Abstract
During the past couple of decades we have witnessed important reforms in the public sector in OECD countries. Various forms of performance management systems have been introduced. This had an impact on the role and position of senior civil servants. The traditional public service bargain (PSB) came under pressure and was replaced by – at least partially – a managerial PSB. This article looks at the performance management systems that have been introduced for senior civil servants in five countries (Canada, the UK, the Netherlands, Denmark, and Belgium) and their impact on the managerial PSB. When there is strong result-based control in the different phases of the performance management cycle, we assume a strong managerial PSB. The finding is that the Anglo-Saxon countries have stronger result-based control systems than the European continental countries. Therefore we conclude that in the former countries the managerial PSBs are stronger than in the latter countries.
Performance management systems for senior civil servants show a different degree of result-based control. Following the four steps in a control process (plan, do, check, act) we can divide the performance management cycle into four phases: planning, monitoring, evaluation, and acting. When there is strong result-based control in each stage, we can speak of a strong managerial public service bargain (PSB). When there is moderate or weak result-based control, there is a moderate or weak managerial PSB, respectively. Each country can make its own choices regarding the performance management system for senior civil servants, but one must be aware that this has an impact on political-administrative relations.
Keywords
Introduction
Relations between senior civil servants and their ministers, and by extension the government and the wider society, has been a classic theme in the field of public administration. A number of authors have sought to describe and analyze these relationships (e.g. Aberbach et al., 1981; Svara, 1985, 1999; Peters, 1987). The most recent work approaches this relationship as a bargain and introduces the concept of Public Service Bargain (PSB) (Hood, 2000b, 2001; Hood and Lodge, 2006).
Political–administrative relations are the result of numerous variables. During the past couple of decades, we have witnessed important reforms in the public sector in OECD countries; many of these are inspired by the New Public Management movement (Pollitt and Bouckaert, 2011). Various forms of performance management systems and devices have been introduced, agencies have been created and put at arm’s length, and contracting has been introduced for senior civil servants in some countries. As a result, the traditional PSB came under pressure, and has been replaced by – at least partially – a managerial PSB (Hood, 2000a).
The objective of this article is to analyse how changes in performance management systems for senior civil servants have impacted on the PSB between those senior civil servants and their political masters. The PSB framework has thus far been used largely in a descriptive way. This article aims to add to the literature on politico-administrative relations by linking performance management systems to PSBs of senior civil servants. In this way, we aim to add to the PSB typology, and possibly refine the ‘managerial’ category of PSBs.
This article compares case studies of Canada, the United Kingdom, the Netherlands, Denmark and Belgium, representing a variety of civil service systems for senior civil servants (Van der Meer, 2011). The empirical analysis focuses on the performance management systems for senior civil servants in central administration at the federal/national level. These are the Deputy Ministers in Canada; members of the Senior Civil Service (SCS) in the UK (these are, among others, the permanent secretaries and the directors-general); members of the Top Management Group in the Netherlands (these are, among others, the secretaries-general and directors-general); the permanent secretaries in Denmark; and the presidents and directors-general of the Federal Public Services in Belgium. Although this group of senior civil servants occupy a similar position in the central administration, it is important to mention that their role is not entirely similar, due to differences in the institutional system (e.g. the degree of agencification, the degree of politicization of their role). This might explain some of the differences which are found in this study.
We will start this article with the conceptualization of PSBs as well as the performance management system. Then we will describe the methodology used for this research. The empirical analysis of our cases is presented in the next section. And finally, we will discuss our findings and their implication for the PSB typology.
Conceptualization of the Public Service Bargain and the performance management system
A PSB refers to ‘any explicit or implicit understanding between (senior) civil servants and other actors in the political system over their duties and entitlements relating to responsibility, autonomy, and political identity and expressed in convention or formal law or a mixture of both’ (Hood, 2000a: 8). A bargain implies an exchange between two or more parties, which in this case is between senior civil servants and ministers, who exchange some ‘x’ for some ‘y’. The exchange will vary depending on the type of public service system. Also, different bargains may exist within one particular system and bargains can change over time (Hood and Lodge, 2006). The traditional PSB in the UK for example is said to be an example of a Schafferian bargain. 1 In this type of bargain public servants provide loyalty and competent service to the government of the day in exchange for trust, anonymity, merit selection and permanent tenure.
Major reform programs during recent years have greatly changed the organization of the civil service in many countries. As mentioned earlier, Hood (2000a) suggests that these reforms have led to the installation of a managerial bargain. In this type of bargain, senior civil servants typically receive more autonomy in organizing their services, accepting in exchange the blame in cases of unsatisfactory results or mistakes. In line with the private sector situation, conditions and working relationships for senior civil servants are person-specific and individually negotiable. Senior civil servants in such a managerial bargain are said to be ‘managers’, often coming from the wider public or private sector.
In the literature on public sector organizations, the concept of autonomy is strongly linked to the concept of control (Verhoest et al., 2010; Demuzere, 2012). ‘Control refers to the mechanisms and instruments that are used by a controlling actor in order to influence the decisions and the behavior of the controlled actor with the aim of fulfilling the objectives of the controlling actor’ (Verhoest et al., 2010: 24). In our case the ministers can be considered the controlling actors while the senior civil servants are the controlled actors. A distinction can be made between input control and result control, referring to input and processes, outputs and effects, respectively (Verhoest et al., 2010). Research on the relationship between autonomy and control has yielded mixed results. Some studies show a positive relationship between the two concepts, while other studies indicate a negative relationship (Demuzere, 2012). Based on principal–agent theory, management autonomy is linked to result-based control (Christensen and Lagreid, 2001). The idea is that an effective and efficient service delivery is dependent on a high degree of management autonomy as well as result-based control. Also, the NPM philosophy puts forward a combination of a high level of management autonomy as well as result-based control in order to increase efficiency and effectiveness. Based on these arguments, we propose that a managerial PSB, which implies autonomy for the senior civil servant, is also characterized by result-based control.
The control of senior civil servants is organized through performance management systems. While the classic public administration literature on performance management focuses on government at large or on the organizational level (Bouckaert and Halligan, 2008), the literature on human resource management mostly deals with performance management at the individual level (Armstrong and Baron, 2005). Performance management is considered an important tool of strategic HRM (Boselie, 2010). In the OECD report ‘Performance based arrangements for senior civil servants’ (Ketelaar et al., 2007: 8), we find the following definition: ‘Performance management is typically used to refer to a set of procedures for defining performance, measuring it, and linking it to incentives or sanctions provided by a superior agency or person to a subordinate one, generally coupled with some increased degree of managerial autonomy for the subordinate’.
From this definition we can derive that performance management is a cyclic and iterative process which consists of various phases (defining, measuring, and linking performance). Also, the literature on result-based control refers to a cyclic process of control; Verhoest (2002), for example, makes a distinction between four stages: defining objectives, measuring results, evaluation linked with rewards and/or sanctions, and audit. An important model which is widely used in the public sector is the Deming cycle (Deming, 1986). Here, four steps are defined in a control process (plan–do–check–act). Based on this model, we have distinguished four phases in the performance management cycle for senior civil servants (Van Dorpe et al., 2011). These four phases are: planning (plan), monitoring (do), evaluation (check), and acting (act). In the first phase (planning), the objectives to be attained, and the resources put to this end, are agreed upon. Performance indicators are usually included to make these objectives measurable. In the second phase (monitoring), progress is monitored against these objectives. When necessary, these objectives can be modified due to changes in the policy priorities or unexpected circumstances. In the third (evaluation) phase, the senior civil servant is assessed on the targets met. In the final phase (acting), the senior civil servant bears the consequences of the evaluation: this can be a reward in cases of positive evaluation, or some kind of punishment in cases of negative evaluation.
Degree of result-based control in each phase of the performance management cycle
In cases of strong result-based control in most of the phases of the performance management cycle, we conclude that there is ‘strong’ managerial PSB. In cases of weak result-based control in most of the phases of the performance management cycle, we speak of ‘weak’ managerial PSB. In cases of mixed assessment, we conclude that there is ‘moderate’ managerial PSB.
Research design and methods
For our research we have selected five cases: Canada, the United Kingdom, the Netherlands, Denmark and Belgium. These cases have been chosen because they represent a variety of civil service systems for senior civil servants. In all of these countries we find a trend toward contractualization of top civil servants. This contractualization refers to the control relation as well as to the labor relation (Hondeghem and Putseys, 2003). As to the control relation, we notice that the hierarchical relationship between the principal (politician) and the agent (the top civil servant) is replaced by a contractual relationship in which the objectives and the means available to the agent are defined by means of performance contracts. As to the labor relation, we notice a transition from a traditional civil servant statute to a private labor contract (short or indefinite term). The situation in the different cases with regard to the control relation will be dealt with in detail in the next section. Below, we discuss briefly the situation with regard to the labor relation.
In the Canadian and Danish civil service, senior civil servants still work in a career system, with permanent tenure (in principle). In the Danish case, fixed-term appointments for senior civil servants are possible, but this is rarely the case. Senior civil servants in the UK have individual employment contracts which are, in principle, permanent. Fixed-term contracts are also possible, in particular for tasks which are limited in time, such as the organization of the Olympic Games. In the Netherlands, senior civil servants are still statutory employees, but they no longer enjoy permanence. They work for the State, with an appointment for five years (in principle). For the top civil servants – these are the members of the Top Management Group 2 – the situation is somewhat different: they work for the Ministry of the Interior and Kingdom Relations, and are detached for the time of their appointment (in principle this is seven years) to another department. In Belgium, a mandate system for senior civil servants was introduced in 2001: senior civil servants work on a ‘temporary statutory basis’. For career civil servants, this means a temporary detachment to a mandate function, for externally recruited senior civil servants, these are renewable short-term contracts (in principle for six years 3 ).
Participation in the surveys and the interviews
For the UK case study, we have relied largely on secondary sources, as it was not possible to send a survey or to do interviews at the time of the research.
As well as the survey and interviews with the group of top civil servants, elite interviews with civil service department officials and other experts were conducted by the research team in all of the countries. Also, a thorough desk research and document analysis was carried out for each case. 4 The research was carried out in the period 2009–10. This means that changes which have been introduced since then are not integrated into the analysis.
The performance management systems for senior civil servants: an empirical analysis
In the following section, we discuss the four phases of the performance management cycle in each country. We discuss to what extent we find a result-based control in the management of senior civil servants’ performance.
Planning (Plan)
In Canada, performance planning is based on the ‘mandate letters’ for the deputy ministers. They are written for two years in principle, but can be adjusted in cases of crises or other circumstances. The minister is generally not involved in the drawing up of these letters as this is the task of the Clerk of the Privy Council. 5 Three outcome domains are included: policy and programs, management, and leadership. Within these fields, both permanent and key objectives are defined. Within the field of policy and programs, the realization of government objectives is captured. The field of management concerns the management of the organization for which the deputy minister is responsible, in line with the Management Accountability Framework (MAF). 6 Not only the ‘what’, but also the ‘how’ is defined in each domain. Also, in all three fields key performance indicators are included.
The yearly performance cycle in the United Kingdom works with individual performance agreements for senior civil servants, in which performance objectives are defined in consultation with the minister 7 and with the agreement of the Cabinet Secretary. A standard format is provided by the Cabinet Office. By means of performance indicators, objectives are fixed in four domains: the business delivery objectives, the corporate objectives, the capability objectives, and the personal objectives. The permanent secretary identifies five or six key targets on which he/she chooses to be evaluated by the end of the year. As a general guideline, the proportion between the domain objectives is 60:20:20, meaning 60 percent business, 20 percent corporate, and 20 percent capability objectives (Cabinet Office, 2009). These documents are not open to the public.
The performance agreements in the Netherlands are called ‘werkafspraken’. These are relatively informal and flexible documents, concluded between the minister and the senior civil servant, and are not made public. In general, the minister fixes the political content and the senior civil servant the managerial content. Although the work agreements result from the minister’s policy agenda, the department’s budget and the department’s work program, they are less concrete than the aforementioned documents. Furthermore, they can contain targets, but there are hardly any quantifiable performance indicators included in these documents. Emphasis is more on ‘how’ the results should be achieved than on the actual objectives themselves. Both a focus on ‘delivery’ and the secretary-general’s role in the current public service modernization program are explicitly stated as part of the performance appraisal (Steen and Van der Meer, 2011).
In Denmark, the performance management cycle for senior civil servants is not extensively developed. All permanent secretaries have a performance contract, but this is not a performance contract in the usual sense of the word. The content of this contract is kept secret. Every year, the permanent secretaries receive a letter from the Ministry of Finances indicating the priorities to be realized in the following year. The content of the contracts varies greatly between different ministers and organizations. According to the permanent secretaries themselves, pre-set efficiency criteria and similar tools are not suitable for their specific jobs, and are more appropriate for heads of executive agencies (Hansen and Salomonsen, 2011).
Senior civil servants in Belgium work with management plans (with a six-year horizon) and operational plans (one year), concluded with their minister or line manager, depending on their place in the hierarchy. In practice, most senior civil servants integrate both plans, and in approximately half of the organizations, there is one common performance plan for all senior civil servants in that organization. Quantifiable performance indicators are included in the majority of the plans (de Visscher et al., 2011a). It is important to mention that the degree to which ministers are involved or interested in these performance plans varies widely.
Based on these data, we conclude that the result-based control in the planning phase of the performance management cycle for top civil servants can be assessed as follows: Canada (strong), UK (strong), the Netherlands (moderate), Denmark (weak), Belgium (moderate to strong).
Monitoring (Do)
The monitoring phase in Canada is rather informally organized. The Clerk of the Privy Council acts as a ‘coach’ to ensure good functioning in the ministries. Follow-up is ensured, among others, by means of breakfast and lunch meetings of deputy ministers and the Clerk.
In the UK, after six months a first official follow-up is organized to monitor progress over the past half-year. At this point, objectives can be changed in cases of changed context. The monitoring of the realization of departmental objectives is the responsibility of Her Majesty’s Treasury (James, 2004). Instruments used for the follow-up are – among others – the balanced scorecard and the EFQM 8 Excellence Model. The emphasis is on output and, even more, on outcome indicators, instead of on input and process indicators (Bouckaert and Halligan, 2008; National Audit Office, 2009).
Until recently, the monitoring phase in the Netherlands was quite informal, on a day-to-day basis. The introduction of ‘delivery scorecards’, in line with the British example, changed this somewhat. The frequency with which senior civil servants report on their performance varies greatly among departments: some of them report several times a year on their progress and results, others not more than once a year. In this case, there is no formal follow-up apart from the evaluation. The degree of formalization of such performance reports is also variable.
The individual performance plans of Danish permanent secretaries are not lengthy, detailed documents, and their performance management system is not extensively developed. This explains why little data can be found on the monitoring phase. It seems that a thorough performance management cycle is in place for Danish agency heads, but much less so for permanent secretaries.
In Belgium, the most important form of ‘monitoring’ is the intermediate evaluation, which takes place every two/three years. 9 Almost 30 percent of the senior civil servants studied state that official evaluations are the only follow-up of their performance plans (Van Dorpe et al., 2011). This may indicate that the performance plan is not an instrument for the daily practice of a senior civil servant. Those senior civil servants whose management plans are monitored use a broad range of instruments: board tables, balanced scorecards, management dashboards, business reviews, etc. This monitoring happens predominantly within organizations, not by some central organization and rarely by the political master. Furthermore, there is great variation in the frequency of this follow-up; in some organizations the management plan is monitored twice a month, in others once a year.
Based on these data, we conclude that result-based control in the monitoring phase of the performance management cycle for top civil servants can be assessed as follows: Canada (moderate because rather informal), UK (strong), the Netherlands (moderate, but variable), Denmark (weak), Belgium (moderate to strong, but variable).
Evaluation (Check)
The evaluation of deputy ministers in Canada is organized centrally and is based on three documents: the MAF, the letter of mandate and the corporate objectives convened with the Clerk. To begin, the deputy minister prepares annually an auto-evaluation and a proposal for next year’s performance agreement. A broad range of actors subsequently deliver input and advice on the deputy minister’s performance: the Committee of Senior Officials (COSO), 10 and also former deputy ministers, the minister, the Clerk of the Privy Council, and a number of other actors. In a two-day COSO meeting, each deputy minister’s performance is discussed for 45 minutes, resulting in a mark 11 for each deputy minister. At this point, the Clerk requests the advice of the minister, but he/she is not obliged to take this advice into account. Finally, the Clerk renders his/her recommendations to the Prime Minister. Every deputy minister receives feedback on his/her performance and final mark in a discussion with the Clerk. As such, the results as well as the way in which these results are obtained and the leadership capacities of deputy ministers are assessed in comparison with other deputy ministers’ performance.
In the United Kingdom, a separate evaluation procedure is in place for permanent secretaries compared to the rest of the SCS. The performance of the latter is assessed in a discussion with their line manager, who places the SCS member in one of four ‘performance groups’. Subsequently, all line managers of one department join in a meeting for a ‘relative assessment against peers’. They investigate, for example, whether or not too many SCS members were ranked in the highest performance group, and if so, who does or does not deserve this. Finally, the decisions are forwarded to the HR division of the department, who processes the pay decisions. The permanent secretaries’ evaluation starts with a written self-evaluation. Both the performance against the original objectives and the way in which this was achieved are assessed, the latter by means of the Leadership Framework. This is the responsibility of the Permanent Secretary Remuneration Committee. Many actors are involved in the process: first the minister and the Head of the Civil Service, and then a third party – this can be a permanent secretary from another department, or a non-executive director of the same or another department, etc. – Her Majesty’s Treasury, the Prime Minister’s Delivery Unit, the Office of Government Commerce, No. 10, and the Cabinet Office’s Capabilities Reviews Team (Cabinet Office, 2009).
The evaluation procedure of secretaries-general (and senior civil servants in general) in the Netherlands is much less formalized compared to the previous cases. The secretary-general discusses his/her performance with the minister, in a formal or informal meeting, or as part of the management meeting, where the organization’s general performance is discussed. As few actors are involved in the procedure, 12 the judgement of the minister is quite decisive. Not only is the performance assessed against the pre-set objectives, but also the general competencies of the secretary-general and his/her day-to-day performance. Between different departments, there is little consistency in the use of the performance agreements. Members of the Top Management Group themselves acknowledge this, and complain about the lack of a shared and unambiguous normative frame of reference.
In Denmark, the permanent secretaries prepare for their evaluation by means of an auto-evaluation on their performance. This information is complemented with survey results on remuneration policy and ethnic minority policy, and with reports from the National Audit Bureau. The role of the minister in this evaluation procedure is not as great as in the other cases. The yearly evaluation of permanent secretaries is performed by the secretaries-general of both the Ministry of Finances and of the Prime Minister. The minister is allowed to express his/her opinion on the permanent secretary’s performance, but this is not a systematic practice. There is a formal part of the evaluation, which focuses on the realization of the priorities defined by the Ministry of Finances, in accordance with pre-set criteria. The most important fields are human resources management – including remuneration policy and ethnic minorities policy – and contract management and the management of agency relations. The policy advice provided by the permanent secretaries is assessed in a quite informal way. This is part of the evaluation, though not part of the performance contract. According to Hansen and Salomonsen, ‘the decision concerning performance pay is not based on specific, ‘objective’ or transparent performance criteria, but rather it is based upon a general assessment of the relative workload of the permanent secretaries in the past year, including reflections on their involvement in preparing and implementing major reforms’ (Hansen and Salomonsen, 2011: 195–196).
As mentioned earlier, Belgian senior civil servants’ performance is evaluated every two years during their mandate, 13 with a final evaluation six months before the end of their mandate. The presidents of the board of directors are assessed by their minister, and the directors-general by their president and their minister. The role of the minister in the assessment of a director-general is by and large rather limited. Each senior civil servant prepares for his/her evaluation by means of an auto-evaluation. Based on their management plan, they are assessed on the realization of their objectives, the way in which they achieved them, their personal contribution to the realization of these objectives, and the efforts they took to develop their competencies. There are three possible marks: ‘insufficient’, ‘sufficient’, and ‘very good’. However, there is some criticism about the evaluation procedure: according to the Court of Audit, there are large differences in the quality of the evaluations (Rekenhof, 2011). Also, in 2009, 82.5 percent of senior civil servants were marked ‘very good’, leading some senior civil servants to doubt the objectivity and value of the evaluation. It is generally acknowledged that the goodwill of the minister is a factor of major importance, and that evaluations on the director-general level are on average more thoroughly executed. The guarantee of objectivity in the evaluation procedure remains one of the major concerns for senior civil servants (Van Dorpe et al., 2011).
Based on these data, we conclude that the result-based control in the evaluation phase of the performance management cycle for top civil servants can be assessed as follows: Canada (strong), UK (strong), the Netherlands (moderate), Denmark (weak to moderate), Belgium (on paper: strong, but in practice: moderate because of questions about objectivity and added value).
Acting (Act)
In the Canadian system, it is the responsibility of the Clerk of the Privy Council to decide upon the consequences of the evaluation, by applying the performance-related pay (PRP) system (Bureau du Conseil Privé, 2008). 14 There are four elements to rewards: an economic increase as a percentage of the base salary, progression within the salary scale, a risk reward, and a bonus. The higher the mark the deputy minister receives, the more elements of the PRP system are applied. A distributive curve is in place, which among other things defines that only a limited number of persons can receive a bonus. If the performance is deemed insufficient, the sanctions vary from a reprimand or a reduction of rewards, toward a demotion or dismissal. We must add that – although they still are permanent civil servants – the prime minister is entitled to move or dismiss deputy ministers by means of a ‘simple non-motivated decision’, in cases of insufficient performance. In order to achieve transparency, anonymous data on the evaluation and its consequences are published on a website. Overall, the deputy ministers consider both the evaluation procedure and the remuneration system to be fair and satisfying (Bourgault, 2011).
The consequences of the evaluation for senior civil servants in the UK are similar to those applied in Canada: in cases of good performance bonuses are awarded, in cases of bad performance, one can receive no bonus at all, or can be demoted or dismissed. The size of the PRP is dependent on the performance group in which one is put: the lowest receive no bonus at all, the highest a rather large bonus. On average, the PRP for the entire SCS in 2008 was 8.6 percent, and this percentage continues to increase year after year. Furthermore, in 2007 the Cabinet Office increased the maximum amount of PRP (which was 20 percent), up to 40 to 50 percent; this allows for some functions to have a rather low base salary, as this can be compensated with performance pay (Van Dorpe and Horton, 2011). Although dismissal is possible in theory, practice shows that this almost never happens. Our interviews revealed that in many cases, after one or more negative evaluations, one chooses to resign. Despite this seemingly efficient system, the SCS survey in 2006 showed that dealing with bad performance was one of the major concerns among SCS members (ORC International, 2006). For permanent secretaries, the reward decision is the responsibility of the Prime Minister, with advice from the Permanent Secretary Remuneration Committee.
As the Dutch secretaries-general still are statutory civil servants, the ‘werkafspraken’ are not closely linked to an extension of the appointment or dismissal. However, ‘dismissal’ of senior civil servants is possible in cases of incompatibility or conflict with the minister. In such a case, the senior civil servant is moved to another position. A positive evaluation can and usually does lead to a bonus, up to a maximum of one month’s salary. This decision is the responsibility of the minister, and within departments specific arrangements can be made. One example of such an arrangement is that such a bonus is awarded in no more than two consecutive years to one senior civil servant. Externally recruited senior civil servants receive a labor market bonus, which is not linked to their performance, but it can be terminated in cases of negative evaluation (Steen and Van der Meer, 2011). Furthermore, secretaries-general receive a 5 percent fixed management bonus on top of their base salary as compensation for leading their ministry. According to the report of the Dijkstal Commission, 15 the relation between performance and rewards is underdeveloped. The commission thus proposed a new reward system, in which 10 percent of the base salary would be linked to the realization of the performance agreements.
For the Danish permanent secretaries, no individual performance system has been implemented in the sense of clear ex post criteria for evaluating performance (Hansen and Salomonsen, 2011). They receive no feedback on their evaluation, and thus it is not a transparent system. There is an impact of the evaluation on their salary and bonuses, though this impact is relatively small. It is the decision of the Minister of Finances to grant these bonuses, based on the advice of the permanent secretaries of the bureau of the Prime Minister and the Ministry of Finances. We can thus state that the impact of performance agreements is of rather small importance, except for the extension of short-term contracts. Removal of a permanent secretary is possible, but this is in cases of a dispute between the latter and the minister, rather than due to performance (Christensen, 2006). As in the Dutch case, Danish permanent secretaries receive a permanent top manager bonus, which is not related to their performance.
In Belgium, no PRP in the narrow sense of the word has been introduced. In an informal way, extra holidays are distributed off the record, or a car is put at one’s disposal. While the evaluation has no direct impact on the salaries of senior civil servants, it does have repercussions for contract renewal. When a senior civil servant is assessed as ‘very good’, the mandate is renewed automatically. If the evaluation is ‘sufficient’, the senior civil servant must reapply for his/her function. When the assessment is ‘insufficient’, renewal of the mandate is not possible. Senior civil servants, who are statutory, may return to their old position in the administration. Externally recruited senior civil servants leave the administration. In practice, there are very few negative evaluations. There is no official conflict resolution in place, as is the case in some of the other countries. Some senior civil servants are rather critical about the objectivity of the whole procedure and point to the arbitrary renewal of mandates (de Visscher et al., 2011).
Based on these data, we conclude that result-based control in the acting phase of the performance management cycle for top civil servants can be assessed as follows: Canada (strong), UK (strong), the Netherlands (moderate because there is a link but not strongly developed), Denmark (weak to moderate), Belgium (moderate).
Conclusion: How strong is the managerial PSB?
Result-based control in each phase of the performance management cycle
Overall, the Canadian case shows strong result-based control throughout the performance management cycle. There are clear objectives with quantifiable indicators in the planning phase, with central steering of the process by the Clerk of the Privy Council. The monitoring phase is a rather informal event, which is why we classified this phase as ‘moderate’. In both the evaluation and act phases, performance is the most important criterion for action. The entire process is formal and centrally organized, which leaves little room for arbitrariness.
The strongest result-based control was found in the British case: all phases of the cycle were characterized by strong result-based control. According to a government official, the UK performance management system is quite strong with regard to the importance of achieving results for remuneration and career advancement. In the UK, as in the Canadian case, several central actors are involved in the process, leaving less room for variation or arbitrariness.
The research on the Dutch case revealed a more moderate result-based control. The entire performance management cycle is organized in a rather informal way, and as such, there is considerable variation between different organizations, ministers and senior civil servants. There is very little involvement of other actors in the entire cycle and the system is not transparent. Furthermore, the link between performance and rewards is not very strongly developed.
In the Danish case, the result-based control was the least observable as compared to the other cases: we found a weak to moderate result-based control. The performance contract for permanent secretaries is not a performance contract in the usual sense of the word; the content is kept secret and varies widely among different ministers and permanent secretaries. The performance management cycle is not extensively developed, and the evaluation and its consequences are not transparent. The link between performance and reward is not very strong.
The Belgian case showed a moderate to strong result-based control but in some cases there was considerable difference between theory and practice as well as between different organizations, ministers and senior civil servants. The planning and monitoring phases are relatively formalized, but the subsequent phases are much less formal, leaving room for variation and possible arbitrariness. Although there is no PRP, result-based control is ‘moderate’ in the act phase, because the renewal of the mandate depends on the outcome of the performance evaluation.
We can conclude that in all our case studies there is result-based control in the different phases of the performance management cycle for senior civil servants, although to varying degrees. This means that we can speak in all cases of a managerial public service bargain, but with different levels of strength. For both the Canadian and British cases, we conclude that a strong managerial PSB has been introduced for senior civil servants. In the Dutch and Belgian cases, we conclude that a moderate managerial bargain is in place. The Belgian case shows strong managerial PSB on paper, but the implementation in practice remains imperfect. We conclude that the current managerial PSB in the Danish case is rather weak. In Denmark, there is a tradition of close cooperation between permanent secretaries and ministers which leads to a less ‘managerial’ type of steering in the departments; managerial PSBs, however, are stronger for the heads of agencies (Hansen and Salomonsen, 2011).
Our findings indicate that the characteristics of a performance management system for top civil servants have implications for the type of PSB in place. It seems that the Anglo-Saxon cases have moved toward a stronger managerial bargain than the small European continental countries. Different types of performance contracts may account for this variation: in the Anglo-Saxon countries ‘hard’ performance contracts have been introduced, as opposed to ‘soft’ performance contracts in other cases (Putseys and Hondeghem, 2002). The difference between a ‘hard’ and ‘soft’ performance contract is linked to a different approach to human behavior in organizations. Hard contracts are based on the idea that people act in their own interests. As a consequence, the contracts are formal, concrete, focused, and with clear incentives. Soft contracts are implicit contracts that continue to exist as a result of common needs of both parties, and not because of legal sanctions (Christensen and Laegreid, 2001). Confidence and collaboration are the main principles of these agreements.
More research is necessary to explain the differences among the cases, but we should certainly take into account the history of the political and institutional systems in each country, including the traditional relationship between politicians and the administration. Other publications based on the same research have shown that the initial PSB, which was in place before the public management reform movement in each country, still has an important impact on the current PSB (Bourgault, 2011; de Visscher et al, 2011a; Hansen and Salomonsen, 2011; Steen and Van der Meer, 2011; Van Dorpe and Horton, 2011).
With regard to the PSB typology of Hood (2000b, 2001), it seems that the ‘managerial’ category should be further refined, making a distinction between ‘strong’, ‘moderate’, and ‘weak’ managerial bargains. In doing so, fewer cases would end up in the ‘hybrid’ category as a residual category, and this category could thus retain its value.
As set out in the introduction, this research aimed to further develop the concept of the Public Service Bargain. We have done this by linking public service bargains to the concept of result-based control in the performance management cycle for top civil servants. Control is of course only one characteristic of a public service bargain. Hood and Lodge (2006) have linked public service bargains to other important variables such as reward, competency, and loyalty/responsibility. These variables can support each other, but might also be contradictory. We have drawn conclusions about the strength of the managerial PSB based on the strength of result-based control in the performance management cycle. It is important, however, to stress that this is only one variable in a whole set of variables that might have an impact on the political–administrative relations.
With an eye toward further research, it would be interesting to look at the public service bargain from the viewpoint of the two actors that play a part in it: the principal (the minister) as well as the agent (the senior civil servant). How does each actor perceive their relationship in general, and the result-based control in the performance management cycle specifically. It might well be that both actors experience this relationship, and the public service bargain, in another way. Future studies might provide insights into the specific perceptions of each actor.