Credibility Versus Control

Abstract

There is a natural tension between theories of party government and theories of regulatory politics. Whereas effective party government requires that politicians have firm control over public policy, the need for credible commitment in regulation stipulates that policy-making capacities are delegated to independent agencies. While the theoretical dimension of this tension is well established, there is little research that examines its empirical implications. To narrow this gap, the analysis assesses whether legal agency independence limits the influence of parties on agency executives. To that end, it investigates the careers of 300 CEOs in 100 West European regulatory agencies. The analysis shows that high levels of agency independence protect appointees with opposition ties from early removal. This presents some of the first evidence to suggest that the institutional response to credibility pressures limits the political use of the appointment channel and, thus, has the potential to constrain party control in regulatory politics.

Keywords

political parties credibility agency independence patronage

Introduction

On December 15, 2006, the socialist government of Portugal ratified a decree¹ that limited the increase in electricity network tariffs for the year 2007 to 6%, thereby overruling a proposal by ERSE, the independent Portuguese energy regulator, to allow for an increase of up to 16%. While the government’s modification of the regulator’s proposal aimed at protecting consumers from a substantial price hike, it was also widely believed to be in violation of European Law (Directive 2003/54/EC). The head of ERSE, Jorge Vasconcelos, tendered his resignation in protest of government interference in regulatory matters on the day the decree was issued (he was dismissed immediately so as to prevent his appearance before a parliamentary committee). In response, the share price of the partly state-owned Energias de Portugal (EDP) rose by 2.4%. Only 2 weeks later, the government appointed Vítor Santos as the new head of ERSE, an economics professor who had served as a state secretary in previous socialist cabinets.

This anecdote illustrates one of the central conflicts of modern policy making: the tension between the partisan agenda of democratically elected governments and the pressures to isolate certain public sector institutions from political interference. It also highlights the fact that this tension often plays out in conflicts over the top personnel of independent agencies.

It is a central requirement of representative democracy that the preferences of the electorate have some influence on public policy. Especially in parliamentary democracies, political parties are crucial intermediaries in this process. They compete in the electoral arena, structure government formation and legislative decision making, and drive the formulation and implementation of policy. Parties are what make the chain of delegation and accountability in parliamentary systems work (Müller, 2000b).

However, all modern democratic constitutions also isolate certain actors within the state apparatus from the direct influence of politicians, most importantly, judges and justices. In the more recent past, however, an increasing number of public sector institutions outside the judiciary have been made independent. Most prominently, central bank independence has become a cornerstone of many monetary regimes (Polillo & Guillén, 2005). In regulatory politics, the rise of the agency model has left elected officials with more fragmented and diluted powers (Gilardi, 2005b). In all these cases, weaker political influence has been demanded to strengthen the credibility of certain policy arrangements and thus produce better policy outcomes.

The tension that emerges between input-oriented democratic imperatives and output-oriented performance demands has long been recognized by scholars of democratic politics. In a 2008 essay, Peter Mair (2008) even identified the fragmentation of the public sector and the growing number of independent government agencies as one of the main challenges to effective party government and thus political legitimacy in representative democracies. By contrast, Scharpf (1970, 1999) and Majone (1998, 2001) have stressed that nonmajoritarian institutions can increase democratic legitimacy by making government more effective and producing better policies, thus compensating for losses in input legitimacy by enhancing the system’s output legitimacy.

Although the theoretical dimension of this problem is thus well established, there is little work that systematically analyzes its empirical implications. The present article aims to address this gap by examining the interplay between party influence and agency independence in regulatory politics. The theoretical section discusses the concept of party government, outlines the case for delegation to nonmajoritarian institutions to establish policy credibility, and highlights the potential conflict between the two approaches. The empirical section then examines the influence of party ties and formal independence on the tenure of 300 CEOs in 100 regulatory agencies (RAs) in 16 West European countries. It focuses especially on the interaction between partisanship and agency independence. The final section concludes.

Theoretical Framework

The Strain on Party Government

Theorists of democratic politics have long recognized that modern mass democracies are unthinkable without political parties that function as intermediaries between the demands of individual citizens and the actions of the state apparatus (Schattschneider, 1942). Virtually all modern democratic societies have therefore established systems of governance that resemble the concept of party government (although exceptions exist, see Veenendaal, 2013).

Party government is typically defined as the process by which executive office is awarded to party politicians through competitive elections with clear policy alternatives, policy is determined by the parties controlling the executive, and the executive is held accountable through parties (Blondel & Cotta, 1996, 2000; Katz, 1986; Mair, 2008; Rose, 1969, 1974; Thomassen, 1994). A well-functioning system of party government is therefore able to deliver public policy that is responsive to the preferences of voters.

Crucially, however, all conceptualizations of party government require that the parties in government have control over the bureaucratic apparatus so as to ensure proper and effective implementation of their policies. Without such a mechanism in place, it becomes impossible for politicians to effectively translate voter (or party) preferences into public policy. This was already acknowledged by Rose (1969) in the first comprehensive account of the conditions for party government. Rose (1969) states as one central requirement that the policies put forward by the parties controlling the executive need to be carried out by the administration. This necessitates that the politicians in government are able to “secure compliance to their own directives from a massive but passive bureaucracy” (p. 418).

There are essentially two reasons why party politicians depend on bureaucrats to translate the government’s decisions into policy outcomes (Thies, 2000). First, it is practically impossible for the core executive to administer the implementation of all of its statutes. Second, although policy-oriented politicians often have clear preferences about policy outcomes, they may at times lack the expertise to ensure that a specific outcome is, in fact, produced. The necessity of a division of labor and the need for expert knowledge are thus the prime reasons why party government hinges on the effective delegation of tasks from the government to the bureaucracy.

How do parties in government ensure that their policy decisions are accurately carried out by the bureaucracy? In parliamentary democracies, government departments are typically headed by party politicians, and civil servants are legally bound to follow the instructions of their ministers. Bureaucratic drift can be limited by statutory or budgetary means (Huber, 2000), but also by strategically appointing partisans to administrative elite positions (Dahlström & Niklasson, 2013; Ennser-Jedenastik, 2014a).

Indeed, party patronage can be conceived of as a linkage mechanism between parties and the government (Blondel, 2002), thus promoting party government. This is consistent with the view that party patronage in many European democracies has been transformed from an electoral to an organizational resource (Kopecký, Mair, & Spirova, 2012). Rather than handing out jobs to loyal followers, its purpose has shifted to allow parties to exert control over the increasingly fragmented governing institutions of the state.

Empirical work on the partisan and professional background of monetary policy makers and other senior bureaucrats supports the notion that politicians can through their appointment powers influence the output of institutions that are otherwise removed from their control (Adolph, 2013; Chang, 2003; Lewis, 2008; Neuenkirch & Neumeier, 2013; Vuletin & Zhu, 2011), even though there may be detrimental effects of politicization on the performance of public sector entities (Lewis, 2007).

As already noted by Katz (1986), changes in the relationships between the government executive and the state apparatus have the potential to make policy delegation difficult and thus hamper comprehensive partisan control over the implementation of policy. This concern has recently been reiterated by Mair (2008) who identifies as the primary challenge to party government the rise of the regulatory state and the tendency toward delegation of decision-making powers to nonmajoritarian authorities. These processes remove policy-making capacities from the direct influence of elected politicians and thus undermine one of the core conditions for effective party control over public policy.

The delegation of powers to independent agencies has been most pronounced in regulatory politics. The massive increase in the number of RAs in policy domains such as utilities, competition, financial markets, health care, food safety, consumer protection, or the environment constitutes one of the most profound transformations in the institutional make-up of the public sector during the past decades (Gilardi, 2005b, 2008; Jordana, Levi-Faur, & Fernández-i-Marín, 2011; Levi-Faur, 2005). As Mair (2008) constitutes, it is in this arena “that we see the conditions for the maintenance of party government slipping away.”

The Problem of Credible Commitment

One of the main drivers behind the trend toward the delegation of regulatory policy to independent agencies is the fundamental difficulty to make credible commitments about one’s own behavior in the future.² As elaborated in greater detail by Shepsle (1991), commitments can be either motivationally or imperatively credible. Motivational credibility means that it can be anticipated that the preferences of the actor who commits (e.g., the government) will not change between t (the time the commitment is made) and t + 1 (the time the commitment must be carried out). Such commitments are “incentive compatible and hence self-enforcing” (Shepsle, 1991, p. 247).

In other cases, preferences may be time-inconsistent. When it can be foreseen that an actor’s preferences at t + 1 will diverge from the commitment made at t, credibility can only be established imperatively by limiting the actor’s discretion to deviate from the originally stated policy. If policy makers are believed to have time-inconsistent preferences, other actors will adapt their behavior not to the policy announced by the government at t but to their expectation of what the government’s course of action at t + 1 is going to be (Kydland & Prescott, 1977). As a consequence, the inability to credibly commit diminishes the government’s capacity to effectively set policy.

The problem of credible commitment has been discussed most extensively in the literature on monetary policy (Cukierman, 1992; Keefer & Stasavage, 2003; Rogoff, 1985) where credibility in the imperative sense can be established by delegating the control over interest rates and the money supply to an independent central bank whose preferences are more hawkish than those of the government (Rogoff, 1985). The empirical evidence suggests that in some cases this institutional mechanism helps to keep inflation low. However, the results are quite mixed. The strength of the effect varies considerably across time and space, and the relationship between independence and inflation appears to be stronger for Organisation for Economic Co-operation and Development (OECD) countries than for developing economies (Alesina & Summers, 1993; Berger, de Haan, & Eijffinger, 2001; de Haan & Kooi, 2000; Klomp & de Haan, 2010a, 2010b).³

The difficulty of making credible commitments is not limited to monetary policy, but represents a problem in many spheres of government regulation (Majone, 1997). Environmental standards are ineffective if governments are expected to loosen them in the future (Helm, Hepburn, & Mash, 2003). Firms will engage in irresponsible behavior if they anticipate that governments will, against all prior announcements, prefer to rescue them in case of an emergency (Hardy, 1992; Kornai, 1986; Panageas, 2010; Schaffer, 1989). Moreover, businesses are more likely to invest if they perceive government policy to be backed up by credible law enforcement institutions (Frye, 2004).

This is why using delegation to insulate parts of the state apparatus from direct political interference is not just viewed as a necessary evil to overcome a rational choice problem. Rather, it is conceived of as an important tool in ensuring that governments can deliver sound policies. As a government’s legitimacy is not only a function of its democratically constituted mandate but also depends on the quality of the public policies it delivers (Scharpf, 1970, 1999), creating independent nonmajoritarian institutions may not undermine democratic legitimacy, but strengthen it.

The problem of credible commitment and the related question of policy effectiveness have thus emerged as one of the central theoretical explanations for the “rise of the regulatory state” (Majone, 1994). Over the past decades, a large number of countries have witnessed a substantial increase in the number of independent RAs. The rise of the agency model in government regulation is the institutional manifestation of the recalibration of state-market-relations in the outgoing 20th century—a development that some scholars have termed the advent of “regulatory capitalism” (Levi-Faur, 2005, 2006).

As in the case of independent central banks, it has been widely argued that RAs are created to signal the government’s commitment to a specific policy course of action (Gilardi, 2002, 2008; Majone, 1994, 1997; Thatcher, 2002; Thatcher & Stone Sweet, 2002). Governments that seek to lend credibility to their policies can do so by removing regulatory powers from the direct influence of elected officials and delegating them to agencies placed outside the core bureaucracy. It thus becomes more difficult for governments to deviate from their originally stated policy plans at a later point in time. The formal independence of these regulators thus serves as a commitment device that lends credibility to a certain policy course of action.

The need to create such institutional commitment devices is particularly large in areas where governments suffer from an especially severe lack of credibility. Such variation in “credibility deficits” may depend on features of a specific government (e.g., ideology, see Tavares, 2004) or features of a specific policy domain. After the privatization of utility corporations, for instance, the loyalty of governments may often be perceived to lie with the former state monopolist. If outside actors thus come to believe that they cannot expect to compete with former state-owned enterprises on a level playing field, they will hesitate to invest in that market (Levy & Spiller, 1994, 1996; Spiller, 1993). It is therefore hardly surprising that the most marked trend toward the agency model (and much of the scholarly focus) has been in utility sectors after privatization (Jordana et al., 2011; Levi-Faur, 2003).

Further empirical support for the credibility thesis is presented by Gilardi (2002, 2005a, 2008) who shows that utility regulators are typically granted greater levels of independence than RAs in other economic or social domains. Similarly, Elgie and McMenamin (2005) report that agency autonomy is higher in areas that have recently been subject to market opening. At the European level, Wonka and Rittberger (2010) find greater levels of independence for agencies involved in economic regulation, where the need for credible commitment is especially high.

To be sure, it is largely unclear whether formal independence directly increases agency performance. Yet, once we assume that independence improves credibility and thus enables better performance, the potential conflict between output-related criteria (performance) and input-oriented (democracy) becomes obvious.

Agency Independence and Party Government

The tension between the conditions for party government and the need to establish credibility is profound. Whereas the former requires effective control by governments over the bureaucratic apparatus and policy outputs, the latter demands insulation of regulatory policy from the interference of elected officials. As noted above, scholars of party politics have long recognized the challenges to party government that arise from the increased delegation of authority to arm’s length institutions. But the questions of legitimacy and accountability of independent agencies have also been raised by researchers within the domain of regulatory politics (e.g., Majone, 1998). As Maggetti (2010) argues, delegation outside the parliamentary chain of delegation leads to a “democratic deficit” which may require “new forms of accountability” (see also Maggetti, Ingold, & Varone, 2013). However, rather than theorizing about the normative dimension of this problem, the present article seeks to provide some evidence as to its real-world implications.

To be sure, the extant literature does not provide a great deal of empirical insight into the partisan drivers of RA politics (see, however, Kleibl, 2013; Lewis, 2002). While, at least in Europe, scholars of party politics have pointed out that the “rise of the regulatory state” poses a severe challenge to party government, it must also be recognized that this development is a fundamentally political process, designed and implemented by politicians with strong electoral and partisan interests. It involves a vast array of policy tools, such as the creation, reform, merger, or termination of agencies, the specification of agency tasks and the amount of discretion granted—all of which are certainly subject to partisan considerations.

The empirical focus of the analysis below is on parliamentary democracies in Western Europe. While this perspective means that generalizations to other contexts must be made with care, there are some interesting implications of the case selection. First, West European parliamentary systems are characterized by strong and cohesive political parties, thus providing generally favorable conditions for party government. Second, many of these countries have no significant tradition of arm’s length institutions making regulatory policy, thus lowering the expectations about the impact of regulatory independence on party reach in regulatory institutions.

The analysis examines the degree to which agency independence affects the conditions for party control in regulatory politics. Of course, there are many ways to go about operationalizing these concepts, and a single study can never be exhaustive but needs to focus on a specific empirical phenomenon. The method of choice here is to examine career patterns of top-level agency officials. Using biographical information to track the partisan affiliation of these individuals and existing data on agency independence, the interaction between the formal autonomy of an agency and the informal influence of parties in the form of political appointments can be gauged (for earlier applications of this method, see Dahlström & Niklasson, 2013; Ennser-Jedenastik, 2013, 2014a, 2014b; Lewis, 2008). To be sure, partisan appointments are only one channel through which political influence can be exerted. Indeed, politicians have a large toolkit of mechanisms at their disposal to reign in nonmajoritarian state institutions. Yet, especially in public sector entities that are removed from the bureaucratic chain of command, the appointment channel may be among the most important ones (Chang, 2003; Chappell, Havrilesky, & McGregor, 1993; Falaschetti, 2002; Flinders & Matthews, 2010). Moreover, as evidenced by the discussion above, party patronage constitutes an important feature of party government (Blondel, 2002; Müller, 2000a), and has been gaining relevance as an organizational and governing resource as the public sector becomes more fragmented and decentralized (Kopecký et al., 2012).

While the empirical question of party influence in RAs has not received much systematic scholarly attention, some extant research has tackled a similar question—the link between formal and de facto independence. In an in-depth qualitative comparative analysis of 16 regulatory authorities, Maggetti (2007, 2012) finds only a weak (and highly conditional) relationship between the two. High de-jure independence only explains de facto independence for long-established agencies and in the presence of many veto players. As other causal combinations also lead to high de facto independence, Maggetti (2007) concludes that there is a “disjuncture between formal and de facto independence” (p. 279).

By contrast, Hanretty and Koop (2012b) find that de-jure independence is a statistically significant predictor of “actual” independence. Their analysis of 87 RAs demonstrates that high formal independence decreases the vulnerability of agency executives to government change and reduces their overall turnover rate.

Some further evidence is presented by Thatcher (2005) who analyzes a sample of two dozen agencies in four large European economies to show that elected officials have largely refrained from politicizing management boards, curbing agency powers or resources, and overturning agency decisions. While political interference in these cases seems to be low, it is not clear whether this finding can be attributed to variation in the levels of agency independence.

The present article seeks to add to these analyses by offering a research design with a more explicit focus on the conditions under which parties can influence RAs. To that end, it examines the impact of agency independence and party affiliation as well as the interaction between these two factors on the survival of agency heads. Thus, it can not only be evaluated whether formal independence and party ties have an impact on the career prospects of administrative elites in RAs but also whether institutional mechanisms to establish credibility (i.e., agency independence) can serve to curb the influence of parties in regulatory politics. The analysis thus presents one of the first systematic empirical investigations of the “challenge to party government” (Mair, 2008) posed by the credibility pressures that have led to the removal of regulatory authority from the direct influence of elected politicians. In so doing, the article not only examines the relationship between formal and actual independence of state institutions but also links the literatures on regulatory politics and party government—two strands of research that have both addressed the potential conflict between policy effectiveness and the democratic mandate (albeit from different angles), but have rarely talked to each other.

Empirical Strategy and Data Overview

Gilardi (2008) has thus far provided the most comprehensive publicly available data set on agency independence. His measure draws on earlier indices developed to capture central bank independence (Alesina & Summers, 1993; Cukierman, 1992; Cukierman, Webb, & Neyapti, 1992). It is a composite index comprised of five subindices and takes into account the rules governing the appointment of leading agency personnel, accountability requirements, financial and organizational autonomy, and regulatory competencies.⁴ For the present analysis, only the subindex referring to the appointment and removal of agency heads is used (column A in Gilardi, 2008, p. 144-146), as it is the feature most directly applicable to the data at hand. The index is a measure that incorporates information on the characteristics of the appointer (e.g., ministers, governments, or parliament), term length, the possibility of reappointment, conditions for early dismissal, and incompatibility requirements. As shown by Bianculli, Fernández-i-Marín, and Jordana (2013), these characteristics are among the strongest correlates of the political autonomy of RAs.

The starting point for the data collection is thus the set of agencies covered by Gilardi (2008) whose data include agencies in seven policy domains (competition, financial markets, energy, telecommunications, medicines, food safety, and the environment) across 17 countries (EU-15 plus Norway and Switzerland). As far as possible, data on the independence of agencies that were created more recently and are thus not covered by Gilardi were added by coding laws and statutes.

All agency CEOs serving between January 2000 and November 2013 were recorded along with their entry and exit dates. This information was generated by examining more than 1,000 annual reports, several hundred press releases, as well as a large number of governmental decrees and newspaper accounts. As it proved extremely difficult to obtain valid information on most Greek agencies, the country was dropped from the sample.⁵ Figure 1 plots variation in CEO tenure across countries and regulatory sectors. Chief regulators in the food safety and financial markets sectors have the shortest tenures, whereas agency heads in energy and environmental regulation survive considerably longer in their positions. With regard to cross-national variation, agency CEOs in Southern Europe (Spain, Portugal, Italy) are the most vulnerable, whereas average tenures in Central European and Nordic countries are notably longer. Figure 1 clearly shows that variation across countries and sectors needs to be controlled for in the multivariate analysis.

Figure 1.

Average CEO tenure by sector and country.

In sum, the data set comprises a list of just more than 300 agency executives. The final step was to obtain curricula vitae (CVs) of all individuals and use this information to code their partisan affiliation and other individual-level data. In most cases, such biographical information can be obtained from agency or government websites, media archives, biographical databases, or professional social networking sites (e.g., www-linkedin-com.web.bisu.edu.cn). In the few remaining cases, (former) executives were asked by email to provide CVs. In the end, the number of cases with missing information could be reduced to the low single digits.

The biographical information was used to code the party affiliation of agency CEOs according to the following criteria: (a) having held public political or party office (e.g., minister, member of parliament, party leader), (b) having stood as a party candidate in elections, (c) having worked as aide to party politicians (e.g., in a cabinet ministeriél or as an aide to a parliamentary party group), (d) being a party member, (e) being portrayed in media accounts as “close to a party.” While the latter criterion is less objective, it applies only to a handful of observations. Recoding these cases to the nonpartisan category does not alter any of the substantive conclusions below. Moreover, it is quite likely that these individuals are, in fact, party members whose “true” degree of affiliation is inaccurately observed.

Party affiliation is then combined with information on the cabinet composition in the respective countries to produce two time-varying dummy variables that indicate whether an individual is affiliated to a party in government or in opposition.

One third of all individuals in the data set (105 out of 316) have discernible party ties, with great variation across countries. The large majority of partisans (73 out of 105) are appointed at times during which their parties are in government. Figure 2 plots the percentage of partisans across countries, with affiliation to government and opposition parties at the time of appointment represented by different shades of gray.

Figure 2.

Percentage of partisans appointed to head regulatory agencies, 2000-2013.

Some of these numbers are less surprising, for instance, the high values for Belgium, Spain, France, and Austria, where politicization of the public sector is often assumed to be substantial (Kopecký et al., 2012; Müller, 2006; Page & Wright, 1999). Moreover, the conventional wisdom on the Nordic democracies is borne out by the data. Sweden, Norway, Denmark, and Finland have partisan percentages below average, although there is some interesting variation between the four countries. However, some of the figures are more counterintuitive. Switzerland’s high share of partisans may be due to the fact that the consociational requirement to have all linguistic groups adequately represented favors political (and thus partisan) bargaining over top-level jobs in the public sector. At the lower end, the value of zero for Ireland is somewhat surprising, yet it chimes with recent research that describes public appointments in Ireland as becoming less partisan and more based on personal networks (O’Malley, Quinlan, & Mair, 2012).

Moreover, there is interesting cross-national variation in the extent to which individuals from the government or opposition camp are appointed. Switzerland, Austria, and Portugal display largely majoritarian patterns of partisan appointments (although in the Swiss case this is due to the fact that the four largest parties are almost always represented in the executive), whereas Belgium, France, and Germany exhibit more consociational arrangements that allow for a substantial number of posts to be filled with opposition affiliates.

A first glance at the relationship between agency independence and partisanship is provided by Table 1. The bivariate breakdown suggests that agency independence does not have much influence on the partisanship of appointees.

Table 1.

Party Affiliation by Agency Independence (Column Percentages).

	Level of agency independence
Affiliation	Low	Medium	High	Total
Opposition party	9.3% (10)	7.2% (7)	15.2% (15)	10.6% (32)
No party	71.0% (76)	63.9% (62)	61.6% (61)	65.7% (199)
Government party	19.6% (21)	28.9% (28)	23.2% (23)	23.8% (72)
All partisans	29.1% (31)	36.1% (35)	38.4% (38)	34.4% (104)
Total	100% (107)	100% (97)	100% (99)	100% (303)

Party affiliation coded at date of entry. Agency independence categorized into tertiles. Several observations missing due to limited data availability on agency independence.

Government-affiliated appointees make up 20% of agency heads in agencies with low independence, 29% in moderately independent RAs, and 23% in highly independent authorities. The shares for opposition affiliates are 9%, 7%, and 15%, respectively, suggesting that there is little systematic variation in the appointment patterns that can be linked to partisanship. However, when looking at government and opposition affiliates combined, there seems to be a slight tendency toward partisan appointees in more independent agencies. Overall, partisans account for about one third of all agency heads, yet they make up only 29% of appointees in less-independent agencies and 38% among CEOs in highly independent RAs. Still, even after collapsing the government and opposition categories, a Cramér’s V of 0.09 and a p value of .33 indicate that the relationship is far from significant, both in the substantive and the statistical sense. Agency independence thus does not seem to have much effect on the appointment of partisans, especially when compared with the variation that exists across countries (see Figure 1). This also suggests that governments do not use their powers of appointment to compensate for their lack of direct influence in more independent RAs on a larger scale.

In addition to the data on partisanship and agency independence, the multivariate analysis will control for country and sector effects, variation in the rule of law (following Hanretty & Koop, 2012b) as well as individual-level characteristics such as prior experience in the private and public sector and gender which were all coded from the professional CVs of the agency heads. Table 2 presents the summary statistics for all independent variables. While not all of these variables come with strong expectations as to their impact, it is noncontroversial to assume that CEOs survive longer in countries with higher rule of law standards. Moreover, tenures may be longer for individuals with public sector experience.

Table 2.

Descriptive Statistics for Independent Variables.

Variable	Source	n	M	SD	Minimum	Maximum
Government affiliation	Coded from CVs	1,658	0.22	0.41	0	1
Government affiliation × Agency independence	Interaction	1,658	8.50	18.71	0	84
Opposition affiliation	Coded from CVs	1,658	0.11	0.31	0	1
Opposition affiliation × Agency independence	Interaction	1,658	5.05	16.07	0	84
Agency independence	Gilardi (2008)	1,658	39.26	20.27	0	87
Public sector experience	Coded from CVs	1,635	0.71	0.45	0	1
Private sector experience	Coded from CVs	1,641	0.45	0.50	0	1
Gender (0 = male, 1 = female)	Coded from CVs	1,658	0.16	0.37	0	1
Rule of law (exponentiated)	Kaufmann, Kraay, and Mastruzzi (2009)	1,658	5.27	1.51	1.40	7.69

Several observations missing for prior experience due to incomplete information in CVs. Rule of law index exponentiated to reduce skew of variable. The agency independence index has been multiplied by 100 to allow for easier interpretation of the regression coefficients.

Analysis

To gauge the effects of party ties and agency independence on the survival of agency CEOs, a number of Cox proportional hazards regressions (Cox, 1972) with shared frailties at the country level are specified. Cox models are semiparametric and have become the preferred choice in many analyses of political survival as they require no specific assumptions about the underlying baseline function. Shared frailties can be thought of as an equivalent of random effects. The hazard rate for individual i in country j at time t is thus:

h_{i j} (t) = h_{0} (t) α_{j} e^{(x_{i j} β_{x})},

where h₀ is the unspecified baseline hazard, α is the shared frailty parameter that varies across countries j with mean 1 and variance θ, x is a set of covariates, and β is a vector of regression coefficients.

The analysis also requires the specification of a censoring regime. The models will hence treat all individuals as “failed” if their term ends through circumstances other than death or illness, term or age limits, and promotion to higher office (e.g., to head RAs or networks at the European level). As several covariates change over time (e.g., government/opposition affiliation as a result of cabinet turnover), the data are organized into yearly spells. This results in more than 1,600 observations nested within the 300 appointees.

Before moving to the multivariate models, Figures 3 and 4 report Kaplan–Meier estimates by party affiliation and agency independence. These step curves can be interpreted as probabilities of surviving in office as a function of time. In Figure 3, it becomes apparent that ties to an opposition party increase the risk of removal substantially. Compared with the nonpartisan group of agency heads, however, government affiliates do not seem to benefit. The predominant partisan logic thus appears to be the early removal of “hostile” appointees rather than the active promotion of loyal partisans. With respect to agency independence, the picture is less clear. Low independence seems to lead to somewhat shorter tenure, but this effect is not present at all times. For much of the period of observation, agency independence does not appear to have a large effect.

Figure 3.

Kaplan–Meier estimates by party affiliation.

Figure 4.

Kaplan–Meier estimates by agency independence.

However, this picture changes dramatically in the Cox regressions (see Table 3). The main assumption in these models is that the hazards for any individual do not change over time. For each variable that violates this assumption an interaction with some function of time (typically the natural log) should be included in the regression equation (Box-Steffensmeier & Jones, 2004). Based on a test of Schoenfeld (1982) residuals (Grambsch & Therneau, 1994), the agency independence variable displays nonproportional hazards in all specifications. Therefore, all models also include an interaction term between the agency independence covariate and the log of time. The coefficients for these two variables are highly significant and very similar in size in all models, providing strong evidence that the impact of agency independence is not constant but changes over time. To evaluate this time-dependent effect, Figure 5 plots the joint effect of agency independence and agency independence × ln(time) based on Model V.

Table 3.

Explaining the Survival of Agency Heads in Regulatory Agencies.

	I	II	III	IV	V
Agency independence	1.097***	1.091***	1.096***	1.180***	1.225***
Agency independence	4.27	3.92	3.91	4.18	4.37
Agency independence × ln(time)	0.987***	0.988***	0.987***	0.978***	0.973***
Agency independence × ln(time)	−4.36	−4.19	−4.21	−4.26	−4.45
Government affiliation	1.036	0.906	13.273		2,637**
Government affiliation	0.16	−0.47	0.63		1.99
Government affiliation × ln(time)			0.668		0.338**
Government affiliation × ln(time)			−0.73		−2.04
Government affiliation × Agency independence			0.963		0.859**
Government affiliation × Agency independence			−0.49		−1.97
Government affiliation × Agency independence × ln(time)			1.006		1.021**
Government affiliation × Agency independence × ln(time)			0.56		1.99
Opposition affiliation	1.544*	1.266		1,491,852***	13,389,564***
Opposition affiliation	1.83	0.96		3.74	4.01
Opposition affiliation × ln(time)				0.167***	0.124***
Opposition affiliation × ln(time)				−3.41	−3.72
Opposition affiliation × Agency independence				0.796***	0.758***
Opposition affiliation × Agency independence				−3.56	−3.91
Opposition affiliation × Agency independence × ln(time)				1.029***	1.036***
Opposition affiliation × Agency independence × ln(time)				3.20	3.59
Public sector experience		0.931	0.929	0.932	0.906
Public sector experience		−0.39	−0.40	−0.39	−0.54
Private sector experience		0.830	0.800	0.785	0.745*
Private sector experience		−1.15	−1.35	−1.49	−1.76
Gender (0 = male, 1 = female)		1.128	1.110	1.135	1.130
Gender (0 = male, 1 = female)		0.53	0.46	0.56	0.54
Rule of law (exp.)		0.756***	0.744***	0.756***	0.755***
Rule of law (exp.)		−4.38	−4.88	−4.45	−4.46
Sector dummies	No	Yes	Yes	Yes	Yes
n (subjects)	303	298	298	298	298
n (failures)	187	182	182	182	182
n (spells)	1,658	1,631	1,631	1,631	1,631
θ (estimated frailty variance)	0.147***	0.024	0.018	0.017	0.015

Cell entries are hazard ratios from Cox proportional hazards regressions with shared frailties at the country level; t values in italics.

p < .1. **p < .05. ***p < .01.

Figure 5.

Joint effect of agency independence and agency independence × ln(time).

The joint effect is positive at first, thus indicating that higher independence leads to a greater risk of removal, but turns insignificant after 2.8 years. After 6.5 years, the effect has become negative at the 95% level. Whereas agency independence thus makes CEOs more vulnerable early in their term, this effect becomes small rather quickly and eventually turns into the opposite direction. While the data cannot provide a definite explanation of this finding, the higher levels of vulnerability in the first years may be linked to greater visibility, media exposure, or political scrutiny of highly independent agencies.

Turning to the party affiliation variables, Model I suggests that ties to the opposition have the expected negative effect on survival (larger hazard ratios indicate a higher probability of being removed), whereas government affiliation does not seem to have a great impact. However, the opposition penalty largely disappears when other factors are controlled for in Model II. Once the interaction terms between the party affiliation predictors and agency independence are introduced in Models III to V, it turns out that the effects of these predictors also vary over time (although the effects for government affiliation are practically always zero, see below). The Cox models therefore specify all relevant constituent terms, including two three-way interactions (Brambor, Clark, & Golder, 2006).

As three-way interactions are difficult to interpret from only looking at the regression table, the analysis will draw on graphical representations of how the impact of party affiliation varies with agency independence and over time. As Figure 6 shows, there is some variation in the effect of government affiliation with agency independence and time, but for most of the combinations between these two variables, the effect is not statistically significant. The statistically significant coefficients relating to government ties in Model V are thus a result of including them in the same model with the predictors of opposition affiliation.

Figure 6.

Effect of government affiliation conditional on agency independence and time.

Figure 6 thus confirms the insignificant results reported in Model III and in the Kaplan–Meier plots above (Figure 3). There is thus no statistically discernible difference in the risk of being removed between government affiliates and nonpartisan agency heads.

However, affiliation with a party of the opposition has a substantial impact. A hazard ratio of 1.54 in Model I suggests that, relative to nonpartisan CEOs, the odds of losing office within a specific time period are 54% higher for opposition affiliates. This finding indicates that there is a substantive penalty for people with ties to an opposition party. Further analysis (not reported) indicates that there is no systematic difference between individuals whose party was in opposition at the time of the appointment and those partisans who were affiliated to the government at first but became “hostile” appointees due to a change in the cabinet composition.

The interaction terms in Models IV and V further suggest that the effect of being affiliated with a party of the opposition is larger in less-independent agencies and early in an appointee’s term, and weakens with higher levels of agency independence and over time.

Figure 7 plots the changes in the effect of opposition affiliation conditional on the level of agency independence and time. The graphs clearly show that the effect of opposition affiliation is very pronounced in the first years after appointment and in agencies with low levels of independence (at very high levels of independence there is even a negative effect, meaning lower risk of removal, at first, although this only affects a handful of individuals in the data set). As time goes by, the strength of the opposition penalty and its interaction with agency independence becomes weaker. After 5 years, it has become statistically indistinguishable from zero and remains so at later points in time. Another way to assess the empirical relevance of this effect is to examine over time the levels of agency independence at which the lower bound of the confidence interval hits zero and then look at the share of appointees who are affected. Table 4 presents the respective figures.

Figure 7.

Effect of opposition affiliation conditional on agency independence and time.

Table 4.

Appointees Affected by Opposition Penalty.

t (years after appointment)	Independence level below which opposition penalty applies	Appointees meeting criteria from columns 1 and 2	Of which have opposition affiliation
1	49	188 (63%)	32 (11%)
2	46	164 (55%)	25 (8%)
3	41	111 (37%)	15 (5%)
4	30	44 (15%)	5 (2%)

Percentages refer to overall number of appointees (N = 298).

At t = 1 (1 year after appointment), the opposition penalty applies to agencies with independence levels below 49, encompassing almost two thirds of all appointees in the data set, of which about a sixth have opposition ties. After 2 years, still more than half of all agency heads are in the group of appointees to which the opposition penalty potentially applies (although only 25 of them actually have opposition affiliation). At t = 3, only 111 individuals (about 37%) in agencies with independence levels below 41 are affected, of which 15 are affiliated with an opposition party. After 4 years, the level of independence below which the effect is significant has fallen to 30, thus only affecting 44 appointees of which a mere 5 have opposition ties.

These figures and the graphs in Figure 7 clearly show that opposition affiliation is an important driver of CEO survival—but this applies only to individuals in less-independent agencies during the first 4.5 years after appointment. Model V suggests that, in an agency with average levels of independence (i.e., an index value of about 39), the hazard ratio for opposition affiliation is 3.8 after 1 year, 2.3 after 2 years, and 1.7 after 3 years (all statistically significant), indicating that the increase in the risk of removal for “hostile” CEOs halves between the first and the third year in office. While it is thus true that CEOs with opposition ties are removed earlier than their nonpartisan and government-affiliated counterparts, this is only true in the first years after appointment.

One potential explanation for why the opposition penalty decreases over time may be that opposition-affiliated CEOs realize their comparatively higher hazards and therefore actively engage in reputation building. This, of course, applies more to individuals who were opposition affiliates at the time of their appointment (of which there are quite a few, see Figure 1). Over time, this strategy provides governments with more information about an agency head and his or her abilities. As more information about a CEO accumulates, the impact of the party label may be overridden. Moreover, appointees may anticipate turnover in government and therefore refrain from acting in an overly partisan manner, thus lowering their hazards in the medium to long term.

More importantly for the present purpose, however, higher levels of agency independence award protection to appointees with ties to the “wrong” party. The potential for partisan influence through the appointment channel is thus limited by institutional barriers that make it more difficult to remove administrative elites at will. To the extent that decision making can be affected by appointing and removing agency heads, the establishment of highly independent agencies hence binds the hands of (future) governments to influence public policy.

The empirical evidence from the event history models suggests that, while parties are neither absent nor without influence in the domain of regulatory politics, the conditions for party control through changes in top administrative personnel are diminishing as a consequence of the rise of nonmajoritarian institutions with high levels of institutional protection from political interference (Mair, 2008). To be sure, partisan influence over top-level appointments in RAs is not the same as party control over regulatory policies. Politicians can (and, in practice, do) find other ways to reign in noncooperative agency heads, such as manipulating compensation and budget levels, or changing agency statutes altogether. In many cases, the implicit threat of such measures will be enough to induce compliance. In more extreme cases, agencies may be reformed, merged, or even terminated to make regulatory policy more responsive to government preferences. If agency independence reduces the effectiveness of the appointment channel, it seems plausible that governments will resort to other means to make their preferences heard.⁶

The flip side of this argument is, of course, the proposition that “credibility works.” Making agencies more independent by limiting politicians’ appointment and dismissal powers reduces partisan influence over the careers of top-level regulators can thus be regarded as a potential strategy to mitigate the problem of time-inconsistency. The fact that party affiliation becomes much less relevant as a determinant of career prospects in highly independent agencies can be seen as promoting a healthy incentive structure for agency heads vis-à-vis the sitting government.

A cursory look at the remaining independent variables reveals that none of the individual-level characteristics included in the regression have a substantial effect (although private sector experience is significant at the 10% level in Model V). The hazard ratios for public sector experience as well as gender are close to one (thus indicating a zero effect), and not statistically significant.

Finally, the measure for the rule of law has a significant effect. Individuals in countries with high values on this index survive significantly longer (most of the variation in this variable is between rather than within countries). The three lowest scoring countries on the rule of law index (Italy, Portugal, and Spain) are also those with the shortest average tenure for agency CEOs (see Figure 1), whereas the highest average tenures (north of 9 years) are found in countries that display index values noticeably above average (Austria, Finland, Luxembourg). The rule of law index thus picks up important variation across countries. In fact, it is the inclusion of this variable that renders θ, the estimated frailty variance, insignificant and thus effectively zero in Models II to V. Remember that the frailty parameter α which accounts for cross-national variation in the Cox models is specified to have a mean of one and variance θ. If θ equals zero, all of the relevant variation across countries by the frailty parameter is captured by the independent variables. This is exactly what happens in Models II to V.

As a caveat to the results presented here, it should be noted that some potentially important drivers of CEO survival are difficult (if not impossible) to observe on a large scale and thus not included in the analysis. Chief among them would be the quality of the individual CEO’s performance in office. While proxies such as indicators for agency performance exist for a limited number of cases (e.g., the Global Competition Review’s rating of competition regulators), these only cover a minority of observations in the data set. There are some reasons to assume that politically affiliated appointees may deliver worse performances than other CEOs (Lewis, 2007) and thus be more vulnerable. However, if it were simply the mediocre quality of partisan appointees that would affect their survival in office, we should not see a divide between government and opposition affiliates.

Another relevant predictor of CEO tenure may be the occurrence of crises in certain industries, such as food scandals, environmental disasters, or financial meltdowns. While such shocks can cause severe upheaval in the regulatory landscape, their impact remains difficult to quantify and test systematically within a large-N design.⁷

Finally, as with all observational studies, causal inference is difficult because the statistical relationships identified by the regression models may be driven by a confounding variable. Indeed, it is not difficult to think of third variables that could be driving both, agency independence levels and partisan patterns of appointment and removal, thus rendering the correlations observed in the analysis spurious. One such factor is the overall level of policy conflict in a domain. Low levels of policy conflict may lead to politicians granting agencies higher levels of discretion while at the same time reducing the urge to remove “hostile” appointees. Another potential confounder is the strength of the rule of law. Strong rule of law may cause both, higher levels of agency independence and noninterference in personnel matters by political parties. The online appendix to this article discusses these two threats to causal identification in more detail and demonstrates that neither appears to be a major problem in this case. Specifically, it is shown that variation in policy conflict does not explain CEO turnover, and that agency independence is negatively correlated with the rule of law, thus invalidating the second concern.

Concluding Remarks

The tensions between the requirements for effective party government and the demands for policy credibility are especially obvious in regulatory politics. Highly independent nonmajoritarian institutions in the public sector pose a profound challenge to partisan control over public policy. While the present article cannot do much to move forward the normative debate about the trade-off between party control and regulatory effectiveness, it can contribute empirically by presenting some of the first evidence to show that this tension has real-world implications. Institutional provisions that reduce the discretion of politicians to appoint and remove agency CEOs at will make administrative elites considerably less vulnerable to partisan turnover in government.

Of course, the empirical strategy adopted here does not allow for a direct evaluation of party influence on the actual policies promoted by RAs, neither does it cover multiple other mechanisms by which governments can ensure compliance by regulators. However, the results presented here indicate that it is well worth speculating whether the behavior of agency heads vis-à-vis the government will be conditioned by their vulnerability to political interference—especially in situations where the agency’s and the government’s preferences diverge.

Indeed, if agency independence rules do not only limit partisan influence on matters of personnel but also on matters of policy substance, this will have profound implications for democratic legitimacy. In the words of Fritz Scharpf (1999), it will lead to a decrease in input legitimacy, while providing a potential increase in output legitimacy—provided that agency independence helps create more effective policies and thus greater social welfare. Indeed, one tangible benefit of agency independence that the above analysis highlights is to partly constrain the undue influence of parties and politicians through patronage and politicization.

However, more independent agencies may also be more vulnerable to undue influence by corporate actors, thus increasing the risk of regulatory capture. The move toward greater regulatory independence may thus very well be justified by better policy outcomes and a cleaner governance of the public sector, yet it comes at a cost in terms of democratic control and may, in addition, provide the regulated industries with incentives to capture agencies.

These speculations point to a number of questions worth addressing in future research, ranging from the combined impact of party politics and agency independence on actual agency behavior to the quality of regulatory policy under varying independence regimes and further to the relationship between agency independence and incentives for regulatory capture.

Footnotes

Author’s Note

An earlier version of this article was presented at the workshop “Clientelism and the Quality of Public Policy” at the ECPR Joint Sessions 2014, University of Salamanca, Spain, April 10-15. Data replication materials are available at

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was conducted within the project “Party Government, Patronage, and the Regulatory State” (Grant J 3409-G11), funded by the Austrian Science Fund (FWF).

Notes

Author Biography

Laurenz Ennser-Jedenastik is an assistant professor at the University of Vienna’s Department of Government. His research centers on political parties, their internal organization, behavior, and personnel, with a special focus on the partisan penetration of state institutions. His works have appeared in journals including European Journal of Political Research, West European Politics, Governance, Party Politics, Political Studies, and Regulation & Governance.

References

Adolph

(2013). Bankers, bureaucrats, and central bank politics: The myth of neutrality. Cambridge, UK: Cambridge University Press.

Alesina

Summers

L. H.

(1993). Central bank independence and macroeconomic performance: Some comparative evidence. Journal of Money, Credit and Banking, 25(2), 151-162.

Berger

de Haan

Eijffinger

S. C. W.

(2001). Central bank independence: An update of theory and evidence. Journal of Economic Surveys, 15, 3-40.

Bianculli

Fernández-i-Marín

Jordana

(2013, October). The world of regulatory agencies: Institutional varieties and administrative traditions (Jerusalem Papers in Regulation & Governance, Working Paper No. 58). Retreived from http://regulation.huji.ac.il/papers/jp%2058.pdf

Blondel

(2002). Party government, patronage, and party decline in Western Europe. In Gunther

Montero

J. R.

Linz

J. J.

(Eds.), Political parties: Old concepts and new challenges (pp. 233-256). Oxford, UK: Oxford University Press.

Blondel

Cotta

(Eds.). (1996). Party and government: An inquiry into the relationship between governments and supporting parties in liberal democracies. Basingstoke, UK: Palgrave.

Blondel

Cotta

(Eds.). (2000). The nature of party government. Basingstoke, UK: Macmillan.

Box-Steffensmeier

J. M.

Jones

B. S.

(2004). Event history modeling: A guide for social scientists. Cambridge, UK: Cambridge University Press.

Brambor

Clark

W. R.

Golder

(2006). Understanding interaction models: Improving empirical analyses. Political Analysis, 14, 63-82.

10.

Chang

K. H.

(2003). Appointing central bankers: The politics of monetary policy in the United States and the European monetary union. Cambridge, UK: Cambridge University Press.

11.

Chappell

H. W.

Jr. Havrilesky

T. M.

McGregor

R. R.

(1993). Partisan monetary policies: Presidential influence through the power of appointment. The Quarterly Journal of Economics, 108, 185-218.

12.

Cox

D. R.

(1972). Regression models and life-tables. Journal of the Royal Statistical Society, Series B (Methodological), 34, 187-220.

13.

Cukierman

(1992). Central bank strategy, credibility, and independence: Theory and evidence. Cambridge, MA: MIT Press.

14.

Cukierman

Webb

S. B.

Neyapti

(1992). Measuring the independence of central banks and its effect on policy outcomes. The World Bank Economic Review, 4, 353-398.

15.

Dahlström

Niklasson

(2013). The politics of politicization in Sweden. Public Administration, 91, 891-907. doi:10.1111/j.1467-9299.2012.02113.x

16.

de Haan

Kooi

W. J.

(2000). Does central bank independence really matter? New evidence for developing countries using a new indicator. Journal of Banking & Finance, 24, 643-664.

17.

Elgie

McMenamin

(2005). Credible commitment, political uncertainty or policy complexity? Explaining variations in the independence of non majoritarian institutions in France. British Journal of Political Science, 35, 531-548.

18.

Ennser-Jedenastik

(2013). Party politics and the survival of central bank governors. European Journal of Political Research, 53, 500-519.

19.

Ennser-Jedenastik

(2014a). Political control and managerial survival in state-owned enterprises. Governance, 27, 135-161.

20.

Ennser-Jedenastik

(2014b). The politics of patronage and coalition: How parties allocate managerial posts in state-owned enterprises. Political Studies, 62, 398-417.

21.

Falaschetti

(2002). Does partisan heritage matter? The case of the federal reserve. Journal of Law, Economics, & Organization, 18, 488-510.

22.

Flinders

Matthews

(2010). Think again: Patronage, governance and the smarter state. Policy & Politics, 38, 639-656.

23.

Frye

(2004). Credible commitment and property rights: Evidence from Russia. American Political Science Review, 98, 453-466.

24.

Gilardi

(2002). Policy credibility and delegation to independent regulatory agencies: A comparative empirical analysis. Journal of European Public Policy, 9, 873-893.

25.

Gilardi

(2005a). The formal independence of regulators: A comparison of 17 countries and 7 sectors. Swiss Political Science Review, 11, 139-167.

26.

Gilardi

(2005b). The institutional foundations of regulatory capitalism: The diffusion of independent regulatory agencies in Western Europe. The Annals of the American Academy of Political and Social Science, 598(1), 84-101.

27.

Gilardi

(2008). Delegation in the regulatory state: Independent regulatory agencies in Western Europe. Cheltenham, UK: Edward Elgar.

28.

Golub

Steunenberg

(2007). How time affects EU decision-making. European Union Politics, 8, 555-566.

29.

Grambsch

P. M.

Therneau

T. M.

(1994). Proportional hazards tests and diagnostics based on weighted residuals. Biometrika, 81, 515-526.

30.

Hanretty

Koop

(2012a). Measuring the formal independence of regulatory agencies. Journal of European Public Policy, 19, 198-216.

31.

Hanretty

Koop

(2012b). Shall the law set them free? The formal and actual independence of regulatory agencies. Regulation & Governance, 7, 195-214.

32.

Hardy

D. C.

(1992). Soft budget constraints, firm commitments, and the social safety net. IMF Staff Papers, 39, 310-329.

33.

Helm

Hepburn

Mash

(2003). Credible carbon policy. Oxford Review of Economic Policy, 19, 438-450.

34.

Huber

J. D.

(2000). Delegation to civil servants in parliamentary democracies. European Journal of Political Research, 37, 397-413.

35.

Jordana

Levi-Faur

Fernández-i-Marín

(2011). The global diffusion of regulatory agencies channels of transfer and stages of diffusion. Comparative Political Studies, 44, 1343-1369.

36.

Katz

R. S.

(1986). Party government: A rationalistic conception. In Castles

F. G.

Wildenmann

(Eds.), Visions and realities of party government (pp. 32-71). Berlin, Germany: Walter de Gruyter.

37.

Kaufmann

Kraay

Mastruzzi

(2009). Governance matters VIII: Aggregate and individual governance indicators 1996-2008 (World Bank Policy Research Working Paper Series, No. 4978). Washington DC: World Bank.

38.

Keefer

Stasavage

(2003). The limits of delegation: Veto players, central bank independence, and the credibility of monetary policy. American Political Science Review, 97, 407-423.

39.

Kleibl

(2013). The politics of financial regulatory agency replacement. Journal of Politics, 75, 552-566.

40.

Klomp

de Haan

(2010a). Central bank independence and inflation revisited. Public Choice, 144, 445-457.

41.

Klomp

de Haan

(2010b). Inflation and central bank independence: A meta-regression analysis. Journal of Economic Surveys, 24, 593-621.

42.

Kopecký

Mair

Spirova

(2012). Party patronage and party government in European democracies. Oxford, UK: Oxford University Press.

43.

Kornai

(1986). The soft budget constraint. Kyklos, 39, 3-30.

44.

Kydland

F. E.

Prescott

E. C.

(1977). Rules rather than discretion: The inconsistency of optimal plans. The Journal of Political Economy, 85, 473-491.

45.

Levi-Faur

(2003). The politics of liberalisation: Privatisation and regulation-for-competition in Europe’s and Latin America’s telecoms and electricity industries. European Journal of Political Research, 42, 705-740.

46.

Levi-Faur

(2005). The global diffusion of regulatory capitalism. The Annals of the American Academy of Political and Social Science, 598(1), 12-32.

47.

Levi-Faur

(2006). Regulatory capitalism: The dynamics of change beyond telecoms and electricity. Governance, 19, 497-525.

48.

Levy

Spiller

P. T.

(1994). Institutional foundations of regulatory commitment: A comparative analysis of telecommunications regulation. Journal of Law, Economics, & Organization, 10, 201-246.

49.

Levy

Spiller

P. T.

(1996). Regulations, institutions, and commitment: Comparative studies of telecommunications. Cambridge, UK: Cambridge University Press.

50.

Lewis

D. E.

(2002). The politics of agency termination: Confronting the myth of agency immortality. The Journal of Politics, 64, 89-107.

51.

Lewis

D. E.

(2007). Testing Pendleton’s premise: Do political appointees make worse bureaucrats? Journal of Politics, 69, 1073-1088.

52.

Lewis

D. E.

(2008). The politics of presidential appointments: Political control and bureaucratic performance. Princeton, NJ: Princeton University Press.

53.

Maggetti

(2007). De facto independence after delegation: A fuzzy-set analysis. Regulation & Governance, 1, 271-294.

54.

Maggetti

(2010). Legitimacy and accountability of independent regulatory agencies: A critical review. Living Reviews in Democracy, 2, 1-9.

55.

Maggetti

(2012). Regulation in practice: The de facto independence of regulatory agencies. Colchester, UK: ECPR Press.

56.

Maggetti

Ingold

Varone

(2013). Having your cake and eating it, too: Can regulatory agencies be both independent and accountable? Swiss Political Science Review, 19, 1-25.

57.

Mair

(2008). The challenge to party government. West European Politics, 31, 211-234.

58.

Majone

(1994). The rise of the regulatory state in Europe. West European Politics, 17, 77-101.

59.

Majone

(1997). Independent agencies and the delegation problem: Theoretical and normative dimensions. In Steuenberg

van Vught

(Eds.), Political institutions and public policy (pp. 139-156). Dordrecht, The Netherlands: Kluwer Academic.

60.

Majone

(1998). Europe’s “democratic deficit”: The question of standards. European Law Journal, 4, 5-28.

61.

Majone

(2001). Non majoritarian institutions and the limits of democratic governance: A political transaction-cost approach. Journal of Institutional and Theoretical Economics, 157(1), 57-78.

62.

Müller

W. C.

(2000a). Patronage by national governments. In Blondel

Cotta

(Eds.), The nature of party government (pp. 141-160). Basingstoke, UK: Macmillan.

63.

Müller

W. C.

(2000b). Political parties in parliamentary democracies: Making delegation and accountability work. European Journal of Political Research, 37, 309-333.

64.

Müller

W. C.

(2006). Party patronage and party colonialization of the state. In Katz

R. S.

Crotty

(Eds.), Handbook of party politics (pp. 189-195). London, England: SAGE.

65.

Neuenkirch

Neumeier

(2013). Party affiliation rather than former occupation: The background of central bank governors and its effect on monetary policy (Joint Discussion Paper Series in Economics). Marburg, Germany: Marburg University.

66.

O’Malley

Quinlan

Mair

(2012). Party patronage in Ireland: Changing parameters. In Kopecký

Mair

Spirova

(Eds.), Party patronage and party government in European democracies (pp. 206-249). Oxford, UK: Oxford University Press.

67.

Page

E. C.

Wright

(Eds.). (1999). Bureaucratic elites in Western European states. Oxford, UK: Oxford University Press.

68.

Panageas

(2010). Bailouts, the incentive to manage risk, and financial crises. Journal of Financial Economics, 95, 296-311.

69.

Polillo

Guillén

M. F.

(2005). Globalization pressures and the State: The worldwide spread of central bank independence. American Journal of Sociology, 110, 1764-1802.

70.

Rogoff

(1985). The optimal degree of commitment to an intermediate monetary target. The Quarterly Journal of Economics, 100, 1169-1189.

71.

Rose

(1969). The variability of party government: A theoretical and empirical critique. Political Studies, 17, 413-445.

72.

Rose

(1974). The problem of party government. London, England: Macmillan.

73.

Schaffer

M. E.

(1989). The credible-commitment problem in the center-enterprise relationship. Journal of Comparative Economics, 13, 359-382.

74.

Scharpf

F. W.

(1970). Demokratietheorie zwischen Utopie und Anpassung. [Democratic Theory between Utopia and Adjustment]. Constanz, Germany: Universitätsverlag.

75.

Scharpf

F. W.

(1999). Governing in Europe: Effective and democratic? London, England: Oxford University Press.

76.

Schattschneider

E. E.

(1942). Party government. New York, NY: Rinehart.

77.

Schoenfeld

(1982). Partial residuals for the proportional hazards regression model. Biometrika, 69, 239-241.

78.

Shepsle

K. A.

(1991). Discretion, institutions, and the problem of government commitment. In Bourdieu

Coleman

(Eds.), Social theory for a changing society (pp. 245-265). Boulder, CO: Westview Press.

79.

Spiller

P. T.

(1993). Institutions and regulatory commitment in utilities’ privatization. Industrial and Corporate Change, 2, 387-450.

80.

Tavares

(2004). Does right or left matter? Cabinets, credibility and fiscal adjustments. Journal of Public Economics, 88, 2447-2468.

81.

Thatcher

(2002). Delegation to independent regulatory agencies: Pressures, functions and contextual mediation. West European Politics, 25(1), 125-147.

82.

Thatcher

(2005). The third force? Independent regulatory agencies and elected politicians in Europe. Governance, 18, 347-373.

83.

Thatcher

Stone Sweet

(2002). Theory and practice of delegation to non-majoritarian institutions. West European Politics, 25(1), 1-22.

84.

Thies

M. F.

(2000). On the primacy of party in government: Why legislative parties can survive party decline in the electorate. In Dalton

R. J.

Wattenberg

M. P.

(Eds.), Parties without partisans: Political chance in advanced industrial democracies (pp. 238-257). Oxford, UK: Oxford University Press.

85.

Thomassen

J. J. A.

(1994). Empirical research into political representation: Failing democracy or failing models? In Jennings

M. K.

Mann

T. E.

(Eds.), Elections at home and abroad: Essays in honor of Warren Miller (pp. 237-265). Ann Arbor: Michigan University Press.

86.

Veenendaal

(2013). How democracy functions without parties: The republic of Palau. Party Politics. Advance online publication. doi:10.1177/1354068813509524

87.

Vuletin

Zhu

(2011). Replacing a “disobedient” central bank governor with a “docile” one: A novel measure of central bank independence and its effect on inflation. Journal of Money, Credit and Banking, 43, 1185-1215.

88.

Wonka

Rittberger

(2010). Credibility, complexity and uncertainty: Explaining the institutional independence of 29 EU agencies. West European Politics, 33, 730-752.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.04 MB

0.00 MB