Arm's-length agencies and authorities are a core feature of Canada's administrative state, with governments at all levels and stripes routinely establishing entities that act with some degree of autonomy from traditional departments and ministerial leadership. From financing infrastructure, to managing energy development, to coordinating health delivery, to building and operating mass transit, or regulating public procurement, these organizations are now instrumental to how public power is exercised. Though no agencies or authorities are entirely free from the oversight and ultimate control of a responsible minister in Canada, in practice many are presented to the public as having considerable operational control of their mandate. The federal government today has over 200 arm's-length agencies and boards delivering a wide array of programs or regulating sectors (Doberstein, 2022), and in some provinces these bodies account for up to half of all government operational expenditures (McCrank et al., 2007). Why do governments create arm's-length agencies and authorities when it means they are delegating power they have earned through election? For some governments it is part of ideological commitments (i.e., new public management ideas that agencies responsible for specific tasks would be more efficiently run than the traditional bureaucracy) (Lapuente and Van de Walle, 2020). In other cases, it is a stakeholder or interest group victory or preference to be regulated via a purpose-built agency or authority (Carpenter, 2010a) or due to special expertise required (Bernier et al., 2022). They can also be used to signal “credible commitment” of a stable regulatory policy environment (Alon-Barkat and Busuioc, 2025), or can be created as part of an effort to show to the public that an issue is a priority of the current government (Carpenter, 2010b). Recently created examples—such as the Canada Infrastructure Bank, Supply Ontario, InBC Investment Corporation, Recovery Alberta, and Québec's L'autorité des marchés publics—illustrate how agencification is a reflexive policy instrument for all sorts of purposes across jurisdictions. For scholars and practitioners, the proliferation of agencies and authorities raises classic but unsettled questions, such as: who is accountable when outcomes are contested; how should performance be judged; and when does autonomy advance or undermine democratic legitimacy? There has been a small core of scholars in Canada focused on types of arm's-length bodies over the years, with Luc Bernier (cf. Bernier et al., 2018) and Malcolm Bird (2022) focused on crown corporations, Lorne Sossin (cf. Houle and Sossin, 2006) and Sule Tomkinson (2018) on administrative tribunals, Charles Conteh (2025) on economic development agencies, and this author (Doberstein, 2022, 2023) more recently on agencies and authorities writ large. Yet Canadian scholarship has tended to surface these dynamics episodically, typically through scandal or reform moments, rather than as a cumulative research program connecting questions of institutional design, accountability, citizen perceptions, and outcomes across sectors and jurisdictions. This essay first situates Canadian work within international debates and then advances a vision for a research agenda focused on (i) agency performance, (ii) relational accountability and public trust, and (iii) politicization and appointments, and concludes by pointing to methodological innovations suited to the Canadian federal context. A principal–agent lens is the classic and foundational approach to the study of agencies: elected principals delegate authority to agents, design ex ante and ex post controls, and monitor to align behavior with political preferences, foregrounding risks of bureaucratic drift and capture (Kiewiet Moore, 1995). Reputation-based approaches similarly highlight how agencies cultivate legitimacy, but across multiple audiences—political overseers, industries, professions, and publics—to secure autonomy (Carpenter, 2010a). Finally, post-NPM paradigms such as New Public Governance emphasize agencies as nodes in collaborative networks, where authority is dispersed and coordination challenges surpass simple principal–agent metaphors (Osborne, 2006; Sørensen and Torfing, 2009). Together, these theoretical traditions underscore that the study of agencies in Canada must straddle both relational and structural perspectives, situating them in a comparative conversation that has moved well beyond classical delegation models. In terms of dominant themes of study in the literature, democratic accountability has been the through-line of Canadian and international debates on agencification, with concerns that structural disaggregation fragments lines of answerability and weakens ministerial responsibility unless alternative control and transparency mechanisms are strongly embedded (Flinders, 2017). Bernier and Farinas (2010) reviewed federal reforms and argue that accountability architectures evolved across sectors and have focused on mitigating these risks. In Québec, for instance, accountability statutes for state-owned enterprises were tightened in the mid-2000s, signaling a more formalized regime for Crown governance (Bernier et al., 2018). Yet other Canadian research at the provincial level shows how autonomy and dispersion complicate answerability and sanctioning, especially when lines between ministers and boards are blurred (Hall, 2021). Empirically, accountability in agency settings is multi-sited: legal, bureaucratic, professional, and political logics intersect and often collide in practice, complicating assessments of who should explain, who can sanction, and on what basis (Romzek (ii) accountability; and (iii) politicization. There are also methodological innovations to pursue, including exploiting experimental designs and new computational social science approaches that can leverage large amounts of discourse and administrative data. I will outline each of these opportunities below. Performance and public value: Significant effort has been put into research to answer a basic, but important, question about agencies: do they perform better than if those functions were delivered via a traditional government department or some other organizational design? Some scholars have found an agency dividend of sorts (either in innovation or efficiency superiority) (Lægreid et al., 2011; Wynen et al., 2014), whereas others find in comparative national studies that public sector efficiency decreases with greater agencification (Overman James, 2003; Yamamoto, 2006). There may be limited additional insight to gain from further broad-scale performance comparisons—the literature is pretty clear that, though difficult to study, the story is mixed and contextual. Yet there are opportunities to provide new insights by embracing performance frameworks that move past narrow efficiency or output counts to align with public value metrics, like legitimacy, trust, and equitable impact. For regulators and utilities, this would mean integrating measures of perceived procedural fairness, transparency of decision-making or enforcement rationales, and distributional impacts across regions and income groups into mandated reports. For service-delivery agencies, client experience and equity metrics, such as time to access, accessibility for rural and Indigenous communities, and culturally competent service, should be measured alongside cost and throughput. Canadian scholars can lead by designing open-source indicator sets and piloting them with agencies and departments willing to experiment, generating evidence on trade-offs and best practices. Accountability and public trust: Accountability is central to democracy, yet the proliferation of arm's-length agencies complicates the Westminster principle of ministerial responsibility by blurring who is answerable for decisions. We see this routinely manifest across Canada when policy controversies arise in health care, transportation, energy regulation, public safety, parole boards, when ministers and agency leadership angle to deflect blame. With fragmented authority, the result can be confusion among citizens about who is responsible, and risks to trust and legitimacy. Understanding how accountability is structured, contested, and perceived in this ambiguous terrain is thus essential for evaluating the democratic role of agencies. Moving beyond principal-agent frameworks of accountability, more recent scholarship has shown how accountability regimes are evolving in more networked, reputation-sensitive, and politically responsive directions (Overman et al., 2021; Papadopoulos, 2023; Pérez-Durán, 2024). Scholars have shown how regulators calibrate their accountability regimes depending on the policy area: agencies in politically salient or high-risk sectors are often subject to stricter political and legal accountability (e.g., Public Prosecution Service of Canada; Canada Energy Regulator), while those in more technical or insulated domains rely more on professional accountability (e.g., Canadian Nuclear Safety Commission; Canadian Food Inspection Agency). These competing demands are not always easily balanced, and scholars caution that overemphasizing one type, such as political responsiveness, may compromise technical integrity or public trust (Maggetti & Verhoest, 2014). Thus, understanding how these modes of accountability interact is critical to evaluating agency legitimacy and responsiveness in Canada. Therefore, accountability in agency governance should be conceptualized and studied as a negotiated, relational practice, not merely a statutory allocation of powers, with attention to the backstage bargains that define boundaries of interference, deference, and mutual protection between ministers and agency heads (Maggetti & Papadopoulos, 2023). Elite interviews with former ministers and CEOs in key agencies can uncover tacit rules of the game—what triggers intervention, how disagreements are arbitrated, and how both sides coordinate public narratives in moments of controversy. Paired process-tracing of specific controversies, triangulating cabinet documents where accessible, board minutes, media statements, and legislative transcripts, can then surface how these tacit bargains are operationalized when stakes are high. Relatedly, a sustained Canadian program on trust in agencies is overdue, extending beyond trust in government writ large to sector-specific measures that distinguish ministries from agencies and regulators from service-delivery authorities. Original surveys should field validated trust batteries that tap perceptions of competence, integrity, benevolence, and fairness, and explicitly test whether the mere identity of the decision-making body shifts trust separate from outcome evaluations. Panel designs around salient controversies—drug pricing, emergency room closures, utility rate hikes—can estimate how trust dynamics evolve when ministers and agencies advance competing narratives, providing evidence for reforms that target transparency versus authority clarity. Populism and agency politicization: Governments of all partisan stripes have not been shy about making sure at least some of their agency leadership appointees are aligned with their political agenda (Doberstein et al., 2025). Many view this as a crucial connection between democratic politics and agencies, yet the populism of the current era may represent new levels and types of politicization worthy of study (c.f. Cohen & Duhl, 2025 in Israel). Populism intensifies pressures on governments to assert visible political control, often clashing with the purported autonomy of arm's-length agencies. While some agencies are designed to buffer decisions from partisanship and safeguard expertise, populist rhetoric often frames unelected bodies as unaccountable “elites” and pushes ministers to intervene directly. This politicization risks undermining professional integrity, distorting policy decisions, and eroding public trust of agencies with mandates designed with impartiality and non-partisanship as key components of their effectiveness. In Canada, this dynamic is evident in recent political attacks on the Alberta Energy Regulator and the restructuring of Ontario's tribunals, for example, where populist leaders intervened aggressively to align agencies with partisan priorities. Such politicization can undermine agencies’ expertise and impartiality, blur accountability lines, and erode public trust. Understanding how populism reshapes the balance between autonomy and control is critical to evaluating the resilience of Canada's administrative state. Populist dynamics invite explicit theorizing of how parties narrate agency autonomy and expertise, with research asking whether leaders attack agencies as unaccountable elites, repurpose them for ideological projects, or engage in personnel-led politicization via appointments (Hinterleitner, 2018). Comparative designs across provinces and parties can test whether populist cabinets centralize control rhetorically while insulating themselves institutionally—e.g., by delegating contentious decisions to technocratic bodies yet criticizing them in public to maintain partisan distance. Discourse-analytic methods can quantify shifts in frames across electoral cycles to reveal how delegative architectures are legitimated or undermined in public argument. Ministers can generally issue direction and orders to agencies as they wish, but appointments are also an opportunity to shape the policy and reputational trajectory of agencies. Yet in Canada we lack longitudinal evidence on and partisan across boards and agency leadership. and (2022) appointment at the federal level in Harper and years, but not partisan political of provincial appointments that expertise, and would of politicization while public debates on and research by Doberstein et al. (2025) has this with Alberta, with to work should appointment and ministerial as instruments that expectations of responsiveness or asking whether politicization or merely methodological advance the study of agency public administration research must methods at the of treating them as that up new of research questions to be than on structural or of agencies, scholars should and comparative designs that capture the dynamics of performance, accountability, and politicization. Canada's provinces provide a terrain for how institutional shape governance over policy regulation, health service delivery, at have agency oversight and political these scholars to how design performance and public not in like or but also in of trust, legitimacy, and and as Alberta and approaches to energy regulators or Ontario and leverage to see how institutional reforms and a to effects of institutional design on citizen institutional outcome and partisan we can directly test how or partisan and media frames would by the environment in which This directly to accountability and public trust, asking whether citizens to ministers responsible in a Westminster or whether arm's-length structures perceptions of in computational discourse such as structural and or systematic analysis of how perform accountability and politicization in public Ministers and agency heads perform accountability through and systematic analysis of their discourse in media can reveal technocratic and designs can shifts in and This approach also be especially for populism and agency whether leaders agencies as or elites, and how such narratives perceptions of legitimacy. Arm's-length agencies and authorities are central to how Canada is tensions between efficiency and democracy, expertise and and autonomy and Canadian scholarship has essential building but the demands a agenda that relational accountability, politicization and appointments, and citizen of by methodological Comparative provincial designs can uncover how institutional over survey can reveal how citizens responsibility in a Westminster and computational discourse analysis can the through which or deflect blame. accountability lines performance or politicization trust in public will In an era of fragile trust in public the is scholars do not the opportunity to study these dynamics with greater and reform debates will the of governments to by or politically in and computational the Canadian public administration can theory of accountability and legitimacy in the administrative state, while for designing arm's-length agencies and authorities that are and worthy of public trust.
Carey Doberstein (Mon,) studied this question.