy, each submits a report to the parliament to explain its position. Once MPs have considered the two positions, the parliament chooses which policy to support. Interestingly, the parliament has never been called upon to resolve a dispute. Rather than risk a parliamentary vote—which would almost inevitably result in the resignation of the defeated actor—the government and the bank have resolved their differences behind closed doors.
I solved the model to determine the conditions under which legislators and coalition partners can infer the consequences of the government’s policy choices. These conditions reflect (1) the amount of preference agreement between the government, the central bank, and the legislator and (2) the bank’s formal structure—in particular, whether the government has the authority to punish the central bank. The following two results are particularly relevant.
The Legislator Trusts the Government’s Proposal
In the model, the legislator trusts the government’s recommendation if the government shares its preference ordering over policy outcomes. The government’s ex-post authority over the central bank does not affect the result. In other words, if party legislators and coalition partners share the government’s policy objectives, they accept its policy evaluation regardless of the central bank’s structure.
The Legislator Trusts the Central Bank’s Evaluation
According to the results, the legislator finds the central bank’s policy evaluation informative under two conditions: (1) the central bank and the legislator share the same policy objectives and (2) the government does not have ex-post authority over the central bank. These conditions, I contend, describe facets of an independent central bank.
The model implies that the institutional features of central bank independence provide central bank bureaucrats with credibility among party legislators, coalition partners, and the public. In particular, the appointments procedure and lack of ex-post authority enhance the bank’s trustworthiness. Consequently, an independent central bank may act as a check on the cabinet’s discretion and help prevent intraparty conflicts over monetary policy. In turn, political parties may grant central banks more independence in order to capitalize on the bank’s political credibility.
According to the results, the central bank’s policy preferences determine the possibility of information transmission. If the cabinet cannot punish the central bank, preference agreement between the legislator and the central bank is sufficient to ensure that the legislator will be informed about the consequences of a policy change. If the central bank’s (and the cabinet’s) preferences differ from those of the legislator, then the legislator has at best only the possibility of knowing the effects of a policy proposal.
The model also implies that the legislator and the cabinet rarely agree about the type of central banker to appoint. If the cabinet cannot punish the central bank, the legislator naturally prefers a central bank that mirrors its own preferences. The cabinet, on the other hand, prefers to increase its discretion and will want a central bank that does not share the legislator’s preferences. In contrast, if the cabinet and the legislator share the same preferences or if the cabinet can dismiss the bureaucrat, the legislator is indifferent to the bureaucrat’s preferences. Institutional control over the appointments procedure therefore critically shapes the central bank’s ability to influence the policy process.
Institutional control of the appointments procedure distinguishes dependent and independent central banks. In dependent central banks, the cabinet controls all appointments to the bank’s governing board. Independent central banks, in contrast, have only a limited number of cabinet appointees on the bank’s board, which instead includes a variety of political, regional, or sectoral representatives. These appointment procedures produce central bank councils that reflect the major divisions over monetary policy within the society, enhancing the central bank’s credibility across constituents with a variety of economic policy demands. Because of these appointments procedures, the preferences of an independent central bank roughly mirror the preferences of the legislature (except that the bank does not have a short-term electoral imperative).
In Germany, for example, differences in monetary-policy preferences usually follow party lines. Until recently, each Land government appointed a representative to the central bank council, ensuring that the political composition of the Bundesbank’s central council echoed the balance of political representation in the country as a whole. In the United States, differences over monetary policy are traditionally regional rather than partisan, pitting conservative easterners against southern and western monetary populists. Placing Reserve Bank governors, elected by regional stockholders, on the FOMC ensures that it has a balance of regional interests over monetary policy similar to those found in Congress. In Austria, representatives of the employers’ associations and the labor unions sit on the bank’s governing board, reflecting the corporatist structure of Austrian society. These appointments procedures ensure that the central bank will possess preferences similar to those found within the legislature, allowing the bank to provide credible information to the legislature and to the public about the cabinet’s monetary-policy choices.
The cabinet’s ex-post power also affects the central bank’s ability to provide information about the consequences of the cabinet’s policy choices. According to the model, unless the cabinet and the legislator fully agree, the legislature has a possibility of receiving new information during the policy process only if the cabinet can impose a penalty on the central bank. Therefore, when the central bank shares the legislator’s preferences, the legislator loses information by granting ex-post power to the cabinet. If the central bank does not share the legislator’s preferences, however, the legislator is indifferent about granting expost authority to the cabinet or in some cases may actually prefer the cabinet to have that authority.
The cabinet’s lack of ability to punish the central bank is another characteristic of an independent central bank. The formal rules of an independent central bank prevent the government from vetoing the bank’s policy decision, dismissing central bankers, or cutting the bank’s budget. In contrast, the cabinet can punish a dependent central bank, severely limiting central bankers’ ability to reveal information about the cabinet’s policies; criticism of cabinet policy may result in retaliation through dismissal or budget cuts. Consequently, backbench legislators and coalition partners place less weight on the bank’s policy evaluation because they know that the bank must tailor its statements to ensure the cabinet’s approval. Without a credible alternative source of information, legislators and coalition partners must instead rely on the cabinet’s policy recommendations. Consequently, the cabinet has more discretion over policy.
The conditions under which the legislator trusts the central bank’s policy evaluation depict the institutional structure of an independent central bank. Party legislators and coalition partners will rely on an independent central bank’s credibility to remain informed about monetary policy. In turn, this information can help prevent potential intraparty conflicts over monetary policy.
If independent central bankers send approving signals about the cabinet’s policy, the cabinet’s legislative supporters and coalition partners can have confidence that policy reflects their interests. If the bank criticizes the cabinet’s monetary policy, party legislators can infer that the cabinet’s policy has consequences that may hurt them. Armed with better information, the cabinet’s legislative and coalition supporters can then accept or veto the cabinet’s choices.
The policy information can also enhance the cabinet’s credibility with its legislative and coalition principals. An independent central bank can reassure legislators that the cabinet had attempted to pursue policies in their interests, even if outcomes are less than satisfactory. Consequently, party legislators and coalition partners will be less likely to withdraw their support from the cabinet.
The argument yields three sets of observable implications concerning the patterns of policy disputes between governments and independent central bankers, the cross-national variation of central bank independence, and the durability of cabinets in systems with an independent central bank.
The model emphasizes the role of the legislature in the policy process. Indeed, the results indicate that the legislature strongly shapes monetary-policy choices. Nevertheless, analysts have noted the relative lack of legislative involvement in monetary policy throughout the industrial countries (LeLoup and Woolley 1991; Woolley and LeLoup 1989). In fact, the model suggests that legislators will not act on monetary policy. In equilibrium, both the cabinet and the central bank make the same policy recommendation. The costs associated with a veto provide an incentive for both the cabinet and the central bank to appear united about the potential consequences of a policy change. As a result, the legislature does not need to decide between the alternative recommendations. Instead, it takes only a passive action, simply supporting the cabinet’s proposal.
The threat of a legislative veto, however, shapes the cabinet’s and the central bank’s incentives. Both cabinet ministers and central bankers recognize that the legislature ultimately determines policy in the event of a dispute. Each anticipates the legislature’s reaction before making a recommendation. If either the cabinet or the central bank recognizes that it will lose a dispute in the legislature, it will accept the other’s recommendation rather than suffer a legislative veto.
An independent central bank’s political credibility, therefore, changes the cabinet’s strategic incentives. If the cabinet announces a policy change that earns a negative evaluation from an independent bank, it risks punishment from party legislators and coalition partners. Rather than chance a legislative veto, cabinets will choose policies that the independent central bank will not criticize. Although the cabinet retains control over policy goals, the bank forces cabinet ministers to be more truthful about economic and monetary policy, checking the cabinet’s ability to manipulate policy for its own purposes. As a result, the bank may appear to control monetary policy.
The possibility that party legislators and coalition partners may support the cabinet’s policy position also affects the central bank’s incentives. If backbench legislators and coalition partners support the cabinet, the central bank must make the same recommendation as the cabinet or suffer the costs of a policy veto, even if the central bank disagrees about the desirability of the policy. Therefore, independent central banks are constrained when the cabinet and a legislative majority agree about policy. Independent central bankers anticipate the cabinet’s level of support in the legislature. If that support is strong, then they are likely to temper their comments in order to avoid the costs of a policy veto. On the other hand, if the cabinet cannot muster a legislative majority to punish an independent central bank, then the bank can criticize cabinet policy.
This strategic interaction in the formulation of monetary policy therefore implies a discernable pattern to disputes between governments and independent central bankers. First, public disagreements between the cabinet and an independent central bank will be relatively rare. Each actor would rather work out a solution privately rather than risk a legislative veto. Second, independent central bankers will be critical of cabinet policy only when they anticipate that the cabinet will be unable to secure a legislative majority to support its position. Third, independent central bankers will not criticize the cabinet, even if they disagree with its policy prescription, when it is clear that the cabinet’s position enjoys widespread support among politicians in the governing party(ies).
The model’s results indicate that politicians can structure the central bank to help them overcome the conflicts created by the informational asymmetries of the policy process. Under what conditions, then, will party legislators, coalition partners, and the cabinet be willing to limit the cabinet’s control over the central bank?
Party legislators and coalition partners will prefer an independent central bank if they cannot trust the cabinet—that is, if the cabinet possesses monetary-policy incentives that differ from their legislative supporters and coalition partners. In this situation, backbench legislators and coalition partners recognize that the cabinet will likely exploit its informational advantages to pursue policies that do not reflect their interests. On the other hand, if party legislators and coalition partners have confidence that the cabinet shares their monetary-policy goals, then they can trust cabinet ministers. They have no incentive to check the cabinet’s discretion with an independent central bank. Instead, they will support a dependent bank.
The cabinet must also support the choice of bureaucratic structure. The priority of government ministers is to maintain their positions in office. If party legislators and coalition partners can credibly threaten to punish the cabinet for its monetary-policy performance, then the cabinet will want them to have information that demonstrates the cabinet’s policy choices were intended to benefit the party. In this situation, the cabinet has an incentive to support an independent central bank. If party legislators and coalition partners cannot credibly threaten to punish the cabinet, however, then the cabinet has little incentive to restrict its monetary-policy discretion with an independent central bank.
The choice of central bank institutions therefore reflects (1) the divergence of policy incentives between the cabinet and its legislative and coalition supporters and (2) the credibility of legislators’ and coalition partners’ threat to punish cabinet ministers. In situations where the cabinet faces policy incentives that differ from party legislators and coalition partners and where those backbench legislators or coalition partners can punish the cabinet, politicians will agree to an independent central bank. If one of those conditions is not present, the bank will be less independent of cabinet control. Differences in these two variables across systems should account for the cross-national variation of central bank institutions.
The informational asymmetries of the monetary-policy process are likely to lead to intraparty conflicts, especially where party politicians face different incentives over monetary policy. Without the information necessary to evaluate the cabinet’s monetary-policy choices, backbench legislators and coalition partners will suspect that the cabinet is manipulating monetary policy away from their ideal points, hurting their electoral fortunes and the party’s long-term viability. Consequently, backbench legislators and coalition partners may be quick to withdraw their support from the cabinet over a monetary-policy dispute.
The credibility of an independent central bank helps prevent some of these potential conflicts. The bank can provide information about the cabinet’s policy behavior to reassure party legislators and coalition partners. Further, the bank can help government ministers justify monetary policy to legislators, coalition partners, and the public.
Therefore, in systems with an independent central bank, one should expect fewer political disputes over monetary policy between party legislators, coalition partners, and government ministers. One way to measure these conflicts is to examine cabinet durability in parliamentary systems. If an independent central bank does prevent conflict over monetary policy, cabinets in parliamentary systems with an independent central bank should have a higher expected duration than cabinets in systems with a dependent central bank, holding all other influences on cabinet durability constant.
Linking central bank independence and cabinet durability can also help explain patterns of central bank reform. Politicians will reform the central bank when it is in their interest to do so—that is, when central bank reform is likely to prolong the tenure of incumbent politicians. Over time, changes in constituent preferences and decreased policy effectiveness may increase the possibility of conflict over monetary policy within the governing party(ies). Where these intraparty conflicts threaten the ability of these parties to stay in office, politicians will adopt an independent central bank to prevent the conflicts from hurting the party’s political fortunes.
The following four chapters explore these implications more fully. The next chapter examines patterns of conflict between the Bundesbank and the government in Germany. Chapter 5 shows how the divergence of policy incentives and the threat of punishment condition the choice of central bank institutions across systems. Chapter 6 demonstrates that an independent central bank increases cabinet durability under conditions of economic openness—precisely where we would expect intraparty conflict over monetary policy to be prominent. Chapter 7 explores central bank reform in Italy and Britain, illustrating how political and economic developments increased the political value of an independent central bank.
