Senior civil servants in Quebec and strategic information transmission: A game theory-driven approach.
Politicians issue major government orientations while public administration ensures their operationalization. From the 1960s onward in Quebec, the development of a technocracy and the introduction of unionism into the public service have reinforced the administrative system and contributed to the bureaucratization of society (Gow 1986). Senior civil servants in Quebec have repeated contact with politicians, as both engage in crafting public policies. According to Bourgault (1983), senior civil servants develop strategies that institutionalize their strength: in preparing, establishing, and ensuring the execution of laws, they also transform the distribution of power in government, for formulating public policies. Still, the primary responsibility of the civil service is providing information and advice to ministers (Savoie 2003). The interrelation of politicians and senior administrators implies that "both sets of actors need one another, each bringing special skills, capabilities, yet limitations to the decision-making process" (Johnson 2002).
This "classical" conceptualization (1) of the Quebec political-administrative interface has recently been called into question. On September 14, 2012, Jean-Marc Salvet, journalist at Le Soleil, reported on the actions of the Assistant Deputy Minister of Health who, while serving as the Deputy Chief of Staff of the Prime Minister, recommended judicial appointments (judgeships) based on the partisan etiquette of the candidates interviewed. (2) More recently, on May 18, 2016, Le Soleil's reporter Simon Boivin, flagged that the Deputy Minister of Transport had refused to comply with an important request of his Minister. As the latter wanted to correct certain irregularities in awarding public contracts, the deputy minister decided not to review the organizational chart of the ministry. (3) These situations raise questions about the exchange of information within the politicoadministrative interface. In this working environment reputed for time compression and limitation, if strategic behaviour is indeed occurring, what form could it take between senior civil servants and ministers? How does political advice get done?
On May 18, 2016, Le Soleil's reporter, Simon Boivin, flagged that the Deputy Minister of Transport had refused to comply with an important request of his Minister. As the latter wanted to correct certain irregularities in awarding public contracts, the deputy minister decided not to review the organizational chart of the ministry.
Much work has been devoted to studying the relationship between senior civil servants and politicians in the Canadian public administration. Some notable examples are Dion's (1993) work on the origins of Canadian bureaucracy, its difficulties, and the solutions to remedy some of its "excessive" characteristics, Blais et al.'s (1997) study of the influence of political ideology on the growth of the public sector, Good's (2003) research on the then still-recent implementation of New Public Management, and its effects on the relationships between public servants and politicians, and Bourgault and Savoie's (2009) work on the evolving role of senior civil servants and their adaptability to changes in Canadian political life. Bakvis (1997) studied advice received by the executive branch coming from bureaucracy, but also from other sources, including external bodies to the government such as think tanks or other policy institutes and private consultants. Savoie's (2003) work highlighted how difficult it can be for administrators to reconcile their subordinate position (with respect to political officials) with the impetus to preserve professional independence. Building on this foundation, Bourgault (2009) elaborated on the special links between administrators and parliamentarians given the Canadian regime.
However, less attention has been devoted to the relationship between elected officials and civil servants in Quebec's public service. Some studies have shown that the two systems are different and that researchers need to better understand distinctive qualities of Quebec and Canada's political-administrative interface (e.g. Clark 2002). According to Fortier et al. (2016), Quebec's administrative appointment process points to a certain level of distrust in bureaucracy. Formal politicization strategies (e.g. using partisan criteria instead of merit to select, retain, promote or sanction civil servants (Peters and Pierre 2004)), is evidence of a distrust. As Fortier et al. (2016) stated, today, formal politicization is considered "normal" in Quebec, even though undesirable. (4) This practice highlights the need for empirical studies about the relationship between public servants and politicians in Quebec that is based on updated, first-hand data.
Although elected officials establish the broad guidelines for the day-to-day tasks of civil servants, the latter seem aware of their strategic position with respect to influencing political decision-makers.
What follows focuses on the concept of strategic information transmission, using a game-theoretic perspective. We have do so for two purposes. First, help to grasp what is considered a "strategic" behaviour between a minister and a senior civil servant as part of an informational exchange while considering the various ways scholars have approached the concept of strategic information transmission. Second, it suggests questions for interviews (see Appendix 1). It also prevents the researcher from anticipating strategy not necessarily born out and allows us to better understand what to expect when looking for strategic behaviours, and to look for behaviours not associated with the traditional conceptualization of public officials executing rational-legal duties (see Bourgault 2009).
By documenting current practices of civil servants, this study seeks to better understand their behaviour when called upon to assist politicians. Relying on semi-structured interviews with administrators in the Quebec government, the study scans the discourses of incumbent senior civil servants for indications of strategic behaviour in information transmission. Relying on insights from current public officials is the most effective way to obtain descriptive data that accurately depict their day-to-day activities. While interviews with former civil servants can provide reliable information about certain aspects of public service normally kept quiet because of the duty of confidentiality and reserve, interviewing incumbent civil servants was the method viewed here as the most capable of minimizing retrospective biases. The study has three specific objectives: 1) develop a descriptive sketch of key traits of senior civil servants' from the interviews, 2) explore the ways strategic behaviour could manifests itself in Quebec's political administrative interface and 3) refine understanding of contemporary informational arrangement in Quebec senior public administration. This study will also foster better understanding how recent publicized incidents (See Endnotes 2 and 3) that have affected public perception of public administration.
Building a survey with formal models
The study draws on the formal literature on strategic information transmission to derive game-theoretic expectations (see Table 1) and shape an interview questionnaire. Even though the theoretical literature on strategic information transmission might be hard to understand, Bechard (2017) argues it offers insightful foundation for practical implications. Gailmard and Patty (2012) argue that, while previous research of formal modeling on public administration focused on bureaucracy as a tool to implement policy commitments, recent literature has instead emphasized "information asymmetries" between bureaucracies and political authorities as a productive focus for understanding public decision-making. Decision-makers (administrative or political) very rarely possess all potential and desirable policy-relevant information. As a consequence, administration in many circumstances can be accurately described as an agency-problem (Gailmard and Patty 2012), making formal models a useful way to capture the logic of civil service and policy advising.
Other frameworks could have been used to assess strategic information transmission in the political administrative interface. For example, the bureau-shaping model (Dunleavy 1991) questions the archetypal budget-maximizing bureaucrat claiming that senior civil servants invest in their status and the quality of their work as individuals. Such utility maximization is best achieved through bureau-shaping rather than budget maximization (Marsh et al. 2000). Hood and Lodge (2006) propose the "public service bargains" model, a typology with four bargaining positions (fatalist, hierarchist, individualist, egalitarian) between senior civil servants and the political staff, analyzed against three differentiating parameters: reward, competency and loyalty. This model has been used to evaluate changes in the public service as a result of New Public Management (see Bourgault 2011). Another notable example is Svara's (2001) conceptual model of "complementarity," which characterizes the interrelation of elected officials and administration as "the interaction between political control and professional independence," where politicians value respect to administrators, while the latter implicitly commit to accountability.
Interview questions are primarily based on Gailmard and Patty's (2012) review of formal theories of public administration. A branching procedure on their list of references was employed to capture the broader literature. Theoretical results emerging from the "cheap talk" and "verifiable information" sub-literatures was given preference as they constitute the dominant findings in the field. These models cover a range of heterogeneous situations that can be found in public service. Additional precautions were taken to propose a review consistent with recent writing in the discipline.
Proposed by Crawford and Sobel (1982), this model assumes the status of the issuer of information is not affected by the information sharing and the receiver does not possess the necessary tools to verify transmitted information. The latter acts by considering the information received as an object that he is free to use or not. However, the decision he makes will have an impact on the well-being of information provider and receiver, hence the strategic dimension of the model. This study generated significant subsequent interest about "information asymmetries" in public administration.
Shared language through interactions
Farrell (1993) proposes that shared language might be a suitable to the extent that the decision-maker and the provider of information have similar preferences and sufficient time to develop a common lexicon, the latter would be only very weakly seduced by the idea of lying, given that he would profit from a relationship that guarantees his or her interests. Aumann and Hart (2003) explore the idea of extending the transmission phase and develop a model called "long cheap talk." The authors demonstrate that a temporal window that allows for a sequence of exchanges can lead to substantial information sharing.
Krishna and Morgan (2001) show that when providers are numerous, the receiver of information can interact with them on several occasions. Therefore, if providers of information have dissimilar preferences among each other, the revelation of nearly complete information is guaranteed. Through argumentation, information that otherwise would have been inaccessible emerges.
Research relying on cheap talk to formalize information transmission as a principal-agent relationship led to a major finding in the field: less information is transmitted when the preferences of the provider and the decision-maker diverge. This finding about the impact of the discordance [or accordance] of both issuer and receiver of information's preferences is now widely known as the "ally principle." (5)
The ally principle
Dickhaut et al. (1995) argue that when ideological preferences are opposed, the ability of providers of information to co-ordinate the exchange decreases, resulting in weak information transmission. Bergstrom and Lachmann (1998) demonstrate that when a certain contextual configuration is obtained, the cost of disclosing information might be dissolved. If several providers establish a prior consensus on which information to transmit, any negative or positive externalities resulting from transmission will be allocated among them. Thus, if the interests are the same, it is possible for the providers of information to act strategically by transferring a single response. In doing so, the decision-maker cannot compare and adjudicate between different responses and then must conclude that the information is accurate.
Program vs budget focus
Bendor et al. (1987) refer to the "programmatic" provider of information as an individual with a strong interest in the development of public programs and policies. This conceptualization stands in contrast to the "budgetary" provider, who is functionally motivated by the acquisition of additional budgets to support the development of his or her organization.
Verification and bias
Other authors have identified situations where decision-makers are free to impose a cost on information providers. Grossman and Hart (1980) assumed a state of competition where multiple agents [own varied levels of information] are available for the decision-maker to consult. "Verifiable information" describes situations in which the information can examine the informational contents delivered to him, with the transmitter modulated the transfer according to the nature of the information that he holds. Patty (2009) explains that if a receiver of information seems sufficiently biased, an external threat to the organization might stimulate the provider to collect information in a way that affirms the former's bias. Because a strong bias is viewed as a signal of obstinacy, the provider will collect information according to the likelihood that the decision-maker will use it. Jervis (2015) argues that if the information is collected according to a decision rule favoring a decision-maker's bias, he may simply accept information that corroborates his or her initial expectation of what the answer to the problem at hand should be. Okuno-Fujiwara et al. (1990) evaluate the relevance of a preliminary phase of information exchange between providers of information, before any transmission takes place with the decision-maker. Their results suggest that a provider can mediate the perception of his or her colleagues before the transmission happens.
Game theory is a tool to model interactive decision making. Each formal model implies specific parameters in payoffs-related interaction between players. While game theory assumes common knowledge (i.e. both players know the specification of the game, know that both players know it, know that both know that both know it, and so on) and rationality (i.e. players always choose strategies that maximize their individual payoffs), it is of interest to bridge the gap between formalization and practice in order to better understand when the latter respects the theory and when it departs from it.
Today, no research has explored strategic information transmission using testimonies from incumbent senior civil servants in Quebec. It is difficult to collect data from public servants, held to the highest standards of confidentiality and discretion. Still, in-depth conversational techniques that rely on interviews with elites remain a necessity to illuminate the complex nature of the political administrative interface (Aberbach and Rockman 2006). We used insights from a diversity of game theoretic work to build a composite questionnaire to articulate a semi-structured interview procedure.
Strategic information transmission may be defined as an informational exchange that occurs between two providers of information where strategic effects make the disclosure of all information not the usual most advantageous policy, although a completely self-interested provider will frequently find it advantageous to reveal some information (Crawford and Sobel 1982). A qualitative investigation of strategic information transmission requires quantification (i.e. formal modeling) along two dimensions: the "strategic" nature of the behaviour and the type of information disclosed.
Since the added methodological value of the study is to be inspired [but not limited] by game-theoretic work, some behaviours were regarded as "strategic" in the data coding solely when they seemed to lack the degree of objectivity requested by Quebec Ministry of the Executive Council in the code of ethics of the public service: "[...] an official must disregard his or her personal opinions in order to carry out the task with the necessary objectivity" (MEC 2003). To that extent, strategic behaviours could encompass both the situational (i.e. actions taken to trigger a change or preserve the political situation) and corporate or organizational (i.e. actions designed to benefit the organization) aspects of information transmission.
Participants were interviewed in their working environments. Interviews were conducted between February and April 2016 and lasted about one hour. To qualify, participants had to formally play an advisory role to politicians (i.e. occupying a designated place in the weekly ministerial meetings with the minister and his or her political staff), be accountable before the minister for specific public policy projects, and lead an administrative unit. Taking the hierarchical structure of Quebec's public service into account (for a typical example, see Figure 1), the sample was limited to assistant deputy ministers (ADMs) with the most interaction with ministerial offices (not to mention the entitled deputy minister), and responsibility for particular public policy schemes. We excluded professionals in government bodies focused on consultation, regulation, and administrative management, as well as those working in administrative tribunals, public enterprises, and public institutions, as defined by Garant (1995). Only ADMs from vertical departments (i.e. pursuing objectives relatively independent from other government offices and working for specific dimensions of social life) were retained in the sample because their work was, by definition, related to the development of public policies.
A senior civil servant's entire team must agree on the presented option. Moreover, like in a theatrical production, the administrative team engages in dress rehearsals to come to a decision on the most-preferred option before presentation.
The saturation criterion was selected for interrupting data collection. According to Fusch and Ness (2015), "if one has reached the point of no new data, one has also most likely reached the point of no new themes; therefore, one has reached data saturation." Data saturation is not about securing a certain number of observations (Burmeister and Aitken 2012), but rather, about reaching the point where nothing new emerges from the data. Guest et al. (2006) observe that "data saturation may be attained by as little as six interviews depending on the sample size of the population." Saturation was attained by confirming a sample of 12 assistant deputy ministers (n = 12). A preliminary sample of 80 ADMs met the selection criteria. From those individuals, a group of 30 candidates was selected by randomization and invited by letter. In this group, 16 ADMs had refused to take part in the study, while 12 others had accepted to participate. We did not call back the last two participants since saturation was met with a sample of 12 ADMs. In the selected sample, 2 ADMs had worked previously in a minister's office as political appointees; the other ADMs had no prior experience in an MP's office. Confidential testimonies were collected using a short introductory scenario that contextualized the discussion. At no time could participants discern that the goal of the interviews was to assess strategic behaviours. Instead, they were approached (i.e. by official letters and during interviews) with the more general idea that the study would provide a better understanding of the "internal mechanics" of Quebec's political administration interface.
Data analysis was performed using QDA Miner. Following the labelling of text fragments, quotes were selected to describe strategic use of information transmission. Those excerpts have been translated from their original French. Because translation can influence the content of the qualitative data, additional attention was paid to ensuring congruence between the two languages. Although these illustrations are insufficient for inference, they provide substantial insight into how senior civil servants perceive their roles with respect to politicians and help develop a valuable narrative about the strategic behaviours civil servants engage in within the Quebec government.
An evolving portrait of civil service
According to Bourgault and Dion (1989), the permanence of senior civil servants in Canada is a longstanding convention and thus a trait of the political system in general. Nevertheless, senior civil servants have increasingly shorter professional lifespans. They tend to have many academic degrees from foreign and Canadian universities and managerial capabilities. Civil servants have also become decreasingly partisan over time (Bourgault and Dion 1991). Our descriptive account of the senior civil servant in Quebec corresponds to some extent to these characteristics. Civil servants in Quebec work for several years within the government and rise over time through a hierarchical pyramid. They are career civil servants (i.e. derive from the European tradition, a career system is a working environment organized around a hierarchical structure that provides several levels of employment for the public official to spend most of their professional life).
Bourgault (2011) points out that from 1977 to 2010, the proportion of PhDs amongst federal senior civil servants dropped from 30 per cent to 12 per cent, "but 76 per cent of the 2010 contingent hold two or three university degrees." Nowadays, federal senior civil servants seem to be "general practitioners." In short, there was a significant drop in sector specialists and an increase in the proportion of senior officials with a management background (Bourgault 2013). Data show that Quebec civil servants have post-graduate degrees in social sciences fields and related disciplines (political science, administration, psychology and economics) and are more generalists than specialists.
Given the length of most civil servants' careers, these administrators tend to have worked under various political parties and have a basic apartisan mode of operation. A senior civil servant in Quebec seems to conform to the characteristics of classical bureaucracy. Although elections and the quick thinking they entail do not affect administrators the way they do politicians (Morin 2011), when asked whether administrators bear an active interest in political issues, two thirds of our participants (8 out of 12) answered affirmatively. Taking into account the typology proposed by Bendor et al. (1987), this may be because senior civil servants are motivated by public policy goals:
"[My motivation] is more the promulgation of a certain issue. The benefit of being here is really the opportunity of having an influential role in advancing a certain vision. If you have your own view of an issue, you're well-positioned to influence ministers on the subject, and that goes for politics too."
It seems as though there is an interesting difference between the ADMs' statements and actual practice. The deontological rule of political neutrality as stated by the Quebec Ministry of the Executive Council is as follows: "In the first place, the obligation of political neutrality implies that the official must, in the performance of his duties, refrain from any partisan work" (MEC 2003). Although elected officials establish the broad guidelines for the day-to-day tasks of civil servants, the latter seem aware of their strategic position with respect to influencing political decision-makers.
Learning the Lingua Franca
According to Farrell (1993), a common vocabulary would be a condition for a workable transmission of information in a relation where the provider of information and the decision-maker have similar preferences. Almost all participants (11 out of 12) stressed the importance of sharing a common lexicon with the elected official and his or her political staff. Assessing the testimonies gathered, the "technocratic" vocabulary civil servants and ministers use is complex and almost cryptographic for the uninitiated. This is not surprising, since specialization leads content to become more "jargonistic." In fact, the other way around would be problematic where senior civil servants do not remain in position long enough to get to know their domain in depth. For some civil servants, it seems important to foster an operational language with politicians, a lexicon that appears to reinforce itself over time; by repeatedly encountering common obstacles over the course of their professional practice, senior civil servants and politicians have the opportunity to improve their correspondence.
When elected officials do not share the viewpoints of their ADMs, the personalities of the former can significantly impact the efficacy of the information transmission. Following a change in government, however, if politicians demonstrate some ideological openness, civil servants could have a window of opportunity to use their position to influence the new minister by informing him or her of things they personally consider important. Using words that sometimes might seem derogatory (e.g. "contaminate"), senior civil servants seem aware of the didactic role they can play, especially when it comes to political appointees. Nonetheless, since an open communication channel is a prerequisite for effective information transmission, it does not seem abnormal that senior civil servants conscientiously assume a pedagogical role in informing political staff:
"So, of course, when a new political staff [is brought in], it has to learn the organizational culture, the vocabulary. As such, the more we meet [staff of the minister's office], the [more] we contaminate them when it comes to vocabulary, the realities of government, and all that. I have a member assigned to the tasks I manage. I met her once a week, a regular meeting. Then we review the [work of the week] and I tell her about substantive files, and then the files that the minister must be aware of and make decisions about. Thus, we work on two fronts [...] in the sense that I contaminate her like that in that I teach her what is being done."
A matter of trust
Knowing that the competence of politicians in understanding sophisticated public matters varies greatly between one another, ADMs appear able to intervene when an opportunity arises by way of a clear request from the minister or his or her political staff. Bourgault (1983) points out that the sum of the good or bad moves previously made by the senior civil servant generates either a sympathetic or a cautious attitude toward him from the elected official. Our data illustrates that politicians seem more likely to ask for help when they trust the civil servant. If the former has a good relationship with the elected official, the request for help by the minister's office enables ADMs to determine the nature of their offer:
"I feel completely at liberty to respond to what the minister asks me according to orientations that I've defined myself, but also to ask him about issues that he did not bring up with me at all, and in that way to point things out to him."
"Yes, if a relationship of trust has been established with a political staff and a minister, the situation is certainly one in which all kinds of relationships are made easier and giving advice to the minister and his staff is easier as well. It also facilitates [innovation] and allows us [to propose] specific projects that come from the internal machinery [the department] that we want defended by the minister in front of his colleagues and ultimately at the Council of Ministers."
"Yes, we have personal projects, that we want [the minister] to defend, that are important but that are not at all on his radar. This owes a lot to the fact that, certainly with time we develop close relationships with the minister that allows us to influence things [...]."
Senior civil servants appear well positioned to generate new policy measures. Savoie (2003) states, "the great majority of new policy measures [...] bubble up from departments, and are shaped differently from those defined from the top down." Christensen (2004) observed that top civil servants "gradually go further in providing their political masters with assistance, even if this implies a break with traditional conceptions of a politics-administration dichotomy." Our data support that the amount of discretion varies according to the quality of the relationship with the minister's office. The senior civil servant seems conscious of his or her duty to advise politicians about the consequences of their actions. Some consider it an obligation:
"As a political advisor to the minister, you do not have to only look at the administrative side. You have to take into account other aspects as well. Maybe [politicians] possess this kind of reflex, [that is] to anticipate the consequences [of their decisions]. But if they did not suspect [specific consequences], I think we have an obligation to mention the potential impacts [of their decisions]."
Bergstrom and Lachmann (1998) have suggested that the provider of information would like to promote adherence to common interests among peers to reduce uncertainty costs. A single message should be produced for the decision-maker. According to our data, senior civil servants are deeply invested in their work as advisors to ministers. However, they can withhold information that is not beneficial to them. Our participants mentioned they meticulously plan their proposals to ministers and, in the instances they decide to transmit this information, make sure it will not create conflict in the administrative unit. When describing their typical conduct when their advice is in conflict with a peer's proposal, some ADMs indicate they take additional time for deliberating in depth. For example:
"[Here is] what I would say to the minister: "I would like to discuss with my team again. I would need a 'second turn of the wheel/ since I'm still unsure about [some] elements." I am not saying I am willing to [stop all proceedings] because [I disagree], but simply that I would like to go back [to the drawing board] and return [with fresh ideas]."
Why do some senior civil servants feel the need to revise their proposals in light of intra-agency disagreement? Okuno-Fujiwara et al. (1990) suggest revision serves as a reunification mechanism, where civil servants group together to minimize the individual risk of incurring the costs of potential disagreement with the minister and his or her staff. The need for consensus within an administrative unit prior to transmitting information was obvious to all twelve participants. A senior civil servant's entire team must agree on the presented option. Moreover, like in a theatrical production, the administrative team engages in dress rehearsals to come to a decision on the most-preferred option before presentation:
"We effectively do a dress rehearsal. We typically say: "the minister just asked you for a special briefing. What are you going to suggest? How do we do that? What is our game plan?" We prepare, because time [with a] minister is quite rare. And you don't want to miss your shot, otherwise he is less likely to like your proposal and will consider you incompetent. So you cannot show up there poorly prepared."
"Normally, this works quite well. You know, reaching a consensus with your peers, with people on the team, the general managers of a given portfolio. Putting together a report, coming up with a recommendation, then looking at your teammates squarely in the eyes and saying: ok, we're going to war!"
Avoiding administrative struggles
According to Krishna and Morgan (2001), a debate between several providers of information in front of the principal promotes the development of consistent information. However, debate can occur among ADMs in private, and not in front of the elected official. Among our participants, 9 out of 12 indicated that there should never be a debate nor any kind of conflict between civil servants in front of politicians. Upstream, the deputy minister is supposed to have already endorsed civil servants' work. As such, civil servants resolve disputes in private in order to provide cohesive recommendations. Because ADMs deal with busy schedules, some could argue that it is done that way in order to save time since, without cohesion, the value of the recommendation may be put into question and cause additional time consuming operation. One official said "In front of the minister is not the place to express your feelings." While another said:
"Of course, the deputy minister never "loses face" to the assistant deputy minister. That is for sure. Maybe he comes back with additional information, but we don't air out our dirty laundry in front of the [minister]. As a team, we have to propose [only] one recommendation to our decision-makers."
Our data indicates that typically one cohesive recommendation is given to the political decision-maker, even though this recommendation is itself the product of several considered "behind-the-scenes" by administrators. Adams (2004) noted that briefing notes and submissions for the minister's office involves an assemblage procedure that collects all types of information beforehand. Civil servants then limit the amount of information to a few options which are progressively reduced until a solution is presented. This practice is so deeply embedded in governmental repertoire that some politicians explicitly request this approach. Such practice could sometimes be the result of a strategic behaviour on the part of civil servants or politicians themselves.
Looking at long discussions that encourage stakeholders to engage in exhaustive exchanges, Aumann and Hart (2003) show that information disclosure increases according to the duration of meetings. However, data suggests that long informational exchanges are not found to contribute to the amount of information shared by civil servants. Still, the data illustrates that ADMs could have some leverage over the schedules of politicians:
"We know whether the minister is in Parliament or [anywhere else]. We know if her driver will be waiting in thirty minutes. If I can meet her at five o'clock on Monday and if she is available until 10 or 11 o'dock because there is nothing else on her schedule, we take advantage of this."
Our data suggest that some senior civil servants are aware of the functional advantages of having privileged moments with the minister:
"From the outset, I will prepare his message. You know, his little "sales pitch" that will ensure that a project can be sold smoothly. Often, I will do this myself. Because, basically, I want him to apply what I am applying [on a day-to-day basis]."
"Basically, we want our projects to progress [properly] and be [approved] by the minister's office and ministerial committees [as soon as possible]. As such, it is also our duty to prepare [the minister] well so that he will properly convey the message."
A strong will to "help"
Despite their obligation to neutrality and reserve, do some senior civil servants nevertheless harbor partisan loyalties to political decision-makers? Robertson (1971) observed that the minister's office "is partisan, politically oriented, yet operationally sensitive" while the ADM's office "is non-partisan, operationally oriented yet politically sensitive." Interviewees disavowed any explicit partisan affiliations. Nonetheless, political loyalties could emerge in the fuzzy intersection of the administrative apparatus and the minister's office. A certain level of "politicization" would be considered as normal since "all DM [and ADM] appointments are--'political'--insofar as they are discretionary and made by a political body (minister's office) after a fairly politicized recommendation process" (Bourgault and Viola-Plante 2014). Indeed, as Peter et al. (2004) suggested, because of the appointment process, public service is inherently political and cannot be completely dissociated of its political substance. Within our sample, all twelve participants stressed the importance of acquiring a political "flair." However, developing the ability to help the elected official politically does not systematically mean that the civil servant has become a political partisan:
"We are the link between an administrative and a political machine. These two machines do not think alike. The agendas are not the same. [...] Therefore, when working with the administrative machine, we must be able to filter [information] in order to bring solutions into a political [world] [...] in certain respects, there are positions that are more difficult to support publicly and politically than others."
This illustration suggests that certain senior civil servants conceive of their roles as "creators" of political outcomes with administrative substance. Recent events suggest (e.g. see endnote 2) this could lead to the civil service being involved in promoting specific political interests. One could see such use of the administrative "machine" as conflicting with the imperative of reserve and neutrality associated with the Quebec civil service (see MEC 2003). On the contrary, others might view this as an example of a public servant fully giving the minister "fearless advices." Still, a "grey area" persists:
"We had an important meeting with major financial issues [on the table]. I asked for a meeting with the minister specifically to tell her that we were headed straight for a wall. We had to take a different direction because what she had as information was not up to date. I knew that if we were to go ahead with this public position, the minister would not be [politically] alive at the end of the day. If I know the minister is going down a road where she will have to engage with people who are in a position to put her in a difficult position, I have an obligation to warn her."
Senior civil servants as craftspeople
Patty (2009) argues that providing the minister with the methodology used to collect the data informing recommendation offers assurance to the minister that the process is indeed rigorous. Our data indicates that this practice is unusual. Those that do engage in it do so at the explicit request of the minister. Others see it as a breakdown in the transmission channel:
"But, it would be a shame if I had to justify all of our methods to the minister. At a certain point, this is neither the place nor the time to do this."
According to participants, even when the methodology used to collect data was presented to the minister, he--although potentially reassured with respect to methodological rigor--rarely uses this publicly. However, (10 out of 12) ADMs did indicate that when methods are discussed, the attention the minister pays to proposals increases substantially. Moreover, ADMs do not appear to collect information that skewed information towards an elected official's priors. Participants were fully aware of the professional duty their work entails. Indeed, in interviews, ADMs seized every opportunity to reiterate their faith in democratic principles, which they perceive as necessarily superseding the individual relationship they may have with a politician. In line with Jervis' (2015) conclusion that politicians should only be sympathetic toward information that strengthens their bias, the data show that certain senior civil servants in Quebec realize that well-disposed ministers are more likely to vie for additional resource allocation to their departments than unsatisfied ones:
"Thus, after a few months, knowing the minister's orientations, we are able to figure out where she might want to leave her mark. As such, we ensure that any public announcements that are made are in line with her interests. Ministers stay in office a maximum of two or three years. As such, they cannot reinvent the wheel. Knowing what is of interest to a minister allows us to know where she will want to argue for budgetary increases."
On the "ally principle"
Dickhaut et al. (1995) argue that conflicting ideological preferences among individuals will impede information transmission. We should expect elected officials to prefer to make decisions based on information biased toward their priors. To explore this possibility in Quebec, participants were asked if they had ever received unusually broad discretionary authority to complete a task. Moreover, they were encouraged to reveal any information they may have had about the ideological preferences of the elected official with respect to this task, which might explain the expansion of discretionary prerogatives. Figure 2 illustrates the relationship between two dimensions--the scope of discretionary authority and the congruence of ideological preferences. Below the middle axis, the ally principle is compared with its opposite. A binary coding scheme (where "1" indicates corresponding ideological preferences and "0" indicates opposing preferences) shows the number of participants indicating they had preferences ideologically congruent with those of the politician (or not). In the uppermost section, the figure illustrates the number of participants who had received unusual discretionary authority.
Almost all participants (11 out of 12) indicated that they had previously taken advantage of increased discretionary authority to carry out a specific task. The degree of autonomy varied from limited authority to "carte blanche." Judging from the sample, senior civil servants appear to take advantage of these one-off expansions of their prerogatives. Many believe that they were given increased latitude because of their years of loyal service to the minister and his or her staff:
"I really had free reign. We were working on a specific matter with the same minister [...], and we had gone through an episode together where I had politically saved his life. You say to yourself that he has total trust in you. As such, I accepted the mandate but said 'I'll do it my way, using my own approach'."
The data support the interpretation of the ally principle suggested by Crawford and Sobel (1982), Farrell (1993), Dickhaut et al. (1995) and Bendor and Meirowitz (2004). Half of the participants (6 out of 12) endorsed the notion that similar preferences led to them having greater discretionary bandwidth to execute certain tasks required by the elected official. By contrast, no participant expressed views in line with Seidmann's (1990) assumption that contradictory preferences foster increases in discretionary authority. Our findings suggest that alignment between ADMs and the minister's office preferences leads to additional delegation of discretionary authority from the latter to the former.
This article has reported on the first effort to use game theory to study strategic information transmission in Quebec's political-administrative interface. Relying on formal models lead to a data collection strategy informed by common assumptions and allowing researchers to label strategy and what he/she cannot. The data provided in this study informs closer analysis of the day-to-day practices of senior civil servants with respect to political decision-makers. Moreover, this exploratory research suggests further empirical scrutiny in public sector context is warranted.
The article suggests that it should not be surprising that some senior civil servants sometimes work in such a way that reflects the political stripes of the governing party or avoid sharing certain pieces of information with political officials. Indeed, the boundaries between politics and public administration in Quebec remain quite unclear. Our data show how ADMs devote a good part to politics and the development of public policies. Over time and through frequent interaction with civil servants, a shared lexicon emerges between the political staff and ADMs, which ultimately facilitate the transmission of information. Trust seems to strengthen the relationship between elected officials and senior civil servants and could lead to greater discretionary authority being given to the latter.
Public servants conscientiously uphold the principles of neutrality and reserve that are required of the public service. In addition, they appear to be "politically competent." Occasionally, this competence may give the impression that certain administrators move away from their role of neutrality and reserve, as illustrated by our interviews and recent media events. Indeed, certain civil servants view themselves as "interpreters" of the administrative language supplied to elected officials. However, they interpret only after having reached a consensus within their administrative units, since providing a unified recommendation to political decision-makers and their staff seems a ubiquitous practice among public service.
Our data indicates that typically one cohesive recommendation is given to the political decisionmaker, even though this recommendation is itself the product of several considered "behind-the-scenes" by administrators.
Civil servants do not seem to promulgate research methodology to political leaders for strategic purposes. On the contrary, they do it foremost in order to reassure ministers, and even then, only at the minister's office request. Decision-makers' political biases do not appear to motivate civil servants' information gathering approaches. However, participants acknowledged that a minister who feels his or her issues of interest are valued is more likely to spend time and effort on sustaining their organization. Our results cohere with the "ally principle," which indicates that aligned preferences between civil servants and politicians facilitate the transmission of information and increase the discretionary authority allocated to ADMs.
The article bears four limitations. First, the data are only from the civil service. Future insight would be enriched by considering data emerging from interviews with sitting ministers on information transmission with the senior civil service and identify possible strategic behaviours of both parties. Furthermore, interviews carried out with political advisors at the minister's office level (e.g. Dunn 1997; Esselment et al. 2014) could add substantial depth to the analysis by comparing their behaviours to those described in this study. Second, analyzing other key positions in ministries may provide different results. We need to exhaustively study not only one position in public departments (in our case, the ADM), but several positions in the public sector hierarchy, in order to identify and compare different models of civil service practice. Third, the sample is limited to 12 participants and as such, care must be taken to interpret the results as exploratory illustrations or clues rather than definitive data of the day-to-day practices of Quebec's ADMs. Fourth, the interview questionnaire is based on Gailmard and Patty (2012), which is not exhaustive nor systematic. Thus, newer work may have been overlooked.
We encourage researchers engaged at the crossroads of public administration and information transmission to use this survey as a basis for additional research. Other future research directions should include a comparison of strategic behaviours versus other forms of politicization found in the literature on political-administrative interface and strategic information transmission. In the same comparative spirit, a necessary next step would be to develop an interpretative framework that distinguishes what is meant by "normal" strategic behaviour on the part of senior civil servants and what kind of behaviours should raise concerns and be considered as abnormalities.
Appendix 1: interview guide
The first part is about the civil servant and its working environment strictly speaking (short answers required):
1. What is your academic background?
2. Did your training prepare you well for your job duties?
3. How long have you been working in the Quebec public service?
4. Up to now, with which departments have you worked?
5. What position(s) did you hold during your professional term?
6. Could you give me a brief description (e.g. description of the formal duties according to the employment status) of the duties associated with your position?
7. Did you have to work under different governments during your time in the public service?
8. During your term as a professional, but outside your official public duties, have you ever shown any interest in politics?
9. On a daily basis, what is your work inspiration? Do you feel more successful when you succeed in confirming the budget of your unit or advancing a file with the political officials? Classification of the assistant deputy minister based on the work by Bendor et al. (1987).
This short section looks at the pre-existing relationship between you and your political superior. To answer, you must consider that the minister and its staff have explicitly indicated their need for assistance:
1. Does the relationship developed over time with your political superior enable you to develop a vocabulary that facilitates communication? Common language test based on the work by Farrell (1993).
For this section, it should be considered, as before, that the principal has explicitly addressed a request to the administration to benefit from your expertise. Special attention must be paid to the first three questions, since the nuances are more subtle:
1. When a team of professionals of which you are a member presents the results of an analysis with which you do not quite agree, how do you react during the interview to the extent that the minister welcomes the recommendation? Nestling begging test based on the work by Bergstrom and Lachmann (1998).
2. When you and your team must advise the minister on a specific issue, is it important beforehand that all the officials involved in the process have reached a consensus on the outcome? Monotonic equilibrium test based on the work by Okuno et al. (1990).
3. At the end of a long-term meeting with the minister and his staff, do you consider that the information that has been shared is more complete than it would have been in a shorter meeting? Do you ever prepare your briefing according to the expected duration of a meeting? Test of the Placement of time based on the work by Aumann and Hart (2003).
4. Before undertaking any research on a project commissioned by the Minister, would you say that ensuring adequate funding to complete the work allows you to plan a more comprehensive research process? Sufficient budget test based on the work by Bendor et al. (1987).
5. Have you ever had the obligation to share the methods used to collect your data with the minister's office? If so, afterwards, did you see the minister as more empathetic to the information you were giving him? After this experience, is presenting research-related methodology a more common practice for you? Tactical Methodologxi Test based on the work by Patty (2009).
6. When you collect the data, is the information collected based on the supposed apprehension of the minister? Have you ever noticed a difference in the usage done by politicians of the informational content you provide? Defensive Denial Knowledge Test based on the work by Jervis (2015).
7. Looking back at your career in the civil service, do you remember a time when the minister gave you more flexibility than usual? If so, did this punctual authority was about a matter on which you and your minister were previously in agreement? Test of the "ally principle" based on the work by Dickhaut (1995).
Benoit Bechard is a PhD candidate in the School of Psychology at the Universite Laval, Quebec.
(1) Dion (1993) defines classical bureaucracy as having formalized rules under a hierarchical and specialized structure and codified relations with insiders (that is, the minister and his or her staff) and outsiders (that is, citizens). As part of an implicit agreement, the elected official avoids "politicizing" the public administration and the latter acts faithfully and discreetly, without partisan considerations (Dion 1993).
(2) Jean-Marc Salvet, "Des taches subalternes pour << Mme Post-it>>", Le Soleil, September 14, 2012.
(3) Simon Boivin, "Un sous-ministre n'a pas a recevoir << d'ordres>>, selon Dominique Savoie", Le Soleil, May 18, 2016.
(4) According to Ministry of the Executive council (2003), "an official shall keep secret the facts or information which he or she becomes aware of in the course of, or in connection with, the performance of his or her duties and which are of a confidential nature. This obligation also means that the official adopts a restraining attitude towards all facts or information that, if disclosed, could be detrimental to the public interest [...]" (2003: 4).
(5) Since then, this principle and its assumptions have been the gateway for the development of a wide formal corpus and has later been tested using laboratory experiments (e.g. Cai and Wang 2006; Kawagoe and Takizawa 2009; for a review, see Ouimet et al., under review).
Aberbach, Joel D., and Bert A. Rockman. 2006. "The past and future of political-administrative relations: Research from bureaucrats and politicians to in the web of politics-and beyond." International Journal of Public Administration 29 (12): 977-995.
Adams, David. 2004. "Usable knowledge in public policy." Australian Journal of Public Administration 63 (1): 29-42.
Aumann, Robert J. and Sergiu Hart. 2003. "Long cheap talk." Econometrica 71 (6): 1619-1660.
Bakvis, Herman. 1997. "Advising the executive: Think tanks, consultants, political staff and kitchen cabinets." In The Hollow Crown, Palgrave Macmillan, London.
Bechard, Benoit. 2017. "Approches formelles en transmission strategique de Tinformation: une boite a outils pour les decideurs publics." Administration Publique du Canada 60 (2): 192-214.
Bendor, Jonathan, Serge Taylor, and Roland Van Gaalen. 1987. "Stacking the deck: Bureaucratic missions and policy design." American Political Science Reviera 81 (03): 873-896.
Bendor, Jonathan and Adam Meirowitz. 2004. "Spatial models of delegation." American Political Science Review 98 (02): 293-310.
Bergstrom, Carl T., and Michael Lachmann. 1998. "Signaling among relatives, iii. Talk is cheap." Proceedings of the National Academy of Sciences 95 (9): 5100-5105.
Blais, Andre, Donald E. Blake, and Stephane Dion. 1997. Governments, Parties, and Public Sector employees. University of Pittsburgh Press, McGill-Queen's University Press.
Bourgault, Jacques. 1983. "Les hauts fonctionnaires quebecois : parametres synergiques de puissance et de servitude." Canadian Journal of Political Science 16 (2): 227-256.
--. 2009. "Les relations entre Tadministration publique et le Parlement." In Le parlementa risme canadien, edited by Rejean Pelletier and Manon Tremblay. Presses de l'Universite Laval.
--. 2011. "Canada's Senior Public Service and the Typology of Bargains: from the Hierarchy of Senior Civil Servants to A community of 'Controlled' Entrepreneurs." Public Policy and Administration 26 (2): 253-275.
--. 2013. Les sous-ministres du gouvernement federal: Statut, profit et rdles. Deputy Ministers in Canada. Toronto: University of Toronto Press/IPAC.
Bourgault, Jacques and Stephane Dion. 1989. << Les gouvernements antibureaucratiques face a la haute administration: une comparaison Quebec-Canada. >> Politiques et management public 7 (2): 97-118.
--. 1991. "Haute fonction publique et changement de gouvernement au Quebec: le cas des sous-ministres en titre (1976-1989)." Politique 19: 81-106.
Bourgault, Jacques and Donald J. Savoie. 2009. << Des gestionnaires sous influence : la nouvelle realite des hauts dirigeants du secteur public. >> Telescope 15:1-12.
Bourgault, Jacques and Stephanie Viola-Plante. 2014. Quebec Deputy Ministers: Accent on Program and Service Delivery in Times of Scarcity. Deputy Ministers in Canada: Comparative and Jurisdictional Perspectives. Toronto: University of Toronto Press/IPAC.
Burmeister, Elizabeth and Leanne M. Aitken. 2012. "Sample size: How many is enough?" Australian Critical Care 25 (4): 271-274.
Cai, Hongbin, and Joseph Tao Yi Wang. 2006. "Overcommunication in strategic information transmission games." Games and Economic Behavior 56 (1): 7-36.
Christensen, Jorgen Gronnegard. 2004. "Political responsiveness in a merit bureaucracy: Denmark." In The Politicization of the Civil Service in Comparative Perspective, edited by B. Guy Peters and Jon Pierre. Routledge.
Clark, David. 2002. "Neoliberalism and public service reform: Canada in comparative perspective." Canadian Journal of Political Science/Revue canadienne de science politique 35 (4): 771-793.
Crawford, Vincent P. and Joel Sobel. 1982. "Strategic information transmission." Econometrica 50 (6): 1431-1451.
Dickhaut, John W., Kevin A. McCabe, and Arijit Mukherji. 1995. "An experimental study of strategic information transmission." Economic Theory 6 (3): 389-403.
Dion, Stephane. 1993. "La bureaucratie." In Introduction a Tadministration publique: une approche politique, edited by James Iain Gow. Gaetan Morin Editeur.
Dunleavy, Patrick. 1991. Public Choice--Economic Explanations in Political Science. New York: Harvester Wheatsheaf.
Dunn, Delmer. 1997. Politics and administration at the top: Lessons from down under. University of Pittsburgh Press.
Esselment, Anna Lennox, Jennifer Lees-Marshment, and Alex Marland. 2014. "The nature of political advising to prime ministers in Australia, Canada, New Zealand and the UK." Commonwealth & Comparative Politics 52 (3): 358-375.
Farrell, Joseph. 1993. "Meaning and credibility in cheap-talk games." Games and Economic Behavior 5 (4): 514-531.
Fortier, Isabelle, Emery Yves, and Rachel Roldan. 2016. << Des conditions et enjeux de la politisation de l'administration publique au regard de l'ethos public: une etude comparee (Canada/ Suisse) de l'experience des cadres dirigeants. >> Gestion et management public 4 (2): 75-98.
Fusch, Patricia I. and Lawrence R. Ness. 2015. "Are we there yet? Data saturation in qualitative research." The qualitative report 20 (9): 1408-1416.
Gailmard, Sean and John W. Patty. 2012. "Formal models of bureaucracy." Annual Review of Political Science 15: 353-377.
Garant, Patrice. 1995. Precis de droit des administrations publiques. Quebec, Canada: Editions Yvon Blais.
Good, David A. 2003. The politics of public management: the HRDC audit of grants and contributions. Toronto, Canada: University of Toronto press.
Gow, James Iain. 1986. Histoire de iadministration publique quebecoise 1867-1970. Montreal, Quebec: Les presses de l'Universite de Montreal. Institui d'administration publique du Canada.
Grossman, Sanford J. and Oliver D. Hart. 1980. "Takeover bids, the free-rider problem, and the theory of the corporation." The Bell Journal of Economics 11 (1): 42-64.
Guest, Greg, Arwen Bunce, and Laura Johnson. 2006. "How many interviews are enough? An experiment with data saturation and variability." Field methods 18 (1): 59-82.
Jervis, Robert. 2015. Perception and misperception in international politics. Princeton University Press.
Johnson, David. 2002. Thinking Government. Ideas, Policies, Institutions, and Public-Sector Management in Canada. Toronto, Canada: Broadview Press.
Kawagoe, Toshiji and Hirokazu Takizawa. 2009. "Equilibrium refinement vs. level-k analysis: An experimental study of cheap-talk games with private information." Games and Economic Behavior 66 (1): 238-255.
Krishna, Vijay and John Morgan. 2001. "A model of expertise." The Quaterly Journal of Economics 116 (2): 747-775.
Marsh, David, M. J. Smith and D. Richards. 2000. "Bureaucrats, politicians and reform in Whitehall: analysing the bureau-shaping model." British Journal of Political Science 30 (3): 461-482.
Ministry of the Executive council (MEC). 2003. L'ethique dans la fonction publique quebecoise. Quebec: Ministry of the Executive Council.
Morin, Claude. 2011. "L'interaction politiciens-administrateurs." In Secrets d'Etats? Les principes qui guident l'administration publique et ses enjeux contemporains, edited by Nelson Michaud. Les Presses de TUniversite Laval.
Okuno-Fujiwara, Masahiro, Andrew Postlewaite, and Kotaro Suzumara. 1990. "Strategic information revelation." The Review of Economic Studies 57 (1): 25-47.
Ouimet, Mathieu, Laurence-Brook Gabriel, Joyce Dogba, Christian Dagenais, Pascal Lalancette, Audrey Brennan, Benoit Bechard, Vincent Piquette, and Herve Tchala Vignon Zomahoun. Forthcoming. Communicating research evidence to decision makers through direct interactions: A scoping review of 'cheap talk' experiments. Under review.
Patty, John W. 2009. "The politics of biased information." The Journal of Politics 71 (2): 385-397.
Peters, B. Guy and Jon Pierre. 2004. The Politicization of the Civil Service in Comparative Perspective: A Quest for Control. New York: Routledge.
Robertson, Gordon. 1971. "The changing role of the Privy Council Office." Canadian Public Administration 14 (4): 487-508.
Savoie, Donald J. 2003. Breaking the bargain. Public servants, Ministers, and Parliament. University of Toronto Press.
Seidmann, Daniel J. 1990. "Effective cheap talk with conflicting interests." Journal of Economic Theory 50 (2): 445-58.
Svara, James H. 2001. "The myth of the dichotomy: Complementarity of politics and administration in the past and the future of public administration." Public Administration Review 61 (2): 176-183.
Caption: Figure 1. The Standard Hierarchical Arrangement of Quebec Senior Civil Service
Caption: Figure 2. Discretionary Authority and Ideological Preferences
Table 1. Integrative Table for Building the Interview Guide Author(s) Practical expectations derived from theoretical models Bendor et al. (1987b) "Programmatic" agents who care about the development of public policies Okuno-Fujiwara (1990) Agents search for consensus amongst peers before sharing information Farrell (1993) A common language facilitates the transmission of information between an agent and a principal Dickhaut et al. (1995) Similar preferences foster information sharing between an agent and a principal Bergstrom and Lachmann A debate between agents in front of the (1998) principal promotes the disclosure of new information Khrisna and Morgan (2001) A principal that may rely on multiple agents can gain further information Aumann and Hart (2003) Longer meetings between an agent and a principal foster fuller information sharing Patty (2009); Jervis (2015)A principal strongly biased encourages the agent to gather the information congruent with its bias
Please Note: Illustration(s) are not available due to copyright restrictions.
|Printer friendly Cite/link Email Feedback|
|Publication:||Canadian Public Administration|
|Date:||Mar 1, 2020|
|Previous Article:||L'adoption et le suivi de la performance d'une mesure interministerielle: Le cas du Supplement a la prime au travail.|
|Next Article:||Public administration in the cross-hairs of evidence-based policy and authentic engagement: School closures in Ontario.|