Why Meritocracy May Damage Bureaucratic Quality

The idea of meritocracy as paramount for bureaucratic quality builds on Weber's conception of bureaucracy (see Rauch and Evans Reference Rauch and Evans2000). It is therefore fruitful to restate his criteria of bureaucracy (see Weber Reference Weber1978: 220–21).

1. 1. Bureaucrats are personally free and subject to authority only within a defined area.

2. 2. Offices are hierarchically organized.

3. 3. Each office has a defined sphere of competence.

4. 4. Offices are filled by free contractual relationship.

5. 5. Recruitment of candidates is based on their technical qualifications

6. 6. Bureaucrats receive pensions and employment protection.

7. 7. Bureaucratic office is the sole occupation of the officeholder.

8. 8. Officeholding constitutes a career.

9. 9. Ownership and management of offices are separate.

10. 10. Officials are subject to strict discipline and control.

Some of these criteria have to some extent been loosened in many of today's state bureaucracies,Footnote ⁵ but in criteria 1 through 5, we see the traits of also current meritocratic bureaucracies. These five criteria, as the current literature argues, distinguish meritocracies from other forms of bureaucratic organization in terms of quality. However, important dimensions of criteria 6 through 10 are overlooked. They involve how to ensure bureaucratic compliance with organizational and consequently political goals: pensions and employment protection, prospects of a career, and clear measures of discipline and control should all give incentives for bureaucrats to comply with orders from above. Overlooking these criteria is problematic, but the issue is broader than what Weber's criteria can overcome. As Weber (Reference Weber1978: 224) notes, compliance in fact demands acceptance of political orders beyond what can be obtained by financial benefits, discipline, or control (see also Fukuyama Reference Fukuyama2013; Simon Reference Simon1976). Even the five criteria do not guarantee that it is government rather than bureaucratic policy that will be constructed and implemented.

Based on these considerations, I challenge a core theoretical assumption behind the alleged effects of meritocracy on bureaucratic quality. If meritocracy is to guarantee bureaucratic quality, it must assume that bureaucrats act as mere instruments of government. However, bureaucrats are strategic and self-interested agents. Despite their employment as servants in an ultimately politically controlled hierarchy, they may engage in slack generation, shirking, or sabotage to serve organizational, or even their own personal, interests (see, e.g., Epstein and O'Halloran Reference Epstein and O'Halloran1994).

Attempting to maximize pecuniary rewards is not necessarily the only purpose of such acts. A range of goals can motivate bureaucrats, including societally beneficial outcomes (Downs Reference Downs1967; Gailmard and Patty Reference Gailmard and Patty2007). Ultimately, bureaucrats can harbor convictions that they know better than the government what is good policy for the state, the nation, or society. Such convictions do not have to constitute a large problem. But they can generate unproductive advice and sabotaged implementation when bureaucrats disagree with the government and cannot be brought into line.

What this implies is that meritocracy results in bureaucratic quality only in the presence of some measure of bureaucratic responsiveness to government aims and policies. Responsiveness connotes the willingness of the bureaucrats to serve diverging governments and their policies with equal care. Responsiveness neither equals bureaucrats’ loyalty to Parliament, a wider public, a specific political order or regime, nor loyalty to a particular government. Instead, it is driven by loyalty to any government in its authority as government. This type of loyalty is similar to what David Easton (Reference Easton1975: 437) terms “diffuse support,” something that may be hard to articulate and defend as opposed to specific support but is nevertheless powerful in determining norms of behavior.

Unresponsiveness emerges when the bureaucracy perceives that a government policy or goal opposes its own, third-party, or societal interests. In a given country, unresponsiveness is observed as differences from one policy (or government) to another in the bureaucracy's willingness to serve political orders from above. Unresponsiveness occurs where bureaucrats try to avoid, diverge, or distract policy making or implementation (slack, shirking, or sabotage) and thus hurt the accuracy and swiftness of bureaucratic quality (for a similar definition and application of the concept, see Aberbach and Rockman Reference Aberbach and Rockman1994; West Reference West2005).

Hence, I argue that meritocracy without responsiveness is likely to damage bureaucratic quality—particularly in terms of swiftness and accuracy. But the opposite is equally true: Responsiveness without meritocracy likely damages bureaucratic quality because “blind” adherence to government policy may hurt the bureaucracy's competence and impartiality. Politicized bureaucracies tend to become “blind,” that is, subject to manipulation by the government. This raises the risk of biased and/or incompetent decision making. In other words, there are inherent tensions between bureaucratic values such as obedience, expertise, and principle. Rather than assuming responsiveness from meritocracy, these tensions can only be considered by conceiving of responsiveness and meritocracy as in continuous battle—in the minds of individual politicians and bureaucrats and in political parties and bureaucratic organizations (see ibid.). The next section more formally shows how the conflict between meritocracy and responsiveness unfolds.

Politicization in the Mirror of Unresponsiveness

In realizing the potential damages of meritocracy, politicization is one widely used strategy that selects bureaucrats with the same political convictions. Politicization all in all impairs bureaucratic quality but, to be able to understand the forging of bureaucratic quality, one first needs to recognize that politicization, beyond its function to gain political support, serves the purpose of ensuring that bureaucracies implement government policies in the intended way (see Kopecky et al. Reference Kopecky, Mair and Spirova2012; Peters and Pierre Reference Peters and Pierre2004).

To elaborate on these points, I take inspiration from the PA literature (see Huber and Shipan Reference Huber and Shipan2002). In the PA framework, a principal such as the government delegates authority to an agent, a bureaucrat, due to the complexity of administrating (and changing) society. But this agent has its own interests and information asymmetry of the task at hand, thus increasing the risk that the agent under certain circumstances commits moral hazard. This simple but general problem can be modeled by the concept of zones of acceptance (Simon Reference Simon1976). Consider an abstract case in which two actors interact in a hierarchical relationship. One actor, the government, wants to direct the other actor, the bureaucracy, to faithfully implement its policy. The government and the bureaucracy each have a “zone of acceptance,” that is, a range of policy positions that they consider acceptable. Bureaucratic responsiveness occurs where there is an intersection between the government's and the bureaucracy's zones of acceptance. But if the government policy is outside the bureaucracy's zone of acceptance, disobedience, slack, shirking, or sabotage likely occur as the bureaucracy resists or becomes dissatisfied with the orders it receives. Stated bluntly, bureaucratic quality suffers if political-administrative relations are marked by contention and conflict rather than cooperation.

Monitoring and sanctions can constrain either the process or the outcome of bureaucratic decisions and thus force the bureaucracy to toe the government's line. However, adopting these constraints comes with its own set of problems. First, for some bureaucratic tasks outputs and outcomes are either not visible or not clearly connected, which limits the utility of using controls and sanctions considerably (Wilson Reference Wilson1989). Second, monitoring, sanctioning, and procedural constraints are costly in terms of the agents’ ability to adapt to changing circumstances (e.g., Epstein and O'Halloran Reference Epstein and O'Halloran1994: 701–2; see also Huber and Shipan Reference Huber and Shipan2002: ch. 4). Third, and relatedly, even in situations in which monitoring and sanctions apply perfectly, bureaucrats may rely on rules in an overly rigid fashion by producing exactly what the rules specify without consideration of the details of the individual case or task. As Simon (Reference Simon1976) realized, this is not beneficial to effective fulfilment of goals. Instead, what is often needed is some measure of subordinate acceptance that these goals are worth pursuing meaning that the goals are to some extent internalized and thus subject to the full attention and skill of the bureaucrats.

Provided these arguments, it becomes clear that bureaucratic unresponsiveness is a problem for a political leadership. The ally principle suggests that politicization is one possible remedy to this problem because it ensures obedience by hiring civil servants whose zones of acceptance overlap with that of the government. However, politicization comes at a cost. In a less stylized line of thinking, of course, the government does not replace the whole bureaucracy but replaces managers, department or agency heads, or other key employees, who shape the goals and behavior of their organizations. While this limited politicization strategy is often used to extend the government's overall control into otherwise meritocratic bureaucracies, it merely tends to move the problems of moral hazard and information asymmetry further down the chain of PA relations, now between the politically hired official and the bureaucratic offices she is supposed to manage.

More fundamentally, politicization harms bureaucratic quality for several reasons, even when the zones of acceptance of government and bureaucracy intersect. First, politicization increases the risk that the bureaucracy becomes corrupt as the government has less incentive to oversee and expose it (Dahlström et al. Reference Dahlström, Lapuente and Teorell2012). Second, politicization makes employment in the state bureaucracy less attractive to competent and ambitious careerists (Lewis Reference Lewis2008: 143–45). Third, politicization may demotivate bureaucrats from investing in the acquisition of field specific knowledge and consequently diminishes bureaucratic expertise (Gailmard and Patty Reference Gailmard and Patty2007). Fourth, politicization may permit the government or related interest groups to pursue rents through loyal appointees in the bureaucracy at the cost of the pursuit of more general duties (Miller Reference Miller2000). Finally, the increased job insecurity that politicization entails for the bureaucrats means that they have less incentive to commit to the implementation of long-term projects (Cornell Reference Cornell2014: 195).

Meritocracy and Responsiveness Combined as Solution

Meritocratic recruitment is initiated because, as is well established in the literature, it makes administrations perform better than politicized ones in terms of policy advice and implementation. For a political principal, meritocracy benefits solutions to complex problems and particularly charged issues, such as socioeconomic development and distribution (see Dahlström et al. Reference Dahlström, Lapuente and Teorell2012). But meritocracy also bears political costs that make it hard to escape politicization. I now show why meritocracy needs responsiveness—not only to achieve bureaucratic quality but also to incentivize the establishment of meritocracy in the first place and ensure against rollbacks of meritocratic recruitment rules.

To show this, consider the abstract case in which responsiveness is modeled through the bureaucracy's zones of acceptance. Zones of acceptance are not fixed but can vary in width (Simon Reference Simon1976: 115–17, 130–34). For bureaucrats, wide zones of acceptance mean that they will faithfully assist in policy making and implementation to reach a wide variety of government policies. Consequently, intersections between governments’ and bureaucrats’ zones of acceptance become more likely. Indeed, the acceptance of many government policies is automatic simply because the policies are initiated by the government as the superior political authority—not because the bureaucracy is particularly pragmatic or agnostic when it comes to ideology or policy solutions. In this way, responsiveness prevents problems when government changes: Because responsive bureaucrats see the incumbent government as their ultimate source of authority, they will frequently be as willing to implement the policies of incoming governments as they were willing to implement the policies of its predecessors. This is why responsiveness is central to improve bureaucratic quality but, as indicated, is also why meritocracy is needed to avoid blind adherence to government wishes.

Responsiveness also explains variations in bureaucratic quality within and between countries on a deeper level because it explains the dynamics of contention over meritocratic reforms. Let us begin from the observation that governments considering the adoption of meritocratic systems anticipate that meritocracy may make the goals the bureaucracy pursues divert from their own. In some contexts, the government strategy is then to build a meritocratic bureaucracy with strong organizational or professional interests. Such interests allegedly tie the bureaucracy to serve a given “public interest” (Silberman Reference Silberman1993: 40, 414–15). Yet, even this solution poses challenges to bureaucratic quality because the personnel of these bureaucracies are experts with strong professional opinions that do not necessarily align with government policy. As Weber famously saw, the expertise of professional bureaucrats may divert policy further away from politicians and delegitimize any ideologically motivated counterarguments: “The power position of a fully developed bureaucracy is always great, under normal conditions overtowering” (Weber Reference Weber1978: 991).

If meritocratic recruitment is applied and the bureaucrats are not responsive, politicization becomes attractive. At some point, a government may come into power whose zone of acceptance does not overlap with that of the now permanent bureaucracy. As a result, the government will have reason to roll back meritocratic recruitment, particularly to managerial posts, to permit itself to follow the ally principle and politicize (such rollbacks do occur, see, e.g., Geddes Reference Geddes1991). Rolling back meritocracy is costly for governments because meritocratic, professional bureaucracies often successfully unite to defend their established privileges (Shefter Reference Shefter1994). In turn, reforms may in fact introduce employment protection schemes, which add a shielding of bureaucrats from quick mass layoffs. In some contexts where multiple veto players must agree before reform can be adopted, rolling back meritocracy may be prohibitively costly even for governments facing very large control problems vis-à-vis their bureaucracy. However, in most instances, with less powerful veto players, meritocracy would be rolled back if responsiveness fails. In fact, because rollbacks may be costly, governments may avoid meritocratic reforms in the first place if they fear that their bureaucracies will be unresponsive. Therefore, these reforms often stall and typically only succeed in trial-and-error fashion under substantial political uncertainty (see, e.g., Silberman Reference Silberman1993: chs. 11–12).

When realizing the political worries of unresponsiveness to follow meritocratic reform, it becomes clear that the combination of meritocracy and responsiveness is very hard to achieve but also that forging responsiveness would remove political barriers to initiate and successfully install meritocracy. However, pecuniary incentives only work instrumentally to motivate certain behavior but are highly dependent on year-to-year financing that make them a less durable source of responsiveness (see Olsen Reference Olsen2006: 13). A more effective, and less costly, way of achieving a stable pattern of accurate and swift bureaucratic performance comes through an internalization of norms of behavior. One powerful source that internalizes responsiveness is diffuse support—a sort of historically grounded force that transfers basic legitimacy from one government to the next, even amidst fundamental regime or policy changes (see also Easton Reference Easton1975: 437). For instance, a shared national identity makes it easier for shifting governments to command the loyalty of bureaucrats from different ethnic or socioeconomic strata of society (Englebert Reference Englebert2000; Wimmer Reference Wimmer2013: ch. 5). Responsiveness is more often than not the outcome of such highly complex social processes as nation building unfolding over decades, or even centuries.

Summing up, modeling bureaucratic responsiveness leads to the proposition (1) that a meritocratically recruited but unresponsive bureaucracy is likely to cause low quality in constructing and implementing policies. I thus expect to observe that meritocratically recruited but unresponsive bureaucrats in a given state bureaucracy successfully obstruct government policies—in particular those policies that they perceive to threaten their personal, ideological, and/or institutional interests. I further propose (2) that governments typically value bureaucratic responsiveness just as highly as meritocracy and thus try to roll back meritocratic reforms if bureaucrats are perceived to be unresponsive. From these two propositions, it is clear that responsiveness, even in the presence of meritocracy, is essential in explaining bureaucratic quality.

Bureaucratic Reforms in Late-Seventeenth- and Eighteenth-Century Prussia

I begin in late-seventeenth-century Prussia. At this time, the Hohenzollern kings centralized, refined, and empowered the bureaucracy that became relatively effective—compared to European contemporaries—in terms of public goods provision and tax collection (Fischer and Lundgreen Reference Fischer, Lundgreen and Tilly1975: 510–17). Neither meritocracy nor responsiveness as conceptualized here, but some premature forms of them, formed this bureaucratic quality.

Hiring and firing of bureaucrats was managed politically, by the king personally or by commissions under royal supervision. The Great Elector, Frederick William (1640–88), and his successors, including Frederick II the Great (1740–86), effectively rooted out patrimonial tendencies at the central levels of bureaucracy by recruiting on merit. It was no fully fledged meritocracy, however, because higher-level civil servants were only chosen among noble families and nonnoble but wealthy men (Ertman Reference Ertman1997: 248, 253–54; Fischer and Lundgreen Reference Fischer, Lundgreen and Tilly1975: 521–22).

Meritocracy and responsiveness were relatively well established in most of the eighteenth century, but there was significant variation in responsiveness that concerned the kings. During the reign of at least three kings, the small central bureaucracy was generally unified and obedient to royal policies. This was evident from its success in cost-efficient and well-prepared tax extraction and organization for military conquest in accordance with royal wishes and with minimal corruption. The bureaucracy's militarily organized hierarchy and stable and predictable career paths contributed to this (Braun Reference Braun and Tilly1975: 276–77). Obedience was also nurtured by the Great Elector's “Calvinist-disciplinary revolution from above” connecting symbolically the duty of office and royal service to religiosity (Gorski Reference Gorski2003: 80). Yet, as this bureaucracy excelled in revenue collection and became moneyed and equipped, it developed a self-sufficient attitude very aware that the king was completely dependent upon its effort and expertise (Fischer and Lundgreen Reference Fischer, Lundgreen and Tilly1975: 510–17). So, for Frederick William I (1713–40) the priority was the tying and disciplining of the bureaucracy to serve the crown. In turn, he installed boards to monitor the powerful administrative body, the General Directory (Gorski Reference Gorski2003: 80; Sheehan Reference Sheehan1989: 63).

At the local level, the Hohenzollern kings initially managed to decrease the fusion of economic and administrative powers in the Junker estates (as simultaneously tax collectors and employers) in East Elbia during the seventeenth and eighteenth centuries. However, the Junkers preserved their local-administrative authority and right to nominate candidates for the main office of local bureaucracy, that is, county commissioner in the Landrats (Braun Reference Braun and Tilly1975: 273). In matters of land, agriculture, and taxation, the Junkers’ interests differed substantially from those of the king that led to continuous conflicts over the financing of public goods, most notably a standing army. This concerned how and at what level tax rates should be set and the reach of local autonomy that the Junkers should enjoy in return for extracting taxes (Ertman Reference Ertman1997: 246, 250). Royal reliance on traditional patrimonial institutions thus worried the kings continuously but, given the Junkers’ economic power and position as controllers of peasants, a purging of Junkers was not an option (Sheehan Reference Sheehan1989: 40). PA problems were thus built into the recruitment system because the interests of the agents, the Junkers, differed from those of the principal, the king. The gradual phaseout of Junkers in the administration by university-educated middle-class recruits was a deliberate attempt to solve this PA problem but, as indicated later, only created new issues of controlling professionals.

By the late eighteenth century, although basic reforms were still in the making, the basis of the later popularized “Weberian” bureaucracy was established. However, the organizational strengthening of the bureaucracy and the nurturing of a bureaucratic class had come with their own expenses. From here, the usual presentation of Prussia's bureaucracy as a responsive unit seems particularly misleading.

Bureaucratic quality was, consequently, strained in accuracy and swiftness. By 1780, still under the strict disciplinary eye of Frederick II the Great, the bureaucracy had outgrown the General Directory and become a diversified, professional group with a corporate interest and identity based as it was on recruitment of university candidates. It also acted independently of the aristocracy. The expanding bureaucracy saw the small General Directory as an outdated royal service whose very existence was contradictory to the task of economic modernization and state building. Outright unresponsiveness to the king rarely occurred. However, one prominent example is that the bureaucratic rank managed to force through a codification of rules of merit in recruitment that the king did not want. The king instead wished to maintain hirings and firings at his discretion (Hagen Reference Hagen1976: 42, 116)—a fundamental breach of bureaucratic accuracy in serving government wishes. Further, even though Frederick II insisted on personally viewing all pieces of legislation before enactment, the bureaucracy, due to its expertise, had considerable influence on the framing of problems, their policy solutions, and the implementation thereof to an extent that limited the de facto ability of the king to determine the direction and pace of policy implementation (Dorn Reference Dorn1931: 405, 414). The task of controlling bureaucrats was thus increasingly relevant for the Crown during the latter half of the eighteenth century but also an increasing concern around which serious political disputes circled (Sheehan Reference Sheehan1989: 142).

Bureaucratic Authoritarianism in Nineteenth-Century Prussia

When Frederick II died in 1786 and the lesser capable Frederick William II (1786–97) inherited the throne, the bureaucracy started dominating policy making almost completely (Rosenberg Reference Rosenberg1958: 94). Exploiting the weakness of the king, the bureaucracy managed to establish itself as a class of professionals, the Beamtenstand, in the Allgemeine Landrecht of 1794 that guaranteed tenure by law and limited the formally absolutist, policy-making authority of the Crown. But to many among the bourgeois, university-educated bureaucratic elites, the larger goal was to bring the Prussian state and society in tune with the spirit of the French Revolution—that is, to abolish serfdom and the monarchy as such. This led to continued clashes with the king in the Law Commission—the main policy-making organ (Gray 1986: 42–43). Once again, the king had trouble determining policy directions and implementation pace.

Only a decade later, in 1807, Prussia's defeat in the battle of Jena to Napoleon's France forced Frederick William III (1797–1840) to install an emergency ministry led by the early agitators of constitutional reform, Stein and Hardenberg. While the king gave them wide-ranging authority to reform the political and administrative system, he did not foresee that Stein and Hardenberg would exploit this window of opportunity to strengthen bureaucracy at the expense of the Crown and initiate what has been termed the era of “bureaucratic absolutism” (Beck Reference Beck1995: 125–27). Thus, all state offices were opened to competition based on merit, and the General Directory was dissolved and replaced by ministries (Gillis Reference Gillis1971: 6–7, 11). Whereas the threat to Prussia's survival initially gave the opportunity for bureaucracy to take decisive power in 1807 (Sheehan Reference Sheehan1989: 70), this does not explain the emergence of the Allgemeine Landrecht before 1807. Instead, as indicated, the expertise and autonomy of bureaucrats resulted in much lower bureaucratic quality than expected from a meritocratic system.

The continued bureaucratic dominance after 1807 was also primarily driven by the power of a meritocratic and autonomous bureaucracy. Until 1848, the increasingly meritocratic bureaucracy routinely circumvented royal wishes, weakening the accuracy and swiftness of bureaucratic quality. Policy making in this period was simply dominated by the bureaucracy, even when policies contradicted the will of the monarch, for instance in the abolition of serfdom and initiation of free movement of paupers that initially worried the king on behalf of his powerbase among the Junkers (Beck Reference Beck1995: 1, 152, 157, 168; Eddie Reference Eddie2013: 2–3, 10, 16). The king, despite his still formidable formal powers in legislation and execution, could only passively observe the implementation of unwanted initiatives, not least those that eroded royal power (Rosenberg Reference Rosenberg1958: ch. 9). Bureaucratic unresponsiveness went hand in hand with the nurturing of the bureaucracy's veto player status in policy making and implementation. This was the stuff of what had basically developed into a bureaucratic authoritarian regime (ibid.).

The year 1848 was a watershed for Europe's as well as Prussia's autocrats. In the mid-nineteenth century, the Beamten had come to see itself as “Plato's Guardian Class” (Gillis Reference Gillis1971: 16), deeply engaged with and sure of their prerogative in policy making by virtue of their power of knowledge and organization. In and around the time of liberal revolution in 1848, this proved to insulate the bureaucracy from political changes at the expense of royal power. Specifically, in the period leading up to 1848 the bureaucracy distorted the execution of royal orders and thus contributed to the demise of the Crown. From the 1830s, notably liberally oriented bureaucrats started to oppose the conservative governance paradigm that ensured bureaucracy's power under alleged, though limited, guidance from a supposedly enlightened monarch. This was not only a source of intrabureaucratic conflict but also a source of unresponsiveness to the monarchical ambitions of Frederick William III. The mobilization by liberal bureaucrats of progressive groups for enfranchisement of the middle class and a revolution of the economic order in Prussia and other German states (Sheehan Reference Sheehan1989: 633–34) illustrates a clear and serious divergence between the zones of acceptance of a group of bureaucrats and the king that harmed the swiftness and accuracy of bureaucratic quality considerably.

When revolution approached in 1848, the conservative bureaucratic class closer to the new king, Frederick William IV (1840–61), managed to preserve its power at the center of political decision making by accepting a liberal constitution that ensured gradual reform without altering the basic identity of bureaucracy (Gillis Reference Gillis1971: 172; Mommsen Reference Mommsen1991: 79). Importantly, the bureaucracy knew the king's interest in preventing further constitutional limits on his powers but did not respect it. Instead, the bureaucracy was more concerned with its own survival than supporting royal directives. Despite notable reforms aimed at politicizing bureaucratic careers during the 1840s and 1850s, the bureaucracy only increased its autonomy and power as a politically vital group—much to the king's regret (Sheehan Reference Sheehan1989: 719–20). Against the predictions of meritocracy, the divergent zones of acceptance of the bureaucracy and the king led to bureaucratic unresponsiveness to royal wishes that harmed bureaucratic quality in fundamental matters of civil-service reform and regime traits.

Aversion of Challenges to Bureaucratic Power in Imperial Germany

The Prussian Foreign Minister Bismarck unified the German states through the 1860s. As a result, the Prussian model of state bureaucracy was adopted as model for the German Empire. The spreading of meritocracy went easily because all German states shared legacies of relatively strong meritocratic bureaucracies (Ziblatt Reference Ziblatt2006: 113–15).

Perhaps surprisingly, given the typical assessment of Bismarck as a strong chancellor, the autonomy and power of bureaucracy was only consolidated further under Bismarck (1871–90) (Gillis Reference Gillis1971: 172; Mommsen Reference Mommsen1991: 79). But Bismarck was still highly concerned about issues of political control of the bureaucracy that otherwise was so successful. In turn, he managed to some extent to make bureaucracy toe his line through strict enforcement of disciplinary laws. These were helped along by his political alliance with the Junkers and a common political-bureaucratic concern for the welfare of the nation-state and the forging of heavy industry (Bonham Reference Bonham1983: 650; Caplan Reference Caplan1979: 306). Greater responsiveness was thus obtained but based on monitoring and sanctioning and temporary loyalty to specific policies.

As expected, such responsiveness did not survive government changes. After Bismarck's removal in 1890, the German Imperial bureaucracy remained extremely influential in policy making, but its zone of acceptance did not always overlap with that of shifting conservative chancellors from the Bismarckian through the Wilhelmine era. This resulted in increasing implementation problems such as in designing and executing social protection schemes where government proposals were sometimes delayed or circumvented in bureaucratic agencies (see, e.g., Mann Reference Mann and Ritter1985: 85–86; Mommsen Reference Mommsen1991: 79).

Alongside conflicts between chancellors and bureaucrats, the new challenge to bureaucratic power from Parliament grew and bureaucracy became a more heterogenous organization. But even the more liberal or socialist bureaucrats subordinated themselves to bureaucratic interests. The line of conflict between chancellors and the bureaucracy thus continued. To keep opposition in bureaucracy in check, chancellors from 1890 to 1914 institutionalized in the Reichstag a group of so-called political bureaucrats (politische beamte) whose function was to ensure loyal adherence to conservative policies in policy making and implementation.

This politicization solution to a problem of unresponsiveness was, however, limited in success. Because of constitutional guarantees of bureaucratic autonomy, the politische beamte of the bureaucracy were highly limited in number and power. Moreover, even the politische beamte could hardly be manipulated with by chancellors. These bureaucrats were educated in universities and affected by the ethos of constitutional protection that dominated bureaucratic offices more generally (Gillis Reference Gillis1971: 33; Mann Reference Mann and Ritter1985: 85–86). As loyalty to the chancellor would often imply an ambiguous stance on constitutional provisions, given the hybrid nature of the regime and parliamentary powers, these bureaucrats most often became disillusioned with politics rather than active political assets for the governments.

This pattern only crystallized further from the first years of the twentieth century when the social democratic SPD (Sozialdemokratische Partei Deutschlands) increased its mandates in the Reichstag and later, in 1912, became Germany's most popular party (Mommsen Reference Mommsen1991: 79). Until 1918, SPD was more revolutionary than reformist, and its radical proposals to reform and democratize bureaucracy and strip the king and chancellor of unchecked executive power were tried frequently in the early twentieth century but forcefully circumvented by higher-level bureaucrats in the ministries who, at this point, defended the emperor-chancellor power conundrum in which their own powers were vested (Caplan Reference Caplan1979: 308; Koch Reference Koch1984: 146). Also, SPD was initially against social insurance to workers but nevertheless adopted by Bismarck and expanded by an increasingly activist bureaucracy after 1890 (Steinmetz Reference Steinmetz1993: 126–27). Unresponsiveness and, consequently, low bureaucratic quality were thus particularly strong when bureaucracy was the target of the policy.

Summing up on the Imperial German period from 1848 to 1918, problems of bureaucratic quality were, in contrast to predictions of meritocracy, confined to the periods when the bureaucracy acted most autonomously from government pressure. Bureaucratic quality waxed and waned not primarily with the level of meritocracy but mostly with the alignment of governmental and bureaucratic policy positions. It is characteristic that implementation of economic policies was generally very effective. But civil-service and constitutional reforms were sabotaged and mired in considerable political-bureaucratic conflicts and welfare policies were dominated by bureaucratic interests against significant resistance among parliamentary actors with legislative prerogatives. Consequently, responsiveness remained a major concern for governments but attempts at politicization such as the installation of politische beamte proved unsuccessful because of the bureaucracy's professionalism and veto player status.

At the brink of revolution in 1918–19, what were then the prospects of bureaucratic responsiveness to the coming democratic-republican system?

Bureaucratic Power Preservation in Democratic Weimar Germany

Let us look at three of the most salient issues to assess bureaucratic quality in Weimar Germany: civil-service reform, socioeconomic policies, and law and order. First, civil-service reform negotiations in 1918–19 in many ways were a replay of former political-bureaucratic interaction despite the unusual circumstance of military defeat in World War I. SPD and DDP (Deutsche Demokratische Partei) opted for a radical break with the past by installing genuine social democracy and democratizing the civil service. They did not trust the old monarchical elite bureaucrats and sought to purge them. But they were met with considerable resistance from a unified Beamtenstand, including those bureaucrats with no attachment to the previous monarchy. The Beamtenstand was further supported by the politische beamten in the conservative party, DVP (Deutsche Volkspartei), and higher-level servants in the ministries (Runge Reference Runge1965: 36–38). Furthermore, SDP and DDP quickly realized that a rebuilding of Germany after the war required the old bureaucracy and its expertise (Bockenforde Reference Bockenforde, Jeserich, Pohl and von Unruh1985: 15–16). As a result, the constitution confirmed bureaucrats’ protected status and privileges as installed in the constitution of 1848. Changes to the meritocratic personnel system were sabotaged and delayed, even when SPD governments wanted change.

Some reforms in the early 1920s cut back on salary and rooted out remaining social privileges within medium- and low-level bureaucratic ranks. However, because of bureaucratic resistance several SPD governments failed to implement policies aimed at a democratization of civil service. The bureaucracy organized a trade union, DBB (Deutsche Beamtenbund), with the clear purpose of protecting civil-service interests against parliamentary politics and working-class ideology (Caplan Reference Caplan1988: 59–61). Also, unorganized lower-level civil servants simply circumvented reforms that, for instance, would have changed criteria for career advancement. Otherwise, department heads used their intimate connection with conservative parliamentarians to stall further reform policies (Runge Reference Runge1965: 119). Thus, when the core interests of bureaucrats were at stake, they were both willing and able to sabotage clear government orders.

Cutbacks and rationalization only grew substantial after several efforts and more strict enforcement from the late 1920s and from 1930 under Chancellor Heinrich Brüning (Caplan Reference Caplan1988: 76–77). As expected given the importance of responsiveness, this served to break down the confidence between the bureaucracy and the government institution.

Comparing across periods, the explanation of the failed civil-service reform attempts during the Weimar Republic is strikingly similar to that of the eighteenth- and nineteenth-century monarchs’ insufficient attempts at rolling back the 1794 code: a combination of a political-functional needed to have strong bureaucratic organizations (especially in the immediate aftermath of World War I) and the veto player strength of bureaucracy. The only difference in Weimar and Imperial Germany as opposed to early-nineteenth-century Prussia was that the bureaucracy's power was further empowered by a nationwide corporate organization that served bureaucratic interests.

Second, in contrast to civil-service reform proposals, fiscal and monetary management in Weimar Germany was generally very effective despite the difficult economic circumstances. Some examples of sabotaged implementation did occur, however. For instance, the most radical socialist reforms were tacitly circumvented in the ministries (Mäding Reference Mäding, Jeserich, Pohl and von Unruh1985: 96). An example is the implementation of the eight-hour working day, which was effective at first but gradually hollowed out as the responsible ministries remained passive when employers let working hours creep up to their traditionally high levels (McElligott Reference McElligott2014: 79).

As is shown by Müller's (Reference Müller2014) analysis of the Ministry of Economics, not all ministries were opposed to socialist or Keynesian policies. Nonetheless, even the labor ministry, dominated by socialists and trade unions, was taken over by conservative bureaucrats from 1924 (Liu Reference Liu1997: 363). This augmented the general tendency of the bureaucracy to guide social change toward prewar economic policies benefitting heavy industry as opposed to workers and farmers. This continued across the period of hyperinflation until 1924, the financial stabilization from 1924 to 1929, and the unemployment crisis from 1929 (Mommsen Reference Mommsen1991: 83, 86, 90, 100, 111–12; Petzina Reference Petzina, Jeserich, Pohl and von Unruh1985: 46, 55–56).

Third, judges were among the oldest, most conservative of the bureaucrats, and most unresponsive. True, the judicial sphere was by law relatively more insulated from politics than the remaining bureaucracy, but judges were still obliged to follow the democratic spirit of the Weimar constitution. Nevertheless, many of them saw the shifting governments inherent in a democracy as threatening the integrity of their vocation. Therefore, their court rulings became increasingly biased toward conservative values. This was characteristic of the Länder jurisdiction where Junkers still dominated and practiced regular breaches of administrative protocols. Relatedly, when terrorism appeared in the 1920s, judges were notorious for protecting right-wingers (such as Freikorps and the Sturmabteilung) as opposed to left-wingers (McElligott Reference McElligott2014: 172).Footnote ¹¹ Judges also deliberately used their rulings and public deliberations to undermine the symbols of the republic (Craig Reference Craig1978: 420). In turn, rulings became instances of unresponsiveness to certain (social democratic) governments. Judges also allied with the Beamtenbund in their attempt to protect bureaucrats’ professional autonomy—especially as they were pressured on salary (McElligott Reference McElligott2014: 100–3).

Summing up, the Prussian and German cases largely corroborate the propositions that (1) a meritocratically recruited but unresponsive bureaucracy is ineffective in constructing and implementing policies, and that (2) governments try to roll back meritocratic reforms when bureaucrats are perceived to be unresponsive. Bureaucratic quality was harmed in the periods where unresponsiveness was most pronounced. This problem did not rise from politicization but instead increased continuously over time as bureaucracy became more meritocratic and grew in professionalism and autonomy.

As this suggests, the analysis allows me to go one step deeper. In explaining the level of responsiveness in Prussia and Germany, my cases indicate the importance of diffuse support, in particular national identity. Responsiveness was a primary concern for kings, chancellors, semidemocratic governments, and fully democratic governments alike—and thus across the whole period of analysis. Specifically, it was the meritocratically oriented 1794 code and the successful 1807 reforms that limited the political and economic power of the political executive vis-à-vis bureaucracy. Kings and subsequent governments were thus trapped in a vicious circle whereby political control of bureaucracy through politicization would cause elite defection (under the more autocratic regimes) and public dissatisfaction (under the more democratic regimes). Because responsiveness never reached very high levels after a genuine bureaucratic organization was built in the mid-eighteenth century, responsiveness seemed to lose out if meritocracy was in place. This period also corresponds with the centuries of aggressive and conflicting nation building in the German states. In accordance with classical studies in nation-state formation (see Wimmer Reference Wimmer2013: 114–15), it took the complete defeat of Hitler's Nazi regime in World War II—a major critical juncture—to decisively purge the old bureaucratic elites and reform the German bureaucratic ethos to one of public service through loyalty to shifting governments in the German Federal Republic (Crawford et al. Reference Crawford, Brady, Wiliarty, Brady, Crawford and Wiliarty1999; Mommsen Reference Mommsen1991).