Policy labs, partners and policy effectiveness in Canada

Abstract Upon election in 2015, the Justin Trudeau Liberal government announced its intention to transform government operations by bringing nonprofit and private sector partners into the center of public sector decision making through new structures such as Policy Hubs and Innovation labs. These collaborative arrangements were intended to yield the benefits of Michael Barber’s theory of deliverology by breaking through the public sector aversion to risk and change and by creating new spaces for devising effective solutions to the increasingly complex social and economic challenges facing government. A preliminary examination of the use of policy hubs and innovation labs in Canada between 2015 and 2020 indicates that the results have been mixed for the nonprofit sector partners. Collaborative relations have offered nonprofit sector partners new opportunities and access to influence policy decisions. However, this influence also poses risks to their independence, legitimacy and effectiveness as policy advocates. Both public and nonprofit sector partners in PILs should heed certain cautions in choosing future partnerships or they may find their ability to achieve meaningful policy change is limited.


Introduction: changing government
When the Justin Trudeau Liberal government was elected in 2015, it quickly announced its intention to put into effect the operational theory of deliverology advocated by Michael Barber (2015) to ensure that the public service met the governmentdefined policy goals. This put Canada squarely onto the track of innovation, results and measurement in the policy process. To ensure that any resistance to change was overcome, hubs and policy laboratories were created both in the central machinery of government and in the departments. By the time of its reelection in 2019, the Liberal government's commitment to deliverology was more ambivalent and the fate of the policy innovation labs (PILs) became more uncertain.
PILs were born out of both the realization that the state's ability to handle complex policy problems is limited and the resultant turn toward more collaborative arrangements with the private and nonprofit sectors (Waltzer 1988;Hirst 2002;Torfing and Triantafillou 2016). They are hybrid organizations comprising talent from the three sectors tasked with developing solutions for particular policy problems in a short timespan. Their work may involve defining these discrete policy problems, and testing and assessing the impact of their proposed solutions. While these experiments may be limited in scope, if their solutions appear viable, then governments may scale them up to address broader policy problems. While PILs are created to create changes in the policy process and results necessary to meet current needs and thus have a bias toward an innovative, progressive agenda, they are intended to reflect the public interest and use objective, evidence-based reasoning in their operation. These policy change mechanisms in the innovation agenda of governments (Tõnurist, Kattel, and Lember 2017;Westley, Geobey, and Robinson 2011) are variously known as Public Innovation Labs, Policy Labs, Innovation Hubs, Living Labs, and other similar names. They may be located within the state structures, like the Canadian Innovation and Impact Unit located in the Privy Council Office, or exist as independent organizations within the nonprofit and private sectors with direct ties to the central machinery of government and senior levels of the bureaucracy (Public Policy Forum 2013). Regardless of their name or location, PILs bring nonprofit and private sector actors together with government officials to devise timely solutions to complex social and economic policy challenges.
The Canadian experience with PILs between 2015 and 2020 provides an opportunity to consider their effectiveness in the policy process and their impact on state relations with the other sectors. While PILs are still relatively new, they embody the move toward public sector entrepreneurialism and collaborative tri-sector relations that advanced significantly under the operational public sector philosophies of New Public Management (NPM) and New Public Governance (NPG) in recent years (Osborne and Gabler 1992;Craft and Howlett 2013). This study of PILs yields some insights into the evolution and future of public-nonprofit sector relations, an area of particular interest given the increasing involvement of the nonprofit sector in policy development. While the article notes that collaborative arrangements between the two sectors are being developed at all levels with positive opportunities for nonprofit organizations to influence policy development, it cautions that the theory and operation of PILs may be limiting the ability of the nonprofit and voluntary sector to affect policy in a meaningful way. An examination of the government documents authorizing the use of PILs and the experience of some of the early PILs reveals that through these collaborative arrangements, relations between the two sectors may be politicized and ultimately unsustainable at some levels. Indeed, the ability of both sectors to devise implementable and scalable solutions to policy problems may be adversely affected. While this argument is based on the Canadian experience, it offers cautions pertinent to the debate in the literature regarding the direction of relations between the public and nonprofit sectors and the use of PILs to address complex policy challenges.
Anchored in the literature on PILs, public sector management and state-nonprofit sector relations, the article examines the recent shift in public sector operations focusing on the use of PILs. The article bases the argument on an analysis of the documentation existing in the public realm with regard to the implementation of deliverology and PILS in the policy process. As Hammond and McDermott argue, documentary analysis "can be useful both as a means of framing organisational research and in order to interrogate issues in the field. It provides opportunities for exploring what is taken for granted and unspoken" (2020). As Bowen argues, documentary research provides a useful means of tracking change within organizations, corroborating findings or revealing current lacunae or assumptions in the literature, and suggesting further questions or developments requiring exploration and analysis using alternate research methods (Bowen 2009). While interviews may be a next step in this research, at this point the documents are an important sources of discerning "multiple realities" and the complexities inherent in organizational culture. In the current case, documentary analysis reveals the limitations and biases that were built (perhaps unwittingly) into the approach toward innovation adopted by the Canadian government by examining the internal government machinery constructed to support innovation, the types of PILs and relations created in the process and examples of which ones were or were not successful or sustainable, and examples of the power relations in these relations.

The Barber vision of deliverology
Shortly after forming government, Prime Minister Justin Trudeau invited the former advisor to the Blair government in Britain, Michael Barber, to address Cabinet on how to manage reform initiatives in government (Barber 2015). Styled "deliverology," Barber's approach links policy results to government priorities, and emphasizes evidence-based policymaking, measurement and evaluation throughout the policy process. Its ultimate goal is to ensure government effectiveness in achieving societal and economic change ("delivering results"). By employing less hierarchical and more collaborative means of decisionmaking that embrace non-state actors, deliverology was designed to circumvent the traditional obstacles to change in the public sector that included bureaucratic lethargy, lack of expertise, intransigence and risk aversion. To resolve complex policy problems characterizing modern public life, he recommended combining strong centralized leadership and clear policy priorities at the highest levels of government with clear direction, measurement and evaluation of results at all stages of the policy process, and the use of external expertise from the nonprofit and private sectors in policy design and implementation (Barber 2015).
Two structural changes to the policy process were important for ensuring that an elected government's priorities guided public sector decisions and achievements. First, Policy Development Units (PDUs) or Policy Hubs would be located in the central machinery of government and/or certain government departments to independently monitor the policy process and encourage compliance with government-defined goals by applying timely nudges to public officials. These bodies ensured accountability of public sector managers to the center of government (Thaler and Sunstein 2008). Second, the adoption and adaptation of private sector idea laboratories in the policy process would help inculcate a culture of change through new methods of policy thinking (Barber 2015, 114-116;Richards, Gallo, and Kronick 2017). These Policy Innovation Labs (PILs) would draw on talented, forward-thinking individuals from the public, private and nonprofit sectors to devise policy solutions and plans of actions to defined problems within specified timelines. Using traditional scientific techniques (experimentation, testing, verification) or design thinking (empathize, define, ideate, prototype, test, implement), PILs were intended to reduce the impact of ideological or values-based norms on the resolution of policy challenges (Williamson 2015;Torjman 2012;McGann, Blomkamp, and Lewis 2018). Solutions would be evidence-driven. As "'islands of experimentation' where the public sector can test and scale out public sector innovations," (Schuurman and Tonurist 2017), the PILs would allow new ideas and new performance indicators to break through bureaucratic lethargy, risk aversion and resistance to change. The PILs would deliver results that the Hubs would ensure complied with central priorities. Thus, Canada joined other jurisdictions (Barber 2015;CPI&PE 2018), in using people and ideas from the private and nonprofit sectors working in Hubs and PILs to transform public policy by addressing complex issues with skills that were not readily available in the public sector (McGann, Blomkamp, and Lewis 2018;Puttick, Baeck, and Colligan 2014;Carstenson and Bason 2012).

The Trudeau response: from the innovation hub to the impact and innovation unit
Mirroring the Barber approach, the Trudeau government created a Cabinet Committee on Agenda, Results and Communication in 2015 to define the policy agenda and key government priorities, track progress on its commitments and oversee strategic communications. A dedicated unit on Results and Delivery was created in the Privy Council Office (PCO-the body that supports Cabinet) to support this Committee and coordinate its work with the Treasury Board. Treasury Board was directed to track progress on policy objectives and ensure financial efficiencies in departments in meeting these objectives (Zussman 2016a;Lindquist 2016). The Results and Delivery Unit would track progress on achieving government goals in departments through the mandate letters issued to the ministers of departments. Also created within the Results and Delivery unit of the PCO was the Innovation Hub which became the Impact and Innovation Unit in 2017 to coordinate with departments and help them achieve their mandated objectives. This centralized and streamlined approach was designed to help public servants to break away from the "fog of accountability" and "web of rules" that impeded policy agility and innovation (Dean 2016;Zussman 2016b;Dobell and Zussman 2018). As Barber had recommended, strong centralized control was a key component of the Trudeau approach to policymaking with Cabinet, the PCO and Treasury Board overseeing policy development in departments and holding public sector managers accountable to centrally defined priorities.
The Trudeau approach was not entirely new. The 2014 report of the Clerk of the Privy Council had favored the use of innovative ideas including a central Hub and departmental PILs (Canada 2014, 11). He noted that the "public service uses open and networked approaches to develop innovative, effective solutions to complex problems and emerging issues," and "draws on a diverse range of data and information to develop evidence-based ideas, analysis and advice" (Canada 2014, 11-12). The report recommended the creation of a "central innovation Hub" in the nonpartisan PCO to "ensure that successful innovation is replicated across government" and "change the way the Public Service does business." The central Hub would "support departments in applying new approaches-such as behavioural or "nudge" economics, big data, and social innovation-to complex policy and program challenges" (Canada 2014, 12). Its work would be complemented by change labs established in designated departments like Employment and Social Development Canada (ESDC) to experiment with new approaches bridging policy, program and service perspectives in solving client problems (Canada 2014, 12). On its face, the Trudeau approach seemed consistent with the Clerk's approach to encouraging innovation in the public sector.
Upon closer investigation, there was an important difference between the Clerk's report and the Trudeau government's approach. The Clerk's report made scarce mention of the political-bureaucratic interface, stressing instead the importance of drawing on a wide "range of data and information to develop evidence-based ideas, analysis and advice" (Canada 2014). In this vision, the Hub's activities would be consistent with the traditional obligation of the public service to offer impartial and informed analysis and advice on proposed policy ideas and solutions. In contrast, the Trudeau government designed Hub was committed to linking the use of new techniques for innovation to achieving the elected government's clearly defined priorities and desired results (Dobell and Zussman 2018). The Trudeau government was intent on ensuring that the policy path to implementation taken by the public sector did not stray far from its political commitments and electoral promises. Rather than generating impartial advice on the formation and implementation of government policy objectives, the Trudeau approach encouraged the newly created units to ensure compliance within departments to its government's partisan and ideologically-driven objectives. This shift was not unlike the use of hackathons in the innovation process to coopt partners to an institutional agenda (Zukin and Papadantonakis 2017). To this end, the government brought a Liberal ally and coauthor of the 2015 Liberal Party election platform into the nonpartisan PCO, to head up the PCO Results and Delivery Secretariat and later the Impact and Innovation Unit (May 2016).

The IIU: engaging the "right" people and partners in policy development
The IIU, as the key engine of the government agenda, has a two-pronged approach that also impacts its relationship with the nonprofit sector. Within government, it is pursuing systems-level changes in policy and the public sector culture that foster outcomes-based strategies. This includes identifying and overcoming internal barriers to innovation and experimentation, implementing a new staffing strategy ("recruiting the right people") to address top priority issues, and accelerating the change process by working more "actively" with departments to secure innovative outcomes (Canada IIU 2018b). The new staffing strategy for example, includes a fellowship that "recruits new talent and skills into the public service to help advance the Government's agenda" (Canada IIU 2018b)-not to advance policy in the public interest or improve policy as the Clerk's report envisioned, but to advance the government's agenda. Partners from the nonprofit sector as individuals or as organizations are brought into this milieu of advancing the government's agenda and achieving its goals. Thus according to the documentation, the staffing strategy simultaneously reinforces the hierarchical power of the state while drawing its partners into achieving the government goals and politicizing them in the process.
The second prong of the IIU approach involves developing relations with PILs outside of government and indirectly providing "non-traditional" (Canada IIU 2018b), independent organizations with access to the center of the policy process. In particular, the IIU encourages the engagement of external organizations to policy innovation through a more flexible and targeted grants and contributions program that is driven by the government's priority areas as well as the Impact Canada Initiative that includes a challenge platform to encourage outside organizations to propose innovative solutions to policy dilemmas (Canada IIU 2018b). Projects have included the Smart Cities Challenge, the Clean Tech Challenge, Canada Learning Bond Outcomes, Increasing Gender Diversity in the Armed Forces, encouraging Donations to the Charitable Sector, and responding to Canada's Opioid Crisis (Canada IIU 2018b). Stakeholder engagement is critical to the innovation agenda and the work of PILs. These partnerships reflect the government's political priorities and thus involve agencies supportive of those priorities and the innovation agenda.
Under the Barber deliverology approach to policy as modified by the Canadian government, networks involving people from all sectors, both within and outside the public sector, are key to addressing complex and persistent policy problems in priority areas. Both the IIU and PILs created under its direction can nudge departments and present new ideas and ways of addressing policy problems moving them from traditional practices to more open, dynamic methods of operation. In theory, PILs and other hybrid organizations existing outside of government can help break through the traditional culture by offering scaled experiments using new techniques and specialized expertise to present solutions to persistent problems that the public service does not have the time or talent to address in its daily operations. However, these collaborative arrangements between the public and nonprofit sectors may be less than satisfactory for both sectors as the following examination of the department-level PILs suggests.

New opportunities to influence policy
The Canadian government is offering nonprofit and private organizations new and varied opportunities to influence policy through the PILs as part of its commitment to networked governance and public sector culture change. PILs have included: Living labs like the University of Prince Edward Island's Clinic for Patient Oriented Research and the Mohawk College unit focusing on energy, health and technology supported by the Competition Bureau of Canada; Policy labs and hubs funded by Natural Resources Canada for mining and the environment, health, Arctic research, closing the gap between labs and the market, microfluids (University of Toronto) and automotive innovation; And, the Canada Revenue Agency's internal Accelerated Business Solutions Lab that has partnered with organizations on workshops.
Other labs have addressed a more inclusive public service for people with disabilities, increasing the women's voice in innovation, clean tech solutions and photojournaling. As this list of examples reveal, PILs are an important component of the government's attempt to shift the public sector culture to one of innovation in its priority areas.
External partnering PILs or hybrid PILs internal to departments are created with the objective of recommending policy solutions with limited attention paid to procedural matters or existing rules. For example, the Service Lab at Innovation, Science and Economic Development Canada (ISED) partnered with the Community of Federal Regulators (CFR-comprising 27 regulatory organizations) at the instigation and under the eye of the Treasury Board Secretariat (TBS) to redesign regulatory guidance information and create a client-centred online service within six months (Jones 2016). The TBS had a vested interest in the project succeeding since it was using it as a model for Canada.ca, a centralized information service to be applied across all departments (Jones 2016). This project was achieved and demonstrates the usefulness of PILs and external partners in transforming a service to make it more user-friendly and to offer scaled experiments for pan-government changes. As the literature suggests, PILS might be most effective as scaled experiments for defined services or where the knowledge is quantifiable, policy-relevant or applied rather than qualitative or involving broad social coalitions or for higher level policy changes (Tõnurist, Kattel, and Lember 2017;Fischer 2003;Carstenson and Bason 2012). By extension then, where policy problems are more complex and data is qualitative or less clear, PILs may be less useful. The example here, though, also demonstrates the "unspoken reality" unveiled by documentary research of the importance of the central machinery in initiating and executing the work of PILs given that the TBS used the CFR as an "outside-in" organization to transform services. Where the central agencies are vested in the exercise and ensure resourcing is sufficient, PILs are more likely to achieve their objectives.
In contrast, the literature suggests that solutions devised by PILs are not always adopted so easily or may be adopted without sufficient evidence. Because PILs tend to focus on particular policy problems and solutions under set priorities and do not participate in implementation for the most part (Bridgespan Group 2014; McGann, Blomkamp, and Lewis 2018), "there's very little guarantee that the solutions will actually work" according to Lab participants interviewed by Martin, Dale, and Stoney (2017). The short timelines and targeted focus of PIL work can also mean that while PILs are very good at defining the first order effects of the solution, they are less equipped to gauge the second and third order effects. Because partners are external to regular department structures, PIL solutions may not be tied to budget lines or even budget realities (Tõnurist, Kattel, and Lember 2017). In these cases, often the public sector department is left to devise the means of implementation and to resolve problems that arise from solutions suggested by PILs. If the solutions are not practical or compliant with government needs or resource allocations, then the value-added of PILs and external partners may be questioned, ultimately raising the issue of sustainability of the alliances.

Politicization of partnerships
Solutions may also reflect government priorities rather than objective or impartial analysis. The inclinations of researchers may inadvertently, or by design, coincide with those government priorities. For example, the Accelerated Business Solutions Lab at the Canada Revenue Agency undertook a study of the needs and experiences of homeless people in filing tax returns. The project did not yield implementable results, an analysis of the current system or conclusive results because the sample size was too small (31 interviews) and scope of analysis was too limited (CRA 2018). Despite the inconclusive nature of the study as reported by the researchers, it was used to justify a significant increase in CRA spending to assist nonprofit organizations in helping the homeless population to file tax returns (Press 2018). In the push toward policy innovation and usage of PILs in the CRA mandate, public sector neutrality and quality of decision-making appeared to be compromised. This example suggests that instead of serving in the traditional role of independent policy advocates, nonprofit organizations may become quiescent partners in the legitimation of policy decisions made with questionable evidence to conform to ideological preferences of the government. This is further evidence of policization of the policy process as foreseen by Aucoin (2012) and others (Dobell and Zussman 2018).

Vibrancy and sustainability in the shadow of government
IIUs and PILs foster collaboration and networked governance embedding co-creation and co-management of policy across government ideally. However, the construction of a centralized Hub overseeing departments and PILS reinforces an inequality built into the relations between state and non-state partners. For example, in 2017 ISED Canada announced the movement of its Innovation Lab to the creative innovative space, the nonprofit Bayview Yards (Canada ISED 2017). The Innovation Lab was established to serve entrepreneurs and innovators and create a modern and user-centric public service. In 2018, ISED closed the Innovation Lab while it reset its mission to have a "stronger focus on supporting the Department in its digital transformation and the design and delivery of client-centric services" and help "position the department to deliver on Canada's innovation and skills plan" (Canada ISED 2018). The repurposing of the Innovation Lab to support digitalization was abrupt and significant. Government objectives and priorities are central to the redesign regardless of prior arrangements with Bayview Yard. Former partners are not likely to be included unless they fit with the new mandate. Most importantly, this example suggests that PILs not achieving government objectives or meeting targets set by the central Hub may be shut down or repurposed as the Innovation Lab was.
Somewhat paradoxically, greater access of nonprofit organizations to the policy process may result in less sustainable relations. Recall that under the deliverology approach to governance, priorities are set and department progress is tracked and measured on achieving those objectives. However, monitoring can also be uneven or spotty (Van Acker and Bouckaert 2018) with the criteria of evaluation poorly defined or poorly suited to the purpose (Dobell and Zussman 2018). This may work against nonprofit partners working on innovative solutions that by definition defy traditional evaluation criteria. Also, with the attention of the center on priority areas and obtaining results in a timely fashion, other policy issues will receive less attention and must compete for other, scarce resources. Partnerships of the nonprofits with government may be episodic and depend on the favored priority of the day while others languish (Tõnurist, Kattel, and Lember 2017;Schuurman and Tonurist 2017). For example, under the Impact Canada Initiative, grants and contributions have been reconfigured to encourage networked governance. However, the priority is identifying "Primarily grants and contributions policies and programs and other government initiatives that may benefit from IIU interventions" (Canada IIU 2018a)-a much smaller pool of eligible recipients. By example, ISED invested $25 million into the nonprofit Creative Destruction Lab at the University of Toronto in October 2018 to be spread among similar partners engaged in helping emerging businesses to scale-up (Creative Destruction Lab 2018), a key priority of the Liberal government.
Further, targeted spending means achieving targeted results. Once a target is achieved or if a program is not achieving its objectives in a timely way, then it may be "orphaned" or abandoned. As an example, the Government of Canada's National Digital and Data consultations concluded in October 2018 with a report released in 2019 on over 1900 ideas from over 580 people interviewed but, like the Innovations Lab, there it seems to rest with no further steps taken either in the federal 2019 budget or post-election plans. Similarly, if a Lab does not efficiently achieve the results desired by government or finds an answer that contradicts the government priority, its funding might be discontinued or not renewed or government support may dissipate. As an example, consider the Smart City project for the Toronto Waterfront, an endeavor run by the Google subsidiary Sidewalk Labs with the support of the federal-provincialmunicipal Waterfront Toronto Commission. When it ran into controversy and public opposition over its data collection techniques, government support faded, the project was scaled back and Sidewalk Labs finally walked away (Smart Cities World 2020; Fox 2020).
Funding and administrative arrangements determined by government-defined priorities may have a serious impact on the labs themselves. In their survey of Canadian labs, Martin, Dale, and Stoney (2017) found that the innovative work of the Labs may be slowed down by the necessity to find partners, identify collaborative space among the partners, devise and test solutions, refine the solutions and to sustain the interest of public sector partners. This also requires sustainable funding sources given that fundraising can be time-consuming and burdensome (Martin, Dale, and Stoney 2017). Further, they note that for social innovation projects, timing in the project is important for securing funding. Foundations may be more willing to fund exploratory stages of projects but governments are more likely to fund once results are "more concrete." If Lab results support government priorities, they are more likely to be funded than if they do not. More worrisome is that "governments usually initiate engagement with a lab with a specific outcome in mind," a practice at odds with the lab philosophy of designing creative solutions (Holliday 2000). Politicization of the nonprofit partners or mission drift of organizations are possible consequences from the dive for dollars.
Partners in PILS may find that their longer term relationship with the public sector is affected by these relationships. Consider this. Labs tend to have set time frames to devise solutions calibrated to the life cycle of the government of the day (Schuurman and Tonurist 2017). The public sector has a longer term focus on the public interest. While it serves the government of the day, it must ultimately answer to the public good and not the government. In contrast, PILs serve the government of the day. Nonprofit partners or participants in these networks serve their organizations and clientele. Labs that require additional resources or times to devise solutions may fall out of favor with government and or lose buy-in from partners. Alternatively, new governments may have different priorities and labs may be closed (Guay 2018). For example, participants in the Guaranteed Basic Income project created in Ontario under the previous Liberal government did not expect the Conservative government elected in 2018 to cancel that project as abruptly as it did (Haridy 2020).

Sustainability and accountability in the shadow of government
A final concern involves conflicting norms of accountability between the two sectors. Eva Sørensen suggests that the public sector model of accountability has shifted "because collaborative innovation processes draw on mixed rather than on fixed accountability standards, shift the position of accountability holders and accountability holdees around in the course of the governance process and share rather than divide responsibilities between the involved actors" (Sørensen 2012). In an era of deliverology and collaboration, accountability is no longer just vertical within the traditional lines of the bureaucracy but now has a more direct reporting relationship to Cabinet through the new units (IIUs and PILs) in the PCO and departments charged with overseeing collaboration, and also horizontally to the multiple partners and their respective communities of stakeholders. Multiples lines of accountability in collaborative innovation cloud public sector and nonprofit relations (Donald Savoie 2008.

Conclusion: PILs, partnerships and the changing public sector
Collaborative arrangements with the nonprofit sector are increasing across governments in a wide variety of forms and at all levels of the public sector. Co-production, co-management, co-governance of policy offer the nonprofit sector unprecedented opportunities to influence policy and craft policy solutions for multifaceted, complex challenges facing society and the economy. The Canadian government has engaged new and more nonprofit actors in units within the central machinery of government, like the IIU in the PCO, to engineer cultural change across the public sector. At the department level, government-nonprofit partnerships range from co-delivery of services to co-creation and co-governance of policy addressing health, environmental, social, economic, digital, technological and other important challenges. The Canadian experience, as exemplified by the ISED Service Lab and Canada.ca, confirms that such opportunities are especially successful where policy problems are defined and discrete, require applied knowledge and solutions that may be replicated in other areas, and where funding is secure and relations are built on complementary strengths. Under these limited conditions, Hubs and PILs comprise innovative relationships that are consonant with the policy agility and entrepreneurship required by a more complex and rapidly changing world.
At the same time, the Canadian experience with PILs and Hubs created under the operational theory of deliverology by the federal Liberal government raises concerns regarding the nature of these state and nonprofit sector alliances. The discussion of the IIU and department-level PILs exposed the existence of a hybrid state in which the "shadow of hierarchy" (Dickinson 2014(Dickinson , 2016) characterized by government dominance and an audit culture hinder the ability of the nonprofit sector to affect policy in meaningful ways. Notably, the creation of the IIU and departmental hubs reinforced central control over departments and the external partners: by defining and reinforcing the government's central priorities and objectives; by using soft techniques like nudging units or redefined grants and contributions programs to encourage partnerships consistent with those objectives; by employing firmer techniques like monitoring, performance measures and evaluation to assess progress toward those goals in departments; and, by using hard techniques like mandate letters, repurposing PILs or discontinuing funding to ensure compliance and discourage deviance from central objectives. While at times, relationships may be characterized by mutual trust and respect the vitality and autonomy of nonprofit partners, another pattern of relations suggests that partnerships may limit the meaningful input of nonprofit partners and apply pressure to them to conform to government expectations.
Above all, Canada's experience confirms that nonprofit partners need be wary of the possibility of politicization through engagement in Hubs and PILs. As shown here, the IIU and PILs reveal a shift in the Canadian public sector away from the traditional dichotomy between administration and politics predicated upon the idea of a rationallegal public administration in which public servants offered nonpartisan and evidencebased advice in the public interest on policies and programs to the government. This was evident in the contrast between the vision of a central Hub and PILs in the Clerk's report and the Trudeau government characterization of the IIU and staffing strategies under the operational theory of deliverology. Closing the gap between results and priorities shifts the role of the public sector from serving the public interest to serving the government. As this process of public sector politicization occurs, nonprofit partners also may find themselves closely associated with the government's political objectives in order to maintain alliances and funding relations. Delivering on expected results may compromise the integrity of their expertise and independence, causing mission drift. The source, terms and duration of funding, may affect the ability of PILs to achieve their objectives. Alternatively, as indicated by the CRA sponsored PIL on the homeless, partners may witness their work being used to justify ideologically-driven government decisions lacking a sound evidentiary basis.
This exploration of Policy Hubs and PILs corroborates arguments that although networked and collaborative governance offer more access to the corridors of policy power, these relations may be not be sustainable or desirable. In some cases, these relations are naturally short-lived since PILs are often engaged in discrete, time-bound projects. However, PILs engaging in a suite of projects may also find their relationship with the government unsustainable or undesirable particularly if their autonomy or ability to serve their mission and clients are compromised. Also, if PILs become too closely identified with one government's political agenda, then a different government may terminate relations or discard policy change tools, or worse, the public sector partners may avoid working with politically-tainted nonprofit organizations. How widespread this result may be requires further investigation.
Finally, Hubs and PILs often draw on high level expertise and substantial resources from both sectors to address discrete and specific policy problems and may propose solutions that cannot be implemented or do not work within existing policy or budgetary parameters. Networks may obfuscate lines of accountability in such cases. In the event that they fail value-for-money audits, Hubs and PILs may be discontinued or repurposed and their usage as policy instruments devalued or subverted to different political ends.
Thus, this preliminary examination of the Canadian experience with Hubs and PILS confirms the value of networked governance and collaborations. However, it also exposes the need of public and, especially, nonprofit actors to pick their partners and partnerships wisely and with care, lest they find themselves politicized, compromised, devalued or discarded. Further research is needed to corroborate these preliminary observations and findings.