Public Professional Accountability: A Conditional Approach

In the past decades, profession(al)s have increasingly been called to account. Several authors have reported that this increased public professional accountability, in the form of showing that professional conduct meets predefined standards or rules, has had severe negative consequences for professionals, their clients and society, and call for ‘intelligent’ forms of accountability; forms of accountability that may inform a wider public about professional conduct but do not harm it. In this paper, we propose a form of ‘intelligent’ public professional accountability. Taking Freidson’s (Professionalism. The third logic, Polity Press, Cambridge, 2001) notion of institutional ethics as a point of departure, we develop a form of accountability that seeks to account for the conditions required for professional conduct. The paper first discusses the current ‘dilemma of professional accountability,’ describes ‘ideal-type professional conduct’ and goes into the conditions it requires. Next, it shows what accounting for these conditions entails and that this form of accountability fits the criteria for intelligent accountability, as set by O’Neill (in: Morris and Vines (eds.) Capital failure: rebuilding trust in financial services, Oxford University Press, Oxford, 2014).

. The most important problem seems to be that calculative accountability forces professionals to give an account that abstracts from the specific situations professionals have to respond to. Such accounts do not do justice to and cannot fully capture professional decisions and actions-which require a context-specific interpretation and translation of complex and tacit professional knowledge and experience (see also Tsoukas 1997;Schwartz and Sharpe 2010;Schwartz 2011;O'Neill 2002O'Neill , 2014. Decontextualized indicators or rules are not only a strength of calculative accountability, as they can make things visible to distant others (Roberts 2009) and allow for easy and cheap measurement (O'Neill 2014), but also a weakness, for they do not account for the in situ specifics that professionals face.
Instrumental behavior can come about if one becomes preoccupied with meeting targets or following rules (as individual or institutional evaluation depends on it; e.g., Roberts 2001). Such instrumentalism may lead to poor professional performance if the good thing to do in a specific situation no longer depends on context-specific discretionary professional judgment and dedication, but only on what the rules prescribe or on the targets that need to be reached (cf. O'Neill 2002(cf. O'Neill , 2014Schwartz 2011). Moreover, if calculative accountability does indeed diminish the professional ''autonomy and room for discretion'' (Banks 2004, p. 8), it may come as no surprise that professionals may find it difficult to appreciate their work as professional work, feel alienated from it, and have trouble in upholding their dedication to professionals values (cf. O'Neill 2002;Banks 2004;Schwartz and Sharpe 2010;Schwartz 2011).
However, although calculative accountability has received much criticism, accountability is, of course, not irrelevant to professions. If professionals are supposed to realize some societal value they should be ''accountable for the effectiveness of the services they deliver'' (Banks 2004, p. 151). Moreover, given the critical observations of authors writing on the sociology of professions that profession(al)s may ''have an interest in keeping their work opaque to outsiders in order to safeguard their freedom of discretion'' (as Levay and Waks 2009, p. 510, summarize), some form of transparency seems to be needed. It should somehow become clear to the public that profession(al)s are delivering the services they are supposed to deliver. Clark (2000) even states that it is a professional duty to provide this clarity. As Roberts (2009) writes, some form of transparency is needed ''as an important check on local collusion and as such as an essential source of confidence for distant others' ' (p. 968). Most critics of professional accountability do not deny this. However, they are critical of the form this accountability has taken and of the extent to which it has invaded professions.
So, as O'Neill puts it, one needs ''less distorting forms of accountability' ' (2002, p. 59) and as an alternative, some authors suggest to use so-called 'narrative' forms of accountability (O'Neill 2002;Etchells 2003;Kamuf 2007). In such forms, an account is not given in terms of pre-fixed categories (such as targets, norms, rules or protocols), but in the form of explaining to and discussing with others reasons for conduct in a way that allows for (communicative) freedom. It is, as Etchells (2003) puts it, an account in the form of ''a story rather than in figures'' (p. 14). A typical example of this form of accountability is a doctor who-without referring to binding rules or targetsexplains a diagnosis to a patient, discusses several alternative treatments, listens to possible objections, and arrives at a professional preference (based on his/her knowledge, experience, and vocation). Such narrative accounts, however, may not suffice in creating confidence to a wider public (see also Roberts 2001). Additionally, given the specific knowledge and experience of professionals (and our lack of it), truly understanding the reasons for professional conduct remains problematic, except for other professionals. Narrative accountability may therefore be an insufficient form of public professional accountability.
One could say that we are faced with a dilemma of professional accountability: On the one hand, we need some form of public accountability to make sure that trust in professionals is warranted. On the other hand, the current forms of accountability 1 may either harm professional conduct and/or may not be able to provide the information to satisfy a general public (e.g., Roberts 1991Roberts , 2009). To deal with this dilemma, some authors propose what they call 'intelligent accountability' (Roberts 2009;Sahlberg 2010;Ellison 2012;O'Neill 2002O'Neill , 2013O'Neill , 2014. O'Neill (2014) writes that ''an intelligent approach to accountability should support the intelligent placing-and refusal-of trust' ' (p. 180). In the case of public professional accountability, intelligent accountability systems should support the public by providing it with evidence of professional trustworthiness and in this way help to place (or refuse to place) trust in professionals. However, Although some authors provide some clues (e.g., Sahlberg 2010; Hutchinson andYoung 2011, and, in particular, O'Neill 2014 who formulates criteria for intelligent accountability) it still remains unclear what intelligent accountability systems should look like.
In our paper, we set out to describe an approach to accountability which may fit intelligent accountability and circumvent the dilemma of professional accountability. While calculative and narrative accountability focus on professional conduct and its results, we use notions from the sociology of profession and organizational theory to explore a form of accountability that focuses on the conditions enabling professional conduct and its results. This entails, for instance, showing that professionals have the time, tools, regulatory potential, information, or incentives, to actually and properly apply their specific knowledge and experience and dedicate themselves to realizing some societal value. In fact, as we will argue, accounting for conditions for professional conduct fits, in our view, O'Neill's (2014) criteria for 'intelligent accountability' as it provides relevant evidence (about enabling conditions) that can help us to place (or refuse to place) trust in professionals.
We organize this paper as follows. We first provide an ideal-type description of professional conduct (''Professional Conduct'' section) and use that to explore the notion of accounting for conditions for professional conduct. We introduce this notion in ''Accounting for Professional Conduct: A Conditional Approach'' section. Next, we discuss conditions for professional conduct (''Which Conditions are Relevant for Ideal-Type Professional Conduct?'' section), and explain what accounting for conditions may look like (''How to Account for Conditions Enabling Professional Conduct'' section). In ''Is Accounting for Conditions a Form of Intelligent Accountability?'' section, we argue that accounting for conditions can be seen as a form of 'intelligent accountability.' In ''The Value of Accounting for Conditions'' section, we discuss the (added) value of accounting for conditions and reflect on it. In ''Summary and Conclusion'' section, we conclude and argue that accounting for conditions may be an antidote to 'managerialist' accounting approaches.

Professional Conduct
Before we can discuss accounting for conditions for professional conduct, it makes sense to first delineate professional conduct. This is, unfortunately, not an unproblematic task given the many approaches to and accounts of professions and professionalism (cf. Abbott 1988;Torstendahl 1990;Evetts 2003;or Muzio et al. 2013; for overviews).
For the purpose of this paper, we propose to follow the approach of Freidson (2001) and give several ideal-type characteristics of professional work. That is, in the spirit of Weber's ideal-type definitions of (cultural) phenomena (Weber 1922), we define professional work from our perspective (i.e., understanding and formulating a suitable form of public professional accountability) as consisting of several characteristics without the empirical claim that all work that is called 'professional work' always realizes all characteristics to the same degree. Just as Freidson (2001) gives an ideal-type definition of professions, we give a related one of professional work. This ideal-type refers to three characteristics: (1) the application and development of specific knowledge and skills, (2) 'intensive technology,' and (3) the dedication to a particular societal value. Freidson (2001) argues that professional work is ''specialized work […] grounded in a body of theoretically based, discretionary knowledge and skill'' (p. 127). In our view, this harbors two separate characteristics: first, professional work can be said to need highly specific (tacit) knowledge and skills, and second, it can be characterized as 'intensive technology' (cf. Thompson 1967).
That professional work requires specific (esoteric/abstract) knowledge and skills, acquired through an extensive period of study and practice, is something that is common to most definitions of professions. Implied is that this body of knowledge is an accepted body of knowledge-thus for instance ruling out (in some cultures at least) reading tealeaves as a profession. Characterizing professional work as 'intensive technology' (as Thompson 1967, describes it) often remains implicit. Intensive technology refers to processes in which ''a variety of techniques is drawn upon in order to achieve a change in some specific object; but the selection, combination and application are determined by feedback from the object itself'' (Thompson 1967, p. 17). In later work, intensive technology is also described under the heading of 'value shop '-work (e.g., Stabell and Fjeldstad 1998) or 'solution-shop' work (Christensen et al. 2010). All these notions highlight trial-and-error diagnosis and treatment of unstructured problems. Based on contextspecific feedback from 'objects' (often: clients or patients) professionals diagnose problems and needs, and propose and adjust 'treatments.' Because professional work is characterized as intensive technology, it can be said to require the exercise of discretionary knowledge and skills, which make professional work difficult if not impossible to standardize or rationalize (Abbott 1988;Freidson 2001).
The third characteristic of our ideal-type is that professional work involves a dedication to a particular societal value (e.g., health, justice, and security) for the sake of this value, not because of economic gain or some other reason (cf. Koehn 1994;Freidson 2001). Note that, this value is comparable with an ethical value; it is a value that ought to be valued for its own sake-not just any value (cf. Koehn 1994Koehn , 1995 for an Aristotelian account of professional values). Profession(al)s derive their societal legitimacy as profession(al)s based on the dedication to this kind of societal value. A dedication to such a value rules out, for instance, a mere commitment to 'what the client wants' for this may run counter to values that ought to be valued. Indeed, this dedication is often seen as the moral duty of professionals and professional behavior should be performed so as to realize the societal value the profession stands for, something which is often expressed in professional oaths or moral codes. This idea of moral duty or dedication features in many descriptions of professionals (e.g., Koehn 1994Koehn , 1995Freidson 2001). Koehn (1994Koehn ( , 1995 even holds that the professional pledge to promote the relevant societal value is the most important characteristic of professional work: it is the basis for the trustworthiness of the profession and without it the other characteristics remain meaningless.
Based on these characteristics, we ideal typically describe professional conduct as conduct in which one applies and further develops specific knowledge, skills, and experience to make situation-sensitive judgments in the context of intensive technology and as conduct that is thoroughly based on a dedication to a particular societal value.
Here, we would like to point out that by means of our ideal-type approach to professional conduct we do not mean to offer an 'essentialist' definition of professional conduct. Instead, we use Weber's (1922) approach and select an ideal-type description of professional conduct to describe conditions supporting it and, in the end, to formulate a form of public professional accountability. Despite the fact that professional conduct may be described in several ways, we select this ideal-type description, because we assume that it may help us to find a way out of the current 'dilemma of public professional accountability.' We think that our selection makes sense because, in the first place, other authors also highlight similar characteristics when they talk about professional conduct. Here, we can point at Freidson (2001) who also employs an ideal-type approach to professions with similar characteristics; or Koehn (1994Koehn ( , 1995, Banks (2004), and others, who seem to offer more essentialist descriptions (which include the characteristics of our ideal-type). We can also point at authors who describe 'professional fields' in terms of the different institutional logics they harbor (e.g., Suddaby et al. 2009;Carter and Spence 2014;Spence and Carter 2014). One of these logics is what they call a 'professional logic' which covers almost the same characteristics as our ideal-type.
A second reason for selecting these characteristics in our ideal-type is that many of the problems of calculative professional accountability (see introduction) refer to a corruption of one or more of the three characteristics chosen. A third reason will become apparent as our argumentation unfolds: we set out to argue that based on this ideal-type we may look for conditions supporting it. As it will turn out, in discussing both enhancing and frustrating conditions for professional conduct, commentators seem to refer to characteristics included in our ideal-type. Fourth, we argue that we can also apply our ideal-type prescriptively and formulate conditions supporting professional conduct according to the proposed ideal-type and thus formulate a form of accountability that may suit professional conduct. Here, in fact, we follow other authors on accountability (like O'Neill 2002(like O'Neill , 2014Roberts 2009) or professionalism (e.g., Freidson 2001) who make prescriptive suggestions. So, we employ the ideal-type both descriptively and prescriptively in the hope of finding a way out of the dilemma of accountability.

Accounting for Professional Conduct: A Conditional Approach
The discussion about professional accountability in the literature mainly focuses on accountability for professional conduct itself. In a calculative account one is, for instance, worried about whether professional work (i.e., its outcomes) meets certain targets, or whether professional work itself is carried out according to some set of rules or procedures. Narrative accountability allows for explaining and discussing reasons for particular professional behavior as the professional sees fit. Both forms focus on professional conduct itself.
However, professional work is, of course, always carried out in a particular social/organizational/societal contextwhich conditions professional work. And, in our view, it makes sense to incorporate these 'contextual' conditions in professional accountability.
The context conditioning professional work has been addressed in several ways. For instance, some authors point at the fact that in the past professionals were mainly selfemployed, but nowadays many professionals work in organizations (e.g., hospitals, or law firms, Suddaby et al. 2009;Evetts 2011;Muzio et al. 2013). As these authors indicate, this organizational context clearly marks a difference in how professionals work and how their work is controlled and regulated. Other authors point at the institutional context of professional work, which refers to the ''regulative, normative and cultural-cognitive elements, that […] provide stability and meaning to social life'' (Scott 2008, p. 222). As Muzio et al. (2013, p. 700) remark, professions have relations with ''institutions such as markets, organizational forms, and business practices'' and are ''targets of institutional change.'' Clearly, to paraphrase Scott (2008), changes in the relation between institutions and professions also entail a change in the conditions which ''provide stability and meaning to professional life.'' The context of professional work is also mentioned by Freidson (2001, p. 216 ff.) when he discusses the difference between what he calls 'practice ethics' and 'institutional ethics.' Practice ethics relate to (dealing with) the ethical issues that individual professionals encounter in their work, while institutional ethics ''deal with the economic, political, social, and ideological circumstances which create many of the moral problems of work'' (p. 216, emphasis in original). Freidson's 'institutional ethics' thus concern conditions under which professionals do their work and encounter moral issues. Amongst these conditions, he includes ''the way [the professional] practice is financed, administered, and controlled in the concrete places where professionals work, and the social policies which establish and enforce the broader legal and economic environment within which practice takes place'' (p. 216). In his discussion, Freidson argues that certain conditions should themselves be declared unethical if they give rise to morally problematic professional issues. Ethically problematic, for instance, is ''providing working conditions that prevent the performance of good [professional] workconditions such as over-heavy caseloads and inadequate space, equipment, and support personnel'' (p. 217).
In this paper, we take this Freidsonian notion of institutional ethics (as the conditions for practice ethics) as a point of departure for professional accountability and use it to discuss what it means to account for the conditions for professional work instead of accounting for professional work itself. Put simply, accounting for the conditions for professional work means showing that the conditions enabling ideal-type professional conduct are realized. As we will argue, this form of accountability can be regarded as a form of 'intelligent accountability,' and may help to circumvent the 'dilemma of professional accountability.' Below, we will further elaborate the idea of accounting for conditions by going into the following questions: (1) which conditions are relevant for ideal-type professional conduct?
(2) how can these conditions be accounted for? and (3) how does accounting for conditions fit intelligent accountability?

Which Conditions are Relevant for Ideal-Type Professional Conduct?
Based on the discussions of conditions impinging on professional work by authors on professionalism and professional behavior (cf., Larson 1977;Abbott 1988;Thorstendahl 1990;Freidson 2001;Banks 2004;Lunt 2008;O'Neill 2002O'Neill , 2013O'Neill , 2014Evetts 2011;or Muzio et al. 2013) and based on authors from organizational theory who describe conditions for transformation processes in general (e.g., Galbraith 1977Galbraith , 1995Nadler and Tushman 1997;Daft 2009;Stabell and Fjeldstad 1998), we identify two general influencing conditions: goals and infrastructural arrangements (see Fig. 1). Organization theory authors sometimes refer to these two sets of conditions as the 'organizational architecture' (cf. Nadler and Tushman 1997). These conditions are, in turn, themselves influenced by the profession(al association), the organization they may work for, and the broader societal environment for which professionals ultimately realize a particular value (see Fig. 1).

Goals Conditioning Professional Work
One conditioning factor refers to the goals set for professional conduct. Such goals define the effectiveness of professional conduct. They determine what to pay attention to while carrying out processes, and hence, they have an influence on how the transformation processes are carried out. This seems to be quite obvious, but for professionals this simple logic has far-reaching consequences as goals may enter the professional work-domain that may undermine its ideal-type characteristics. This is the main argument of Freidson (2001) who holds that goals related to the market (profit maximization) and state/bureaucracy (maximizing the ''predictability and reliability of […] services and products,'' p. 217) should not enter the realm of professionalism. Market-and state-related goals are goals from different 'logics' and do not fit the logic of professionalism. As he argues, a focus on predictability and reliability of processes may come at the cost of professional context-specific discretion (pp. 217-218). Similarly, a focus on profit maximization may lead to a decrease in attention to quality, obligation and may even lead to an impoverished state of professional knowledge as this knowledge itself becomes an economic commodity (Freidson 2001, pp. 218-219). As Freidson (2001) and Koehn (1994) argue, the only 'real' professional goal is the realization of the societal value the profession ought to realize. While market or bureaucratic control goals are not problematic goals per se, they should not dominate the logic of the profession. The more they enter the profession, the more they lead to problems in pursuing professional goals. As Freidson puts it, they come at the loss of the ''soul of professionalism' ' (2001, Chap. 9).
Similar lines of thought are also put forward by authors who are informed by institutional theory. For instance, Suddaby et al. (2009), Spence and Carter (2014), or Carter and Spence (2014) argue that a dominant commercial logic may come at the cost of what they call the (technical) professional logic (which refers to calling, societal value, and specific technical expertise). Additionally, the problematic effects of 'non-professional goals' are also apparent in the work of critics of calculative accountability. If such goals are used to arrive at targets for individual professional work, one may introduce incentives to behave in an unprofessional way (cf. O'Neill 2002(cf. O'Neill , 2014Roberts 2009;or Schwartz 2011). Moreover, such individual performance targets (e.g., short-term targets; efficiency targets; profit targets; or targets in terms of rule-following) have been found to lie at the root of much instrumental or even irresponsible behavior (e.g., Banks;2004;Lunt 2008;O'Neill 2002;Treviño and Nelson 2007;Roberts 2009;Schwartz and Sharpe 2010;Schwartz 2011).
In all, the type of goals governing professional work affects that work. In terms of the characteristics of our ideal-type description of professional work, we could, with Freidson and others, argue that market and bureaucratic goals may (1) hinder the application and further development of specialized professional knowledge (e.g., because a focus on profit may favor the application and development of specific (i.e., profitable) knowledge), and (2) hinder professional work as intensive technology (e.g., because the focus on reliability, predictability, and costreduction may come at the cost of discretion and the time needed for dealing with clients; see Schwartz 2011). A third problem (3) is that the more market and bureaucratic goals govern professional work, the more difficult it becomes to realize the profession's dedication to its associated societal value. So, the more market and bureaucratic goals enter the profession, the less professional work is conditioned as ideal-type professional work.

Infrastructural Arrangements Conditioning Professional Work
The second set of conditioning factors are what we term the 'infrastructural arrangements' (cf. (Achterbergh and Vriens 2010;or Galbraith 1995 who uses the term 'design' to refer these arrangements). These arrangements consist of three aspects that directly influence the way professional work is carried out: (1) the way work is structured, (2) the practices and policies used to select, appraise, monitor, reward, and develop (the performance of) professionals, and (3) the technology professionals use to carry out their work (see also Galbraith 1977Galbraith , 1995Nadler and Tushman 1997;Robbins and Barnwell 2006;or Daft 2009). Although Freidson's (2001) idea of institutional ethics included such infrastructural conditions, he did not explicitly elaborate their nature. In this section, we use organization theory to categorize these conditions and we briefly discuss them.

Structure
The first infrastructural aspect concerns the way in which professional work is structured, i.e., how it is broken down into sub-processes and how it is coordinated (cf. Mintzberg 1983). Traditionally, the structure of work can be characterized by the degree of formalization/standardization (the degree to which work is determined by rules and procedures, cf. Donaldson 2001), specialization (the degree to which work is broken down into sub-tasks, cf. Mintzberg 1983), and centralization (the degree to which decision authority rests with only one or a few members of the

Professional conduct
Infrastructural arrangement Goals organizaƟon profession Societal environment: -Market -State -General public Fig. 1 Professional conduct, conditions, and ''institutions'' influencing these conditions organization, cf. Aiken and Hage 1971;Mintzberg 1983). As many authors state, professional work as intensive technology is enabled by organic structures rather than by mechanistic structures (e.g., Burns and Stalker 1963;Thompson 1967;Mintzberg 1983). Structures with tasks covering the complete 'job-to-be-done,' with the decentralized regulatory potential to deal with 'cases' and disturbances as one sees fit, and with a low degree of formalization, better fit ideal-type professional work. So, the more one moves away from such structures, the less professional work can be carried out as ideal-type professional work. As Freidson (2001) argues, bureaucracy, which is typified by high levels of formalization, specialization, and centralization (cf. Donaldson 2001) ''is naturally at odds with professionalism (p. 217)'' as it minimizes discretion (similar claims were summarized by Wallace 1995). It is possible to systematically discuss the influence of both high and low degrees of formalization/standardization, specialization, and centralization on the ideal-type characteristics of professional work. However, that would fall beyond the scope of this paper. Here, we merely want to illustrate that structures may affect professional work positively or negatively.
For instance, with respect to ''professional work as intensive technology-based on discretion and feedback,'' discretion is frustrated by high degrees of formalization and standardization. In fact, rules and procedures are meant to curtail the discretion of professionals forcing them to produce reliable and predictable behavior. However, they cannot cover the complexities of professional conduct as discussed above. Specialization into small tasks may also hinder using feedback to adjust earlier professional decisions and actions-which is often required in a professional setting. For instance, if one assigns professional diagnosis and professional treatment to different persons who even work in different organizations (this threatens to be the case in Dutch mental youth care) it becomes difficult to use the problems encountered in a treatment to adjust the diagnosis (which may also obstruct experiential learning about particular diagnoses and treatments). Centralization may also cause problems. For instance, doubts about the progress of a treatment may cause a professional to run additional tests and adjust the diagnosis or alter the treatment. However, a high degree of centralization may obstruct this as professionals may lack the decision authority to decide that extra tests are to be carried out.
By contrast, low degrees of formalization, specialization, and centralization lead to structures in which professionals have the opportunity to exercise discretion (low formalization) and use feedback related to all aspects of the professional intervention to adjust their decisions and actions (low specialization and centralization). Monsen and de Blok (2013), for example, describe 'Buurtzorg Nederland,' a home-care organization in which professionals work in teams performing all care duties for a number of elderly people in a geographically restricted area. These teams are self-sufficient, do their own planning, and have decision authority to deal with disturbances. And, as it turns out, this organization does not only enable professionals to do their work adequately (as judged by the professionals working for it and by the patients treated), but it also is the most cost-efficient home-care organization of the Netherlands.
In all, if one characterizes professional work as intensive technology it is, following Thompson (1967) a 'structural category mistake' to organize it as long-linked technology (as described by Thompson 1967).

Performance Measurement, Control, Reward, Motivation, and Development of Practices and Policies
A second aspect of infrastructural arrangements influencing professional conduct relates to practices and policies used to select, assess, appraise, monitor, reward, sanction, motivate, and develop professionals and their performance (cf. Merchant and Otley 2007;Ferreira and Otley 2009). Part of these policies and practices translate organizational goals (as discussed earlier) into targets for individual work. Another set of these practices is related to monitoring whether professionals reach the goals set, to the overall assessment of professional performance and to rewarding, sanctioning, and motivating professionals-practices that are closely related to accountability (cf. O'Neill 2014). In this paper, we do not want to treat the influence of all these practices on the three characteristics of professional work. Instead, as we did with structural conditions, we want to illustrate that such practices may condition professional work positively or negatively. To do so, we will treat these practices referring to three issues: (1) the degree to which ideal-type professional goals enter these practices, (2) the degree to which professionals themselves take part in these practices, and (3) the form of these practices.
Practices like monitoring, appraising, and rewarding professionals assume that performance targets are set for individual professional performance. Whether these practices enable ideal-type professional work co-depends, in our view, on the degree to which professional targets govern individual professional performance. Put simply, if professional performance is characterized by applying and developing specialized knowledge, by intensive technology (involving discretion and processing feedback) and by a dedication to a societal good, then it follows that professional performance should be governed by targets reflecting these three characteristics. In line with what we have said in the section on 'goals' as conditions for professional work, then, we argue that monitoring, assessing, and appraising professional behavior start with setting professionally relevant targets for individual professional performance (cf. Hutchinson and Young 2011;O'Neill 2014), i.e., targets related to Freidson's (2001) professional logic.
Although it may seem easy to ward off non-professional goals and targets, Anderson-Cough et al. (2000) show that the incorporation of 'non-ideal-type' professional goals and targets can be a subtle process of socialization. Their study shows how a 'client-is-king' discourse may dominate the life of accounting professionals and leads to the ''internalization of demands and accountabilities within professional identity'' (p. 1165); i.e., by means of socialization, non-ideal-type professional demands are seen as ''part of being an accounting professional'' (p. 1165).
The second distinction concerning performance control and reward practices relates to the degree to which professionals themselves take part in them (a point also made by Levay and Waks 2009). Professionals know what their work entails-better than non-professionals. Therefore, it makes sense to include them in specifying targets for their own work, in monitoring whether their performance is still good, and in whether particular performance can be judged as professional performance. In fact, including professionals in such practices was accepted in a time in which professions were granted the autonomy to select and monitor the behavior of their own members (cf. Freidson 2001;Banks 2004). However, with the increase of market, organizational, and state control, this inclusion is waning off. Nowadays, goals are often set by non-professionals, representing non-professional institutions (e.g., Freidson 2001;O'Neill 2002;Lunt 2008;Schwartz 2011), demanding metrics that are easy to obtain and understand (O'Neill 2002).
The third dimension concerning performance measurement and control practices has to do with the form they take. The practices of setting individual targets, measuring whether these targets are met, and applying rewards or sanctions are of course strongly related to the idea of accountability. In fact, some authors even seem to equate (forms of) accountability to this particular set of practices. O'Neill (2014), for instance, equates ''managerial accountability'' with the sequence of ''setting targets, measuring results, publicizing these results, and sanction and reward'' (p. 175). In the introduction, we discussed two different forms of accountability: calculative accountability and narrative accountability. If applied to control of individual performance, we can now appreciate these forms as particular ways of instantiating a sequence of performance control practices (setting targets, measuring success, publicizing, sanctioning, and rewarding). Seen this way, we can state that the particular form an accountability-system takes (which may comprise more or less calculative/narrative components) to govern professional conduct is itself a condition for professional conduct. Calculative accountability seems to demand that professionals show to non-professionals that their behavior is appropriate in a way these non-professionals understand. Here monitoring and appraisal are cast in terms of the calculative account and leave little room for explanation and nuance beyond the account (cf. Messner 2009), which, may not do justice to professional behavior. Narrative accountability, in contrast, asks professionals to appreciate their behavior in terms of the categories they themselves would use and enables a discussion about their behavior. So, narrative accountability would better fit practices of setting targets, monitoring, and appraisal, doing more justice to ideal-type professional work.

Technological Conditions for Professional Conduct
A third aspect of the infrastructural arrangement enabling professional work relates to the 'technological means,' necessary for performing the work. This includes a rather large set of means-including the equipment they use, the physical lay-out of the space they work in, the ICT supporting their work, etc. Although, perhaps, a conceptually less challenging concept than structure and performance measurement practices, it goes without saying that without the proper equipment, ICT, etc., professionals will have a hard time reaching their goals. In fact, as Freidson (2001)

Organizations, Professional Associations, and Society
As we discussed, professional conduct is conditioned by goals set for it and the infrastructural arrangement professionals work in. These conditions, in turn, are dependent on a broader conditional context: the organizations professionals work for, the associations they are member of and the encompassing society.
That is, goals and infrastructural arrangements are provided by the organization a professional works in and the professional body s/he is member of (see also Fig. 1). The 'organization' can refer to a private professional practice or to a larger organization employing the professional (ranging from small to large-scale national or even multinational professional service firms; see Evetts 2011). In such organizations, the degree to which professionals can set their own goals and provide for their own infrastructure is often much lower than in the self-employed situation. Goals and infrastructural arrangements are also determined by the professional associations. These associations may set goals for professional conduct (e.g., they can specify the nature and the quality of professional services; thus specifying what it means to realize the societal value the profession is dedicated to). Goals for professional conduct are also formulated in professional codes of conduct. Professional associations can also influence the infrastructural arrangements in which professionals do their work. For instance, it may be that a professional association formulates guidelines or procedures to deal with specific cases. In the Netherlands, for instance, the veterinarian association formulated rules with respect to prescribing antibiotics to cattle preventively.
Moreover, the organizations and professional associations are, in turn, influenced by the larger societal context they are part of (see also Fig. 1). For instance, insurance companies, local, and national governmental institutions, special interest groups, the general public, the media, the industry providing equipment and medication for professional diagnosis and treatment all have an influence on the goals and infrastructural arrangements conditioning professional work. That this societal environment has an influence, is of course nothing but a truism. Here, we want to mention it because the influence of this larger environment often means that organizations and professional bodies may not always be free in setting goals and devising infrastructural arrangements.

How to Account for Conditions Enabling Professional Conduct
After discussing the conditions for ideal-type professional conduct, we now want to treat accounting for conditions for professional conduct. In general, it takes the ideal-type description of professional conduct as a prescriptive starting point and seeks to show to outsiders that what is required to perform it, is met. That is, based on what we have discussed so far, accounting for conditions entails showing the degree to which (1) goals set for professionals, and (2) infrastructural arrangements in which they work (a) enable the application and further development of professional knowledge, skills, and experience, (b) secure professional work as intensive technology, thus enabling context-specific diagnosis and treatment based on discretion and feedback, and (c) make sure that professionals are/keep on being dedicated to the societal value the profession they belong to is supposed to realize.
Accounting for conditions, then, amounts to giving a judgment about whether goals and infrastructural arrangements enable/do not hinder professional work. Concerning this judgment, at least two distinctions can be made. First, it may be a simple or a complex judgment. A simple judgment could be a general impression of whether a professional in a specific context can do his or her job as a professional (as in 'yes' or 'no,' or as in a score ranging from 1 to 10). It can also be a complex, detailed judgment, in which one addresses all cells of Table 1-which can be done in a qualitative and/or quantitative way. In order to arrive at such judgments one may devise instruments (questionnaires, diagnostic tools, scales)-see example below.
A second distinction refers to those giving the judgment. A judgment may be given by professionals themselves (e.g., directly, by means of questionnaires or having them rating conditions, or indirectly, by means of some organizational procedure for filing complaints about conditions). These professional judgments may be aggregated into an overall judgment about the state of conditions. Making such information public, in turn, can be regarded as a form of 'conditional accountability,' providing the public with relevant information about conditions for professional work.
Information about conditions could also be provided by organizations employing professionals. In such a case, the organization may present the aggregated judgments of professionals themselves, and/or ask relevant independent professionals to judge these conditions.
A judgment about conditions for professional conduct can also be given by a professional association. In that case the association judges, e.g., by means of an audit, visit, or questionnaire, whether a particular group of professionals can do their work as professionals given the particular conditions they have to work in (e.g., judging the conditions of professionals working in a particular organization, or area where different conditions apply, e.g., state or municipality). A professional association may also set up a complaint procedure or a professional ombudsman where professionals may voice criticism about conditions. Besides professionals themselves, the organizations housing them, or professional associations, governmental bodies (e.g., national health authorities in the case of health professionals) may also employ professionals to arrive at a judgment about conditions for professional work.
Given these distinctions (more can be given, but they fall outside the scope of this paper), accounting for conditions can take many forms, e.g., simple impressions or more complex periodic reports by groups of professionals themselves (e.g., belonging to one organization), professional associations or governmental bodies; see Table 2 for an exemplary scheme.
Here we should note that, even though a judgment about conditions may be arrived at by four parties, we think that the most informed judgment about conditions is given by the professionals who have to work in them (see also below).
To give an example of what such judgments might look like, we want to briefly discuss a procedure one of the authors recently used to research the conditions influencing the work of Dutch youth mental health-care professionals (in particular self-employed psychologists and psychiatrists). In an interview, we first discussed what working as a professional meant to them and what their work as professionals entailed. Here, we wanted to make sure that their descriptions fitted our ideal-type description. After determining that it did, we then introduced our model explaining how goals and infrastructures can influence ideal-type professional work (both positively and negatively) and asked for an overall score (between 1 and 10) as an indication of whether the current conditions enabled their work as professionals. Besides, we discussed each of Reports stating professional judgments on all relations between conditions/aspects of ideal typical work (see Table 1)

Organizations employing professionals
Aggregated impressions/ratings of confidence in conditions concerning the professionals working in the organization

Professional association
Overall impressions/periodic ratings of 'confidence in conditions'

Infrequent audit-reports concerning all relations between conditions and professional work;
Reports by ombudsman related to professional association

Governmental body
Overall impressions/periodic ratings of 'confidence in conditions'

Infrequent audit-reports concerning all relations between conditions and professional work;
Reports by (governmental) ombudsman D. Vriens et al. the cells of Table 1, asking for their opinion about whether certain conditions were enabling or not. We also asked to give a score for each row, indicating the overall supporting effect of a condition. Based on this procedure, it was possible to assemble and aggregate professionals' own impressions about the supporting state of the conditionsboth quantitatively as well as qualitatively. Moreover, based on this procedure it was also possible to pinpoint distinctive (un)supportive conditions. This procedure may yield a 'conditional footprint' for ideal-typical professional work for a specific group of professionals. It should be noted that this procedure is just an example and many variations of arriving at such judgments may be devised. At this point it is also worth noting that professional associations and governmental bodies are already doing research into and producing reports about conditions for professional work. Looking at health-care, we can, for instance, point at the American Nursing Association (ANA 2016) which reports about the working environment of nurses (including conditions as culture, safety, and working hours); the German ''Marburger Bund'' which reports on conditions like working hours of and increased economic pressure on physicians (IQME 2015), the Dutch VvAA (representing several groups of health-care workers) reporting on how increased bureaucracy affects professionals (VvAA 2016); Eurofound-an EU related agency for ''the improvement of living and working conditions'' (Eurofound 2016)-which runs surveys about conditions like 'work organization.' One of their reports discusses how such conditions affect Austrian health-care professionals (Krenn 2010). We can also mention WHO (Europe), which has set up guidelines for investigating healthy working conditions (Wiskow et al. 2010).

Is Accounting for Conditions a Form of Intelligent Accountability?
To see whether judgments about conditions for professional conduct may be regarded as a sensible form of public accountability, we want to use O'Neill's (2014) description of intelligent accountability. As we stated in the introduction, intelligent accountability was introduced as a form of accountability circumnavigating the pitfalls of existing accountability systems. In O'Neill's view, intelligent accountability revolves around providing reasonable evidence of trustworthiness (honesty, reliability, and competence) of obligation bearers-in our case: professionals. Intelligent accountability systems should provide and communicate such evidence of trustworthiness so that we can confidently place trust in professionals. In particular, intelligent accountability (1) ''should begin from an account of what is required of specific obligation bearers,'' (2) should ''provide evidence of trustworthiness or untrustworthiness,'' and (3) needs informed, independent judgment and ''would seek to communicate this evidence […] so enabling the placing or refusal of trust.'' (O'Neill 2014, p.183f). In this section, then, we refer to these three criteria to discuss (and refine) accounting for conditions as a defendable ('intelligent') form of public accountability.
Criterion 1: Intelligent Accountability Should Begin from an Account of What is Required As we see it, accounting for conditions is firmly rooted in ''what is required from obligation bearers.'' That is, we propose to describe what is required as ideal-type professional conduct (consisting of the three characteristics), and accounting for conditions seeks to show that the conditions for this conduct are adequate. The main idea here is 'If society wants X from professionals, accounting for conditions makes sure that conditions for realizing X are met.' We propose to equate X with our ideal-type description of professional conduct (see also ''Professional Conduct'' section).
In arriving at an account of what is expected from professionals, one should make sure that this is a shared account by the obligation bearers (professionals) and those receiving the account. If, for instance, society expects accounting professionals to be dedicated to competently ensuring that (financial and other) information provided by some party can be trusted-as relevant, complete, and true (cf. ICAEW 2016), then accounting professionals themselves should also be dedicated to this value and not to some other value, like, for instance 'serving clients' (cf. Anderson-Cough et al. 2000). To align the expectations of obligation bearers and those receiving the account one could start with having professionals explicitly describe their professional work (in terms of the ideal-type characteristics). More complex instruments, like the questionnaire developed by Suddaby et al. (2009), to find out to what degree the professional logic is dominated by a commercial logic, could also be used.

Criterion 2: Intelligent Accountability Should Provide Evidence of (Un)trustworthiness
Accounting for conditions provides specific evidence that can help to judge the (un)trustworthiness of professionals. In our case, trustworthiness of professionals refers to whether professionals can honestly, reliably, and competently show ideal-type professional conduct. The evidence 'accounting for conditions' provides, is in terms of a judgment about the conditions required for this ideal-type professional conduct, i.e., whether they are met or not. As we described above, such evidence may consist in judgments by professionals themselves about whether conditions enable them to deliver professional conduct. If, for instance, we are told that most professionals in a certain hospital complain that its formalized and centralized structure disables them to do their work as professionals (or more general: if we know that professionals feel that they can't do their work as professionals, given the conditions they work in) this certainly provides us with information that we can use to place or refuse trust. Note that, in such a case we may not refuse to place trust in the professionals themselves, but rather in professionals exhibiting ideal-type professional conduct given the conditions they need to work in.
On the other hand, suppose we learn that professionals in another hospital state that they can do their work as professionals. And if we ask some more, it turns out that this means that in their opinion, the circumstances they do their work in enable them to work in a professional way, i.e., to dedicate themselves to the patient's health, to apply their specialized, tacit knowledge and develop it, to exercise discretion and to make decisions doing justice to the patient's specific circumstances. Would such a statement foster trust? Would it provide us with evidence about the professionals' trustworthiness? We'd say that it does (given some provisions; see below). And based on this idea, we'd say that accounting for the conditions for professional work does make sense.
The case of the Dutch youth mental care referred to earlier may also point at the relevance of conditional evidence for placing (or refusing to place) trust. In this case (see Vriesema 2016), it turned out that some child-psychologists indicated that their work as professionals was seriously threatened because of the way it had been restructured and formalized. To decrease the number of referrals to specialized (expensive) psychologists and psychiatrists, the Dutch government had recently decided to introduce a new system for referral and treatment (related to care paid for by insurance/government). So-called 'easy cases' should first be referred to a 'social neighborhood team'-a team of non-specialized care workers who try to deal with these cases. Sometimes this works out if the psychological problems are superficial (e.g., related to 'standard' behavioral problems). Difficult cases are referred to specialized care. Moreover, general practitioners (GP) who first see most children with psychological problems decide if they are 'easy' or 'difficult' cases. And, for the difficult cases, the GP decides the length of the therapy (short, intermediate or long).
This new system entails a form of specialization in which a task (the first diagnosis of psychological problems) is no longer part of the job of specialists. Moreover, it also means a decrease of decision authority regarding work, as the GP-not the specialist-now decides on the length of the therapy. This has led to severe problematic circumstances in which children were wrongly diagnosed as 'easy cases' and received insufficient care. When this wrong diagnosis was finally recognized, problems often had already become worse (sometimes resulting in a loss of faith in care), making it extra difficult to treat them. Other problems emerged because GPs did not adequately determine the amount of sessions required. In such cases it sometimes turned out to be impossible to treat a patient. An additional reason for psychologists to complain about conditions for their work related to an increase in formalization. As budgets for mental health-care had been decentralized to municipalities, every municipality had set up its own system for screening and monitoring careworkers, for granting budgets, for billing, etc. Having patients from different municipalities, then, dramatically increased workload. Now, would this conditional evidence, provided that it is reliable, help to place or refuse to place trust in professional conduct? We'd say that it does. Indeed, as the described system change does not apply to privately paid care, parents of children with psychological problems use this evidence and are increasingly sending their children to privately paid psychologists. Conditional information, then, is already being used by society-although not in a way that might improve care.

Criterion 3: Reliably Obtaining and Intelligibly Communicating Evidence
The third criterion that can be derived from O'Neill's intelligent accountability is that evidence of (un)trustworthiness of professional conduct should be reliably obtained, i.e., based on an informed (expert) and independent judgment, and it should enable intelligible communication to a wider public (O'Neill 2014, pp. 184-185).
To start with the last issue, we think that information about whether conditions enable professionals to do their job as professionals can be conveyed quite easily to a broader public. Simple judgments about the state of these conditions (as described earlier) are easy to communicate. In the end, there is one overall indicator: the degree to which professionals themselves find that conditions are enabling them to do their work as professionals. More complex judgments (going into the relations between different conditions and aspects of professional work) may require some more time to digest. In the end, however, they also boil down to a judgment of professionals about which conditions are or are not met, and which aspect(s) of professional work is (are) in danger. We think that such information can be intelligibly conveyed to a wider public and provides evidence for placing or refusing to place trust.
In fact, in the mental health-care example discussed above even 'complex information' about the effect of structural conditions was readily communicated to a general, nonprofessional public, as shown by the coverage it received in a national newspaper (Vriesema 2016).
The second issue is whether information about the adequacy of conditions can be reliably obtained. This is the case if such information is based on an informed (expert) and on an independent judgment (O'Neill 2014, pp. 184-185). A judgment by professionals about the adequacy of conditions for professional work could hardly be more informed. Who else would be in a better position to judge the conditions for professional work than professionals performing that work in those conditions? So, the lack of expertise is no reason for possibly unreliable judgments. However, there may be other reasons why such judgments may not be reliable. Such judgments may be biased and can be tainted by opportunism or fear. Selfemployed professionals, for instance, may find it (commercially) undesirable to tell the world that certain pressures make that they cannot do their job properly. Similarly, professionals employed in organizations may (often rightly) fear that their judgments about the state of the organizational conditions influencing their work may have consequences for their own career. At the same time, professionals may blame 'poor conditions' for professional conduct to cover up their own culpable behavior and dodge responsibility for it.
To counter such problems, safeguards should be built in. For instance, by making sure that these judgments can be done anonymously; or by employing independent professionals to gather diagnostic information about conditions and to make it public on behalf of a group of professionals. Such additional information may (dis)confirm statements by professionals themselves. Independently collecting information about conditions for professional conduct might be regarded to be the responsibility of a professional association. One could even think of governmental or professional regulations prescribing the independent and anonymous collection and dissemination of such 'conditional information' (similar to the often required dissemination of information about indicators related to the outcome of professional work). Moreover, reports or audits of professional associations or governmental bodies containing judgments about conditions are relevant as a form of triangulation of judgments by professionals who are working in the conditions themselves.
So, the question whether professional judgments about the conditions for their work are reliable may depend on the safeguards one builds into prevent opportunism, fear, or downright fraud. In our view, these dangers are no reason for refusing to rely on judgments by professionals themselves. Here, we follow O'Neill (2014) who argues that ''[t]he dangers of corruption, producer capture, and professional cosiness are real, but it is absurd to try to remedy these by dispensing with informed judgement' ' (p. 185). In all, we think that, if safeguards are built in, professional judgments about the adequacy of conditions can provide reliable evidence about the trustworthiness of their conduct.

The Value of Accounting for Conditions
In this paper, we explore whether publicly accounting for conditions enabling professional conduct could be regarded as a relevant form of accountability-in addition to calculative and/or narrative forms of accountability. By way of a summary, we compare, in Table 3, the three 'pure' forms of accountability discussed in this paper (note, however, that many systems of accountability are hybrids of calculative and narrative accountability).
As we see it, conditional accountability might be regarded as a form of accountability with its own merits. However, before embracing this form as a relevant form of accountability, we may first need to counter some objections to it.

Objection 1: Accounting for Conditions Does not Show How Professionals Actually Perform
A first objection could be that if we only receive information about conditions, we remain clueless about how well a professional actually performs. This is true, to some extent. Based on the provided information we may place or refuse to place trust because we know whether a professional is enabled to do his/her work as a professional honestly, reliably, and competently. We agree that this information does not show that the overall result is adequate, i.e., that professionals actually do their work honestly, reliably, and competently. After all, there may still be incompetent, sloppy or malevolent professionals performing poorly, showing opportunistic or downright criminal behavior, even if conditions were met. Adequate conditions increase the possibility of ideal-type professional conduct, they do not guarantee it.
Here, we would argue that some form of monitoring and controlling professional conduct itself is of course needed. However, we would also argue that this may not necessarily include (more) calculative accountability, and that it may fall outside the realm of public professional accountability. Whether professional conduct itself is adequate can best be judged by professionals and as O'Neill (2014) holds we should expect professions and organizations housing professions to have installed (1) proper systems of professional peer-review and supervision, which may provide narrative systems of professional accountability, and (2) proper complaint procedures and systems of inspection, examination and punishment of professional misconduct. These latter systems should be designed in such a way that independent and expert judgments about professional conduct are ensured (see also O'Neill 2014, p. 185). Dealing with professional incompetence and malpractice is necessary-but it does not mean that information about indicators of the results of professional conduct should be provided to a wider public. By contrast, what, in our view, would help a wider public to place trust in professionals might be information showing that the systems mentioned above are installed and function properly, e.g., according to independent expert professionals. Such information about the proper functioning of these systems is, in fact, also information about conditions for professional conduct as it shows how a profession or organization housing professions seeks to secure adequate conduct by dealing with professional misbehavior.

Objection 2: Good Calculative Indicators Make Accounting for Conditions Redundant
A second objection might state that once one has good calculative indicators, one does not need to be told that the conditions are met. This may be so, but a major problem with professional accountability is that such calculative indicators are difficult to obtain. As we discussed, they have the tendency to be unable to cover professional conduct; they are simplified abstractions, should be put into context-which often requires professional knowledgeand can be forged.
A related objection is that if calculative indicators are met, we do not need to account for the conditions that led up to these results. We would argue that if such indicators were met, but the conditions for professional conduct were not in place, one might have reasons to be suspicious about these indicators and their values! This situation may point at instrumental behavior, downright fraud, it may imply that the indicators are poor indicators of professional behavior, or it may mean that burnouts are about to be expected, because professionals do a good job despite disabling circumstances. Seen this way, if one chooses to use output-indicators, information about conditions for professional conduct can actually help to put the values of these indicators in context.

Objection 3: Accounting for Conditions Denies that Economic and Bureaucratic Control Goals are Relevant for Professional Conduct
As we discussed above, accounting for conditions for idealtype professional conduct emphasizes Freidson's professional logic. But, by doing so, it seems to downplay other goals-notably economic and bureaucratic control goals like efficiency, profitability, uniformity and equity-which are also relevant for professional work. If one does not pay attention to economic goals, professions may become too expensive. Moreover, goals like uniformity or equity seek to ensure that we can expect to be treated according to Difficult to convey information about professional work to nonprofessionals Need for safeguards to make sure that information is reliably obtained and communicated accepted standards and are a safeguard against unequal treatments and professional whim. A form of accountability which does not include such goals is suspect. Instead, one should admit that there is an inherent tension between economic and bureaucratic goals on the one hand and professional goals on the other hand. So, one may object, accounting for conditions (which emphasizes professional goals) does not balance all relevant goals. We do not deny that economic and bureaucratic goals should enter professional work, and we also agree that one should try to balance all relevant goals. In fact, we think that accounting for conditions can actually support us in finding this balance. In our reply to this objection, we would like to stress that (1) focusing on professional goals should be central if we want professions to realize the societal values we want them to be dedicated to, (2) that such a focus does not mean that other goals are unimportant, and (3) that empirical evidence shows that a focus on conditions for professional work can actually be a starting point for making it more ''accessible and affordable'' (cf. Christensen et al. 2010).
We need to ask ourselves what it is we need professions for. As put forward by authors on professionalism, society depends on professions to deliver certain societal values such as health, justice, or education (e.g., Larson 1977;Abbott 1988;Freidson 2001;Koehn 1994). And, as Freidson (2001) argues, if society appreciates this value as a value a profession is supposed to deliver, it should not emphasize an economic or bureaucratic perspective to control professions. He argues that, if these perspectives become dominant it comes at the cost of professional conduct and hence at the values society wants these professions to realize. Instead, society should allow for a dominant 'professional perspective,' enabling professionals to realize the societal values professions ought to be dedicated to.
But if a 'professional perspective' should be centralwould that mean that one needn't pay attention to economic and bureaucratic goals at all? No. In our view, it means that there is a difference between realizing the primary professional goal (conform the ideal-type) and paying attention to cost and uniformity as secondary goals. This entails, for instance, that a doctor should judge the adequacy of several different treatments on professional grounds. But if, based on these grounds, several equifinal alternatives emerge, cost-criteria should be used to make a decision. Similarly, as a general principle a doctor may be guided by a particular protocol for treatment of some disease, but this doctor may decide to deviate from this protocol based on his/her discretionary professional judgment. In both cases, the professional perspective is central but economic and bureaucratic goals are not ruled out. Such goals can enter professional decision making, but only as secondary goals. If they become primary goals, they will lead to the problems discussed earlier and function as a disabling condition. So, making the professional perspective central does not mean that professionals do not have to take into account economic and bureaucratic goals. It also does not mean that professionals do not have to account for their (non-)inclusion in their judgment. Although the main logic should be the professional logic, the inclusion of such secondary goals can act as a safeguard to ward off waste of resources and inequity.
In addition, a focus on conditions for professional conduct (and hence on professional goals) is by no means at odds with making professional conduct more ''affordable and accessible'' (cf. Christensen et al. 2010;Monsen and de Blok 2013). In line with notions from 'organic organization theorists' (e.g., Donaldson 2001) these authors describe health-and home-care organizations which are geared at optimizing the conditions for professional work. Christensen et al. (2010) describe how they witness the emergence of what they call 'dedicated hospitals,' and how these are able to deliver better and cheaper professional care. Similar organizations also emerge in home-care, for instance Buurtzorg in the Netherlands (cf. Monsen and de Blok 2013), showing that organizations structured around professional work exist that deliver much better and far less expensive care than traditional home-care organizations. What is interesting about these health-care institutions is that they lead to institutions which better condition professional conduct and deliver affordable and accessible professional services.
Objection 4: If Accounting for Conditions Means that There is No Longer a Place for Calculative and Narrative Professional Accountability, Then It is No Good Our suggestion to publicly account for conditions enabling ideal-type professional conduct does not mean that forms of narrative or calculative accountability are no longer relevant. Narrative accountability is still required for micro-accountability between professional and 'client' (e.g., explaining a patient why a certain treatment is preferred) or between professionals (e.g., in peer-consultation, or professional inspection). The latter can be relevant for securing the quality of professional work or for explaining how primary and secondary goals were balanced. However, as we discussed, narrative accountability may be problematic as a form of public accountability. Similarly, some form of calculative accountability can still be relevant. But here, we should be very careful. As a general remark, using calculative indicators for the result of professional conduct can only work if these indicators make sense in the context of professional conduct. Although we do not rule out the Public Professional Accountability: A Conditional Approach existence of such indicators, many authors warn that such indicators are difficult to obtain. Calculative indicators may also be relevant for the secondary goals entering professional decision making (e.g., indicators taken from financial reporting). However, one should make sure that such indicators have their proper place in professional accountability (e.g., they may serve as input in professional narrative accountability on how primary and secondary goals were integrated). Moreover, they can be used as signals indicating that conditions make professional work (too) expensive and trigger a search for less costly conditions (see discussion on previous objection). However, to prevent a subordination of the professional logic, it seems to be better to not use them for judging professional conduct itself. By contrast, indicators can be suitable for conveying information about conditions of professional conduct. As we discussed, the degree to which professionals believe that (their own) work as professional is enabled could, for instance, be a relevant indicator.
An emphasis on accounting for conditions also does not mean that following rules is no longer important (see above). Rules, regulations, or protocols are still important for professional work but only if they are accepted as professionally useful and if, based on discretionary professional judgment, deviations are possible.

Summary and Conclusion
In this paper we explored the possibility of 'accounting for conditions for professional conduct' as a way of fostering public trust in professionals. We argued that current forms of accountability (notably calculative and narrative accountability) may not be suitable as forms of public professional accountability and that accounting for conditions for professional conduct may be a valuable additional form of accountability.
In order to do so, we started with an ideal-typical description of professional conduct as conduct with three characteristics: (1) it applies and further develops specialized knowledge, (2) it is intensive technology based on discretion and feedback, and (3) it is devoted to a societal value. We then explored the idea of accounting for the conditions for professional conduct. To understand this form of accountability, we first presented a model comprising the conditions for professional conduct and next, we discussed how accounting for conditions could be thought of and how it relates to the criteria set for intelligent accountability, set by O'Neill (2014) and we reflected on its (added) value.
We believe that accounting for conditions may have a place in professional public accountability, alongside calculative and narrative accountability, and that it may help to foster public trust. In fact, we think that it can even act as an instrument for the emancipation of profession(al)s. To appreciate this statement, it is relevant to understand the reasons why calculative accountability was introduced. As Roberts (2009) or O'Neill (2002 note, we are often told that such systems need to be introduced to create public trust. But, as O'Neill (2002, p. 52) puts it, they are often nothing more than instruments for controlling professionals (a thought related to Freidson 2001) and need to be in the form of numbers and rules because these are easy to measure and understand (O'Neill 2002, p. 54). In this way, professionals are controlled by the (governmental or organizational) 'paymasters' managing them, i.e., by 'managers' securing other logics than a professional one. However, this management itself is often not included in these calculative figures. This seems to be strange because management has a large responsibility for the conditions enabling professional work. Accounting for conditions does two things at once: it shows whether professionals are enabled to do their work as professionals and it shows whether management, i.e., those responsible for these conditions, has created those conditions. In fact, as the emphasis is on conditions, the attention of accountability shifts from professionals to their management. It is therefore a professionally 'liberating' form of accountability, away from a more managerialist approach (cf. Parker 2002).
Although this paper introduced and explored 'conditional accountability,' some work is still needed to make it into an operational form of accountability. In particular, empirical work is needed to elaborate the set of conditions and to make and validate manageable procedures for producing an account of professional conditions (including questionnaires and safeguards to make sure that reliable accounts are given). Research is also needed to balance the three forms of accountability into one appropriate system for professional accountability-which may well be different for different professions. Although further research is still needed, it is our belief that accounting for conditions may well be a form of accountability that presents us with a way out of the 'dilemma of professional accountability.'

Compliance with Ethical Standards
Conflict of interest The authors declare that they have no conflict of interest.
Ethical approval This article does not contain any studies with human participants or animals performed by any of the authors.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://crea tivecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.