Accountability Agreements for Ontario Universities: The Balancing Character of a Policy Instrument : le rôle politique

This paper demonstrates how the choice of instrument facilitates acceptance of a new accountability requirement in the Ontario university sector as it helps balance the government’s need for control with the universities’ need for independence. The instrument, conceptualized as an agreement, embodies the negotiated character of the relationship between government and universities, and conveys the idea to different actors that their needs are met. Despite the promises of the instrument, when objectives are ambiguous, uncertainty is pervasive, and negotiation is limited, the increase in government control is minimized and the changes in university autonomy are negligible, thus suggesting that symbolic and rhetorical compliance may be the sustainable equilibrium between governments and governed. Nonetheless, some level of transformation is observed in the sector as the new tool contributes to strengthening priority alignment, highlighting the value of sharing stories, and increasing acceptance of reporting requirements.


Introduction
In 2005, the Ontario government introduced Reaching Higher, a $6.2 billion plan to increase accessibility, improve quality and demonstrate accountability in the province's post-secondary education (PSE) system. One key element of the strategy was the inception of bilateral multi-year accountability agreements between the province and universities with annual report-backs (MYA/MYAAs). The new instrument was intended to outline how each university planned to use the incremental provincial funding to accomplish its own mission while simultaneously contributing to the priorities identified by the provincial government in their Reaching Higher plan.
The Ontario PSE sector is generally portrayed as one that enjoys significant autonomy, given that: (i) universities are non-for-profit corporations created by a charter and established separately from the government (Usher and Potter 2006); (ii) university administrators enjoy significant decision-making powers regarding the use of financial, human and other resources (Metcalfe, Fisher, Gingras, Jones, Rubenson and Snee 2011); (iii) universities can make decisions regarding their internal organization and processes (Jones et al. 2001), and (iv) universities have traditionally enjoyed leeway regarding how to spend their funding, due to the use of block grants for allocating funding (Kymlicka 1982). Given that tradition of autonomy, the MYA/MYAAs were immediately perceived as an instrument that could potentially strengthen the government's influence on institutional behaviour. It collapsed "system-level accessibility, accountability and funding goals into one mechanism" (Shanahan 2009: 7) and allowed the government "to exercise a degree of control over post-secondary institutions that did not exist before" (Clark, Moran, Skolnik and Trick 2009: 128). A dissenting opinion that was expressed pointed out that the potential threat to university autonomy could be counterbalanced by instrument design. The proposed approach gave universities the opportunity to present their objectives as individual institutions (Snowdon 2005) and recommended that government demands for accountability be confined to "the collection of data on a limited number of performance indicators" (Skolnik 2005: 9). This paper analyses how Ontario universities were impacted as a result of the new requirement and assesses whether the aforementioned misgivings of reduced autonomy materialized. That is, to what extent did the introduction of the MYA/MYAAs modify the relationship structure between Ontario universities and the provincial government? The analysis focuses on understanding whether the type of instrument used -that of a bilateral agreement -helps explain the effects observed. The paper is organized as follows. First, the theoretical framework guiding the discussion is presented. Second, a description of the Ontario MYA/MYAAs and their evolution is provided. This is followed by an overview of the methodology that was used for this research. The paper continues with a discussion of the effects produced by the MYA/MYAAs, including: priority alignment, recognizing the value of sharing stories, and the acceptability of accountability requirements. The concluding remarks discuss the challenges of implementing a new policy tool without considering its political implications. With respect to the specific instrument studied, the changes in university autonomy are negligible.

The theory of policy instruments and performance agreements in PSE
Given that one core government objective is to steer universities and to ensure priority alignment, the choice of a performance agreement as the instrument to implement such objectives is justified from a technical perspective, as these tools are conceived as conducive to steering and to convergence between organizational and policy objectives (García de Fanelli 2006;Gornitzka, Stensaker, Smeby and De Boer 2004;Jongbloed and Vossensteyn 2001). The introduction of such an agreement-based instrument, suggests a mobilizing relationship between government and universities as the former seeks direct involvement in the definition of objectives (Lascoumes and Le Galès 2004). That is, alignment to the government's priorities is the intended effect, but this is pursued under the underlying ideal of "negotiated governance," where actors can reach mutually satisfying decisions while organizations keep their autonomy vis-à-vis the state (Le . In addition, an agreement or contract, as an instrument of public action, promises to facilitate control in a sector where autonomy is a fundamental value (Lascoumes and Le Galès 2007). It promotes a shared understanding of the issues and solutions supported by an interactive dialogue between government and universities, and it is expected to build trust (Massy 2011).
Furthermore, the introduction of new accountability instruments is often framed within a normative perspective of principal-agent relationships, where the government's objective to exercise control in the sector is considered legitimate (Enders, Boer and Weyer 2012). Moreover, being one example of the introduction of "better management techniques" under the New Public Management (NPM) banner, such instruments are considered politically neutral and are not expected to distort political relationships (Pollitt 1993).
On the other hand, the theory of instruments within the political sociology tradition, questions the assumption of neutrality mentioned above. The approach purports that the analysis of such a process should go beyond the technical considerations of instrument choice by asking how broad are its effects -independently of its stated objectives -how the instrument is chosen and how its implementation structures the relationship between actors (Ferlie, Musselin and Andresani 2008;Le Galès 2004, 2007). For instance, the initial degree of autonomy in the relationship between government and institutions, as well as the autonomy values embedded in the instrument and the negotiation, may limit what the instrument can accomplish (Reale and Seeber 2012). In addition, the extent of priority alignment when the agreement is first introduced is an important determinant of the strategies universities will chose to implement, of their willingness to embrace the new requirement, and of its consequences (Maassen and Gorniztka 1999).
According to Lascoumes and Le Galès (2007), some possible effects resulting from the introduction of an instrument include how the new tool: (i) … will create uncertainties about the balance of power, for instance regarding concerns about university autonomy and government control; (ii) … eventually privileges certain actors and excludes others, since power may be redistributed (Kassim and Le Galès 2010); (iii) … will partly determine the way in which actors behave, as instrument choice helps explain how the actors involved choose to implement new accountability requirements; (iv) … constraints actors while offering them new possibilities which, in turn, will help explain implementation choices by both universities and government; and (v) … drives forward a certain representation of problems. That is, given that instruments shape the political discourse (Ibid.), their introduction may contribute to the institutionalization of how issues in the Ontario PSE sector are represented. In addition to those possible effects, this paper takes into consideration that public performance reporting and equivalent initiatives have often times become an end in themselves, perceived as an indication of good governance (Bovens 2005;Stensaker 2009) used primarily for informational and symbolic accountability purposes (Gregory 2003;McDavid and Huse 2008). Some accountability initiatives in PSE, in particular, have been characterized as symbolic, especially when "shallow and mechanistic approaches to the promotion of responsible behaviour and the enforcement of accountability" are privileged (Thomas 1997: 142) and when the changes they entail are primarily rhetoric and not intended to sanction individuals or their activities (Huisman and Currie 2004). The potential for superficial accountability indicators and measures is exacerbated in

Revue Gouvernance
Volume 13, numéro 1, 2016 54 PSE, where the limited understanding of its "production process" contributes to the appeal of measures that appear good despite their inability to capture the complexity of the higher education process (Brunetto and Farr-Wharton 2005;Lemelin 1999). Nonetheless, even symbolic and ceremonial changes are not necessarily inconsequential (DiMaggio and Powell 1983), particularly when the instrument dictates how actors behave, provides both constraints and opportunities, and formalizes a certain representation of problems (Lascoumes and Le Galès 2007). Using this framework as a guide, the analysis in this paper looks into the effects of introducing the MYA/MYAAs regarding the extent to which autonomy was affected, priorities were aligned, and what new opportunities were created for the sector.

The Ontario MYA/MYAAS
After the 2005 announcement, the government introduced the Quality Improvement Fund (QIF), which provided about $124 million in new funding to universities for the 2005-06 fiscal year. Institutions receiving money from the fund were expected to "sign accountability agreements that set out how the money is to be spent and the expected results from these investments" (MTCU 2005). Therefore, one-year interim accountability agreements (IAA) were requested and produced in the fall 2005. The QIF had specific objectives and encouraged hiring of new faculty and staff, more resources and equipment, and introducing better student services, while supporting each institution's differentiated mission (Ibid.). The interim agreements explicitly requested institutions to report back on those priorities.
In the summer 2006, full blown action plans to support the new funding distribution were requested from universities in the form of a Multi-Year Action plan (MYA) for the threeyear period 2006-07 to 2008-09. The MYAs were "intended to outline the government's commitment to stable funding, articulate each institution's commitment to accessibility, quality improvements and measurements of results, and tie the commitments to results. The agreements define[d] goals and system-wide measurements alongside institutionspecific action plans of indicators and quantifiable targets" (Iacobucci 2009: 1). The Ontario agreements were narrower in focus than similar agreements implemented in other jurisdictions where their broader reach has included research activities, enrolment targets and greater funding allocations (Gornitzka et al. 2004). The measures were elaborated collaboratively between universities and government via consultations. Given the great diversity in information disclosure across Ontario universities and the absence of a sector-wide data infrastructure, one main objective of the new instrument was for the government to capture university information with a view to measure the pulse of Ontario's PSE.

Revue Gouvernance
Volume 13, numéro 1, 2016 55 As of fall 2007, universities were required to produce annual report-backs on their commitments and outcomes (MYAAs), including: (i) institution-specific, multi-year access strategies, indicators and results regarding outreach and access for Aboriginal peoples, persons with disabilities and first-generation students; (ii) information on the Student Access Guarantee; (iii) institution-specific multi-year quality strategies, indicators and results (specifically, information on staff and faculty hiring initiatives); (iv) student engagement and satisfaction information; and (v) undergraduate student retention figures.
Although the general guidelines for the agreement and report-backs were the same for all universities, other than the specific requirements on a few specific data collection tools (e.g., CGPSS, CSRDE, NSSE), 1 the templates offered leeway for universities to decide what strategies to identify as part of their commitments as well as the metrics to report on progress. In addition, given that it was a budget announcement, much emphasis was placed on getting success stories out the door fast. The original MYA and subsequent report backs (MYAAs) are referred to in this paper as "the first-generation." In 2008-09, as the end of the three-year plan neared, the Ontario government extended the original MYAs for another year. A period of uncertainty followed, during which the fate of the annual report backs as well as the renewal of the MYAs was unknown. Nonetheless, the annual report backs were not abandoned and MYAAs were diligently submitted every year. In 2012-2013, with the introduction of strategic mandated agreements (SMAs), the MYAA annual report-backs became institutionalized as a data reporting tool -these are referred to as "the second generation." The timeline and evolution of the Ontario agreements is summarized in the

Implementation and priority alignment
The new instrument appealed to the dominant discourse regarding the relationships between government and universities, particularly since universities perceived that their involvement in bilateral negotiations would help keep government control in check. The pervasiveness of this perception is evident in the choice of an agreement or contract as the preferred tool in all of the proposals that were put on the table in Ontario at the time when the MYA/MYAAs were first being considered (COU 2004;HEQCO 2009;Rae 2005). The agreements were thus expected to structure the relationship between actors in an acceptable manner coherent with the governance paradigm dominating the sector. Contrary to the expected behaviour, the implementation process of the first-generation MYA/MYAAs was characterized by limited negotiation between government and universities that focused mostly on technical consultations, regarding data definitions and other minutia, without much emphasis on a conversation to promote priority alignment and joint determination of adequate goals.
The implementation process was characterized by ambiguity of the government's objectives and pervasive uncertainty. There was widespread uncertainty regarding the funding that was going to be obtained and its timelines; also, although the agreements were institution-specific, the government's overall objective was to demonstrate performance at the system-level. This ambiguity would nuance the degree to which the new requirements were met. Given the lack of clarity on expectations vis-à-vis what was considered an acceptable commitment, universities had significant leeway in choosing what ideas to put forward. Also, universities are willing to give up some autonomy in exchange for more funding. However, reticence is exercised when the funding is not provided in a timely fashion or it is uncertain, thus universities act strategically to protect their core activities.
This translated into great diversity in reporting strategies and commitments observed across implementation sites, particularly regarding the scope of the initiatives and the

Revue Gouvernance
Volume 13, numéro 1, 2016 58 ambitiousness of targets, albeit the general representation of issues at hand was quite similar across universities. During the first two years of implementation, the government put significant pressure on universities to develop stretch targets that demonstrated improvements, even in cases when their performance was already of high caliber. This was an initial effort to be more controlling and directive. Universities, on the other hand, privileged initiatives that were easily attainable or already planned for, and the majority minimized the use of stretch targets to manage risk. This strategy was possible given the general character of the government's objectives and the flexibility allowed to universities to answer to the new requirement. As the process evolved, the government's follow-up strategy focused on monitoring and shifted away from evaluating the objectives and targets specified in the MYAs. Sanctions were not used and the evaluation focused on simplistic and technical matters of the reports.
In addition, the strategy for which indicators were developed was often a shortlist that did not represent fairly all of the activities underway on university campuses that jointly contributed to both internal and Reaching Higher's goals. The measures and strategies selected appeared to be "those most readily available or those that are the easiest to collect, rather than those that are most important to the mission or goals of the institution" (Lewis, Hendel and Kallsen 2007: 210). Thus, universities reported for the most part on activities that they had already initiated or were underway, often extracting from existing strategic, planning and academic plans, therefore, the extent to which priority alignment permeated within the university internal structures was limited.
This weakened effect was counterbalanced by the priority alignment at the outset of the initiative vis-à-vis some objectives, such as hiring more faculty members, improving retention and developing quality initiatives. Continued alignment for the core objectives of quality and accessibility was further facilitated by two factors. First, the MYA/MYAAs' broad, high-level themes are congruent with the mission of universities (e.g. ensuring the quality of education). Such coherence contributed to the acceptance of the new instrument given that policy changes are more likely to be contentious when "they threaten what actors perceive as their organizational core policy mission" (Durant 2008: 291). Second, the Reaching Higher's objectives focused their attention on system-wide priorities that were not interpreted as constraining and allowed room for interpretation. That is, general goals were valuable as they reduced conflict and were more readily accepted by sector stakeholders (Kübler and De Maillard 2009;Matland 1995;Sabatier 2008).
However, the absence of system-wide commitments obliged the government to rely on each institution's identification of their contribution. That is, "for contracts to be successful, governments must be willing and able of stating their policy and program

Revue Gouvernance
Volume 13, numéro 1, 2016 59 goals with sufficient precision to set direction and monitor performance" (Thomas 1997: 157). The potential for political conflict from establishing system-level goals with institutional commitments to contribute to such goals and from the risk arising from committing to multi-year funding plans superseded the rationality implicit in the instrumentality of performance agreements. Given that clear government objectives were never established or remained ambiguous, the government had no choice but to accept what universities proposed and their control was not significantly increased: universities continued with business as usual or as planned. At the end, the relationship between universities and government changed only slightly.
This finding concurs with the documented experience regarding performance agreements in other jurisdictions. For instance, it is reported that internal faculty and staff members have limited knowledge of the initiatives and goals identified in the contracts, therefore limiting their commitment (García de Fanelli 2006; Vilalta and Brugué 2010). This results in part from the challenges to coordinate activities within universities, given that the "distribution of decision-making responsibilities and the degree of institutional fragmentation are important factors conditioning the extent to which co-ordinated change […] is possible or likely" (Maassen and Gornitzka 1999: 302). Complementarily, it is observed that "systems implemented to satisfy external requirements are less likely to influence internal behavior than are those implemented to satisfy the organization's own needs" (Cavalluzzo and Ittner 2004: 244).
There is one exception regarding the limited changes in alignment so far discussed, given that the requirement included new and very specific topics that had not been dealt within an accountability arrangement in the past, such as under-represented students. The focus on specific areas of interest was, in the opinion of some stakeholders, an important change in the way the government had related to universities before. It is why universities took time to reflect on the activities that they were undertaking in the context of the government's more granular priorities. This emphasis, coupled with the provision of targeted funding envelopes, is believed to have contributed to universities articulating more strategies to support first-generation and Aboriginal students, which may not have been a priority before for some institutions. It also encouraged university representatives to be more aware of the work being done internally, and as a result of the additional funding, some existing programs were expanded. As a result, some degree of isomorphism was observed across universities in this area, and resulted in the intangible effect of awareness creation.

Recognizing the value of stories
A significant amount of information regarding accessibility and quality of learning became available via the MYA/MYAAs. However, the majority of the information was narrative, and when quantitative indicators were produced, the lack of common definitions limited its comparability. Although many sector stakeholders indicate that the information collected is unreliable and that it has limited usefulness given unclear definitions, government representatives assess positively the usefulness of the information obtained. Government officials even argue that in the absence of an instrument such as the MYAAs, they would have to create something equivalent, which echoes the observation that government representatives tend to comment positively and strongly support initiatives in which they have a vested interest (McDavid and Hawthorn 2006;Thomas 2007).
Indeed, despite the data comparability limitations mentioned above, ministry-level officials report using the information compiled via the MYAAs to write provincial summaries, to share sector-specific information with other government agencies/ministries, to respond to media enquiries, and to provide background information to senior government officials. Thus, the information obtained contributed to meeting internal accountability objectives, and to answerability from a perspective of ministerial responsibility, that is "the obligation of ministers to provide information and explanations to Parliament concerning activities within their portfolios" (Thomas 1997: 144). The MYAAs, particularly in their second generation form that emphasized more standard data collection, were regarded by ministry representatives as valuable tools to gather information in a consistent fashion across universities. Moreover, their reported utilization was not limited to the factual information communicated within the government, but MTCU staff report that it also informed government activities, by identifying system-wide issues and shaping government thinking that eventually translates into policy. In addition, the perception that the MYA/MYAAs were read by individuals not associated with the government or universities was pervasive in the government representatives' discourse. These agreements were deemed to have contributed to a better public understanding of the Ontario university system and of the differences across universities, thus improving decision-making processes. However, no evidence regarding the use by individual students or parents was found by this study.
Non-governmental stakeholders agree that ministry representatives obtained tangible benefits from the MYA/MYAAs. However, these benefits are not derived from the quantitative or factual information, but rather are found in the narrative and in the stories that are collected in a consistent format across universities that reduce compilation efforts. In particular, the narrative that linked Reaching Higher investments to success

Revue Gouvernance
Volume 13, numéro 1, 2016 61 stories at university sites resonated with the government in the early years of the initiative. Beyond the availability of stories, university stakeholders cast doubts on the usefulness of this exercise, as they have no evidence of whether and how the information provided in the agreements worked its way back to policy, and government media releases using the information were quite uncommon. Even its usefulness as a government monitoring exercise is questioned, as participants in this study have not seen the information collected in government annual reports or other similar public documents. Therefore, there is disagreement in the sector regarding the extent to which the expected uses of the information, both factual and narrative, were actually realized.
Although it is argued that the provision of information as a result of accountability requirements contributes to the legitimacy of post-secondary institutions (Harvey and Stensaker 2011;Maingot and Zeghal 2008;Ryan, Williams, Charles and Waterhouse 2008;Trow 1996), particularly when this information clarifies how post-secondary institutions make an economic contribution and produce value in a knowledge society (Harvey and Stensaker 2011), the provision of information observed in the Ontario case is at a much more primitive level, more concerned with counting things, rather than understanding and explaining the societal value of PSE. Therefore, from a legitimacy perspective, the initiative played a more important role facilitating the legitimization of government initiatives than those of universities. These information collection processes were established to meet external demands of accountability, creating an image of rationality and efficiency without having a real objective of influencing internal decision making (Scott 2008;Taylor 2009).
As a consequence, the implementation process of the MYA/MYAAs is portrayed as one that privileged rhetorical approaches. Nonetheless, even rhetorical changes are not inconsequential as they can reflect shifts in values, assumptions and culture at the organizational level (Boyce 2003). The introduction of particular policies reinforces certain beliefs thus changing the interpretations of actors of their reality and consequently their actions (Watts, McNair and Bard 2010;Rosen 2009). In that regard, as a result of the MYA/MYAAs, Ontario universities appear to be more aware of the value of providing stories to government to justify their investments, as well as of the value of reporting to a broader audience beyond the government. There is thus a greater predisposition to sharing stories and to recognizing the potential benefits from collaborating with the government.
The benefits from collaboration arise from universities and MTCU representatives having a shared agenda of justifying the benefits and outcomes of increased funding to the sector, to jointly produce a positive impression within the government, and to ensure continued access to financial resources. These benefits are compounded when provincial financial resources are limited, as further investments in the PSE sector must be justified relative

Revue Gouvernance
Volume 13, numéro 1, 2016 62 to other competing priorities. Although the value of such collaboration was recognized in the past (Shapiro and Shapiro 1994), the change observed with the introduction of the MYA/MYAAs is that these benefits of collaboration appear to have permeated the administrative structure of universities to the offices having a direct hand in the information shared. University administrators have realized that if they tell their story, they will exert control of the way it is presented and thus contribute to the legitimization of their funding needs and influence access to returns of a financial nature. Thus, the story is constructed in such a way that the solution, in this case, is greater funding to improve quality, and is part of the narrative, taking the shape of a casual story (Radaelli 2010;Stone 1989).
A second motivation for universities to willingly share success stories with the government is that the relationship between universities and government is uneven across Ontario. Some universities, due to their size, geographical location or political leverage, have closer contact with and greater influence on government activities. Therefore, the legitimatization value of the MYA/MYAAs was greater for smaller and less influential universities, as university representatives perceived the original MYA plan as an opportunity to share their story with government officials and as a framework for discussing their needs, their priorities and their plans. In that sense, the new instrument was embraced by universities with a view to advance their own policy goals (Durant 2008) and to communicate a coherent and shared vision of what universities are and do.

Acceptance of accountability requirements
The Ontario MYA/MYAA experience could be qualified as symbolic to the extent that, according to several sector representatives, the mere existence of the report appears to have been enough to satisfy the government requirements. Examples of the limited interest on content include that some activities reported by universities are based on information provided by the government, and that the contextual information that would allow assessing whether performance improvement was observed or not was generally lacking. Furthermore, although universities were required to identify initiatives and define targets for their outcomes, the evaluation of improvement and the use of sanctions for underperformance were minimal. Consequently, sector stakeholders describe the process as mostly rhetorical, and even irrelevant.
Nonetheless, some stakeholders indicate that the institutionalization of the MYAAs facilitated greater acceptance of reporting requirements, both regarding the need for system-level data and the development of new indicators. It also motivated universities to publish and publicize their activities more broadly. This shift in perceptions at the

Revue Gouvernance
Volume 13, numéro 1, 2016 63 university-level, however, is not solely due to the MYA/MYAAs. Forces behind increased levels of information disclosure include the development of e-government that promotes a digital organizational presence (Ahn 2012;Norris 2010), fundraising efforts and the need to respond to the needs of stakeholders (Grosjean and Atkinson-Grosjean 2000;Nelson, Banks and Fisher 2003), the pressure exerted by international competition and international rankings (Maingot and Zeghal 2008), and preventing exposure to the media (Kennedy 1997). In Ontario, it is also seen as a "defensive" strategy to deal with external demands, illustrated by the creation of the Common University Data Ontario (CUDO) information website. This approach echoes the strategies reported by the Canadian Research Granting councils that "deal with the challenge of appeasing demands for greater accountability (…) by stepping efforts to communicate those outcomes that can be reported" (Sá, Kretz and Sigurdson 2013: 112).
The acceptance of such values is not without challenges and can range from an assimilation of "the concept of each university being accountable to society," as observed in Catalonia (Dill and Beerkens 2010: 289) to a less demanding requirement of promoting "a culture of transparency about outcomes" (Salmi 2009: vii). Since transparency consists of the "voluntary disclosure of information that shapes, structures or contributes to attitudes, behaviours, actions or discourse, for the purpose of ensuring that they are totally intelligible" (Bernier 2012), it is a necessary, but not sufficient, condition for accountability, as "being made more accountable [… is] not the same as being made more transparent or publicly available" (Power 2000: 117). Nonetheless, according to central stakeholders and government representatives, the internalization of the transparency requirement at universities has facilitated the transition towards new instruments that could eventually be more demanding.
The MYA/MYAAs are thus both a cause behind the greater acceptance of accountability and a consequence of it. These were introduced as a result of an evolution in the scope and reach of the accountability movement in Ontario and worldwide. However, their internalization on the part of universities illustrates their role as facilitators in the greater acceptance of accountability. The accountability movement, and the introduction and increased utilization of new instruments produced a reinforcement cycle that drove the way in which actors behaved and formalized a certain representation of accountability. In this case, the common representation is that more accountability is required.

Conclusion
This paper shows that as result of the introduction of the MYA/MYAAs, the Ontario government as well as other actors in the PSE field appear to have more access to factual information about universities, which has contributed to the legitimization of government activities. Furthermore, even when accountability is operationalized as a rhetorical tool, the benefits of sharing stories -that is, the power that is brought forward by a clear narrative -is internalized by university administrators, providing an example of organizational value shifts. Finally, some degree of priority alignment is observed between the government and universities, but the alignment was primarily facilitated by generality and ambiguity in government goals as well as the funding associated with the agreements. Furthermore, the value of the instrument as a directive and control tool is determined in part by the specificity of the objectives included, and being such specificity was practically absent in the Ontario experience, the tool was transformed in its second generation into a signaling mechanism of government's interests.
The Ontario case illustrates that both universities and government will not implement performance agreements or contracts as a politically neutral tool. That is, the proposed use of performance agreements to find areas of alignment to improve performance still face the political conflicts of the sector and are not fully implemented where priorities are not clearly established. The limitations of the top-down approach become apparent, as it ignores the politics of policy formulation and design, and overemphasizes the ability of policy proponents to structure implementation. A promising tool in theory is overturned by the politics surrounding its operationalization. An unintended consequence, however, is reduced trust on existing account-giving processes or instruments. This, in turn, reinforces the impetus to introducing additional rational-instrumental measures that are rejected by universities and thus answered to symbolically, entering into a vicious circle of accumulation of instruments that are reduced in their implementation to limited compliance. Nonetheless, the symbolic dimension of the policy contributes to the legitimization of its material aspects (Rosen 2009), illustrated by the greater openness towards additional accountability impositions that is perceived by some sector stakeholders.
The Ontario case is a clear example of how in the study of instruments, a key determinant of their effects is the context in which these are introduced, which may be conducive to a reformulation of its principles or to resistance to its introduction (Lascoumes and Simard 2011). That is, instruments are autonomous of their original goals, and policy transfer results in context-specific implementation (Reale and Seeber 2012). The use of performance agreements in PSE is becoming increasingly common. The idea of a contract in accountability promises to clarify relationships, negotiate objectives, and increase trust (Gornitzka et al. 2004;Massy 2011). However, such a promise overlooks the autonomy and the interests of universities and the impact of an uncertain fiscal environment. The continued introduction of performance agreements and similar tools both in Ontario and in other jurisdictions, not only in PSE but also in other sectors, such as health, provides evidence that the discussion is put to rest, or at least managed, by implementing a tool that conveys the message that an acceptable balance between accountability and control has been established, albeit symbolically, in a process of legitimization that ignores the limitations implementing a new instrument without due attention given to its potential effects.