Public administration in authoritarian regimes: propositions for comparative research

ABSTRACT

Authoritarian regimes share an architecture of power that shores up unilateral control from the top while weakening bottom-up initiatives organised independently of central support. This kind of collective choice environment entails administrative options and policy strategies unanticipated by standard theories that explain public administration where the exercise of power is circumscribed by the separation of powers and political representation. How public administration in authoritarian regimes is shaped by radically different patterns of elite power play and state-society interactions is set out in ten propositions, with a focus on why governance solutions that seem inadequate or even impracticable in liberal democracies would be favoured and how variations in leadership style can cascade into broader disparities in administrative behaviour. The conclusion offers an example of how public policy theories originally developed to account for administrative processes and outcomes in the US and other democracies can be comparatively adapted to non-democratic systems.

KEYWORDS:

• Public administration
• public policy
• authoritarian regimes
• comparative analysis
• governance

Scholars of public administration have shown a growing recognition towards the traditions and practices of non-Western governments as a legitimate topic of research. Part of the shift stems from a broader reflection on the field’s “Western centricity”, which it needs to address through “a deeper understanding of how and why administrative practices differ across regions” (van der Wal et al., 2021, p. 295). Calling the pattern a legacy of “colonial imposition and post-colonial imitation”, Haque (2019) similarly advocates for “more comprehensive, culturally nuanced research” (p. 110). Constrained by historical biases, existing frameworks often lack the conceptual and evidentiary foundations to explain public administration outside of liberal democracies. It is unlikely that current accounts of public engagement can reveal the logic of citizen participation in regimes where the scope of policy discussion and public participation is highly circumscribed (Truex, 2016) and preferences routinely falsified (Kuran, 1997), nor can theoretical descriptions of bureaucratic control where political powers are institutionally separated and regularly transferred extend to regimes where power sharing is limited to cofaction allies and reshaped only through political purges (Svolik, 2009). Current models that assume institutional facilities for political representation and civic action are unlikely to explain policy advocacy well where collective action is stymied (Fu, 2017).

Overcoming these limitations requires nesting public administration in the broader constitutional order of authoritarian regimes. For political scientists, how dictatorial power is accrued, maintained, and exercised forms the central focus of theories of authoritarianism. A large body of scholarship deals with how autocrats co-opt challengers, contain factional rivalry, and censor regime critics (Art, 2012). Public administration research is dominated by topics pertaining to the operative and functional aspects of governing, such as budgetary allocations, bureaucratic oversight, rulemaking and implementation, and civil service performance. Because of the historical split in focus, major theoretical frameworks of public administration draw almost exclusively from liberal democracies and lack some of the theoretical vocabulary for direct application to authoritarian regimes.

Theory development that closes this research gap will have practical implications in the wake of global democratic backsliding. In China, the legacy of collective leadership is being reshaped under President Xi’s “strongman” leadership. His government has undertaken major reforms to raise bureaucratic compliance and responsiveness to top-down intervention (e.g., Chan et al., 2021). The political playbook is also being rewritten in countries going through similar changes. In Hungary and Turkey, populist leaders staged rapid takeover of the administrative apparatus and purged government institutions once controlled by rivals (Böhmelt et al., 2019), while elites in Venezuela and Cuba now have to contend with post-transition politics (Anselmi, 2017; Muno & Briceño, 2023). These changes will affect what kind of advice public administration researchers can give to leaders and the bureaucrats they control and, subsequently, how much their advice can rely on a body of scholarship based mostly on empirical findings from democracies.

In response to these developments, researchers start to apply standard theories of public administration to non-democratic contexts (e.g., W. Li & Weible, 2021; Gromping & Teets, 2023). In a special issue of this journal last year, scholars of public administration emphasise the need to broaden the scope of research on authoritarian regimes beyond the ruling elites (Peters, 2023). The administrative apparatus is a primary target for leaders seeking to swiftly consolidate power (Muno & Briceño, 2023). They also report cross-regime variations in bureaucratic performance (Knox & Janenova, 2023), policy transfer patterns (Duong, 2023), and infringement on citizen rights (Yerramsetti, 2023). At the same time, as Ripoll and Rode (2023) show in a study of public service motivation, the impact of authoritarian rule can be at times more modest then commonly assumed. This study builds on these efforts and offers ten propositions on the nature of public administration in authoritarian regimes.

Public administration and authoritarian politics

Authoritarian regimes differ from the democratic varieties in how political power is shared among the ruling elite and exercised to foster social control (e.g., Guliyev, 2011; Lührmann et al., 2018). How power is shared within the ruling elite sets the terms for elite competition, coalition building, and other high-stakes manoeuvres within the political establishment. In this regard, comparativists argue that authoritarian rule necessarily entails few institutional constraints on the leader, but that also means weaker deterrents against opportunistic behaviour, making authoritarian regimes more vulnerable to bouts of extreme instability (Tsebelis, 2002) and falsified support a primary preoccupation for their leaders (Egorov & Sonin, 2011). Authoritarian regimes also set themselves apart with regards to the centrality of coercion in social control. A mix of repression and grassroots mobilisation helps the regimes manage social forces, direct public opinion, and rein in collective action.

Authoritarian leaders are typically portrayed as capable of mobilising administrative personnel and resources on a scale unimaginable to leaders of democracies. However, they are often strategically constrained to rely on collective leadership and elite co-optation to neutralise potential challenges and stabilise elite support (de Mesquita et al., 2005; Svolik, 2009). Power sharing in the authoritarian setting is complicated by opportunistic behaviour among actors. Elite partners anticipate the autocrat to defect when the tactical advantages the deal originally confers no longer matter, while the autocrat expects them to switch allegiances when offered better terms by challengers (Egorov & Sonin, 2011). Given these expectations, unilateral interventions from the top would intensify the survival dilemma for actors at all levels of the hierarchy, augmenting risk exposure and motivating inefficient choices. Because autocrats rely on a different set of strategies to deal with challengers and shape state-society relationships, different patterns of public administration unanticipated by standard theories can be observed.

In comparative politics, extensive theoretical effort has gone into characterising the nature of authoritarian power, with various definitions settling on the dynamics of elite power play as the key dimension that sets regimes apart. The distinction matters to comparative public administration because the payoffs for public administrators and other actors are intimately tied to, as the first five propositions argue, the pattern of elite power play. For example, bureaucrats under co-optative leaders tend to prioritise domain-specific goals, even if it means not fully aligning with central mandates. Under coercive leaders, bureaucrats use discretion sparingly and focus on central instructions, even when doing so deprioritizes concerns specific to their home policy domain. Comparative research should explain what bureaucrats and other policy actors do in response to the different features and varieties of authoritarian rule and how their choices in turn shape policy agendas, organisational competencies, and other public administrative outcomes.

The next five propositions turn to how the regime’s interactions with society shape public administration. Vehicles of state-society interactions in liberal democracies – competitive elections, policy advocacy, citizen engagement – are either absent or significantly reconstituted in authoritarian states, which instead rely heavily on top-down mobilisation and coercion to enforce policies (Lei, 2018). Whether that entails correspondingly drastic differences in administrative processes and outcomes is unclear. For example, opposition to current policy is typically underrepresented in authoritarian regimes (Wintrobe, 1998), which some argue could cause political leaders to act conservatively in favour of the policy status quo and make the system highly unresponsive (e.g., Lam & Chan, 2015). Yet, studies also show that authoritarian regimes can be responsive to social and political demands (Tsai, 2007; Weiss, 2014), especially in cases of strategic adaptation of representative and participatory institutions (Magaloni, 2006; Manion, 2015; Smyth et al., 2019; Svolik, 2009). Similar mechanisms linking authoritarian social control to public administration could be better specified and their effects ascertained.

Elite power play

In democratic governments, policy actors shield themselves from the uncertainty of electoral contests, leadership change, and institutional competition by expanding autonomy through reputational engineering (Carpenter, 2001; Carpenter & Krause, 2012), playing one institutional rival off against another (Gailmard, 2009, Hammond & Knott, 1996; Kim, 2008), and building subsystems insulated from external influence (Baumgartner & Jones, 1993; Sabatier & Jenkins-Smith, 1993; Wood, 1988). These strategies are likely to be ineffective when decision-making becomes highly centralised and dissent is actively repressed (Zhan & Zhu, 2022). Policy actors are dependent on central leaders in managing complex policies (van der Kamp, 2021), but what leaders can do is limited by other factors, such as built-in bias against information that does not conform with current beliefs (Kuran, 1997). The first set of propositions deals with the effects of elite power play on public administration in authoritarian regimes.

Police-patrol oversight is more practicable than third-party oversight in authoritarian regimes

Autocrats ward off opportunistic challengers by demonstrating their ability to root out opposition, censor views that conflict with their own, and neutralise other threats to regime legitimacy and survival. These actions encourage preference falsification, delaying the detection of imminent threats until they become too big to ignore. Considerable progress has been made in political science to explain how authoritarian leaders wrestle with the problem by expanding information supply through controlled representation, limited citizen participation, and selective media censorship (e.g., Manion, 2015; Lorentzen, 2013). These strategies coincide with the research on agency problems in government bureaucracy. While leaders have the authority to set the policy agenda, the practical knowledge bureaucrats have on the ground creates wiggle room for the latter to manipulate the former. The wider the information gap becomes, the harder it is to detect budget maximisation, shirking, enforcement distortion, mission drift, and other forms of administrative non-compliance (Kerwin & Furlong, 2018).

The literature has identified two general strategies to address information-induced non-compliance. Oversight agencies perform regular audits and performance evaluations to obtain information that agents may deliberately conceal, as well as ad hoc investigations directed at other activities. However, top-down oversight could lead to second-order agency problems and often incur significant organisational costs. Alternatives to top-down oversight require procedures that make third-party engagement integral to rulemaking (Potoski & Woods, 2001). Third-party participation empowers interest groups, individual citizens, and other stakeholders to report instances where bureaucrats fail to implement programmes in accordance with the leader’s preferences (Epstein & O’Halloran, 1995).

Scholars of bureaucratic oversight have argued that third-party oversight is more robust than top-down oversight. Their account explains why political principals push for procedures that mandate public comments and other provisions for stakeholder engagement in programme development and implementation. These arrangements create new access points to information that is otherwise restricted to the bureaucratic insiders, but doing so without incurring significant information retrieval and processing costs. Their account also rationalises the perceived underdevelopment of top-down mechanisms in the presence of third-party monitors as an efficient alternative. Theory development has been dominated by the questions about the organisational designs of third-party oversight (McCubbins & Schwartz, 1984) and the susceptibility of the arrangement to strategic misrepresentation of information (Epstein & O’Halloran, 1995; Potoski & Woods, 2001).

Autocrats are unlikely to realise these benefits using third-party oversight. Far from being a counterweight to state authority, policy actors lack the political independence and organisational capacity to challenge government policy. Threatened with repression, they steer clear of issues over which the regime has strong preferences (Stern & Hassid, 2012). They also “atomize” collective demands, whereby issues that concern the wider community would be deliberately presented as grievances affecting individual citizens alone (Fu, 2017). Rather than the independent, highly mobilised actors that they need to become in order to make an impact in democracies (Mattingly, 2019; Jamal, 2009), non-state actors in authoritarian regimes are systematically absorbed into the government apparatus for surveillance and control. The absence of powerful and independent non-state actors makes top-down inspections that the traditional literature often considers inefficient, underutilised, and therefore unimportant for in-depth theoretical inquiry the only practicable option. More empirical work is required on what autocratic leaders do to rein in the well-documented inefficiencies of top-down oversight and assess compliance increases as intended. An important aspect of this proposition is the extent that asymmetric information affords a measure of advantage to noncompliant bureaucrats despite the efforts to close the gap.

In more coercive forms of authoritarianism, bureaucrats converge on a smaller number of policy issues, leaving the government’s agenda less diversified

Scholars of authoritarian politics report notable variations in the pattern of elite bargains across regimes (Svolik, 2009). In some cases, the autocrat leads only as “first among equals” and takes care of a broad range of interests scattered across the political landscape. To avoid destabilising their coalition, autocrats act in close consultation with the ruling elite. Precautions of this nature are seldom observed in regimes where the autocrat has been firmly established. Compared to autocrats who work co-optatively, established autocrats make decisions independently and enforce them unilaterally, even if that disrupts the policy status quo favoured by other actors (van der Kamp 2012). Established autocrats are also motivated to demonstrate their capacity to manufacture support and drive potential sceptics into self-censorship (Wedeen, 2015).

New research can specify the risks these differences in elite bargains bring to administrators, the strategies they come up in response, and how their responses aggregate into system-level divergence in administrative outcomes. Direct interference is limited under co-optative leaders, allowing bureaucrats to expect broad instructions rather than interference over programme details. In comparison, leaders of established regimes exercise power unilaterally, driven by the fact that their ability to do so is closely watched by potential challengers mulling revolt as well as allies calculating their moves when autocrats become vulnerable (Egorov & Sonin, 2011; Kuran, 1997). As “strongmen” leaders take a direct interest in whether their priorities are followed through on the ground, administrators expect frequent interference by a watchful leader whose reaction to perceived non-compliance can be swift and severe. To avoid being seen as taking independent actions, administrators emulate one another to achieve safety in numbers (Chen & Jia, 2023).

As the proposition suggests, bureaucrats expect less direct intervention under “steward” leaders and tend to focus on priorities within their own policy domains. Although this may foster parochialism, a higher degree of specialisation within specific domains would in turn enhance cross-domain diversification. Under “strongman” leaders, administrators experience heightened pressure to align their actions with the overarching policy vision articulated by the leadership. The bureaucrat’s greater attention towards central priorities and the autocrat’s greater willingness to interfere will undercut domain-specific focus and unify priorities across domains.

Co-optative regimes use trustee-type bargains with the civil service, while coercive regimes use agent-type bargains

According to Hood and Lodge (2006), trustee-style bargains entail a decoupling of career outcomes of public administrators from their commitment to the political principal and their policy agenda, whereas these conditions are tightly connected when the principal directly determines career outcomes in agent-style bargains. Transplanting their framework to authoritarian systems, how exposed the autocrat is to elite challenges could similarly determine how civil service bargains should be structured. Under contested leaders, bureaucrats will exercise a high degree of de facto autonomy not conventionally expected of authoritarian regimes but nonetheless realisable due to the long organisational distance created by the power sharing arrangements that stabilise elite support. Autocrats who manage to consolidate power and rise above factional contestation will find their access to the levers of administrative power uninhibited by fragmented control. A shift in focus from delegation to compliance follows, in response to which bureaucrats will steer away from activities that can be misconstrued as quiet resistance. Linking the pattern of principal-agent interactions to the dynamic of authoritarian elite politics could raise generalisability, as Burns et al. (2013) suggest in an application of Hood and Lodge to colonial Hong Kong.

The proposition extends to other areas. In public policy research, the traditional understanding has been that government processes evolve gradually over time. However, recent research has challenged this notion by consistently revealing that structural redundancies resulting from the separation of powers can give rise to sudden policy shifts amidst prolonged periods of inactivity (Jones & Baumgartner, 2005). While discussions regarding the underlying factors remain inconclusive, replication studies indicate that this pattern of “punctuated” change may be more pronounced in authoritarian regimes because of information undersupply (Chan & Zhao, 2016; Baumgartner et al., 2018). Future work may focus on whether the established autocrat’s preference for tighter bureaucratic control through agent-based bargains – which is likely to further escalate the problem of information undersupply – makes their governments even less capable of churning out incremental changes despite a having higher level of structural centralisation, which scholars have identified as a primary source of “punctuation” in democratic regimes.

At a higher level of abstraction, the contrast between established and contested autocrats captures the divergent logics of coordination in large hierarchies. The costs associated with coordination tend to increase under contested autocrats. The administrative system adapts to fragmented control by distributing larger grants of discretion and empowering bottom-up initiatives, producing elevated levels of disjointedness in government priorities and unevenness in the rates of policy change. With more unified control, the established autocrat can set government priorities centrally, take direct charge of complex initiatives, and make enforcement against non-compliance credible. As risk aversion spreads, bureaucrats become collectively more dependent on top-down coordination.

The way administrative behaviour changes with elite dynamics in authoritarian regimes is comparable to how public administrators react to presidential and legislative intervention in democracies (Moe, 1995). In both cases, what political leaders can do to direct bureaucratic activities is bounded by institutions, be they formal constitutional arrangements specifying the terms of power separation across government branches or informal deals carving out organisational territories split among competing factions. Just as the notion of presidential and legislative bureaucracy helps clarify the sources of structural diversity in government bureaucracy, how the organisational logic of authoritarian power shapes bureaucratic behaviour carries similarly generalisable implications.

The emphasis on political loyalty in personnel decision-making undermines administrative competence

Political leaders rely on competent public administrators to identify and define problems, develop solutions, execute programmes, and perform other important functions of government. Loyalty is another highly-valued attribute. Administrators driven by different values and therefore seen as politically disloyal may direct government attention and resources to other issues, adopt policy programmes in support of rival agendas, and collaborate with competing interests outside of the establishment. To political leaders, how loyal bureaucrats are will decide how much they can rely on them to achieve critical policy goals, which in turn will decide whether they can cater to the policy demands of and therefore secure the continued political support from their mass support base, as well as party elders, major donors, and other members of the selectorate.

Compared to democratic leaders, authoritarian leaders may have unusually powerful incentives to prioritise loyalty. Comparative research on this issue has focused on how the autocrat’s preoccupation with disloyalty may interfere with political selection based on functional competence and how these priorities weigh against each other as context changes. Comparative research could assess the effectiveness of different strategies authoritarian regimes adopt to de-bias personnel decisions. While leaders in democracies also prefer administrators who are politically aligned (Wood & Waterman, 1991), the intensity of their preference is weaker and they also face greater constraint over personnel choices. Without institutional safeguards to prevent purges and coups, autocrats are in constant fear of overthrow by capable administrators, who are just as worried about persecution on misjudged intentions. The problem is escalated by the reduced visibility of political preferences. Policy positions are also easier to observe in democracies, whereas political loyalty is enmeshed in informal affinities and often deliberately falsified in authoritarian regimes (Pan & Chen, 2018), leading to costlier solutions to the loyalty trade-off.

Comparative public administration could start with new propositions about personnel management, acknowledging the prevalence of the tension between loyalty and competence in many decision-making processes in authoritarian regimes. Whereas the field is traditionally focused on functional merits such as operational efficiency and policy responsiveness, political allegiance may take a much more prominent role in the personnel arrangements of authoritarian regimes. That is not to say authoritarian leaders do not care about competence, but great limitations attend their pursuit of meritocracy. How the clash of priorities is reflected in administrative institutions and the extent functional goals are crowded out by political concerns could be indicated by questions such as how government administrators are recruited and inducted and how administrators are vetted for senior appointments. The instability not only means that political leaders struggle to effectively reward functional competence and punish misconduct, it also means that bureaucrats must tread carefully between shifting priorities.

New propositions should guide theoretical discussion of the way bureaucrats react to the autocrat’s conflicted preferences. In particular, they may draw attention to situations where objective measures of competence, rather than traits suggestive of loyalty, becomes more salient in personnel management. In some situations, authoritarian leaders prioritise bureaucratic competence in the hopes of improving governance results, such as more robust policy enforcement and stronger responsiveness to frontline needs. The loyalty criterion may also apply unevenly. Bureaucrats at the lower rungs of the hierarchy may be evaluated on measures of competence, but those located closer to the top would still find loyalty the primary determinant for career success (Landry et al., 2018). In other situations, autocrats may favour officials who are incompetent, building what Shih (2022) calls “coalitions of the weak” to shore up their own survival. The interplay driven by the loyalty-competence trade-off has direct implications for oversight and how leaders keep bureaucrats compliant.

Authoritarian regimes artificially inflate the level of organizational fragmentation of the government bureaucracy to foster top-down control

Administrative processes in authoritarian regimes are often assumed to be highly centralised both structurally and operationally. Structurally, hierarchical control entails extensive rules specifying how power is distributed in the administrative process. These rules would also give political leaders the levers required to monitor and intervene in government activities at all levels. Operationally, this results in policy being devised and implemented under the direction of senior leaders, who do so without the constraints of formal institutions that mediate or separate power typical of democratic systems. The emphasis on structural and operational centralisation sets the authoritarian bureaucracy apart from liberal democracies, where bureaucrats often possess institutionally protected grants of discretion, which they use to make decisions in pursuit of professional values or policy mandates set by previous administrations, even if they may stand in opposition to the preferences signalled by the current leaders. The high levels of bureaucratic autonomy realised under these provisions are incongruent with the political ethos of authoritarian regimes (e.g., W. F. Lam, 2005; Lee, 1999).

At the same time, comparativists have long argued that authoritarian regimes are more fragmented – and the policy choices they make more inchoate – than is commonly suggested. According to Lampton (1987), the authoritarian bureaucracy is “best understood as a bargaining, not a ‘command’, system” defined by “protracted haggling, posturing, conditional outcomes, frequent failure to reach agreement, and issues that rise and fall on the agenda, but are rarely fully resolved or discarded” (p. 12). The rhetoric of unquestioning loyalty and intolerance for dissent obscures the fact that bureaucrats are part of “a conglomerate of large organizations” where actors “differ substantially about what their government should do” and “compete in attempting to affect both government decisions and the actions of their government” (Allison & Halperin, 1972, p. 42). These conditions have been variously fashioned as “fragmented” (Lampton, 1987; Lieberthal & Oksenberg, 1988), “pluralized” (Mertha, 2009), “decentralized” (Landry, 2008), and “steered” (Schubert & Alpermann, 2019) authoritarianism, representing a shared focus on the structural “paradoxes” that generate conflict, opposition, and entrepreneurship rather than shoring up absolute control (C. Li, 2012).

Even more surprising is the observation that the organisational disarticulation could be entirely intentional (Shih, 2022). Authoritarian regimes may actively inflate internal rivalry in government processes to prevent complex policy actions from being coordinated from the bottom up. As the system grows more fragmented, the autocrat becomes the only actor powerful enough to cut through turf wars and strategic bargaining to direct ambitious policy initiatives from the top down. Another source of fragmentation originates with the use of tournament-style competition to drive policy performance. Tying performance metrics to career outcomes, the practice was instrumental to the long-term focus on economic growth among Chinese administrators, who are now responding to the central government’s new initiative to improve sustainability (H. Li & Zhou, 2005; Yao & Zhang, 2015). An important side-effect is the intense competition between rival factions and even among officials from the same faction (T. Chen & Hong, 2021; Fang et al., 2022). In light of Terry Moe’s classic discussion of the “congressional bureaucracy” in the US, structure-induced disarticulation in authoritarian governments could point to the very different political underpinnings of an otherwise similar phenomenon.

The degree of institutional centralisation can vary significantly from one government to another, representing different preferences for the degree of control and government responsiveness. More centralised systems delegate less to maximise top-down control, while more decentralised systems delegate more, enabling independent actions. These choices can fundamentally change how bureaucrats behave, and by implication how policy challenges can be handled at the systemic level. Given similar trade-offs, authoritarian bureaucracies may converge to a different set of institutional formulae that existing frameworks cannot sufficiently capture. Centralised control is not only a matter of giving authoritarian leaders more power and making them more difficult to criticise and challenge, but it also depends on disabling activities that could interfere with top-down intervention. The extent authoritarian institutions are deliberately fragmented – a condition that is likely to be far more widespread than commonly assumed – and how the fragmentation changes administrative practices and outcomes should form a new focus for comparative research.

State-society interactions

Many existing theories of public administration are implicitly set in an environment where power is fragmented by design. They describe how the surge of public attention during “focal events” (Birkland, 1997) or “policy windows” (Kingdon, 1984) is exploited to destabilise the status quo; how policy coalitions divided by opposing beliefs take advantage of checks and balances to sustain their turf in the policy process (Jenkins-Smith et al., 1991; Sabatier & Weible, 2014); and how actors create their preferred narratives and drive out alternatives favoured by rivals (Shanahan et al., 2011). These accounts explain how policy actors exploit democratic institutions and are unlikely to extend to behaviours in the shadow of authoritarian rule. The propositions below concern the latter, where administrative processes and outcomes arise in the absence of the structural and organisational factors highlighted in the existing literature.

Policy advocates favor strategies that minimize confrontation and display greater deference to the government’s position

Advocacy is considered an important catalyst for policy change in the standard literature (Kingdon, 1984). Entrepreneurs recruit like-minded stakeholders as allies and construct alternative narratives that challenge existing beliefs (Birkland, 1998; Stone, 1989), and they move between policymaking venues in search of sympathy and support (Pralle, 2006). Those who oppose change and prefer the policy status quo can counter these activities by portraying the issue in contention as expert-driven, excluding dissenting views from the deliberations within the policy subsystem (Howlett et al., 1995). With their emphasis on checks and balances, stakeholder engagement, and policy participation, democratic institutions are designed to support and even motivate these entrepreneurial manoeuvres. This holds true even when these manoeuvres run against the policy preferences of the majority or the interests of dominant elites.

While policy entrepreneurs in authoritarian regimes can be no less enterprising, what they do to influence policy must be understood in the context of strong aversion to bottom-up mobilisation and challenges to the regime. Advocates align their agenda with the regime’s and steer clear from confrontational tactics. They coordinate with those in power, build ties with established interests, and pursue limited change seen as consistent with current policy (Mattingly, 2020; Stern & Hassid, 2012). In China, LGBT advocacy exemplifies this approach. While advocating for the removal of homosexuality as a mental order designation and other hostile practices (Parkin, 2018), they have done so largely without grassroots mobilisation. Instead, they focus on partnerships with the government in the provision of AIDS and counselling services, framing their activities as part of the government’s public health agenda (Kong et al., 2021). Many strategies drawing on the gay rights movement are unlikely to be favoured, especially when these tactics entail open challenges to current policy or the social values that the regime is expected to preserve. On the other hand, representation within bounds, though often ineffective in the highly contentious environment of democratic systems, could drive concrete if slow-paced concessions (Truex, 2016). Policy advocacy in authoritarian regimes may be characterised by much less direct confrontation and more measured efforts to persuade and convince (e.g., Fu, 2017). Identifying these strategies and mapping them to the constraints of authoritarian politics would make an important contribution (Gromping & Teets, 2023).

The primacy of political contention in advocacy will be challenged in comparative public administration. Longstanding scepticism for liberal democratic politics has led to a troubled relationship with causes associated with collective action and social mobilisation, such as environmental protection and civil rights. The same political fears that led government officials to target collective action at an early stage (J. Chen et al., 2016) also drive them to favour groups that have the incentive to comply (H. Li et al., 2017). In the case of China, non-state actors adapt in a fashion reflected by the transformation of many civil society groups from the rights based advocacy to more service oriented practice as government crackdown intensifies in recent years (H. Li et al., 2017). The pattern of selective partnership replicates in government procurement, service delivery, philanthropy, and other areas where the state seeks input from external actors. Tactical adaptation is also observed among actors aiming to engage the state. Rather than mobilising collective protests with aggrieved citizens, they strategically avoid direct confrontation with the regime and guide individuals to present their claims as individual appeals for redress (Fu, 2017).

Policy participation expands access and representation in authoritarian regimes, but excludes citizen empowerment and disables collective action

In authoritarian regimes, government control extends directly to individual citizens through grassroots organisations designed to monitor and mobilise them. Under Stalin, millions of farmers were grouped into large grassroots organisations called the kolkhozes, which were instrumental to the Soviet Union’s drive to break up traditional land ownership in favour of collectivised management of farmland. In China, health authorities relied on neighbourhood committees (jiedao ban) to coordinate lockdowns of entire cities and administer screening tests to millions during the COVID-19 pandemic. Using grassroots units to mobilise entire segments of society for initiatives often too disruptive to the existing socio-economic order and at tremendous personal cost to individual citizens for democratic governments to undertake is a hallmark of authoritarian rule.

The prominent role of mass participation in authoritarian regimes poses challenges for the application of traditional participation theory. The earliest proponents of citizen participation saw the practice as a solution to the democratic deficit and spoke of it as an innovative attempt to “transform” (Wampler, 2012) and “deepen” (Fung & Olin Wright, 2001) democracy, turning citizens from the victims of manipulation into active policy partners through empowerment beyond the ballot box (Arnstein, 1969). From urban planning and participatory budgeting, effort to expand participation has contributed to community consensus, policy inclusiveness, and legitimacy (Fung, 2006). Some of these benefits are being realised in authoritarian regimes, where officials adopt participatory practices to legitimate decisions (Balla, 2014; Li et al., 2012). They also use engagement to galvanise support (Lo & Leung, 2000) and coordinate stakeholder actions (J. Li et al., 2020). However, these benefits participation confers would not justify to authoritarian leaders any transformative change to the power relations between state and society in the way the concept was originally envisioned.

As authoritarians do not seek to empower citizens but to deepen their control infrastructure, comparative researchers may consider how the legacy of mass mobilisation constrains how policy participation is organised in authoritarian regimes, and the extent officials have been successful in achieving more support and legitimacy while sustaining even greater top-down control. Citizen empowerment is unlikely to feature prominently when adapted to such a political context. Not only is the notion at odds with absolute state control, but citizens may also be accustomed to government-directed participation rather than bargain with the authorities. Policy advocates also show a level of self-restraint atypical of citizen participation for democratic politics, preferring discretion and eschewing open confrontation. Common models of civic engagement involve self-governance, partnership, and strategic bargaining between citizens and officials (Fung, 2009), all of which are presumably irrelevant to if not incompatible with authoritarian rule.

Policy narrative devices that exploit conflict and confrontation are actively avoided in authoritarian regimes

The observation that democratic institutions mediate public choice and policy discourse is central to a suite of theoretical accounts about agenda setting. Kingdon’s multiple-stream framework is one of the first to refer to the strategic deployment of problem definitions as an exercise in policy entrepreneurship. Subsequent models further specify the rhetorical devices entrepreneurs use to present policy preferences. Advocates who consider themselves the “losing side” associate a “diffusion of costs” and a “concentration of benefits” with the status quo to justify their demand for change by showing how their preferred policy will spread benefits more evenly (McBeth et al., 2007). Advocates seeking change will also link the issue to a larger debate beyond the subsystem where they are usually handled (Nie, 2003; McBeth et al., 2007). To defend current policy and the preferences that it embodies, those on the “winning side” come up with competing narratives that underline the scientific certainty of the policy status quo (Leong, 2015).

Policy advocates construct elaborate policy narratives in which they cast themselves as “heroes” deserving trust or “victims” deserving sympathy but their opponents as “villains” whose actions must be curtailed to achieve more equitable distribution of policy benefits and costs (Heikkila et al. 2014; McBeth et al. 2007). They also develop “causal stories” to specify the mechanisms contributing to problematic conditions and justify policy actions aligned with their preferences. These observations are strewn across different models that make sense of agenda change as the competition between strategically developed “narratives” or “causal stories” encapsulating core preferences. How policy agenda evolve in democracies appears to be partly driven by how policy preferences can be expressed.

In authoritarian regimes, state control of communications leaves preference articulation heavily circumscribed (Huang, 2015), potentially diminishing the extent agenda change can be attributed to policy narratives. Contemporary regimes adopt sophisticated messaging techniques incorporating “many of the benefits of free media”, such as enabling “investigative reporting on lower-level officialdom” (Lorentzen, 2013, p. 402), into a governance framework that gives the media censors the capacity to detect and curb the dissemination of sensitive information (King et al., 2013). Similar developments are documented in social media, where technical advancements have enabled authoritarian regimes “to shape the contours of public discussion and to cheaply gather information about falsified public preferences” (Gunitsky, 2015; also Stockmann & Gallagher, 2011; Xu, 2021). These changes have forced advocates to adapt. Compared to the early emphasis on how entrepreneurs in authoritarian regimes are constrained and how that leads to extended inertia in the policy process (e.g., Lo & Leung, 2000), recent research underscores how they switch between assertiveness and acquiescence, occasionally achieving significant increase in responsiveness and representation in policy through “tactical contention” (Gromping & Teets, 2023). Future research should map these alternative strategies and shed light on their effectiveness.

The capacity of authoritarian regimes to capture policy intelligence is excessive for known problems, but inadequate for pre-empting emerging, previously unknown challenges

Both democratic and authoritarian governments can be subjected to serious survival threats for failing to anticipate and manage disruptive events (Rosenthal & Kouzmin, 1997). To pre-empt natural disasters like extreme weather, wildfires, and earthquakes as well as more complex and causally ambiguous events such as social unrest and public health emergencies, governments establish sophisticated bureaucracies to collect and organise vast amounts of relevant intelligence. As authorities are increasingly capable of using new information and communication technologies to conduct large-scale monitoring and feed real-time updates to predictive algorithms (Unlu et al., 2010), policy intelligence work has reconstituted itself from data collection to information processing. That includes shifting the focus from how to collect data to what conditions to prioritise in intelligence work and then establishing how the conditions being monitored contribute to or inhibit the disruptive phenomena. Often, these activities are embedded in an organisational architecture that directs the information flow within and across agencies, where alternative sources of information compete with one another and a variety of roles exists in intelligence work (Wilensky, 1963). How information is circulated and validated in these processes will decide how robust the overall policy system is. The policy system is more robust when alternative causal theories can be discovered and adopted when the current model no longer delivers accurate predictions. Policy intelligence requires deep domain expertise to decipher (Athey, 2017) and the ability to extract reliable information from actors who may strategically distort or withhold what they know (Manion, 2015). A policy system that prioritises political correctness over domain knowledge or curtails the spread of information that runs counter to prevalent beliefs, thus amplifying the incentive to act deceptively, is likely to be also less robust.

Referring to the case of China, some have characterised the combination of information technology with authoritarian politics as a unique innovation in digital and social surveillance unmatched in scale and sophistication by Western practices (Mattingly, 2020; Roberts, 2018), but democratic governments also invest significantly into advanced surveillance technologies to track activities they consider threatening to public security (Sloan & Warner, 2016). More critical is what governments can do with the data they acquire. In authoritarian regimes, the bureaucratic apparatus operates within a much tighter structure for centralised command, and political institutions offer relatively limited protection to the targets of surveillance. These give rise to intelligence practices hard to adopt elsewhere due to privacy concerns and institutional rivalry (Liu & Zhao, 2021). With higher levels of centralisation, data brought in from disparate places can be internally integrated given sufficient support from the top (Sagarik et al., 2018).

Authoritarian regimes have exceptionally powerful incentive to invest in intelligence concerning threats they already prioritised. Paid and voluntary agents are mobilised to police local neighbourhoods, and government censors sift through public and private messages for discussions deemed to be problematic by the security apparatus (Pfaff, 2001; Roberts, 2018). In China, the regime not only targets regime critics who promote anti-regime views, it has also acquired capabilities to locate discontent among ordinary citizens (King et al., 2013). The sustained efforts to expand surveillance in this direction have raised the effectiveness of grassroots control, to the extent that policymakers felt secure enough to cut back on welfare provisions previously thought essential to survival (Xu, 2021). Apart from political disaffection, similar efforts to spot adverse developments early on can be observed for other prioritised risks, such as those in public health (Ding & Lin, 2021) and finance (Gruin, 2019).

This informational advantage conferred by the architecture of authoritarian surveillance could turn into a limitation when it comes to handling risks that fall outside existing problem definitions. In political systems that permit open debates on perspectives that diverge from the dominant policy beliefs and on issues that are marginal to the policy agenda, unfamiliar problems can be frequently brought up. In authoritarian regimes, the limits on political scepticism translate into a systemic underappreciation for risks not already given priority. They may also struggle to address the so-called “wicked problems”, which the public policy literature refers to problems that do not fit standard definitions. The closed information environment that authoritarian regimes actively maintain does not conduce to the pluralistic discourse prescribed for effective solution search in such cases. In other words, authoritarian intelligence systems may be optimised for well-defined problems, but they perform poorly when given more complex tasks, such as problem discovery.

Authoritarian regimes pursue sustainable growth, digitalization, and social inclusion without embracing decentralized and collaborative forms of governance

Political leaders increasingly look to sustainable development, digital transformation, and social inclusion as the new developmental goals, and governance strategies have consequently evolved in support of their efforts to deliver these goals. Rejecting past paternalistic practices (Arnstein, 1969), new approaches emphasise collaborative governance (Ansell & Gash, 2008; Bingham et al., 2005), citizen engagement (Fung, 2006), and distributed decision-making (Olsson et al., 2004). These strategies promote citizen empowerment (Baiocchi, 2005; Fung, 2006), state-society partnership (Salamon, 2002), and joined-up actions (Agranoff, 2006; Hood, 2005). Although authoritarian regimes also embrace sustainability and digitisation (Zhang & Barr, 2013), they are challenged to pivot from governance strategies that the standard literature finds outdated to a new style of public administration where the government “rows” less but “steers” more (Salamon, 2000).

Several lines of inquiry will shape comparative research on this issue. First, the literature should evaluate whether the “blunt force” measures regimes use to curb non-compliance can also stimulate transition towards the new developmental goals. In China, the central government has moved away from GDP growth to support sustainable development in recent years (Rauch & Chi, 2010). This was followed by an abrupt policy change, whereby regional governments are now mobilised to de-emphasise growth-enhancing measures and restrict polluting businesses (van der Kamp 2012), often at significant economic costs (de Groot & Schuitema, 2012). Such a process would not resonate with the general notion of “green politics”, which counts grassroots activism, electoral opposition, and policy advocacy as important drivers of change (Burchell, 2002). The involvement of policy advocates in social movements (Ku, 2010) or partisan competition (P. E. Lam, 1999) deviates from the authoritarian norm, making it difficult to explain how the policy status quo is overturned in favour of the green agenda.

The extent to which market and societal forces operate similarly in non-democratic societies as they do in democracies remains to be assessed. Contemporary developmental goals are understood to be contingent on changes in the broader economic structure (Vachon et al., 2006) and social values (Reilly & Hynan, 2014), rather than state coercion alone (O’Brien & Li, 1999). When governance strategies are modified to accommodate authoritarian rule, reaction from critical stakeholders can vary. While opening up the offshore bond market to overseas investors, Chinese regulators retain the power to suspend any bond issuance, a rare practice that does not appear to deter market participants (Ji, 2022). In contrast, potential investors in China’s offshore Renminbi markets react poorly to the high level of central control imposed by regulators due to the involvement of a more diverse set of regulatory agencies and the competing agendas they represent (Sohn, 2015). Policy stakeholders sometimes accept the state’s formal ability to act unilaterally, but such tolerance takes a long time and special incentives to emerge. Whether the regime’s preference for centralised control can be accommodated to the satisfaction of key policy participants will decide the long-term prospects of authoritarian competitiveness in various domains of future economic and technological development.

These concerns apply to Web3.0 technologies. A prime example is the distributed ledger (blockchain) technology, which has been deployed in the creation of cryptocurrencies and other virtual assets, enabling digitally connected actors to engage in anonymous and “trustless” transactions in the absence of central oversight. These developments have led to drastic policy reactions in China, such as a complete ban of cryptocurrency trades in the country. But the measures leave room for the selective adoption of emerging technologies where control by the state can be advanced, as seen with China’s introduction of the e-CNY digital currency. If widely implemented, this could greatly enhance state control over household and individual spending. Whether authoritarian regimes can adapt their governance strategies – many of which are assumed obsolete by the current literature – to the era of digital innovation and sustainable growth will be one of the biggest questions for comparative public administration to answer.

Concluding notes: pathways towards comparative public administration

However universal the need to deliver effective administration and good governance is, differences in the public choice and power architecture will force policymakers in democratic and authoritarian governments to adopt solutions that seem alien, inadequate, or simply impracticable to one another. My propositions spell out the implication for comparative research from several angles. First, policy tools and practices proven effective for one type of political regimes may become less suitable or even counterproductive for others. For example, oversight strategies that rely on stakeholders to actively expose and report government failure – activities that are commonplace enough in liberal democracies – are unlikely to function well when the public discourse is widely censored. At the same time, advocates who avoid conflict with the government and minimise public attention about their work will get immediately drowned out in the saturated agenda space in a liberal democracy, but the ability to exercise restraint is a shared trait for effective advocates in authoritarian regimes. In these cases, textbook tactics for actors in liberal democracies can be the very thing that counterparts in authoritarian regimes need to avoid.

Second, authoritarian regimes sometimes embrace practices that have fallen out of favour in the current scholarship, but researchers would be wise not to dismiss what they do as unworthy of serious discussion. This applies to the pursuit of sustainable growth and social inclusion, for example. The question of whether post-industrial development can be realised independently from proposals that expand citizen empowerment, political devolution, policy transparency, and collaborative governance is a valid subject for comparative research on regimes where these commonly advocated strategies invite suspicion and scepticism of those in power. Third, similar to how policy advocates concentrate on local institutions in democracies that are federalised while directing their efforts towards national venues in unitary democratic systems, corresponding diversity in structure and power dynamics results in varied behaviour in authoritarian governments. Established leaders act unilaterally and get themselves personally involved in administrative matters. Others require the support from multiple factions to survive, in which case intervention from the top is rare while delegation becomes the norm. The contrast in leadership style translates into broader differences in how the rest of the administrative system operates.

So what should be done for existing theories in the surge of interest in the Global South and non-Western governance traditions (Eiró & Lotta, 2023)? Establishing the link between the activities the theories explain and the public choice and power architecture in which they are nested will be the obvious first step, which helps separate specific behaviours and patterns that are intertwined with the broader political order and therefore most likely to change across regime types from those that are tied to universal, non-regime-specific conditions. Next, researchers should engage in new empirical work prioritising these behaviours and patterns outside the liberal democratic cases. These steps can be illustrated with the history of punctuated equilibrium theory, a major public administration framework that has drawn heavily from comparative politics. Traditionally, scholars of public administration in US were incrementalists, arguing that internal processes that determine the policy agenda and organisational structure of the government moved only gradually because of risk aversion among political actors and formal institutions that prevent unilateral control. This view came into question throughout the 1990s, as new analyses of budgetary allocations, congressional and presidential agendas, and other government activities in the US revealed not only extended periods of inertia but also episodic shifts away from the status quo, leading to the “punctuated equilibrium” characterisation. In a series of replications throughout the 2000s, scholars drawing on data from European and Commonwealth countries report two generalisable findings. First, the pattern is not unique to the US; it is universally applicable to many other liberal democracies whose political institutions are differently organised. Second, the more decentralised the political system, the more punctuated the pattern of change it experiences (Baumgartner et al., 2009; Jones & Baumgartner, 2005).

Jones, Baumgartner, and their collaborators drew direct guidance from regime typology (e.g., Lijphart, 1999) to make sense of the cross-country variation. This allowed them to develop a framework that engages not only factors at the individual and organisation levels (e.g., policy entrepreneurs seeking change through venue shopping or strategic interplay inside a policy subsystem to deter such activities), but also the broader institutions that determine how many veto points are present in the policy process, how much control the legislative majority has, and how well the legislature and executive branches work together. Because the link between the activities the punctuated equilibrium theory describes and the broader political order in which they are nested takes centre stage in their models and analyses, they crafted a theoretical vocabulary for researchers to formulate and then test propositions on authoritarian regimes (Baumgartner et al. 2016; Lam and Chan 2015). The questions they raised have since become part of the mainstream debate covered by public administration journals in China (van den Dool & Li, 2023).

The framework’s transition from an initial focus on the US to the policy dynamics in China reveals just one of many pathways. Further discussion about the other pathways will generate critical value not only for theory but also for practice, as well. From an academic standpoint, comparative findings will connect established theories with the political systems in which it is practiced, thus adding new information about the boundary conditions of relevant propositions, or otherwise increase their generalisability (e.g., W. Li & Weible, 2021; H. Li et al., 2017). As for practice, at a time when decision-makers worldwide grow increasingly divided over what constitutes best practice in governance and public administration, researchers will find it necessary to explain how global challenges are being met by regionally diverse strategies, including those employed outside liberal democracies. As practices from abroad, especially in the Global South and other severely underrepresented regions, begin to receive unprecedented attention, this line of inquiry can guide experimentation in governance in an increasingly polarised environment for international policy learning and transfers.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Research Grants Council, University Grants Committee [17615920].

Notes on contributors

Kwan Nok Chan

Dr. Kwan Nok Chan is Associate Professor of Politics of Public Administration at the University of Hong Kong. He researches public policy in the authoritarian setting, with a special focus on how policy actors organise, govern, and consume information as a resource across administrative and political contexts.