Contemporary policy work in subnational governments and NGOs: Comparing evidence from Australia, Canada and the Czech Republic^*

* The work of Arnošt Veselý was supported by the Czech Science Foundation Grant (P404/12/0725). The authors would like to thank two anonymous reviewers for their valuable comments and suggestions.

Abstract

This article introduces the special issue on policy work at the sub-national government level and non-governmental sector in three countries: Australia, Canada, and the Czech Republic. It consists of three parts. First, new public governance theory is presented as a general theoretical perspective that captures some of the important macro factors influencing the nature of current policy work. Second, an overview of empirical findings on policy work of subnational governments and NGOs is provided. This review discusses different possible interpretations of these findings. Third, main findings from each paper presented in this special issue are summarized.

1 Introduction

The theme of this special issue is policy work at the sub-national government level and non-governmental sector in three countries: Australia, Canada, and the Czech Republic. Our focus is warranted for several basic reasons. First, analysis of the policy process has tended until recently to overlook the mid-level staff on the front-line of policy advisory systems. Second, the sub-national level of government is also largely overlooked in academic research despite being responsible for strategically important areas of public policy and delivery. And third, non-governmental organizations have become, to greater or lesser degrees, more engaged in the policy process.

The aim of this special issue is to present empirically-based, original research on these understudied topics from three different countries. To the best of our knowledge, this is the first forum where policy work in these sectors is explicitly studied in a comparative perspective. We believe that such a comparative view might shed new light on what are generalizable features of current subnational and NGO policy work, and what is idiosyncratic for each country presented. In so doing, we attempt to link the macro perspective (the changing global context of policy work) with the micro perspective (actually policy work practices).

This introductory article consists of three parts. First, we present the changing macro context in which policy work worldwide is conducted. Arguably, numerous macro factors have an impact upon the nature of policy work. Consequently, countless numbers of competing theories can serve as a starting point for analysing policy work in a comparative perspective. In this article, we focus only on theories related to new public governance (NPG). This theme is explicitly or implicitly present across all articles in this special issue, and it seems to be equally relevant for all countries included here. However, we are quite aware that such a theoretical framework is far from comprehensive, and other relevant theoretical approaches are considered as well. Indeed, as argued in several articles presented here, while new public governance theory leads to new insights, taken alone, it remains largely incomplete. Additional theories are thus employed in several of the articles. The second part of this introductory paper provides an overview of empirical findings on policy work in subnational government and NGOs. It also attempts to show how such findings might be interpreted against different “baselines”. And finally, the third part of this paper summarizes main findings from each paper presented in this special issue.

2 New public governance theory and its implications for policy work

Specifically, new public governance is a response to the dispersal effects of new public management (NPM) reforms which created a need for “collaborative governance through networks” (Davies, 2009). The politico-historical origins of this shift can be found with the accession to government of Tony Blair's New Labour in 1997. Confronting the public administrative legacy of the Thatcher era, characterized by an ideological preference for market type solutions to public problems and a fragmented, disjoined state of public institutions and agencies, Blair's government set out to improve horizontal co-ordination by making “joined up” government the administrative goal. Even earlier, in the United States, the Clinton presidency, elements of a similar post-New Right effort to modernize rather than minimize the role of the state had been put in place (Pollitt & Boukaert, 2000, p. 94). NPG theory is an extension of these developments which its proponents characterize as the “transformation of the public sector [involving] ‘less government’ (or less rowing) but ‘more governance’ (or more steering)” (Rhodes, 1996, p. 655). Extrapolating from this framing, these contributions, by examining government and non-government policy work and policy workers, explore the links between macro-level processes encompassing actually existing policy processes and assess these against NPG framings and micro-level processes and practices (how policy work both in the state and in NGOs is conducted). The macro–micro framing leads to specific questions including: what are the consequences of NPM in shaping the tasks and functions of policy workers located in subnational states? Does NPG provide accurate explanatory insight to these working relationships? Is networking now more prevalent in the policy process? Can negotiation and consultation with outside actors be considered as a distinct activity of policy work? Is there any indication that knowledge and information are more shared between subnational governments and NGOs? What has been the effect of competitive contracting regimes on NGO participation in the policy process? If there is more engagement between both state and non-state actors, how is this assessed in terms of genuine power sharing and/or enduring hierarchical control by government? In answering these questions this Special Issue illuminates the link between policy regime and policy practice/work.

How policy workers do their jobs and construct policy is largely unknown terrain (Page & Jenkins, 2005). Recent research about governance arrangements suggests that the policy advice system has opened to the point that a “new range of political practices has emerged between institutional layers of the state and between state institutions and societal organizations” (Hajer & Wagenaar, 2003, p. 1). Consequently the substantive contribution of non-governmental policy actors, their capacity to engage with government, and how they are ‘heard’ by government becomes a serious consideration. Thus, rather than a top-down, internalized to the state function, policy work comes to be seen as a “far more elongated and ramshackle than the words ‘decision-making’ seem to imply” (Page & Jenkins, 2005, p. 81).

Thus, the image is constructed of an interactive policy making process where the government engages with relevant non-governmental policy actors. NPG posits that non-state actors have a role in the production of advice and decisions for government (Government of New Zealand, 2010, pp. 1–2). Policy engagement, thus conceptualized, brings NGO actors into the day-to-day activities of government policy units (Rowe & Frewer, 2005, p. 253). Consequently, government policy analysts are now expected to “engage in greater consultation, consensus building and public dialogue” as a core part of their professional role (Wellstead, Stedman, & Lindquist, 2009, p. 37). The emergence of such collaborative governance practices, defined as “a formal arrangement involving the direct engagement of state and non-state actors in a deliberative and consensus-oriented process for the purpose of developing policy and/or managing resources” (Robertson & Choi, 2012, p. 85), ostensibly signals a new era in government-stakeholder policy engagement. In such arrangements the centre of policy work and deliberation is located not within government policy units but rather in civil society partnerships which provide a site of collaboration for both government and non-government actors (Bradford & Andrew, 2010, p. 5). In this way, collaborative governance re-imagines the policy process as an approach that takes place outside of state structures, at least partially, and where policy is informed by the experience and knowledge of a variety of actors including those outside of the state.

The contestable result is that the policy process has been restructured to include new and more policy actors (Colebatch, 2006a, 2006b). NGOs have their own policy function to facilitate advocacy and consultation with government. And public service policy work now incorporates a much expanded role for political staff in the process. Political steering is carried out through networks involving “overlapping roles of political and societal actors” and characterized by “low institutionalization and a general blurring of bureaucratic demarcations” (Koch, 2013, pp. 397–398). Optimally, NGO actors enter at an early stage in the process so as to have a meaningful role in shaping the final policy product (Edelenbos, 1999). This pluralist policy process, engaging a broad spectrum of non-governmental actors, is understood to be a positive development in the realm of policy praxis (Van der Heijden, 2013, p. 8) resulting in a widely held view that “the more new governance, the better” (Solomon, 2008, p. 862) the policy outcome.

The core characteristics of NPG are collaboration and deliberation (Gunningham, 2009; Hoffmann, 2011; Karkkainen, 2004; Lobel, 2004; Solomon, 2008). In concrete terms collaboration refers to bringing “multiple stakeholders together in common forums with public agencies to engage in consensus-oriented decisions making” (Ansell & Gash, 2008, p. 543). The new paradigm of power sharing thus emerges (Ansell & Gash, 2008). What is distinctive is that “a wider variety of non-governmental organizations are becoming active participants in governing” (Bevir, 2011, p. 2). The proponents of NPG understand this as a new paradigm expressing a “change in the nature of the meaning of government” (Bevir & Rhodes, 2004, p. 4). Policies constructed through this process are not the product of competition and power politics but rather the result of a consensus-oriented process producing integrated policy outcomes. To be effective, collaborative governance processes must openly recognize and mitigate power disparities between actors within a policy community (Purdy, 2012). Thus, policy capacity can be dispersed among actors within each policy community, unlike more traditional forms of policy-making whereby decision-making processes occur within the ‘black box’ of government. The result is a new interpretation of the policy process, which “is not imprisoned in closed institutions and is not the province of professional politicians” (Newman, Barnes, Sullivan, & Knops, 2004, p. 204). A range of actors located not just at the top of the hierarchy or indeed even within the state apparatus (Tenbensel, 2006) are active in policy work. This view, while an orthodoxy in its own right (Davies, 2011, p. 48), is contested by a serious body of critical perspectives grounded in empirical research which has more recently emerged (Goetz, 2008; Hooghe & Marks, 2003; Janicke & Jorgens, 2006).

The deliberative process at the centre of NPG theory assumes a fairly even distribution of political and other resources among non-governmental policy actors. This fails to acknowledge that, within the NPG frame, “talk is disconnected from power” (Noveck, 2011, p. 89). Policy is an expression of power relations in society and, as power is unevenly distributed, the policy process cannot be freed of this constraint by simply becoming more inclusive (Ojha, 2013, p. 244). In this respect it is necessary to question the meaningfulness of non-governmental stakeholder participation (Ford & Condon, 2011). Moreover, NGOs themselves reflect an uneven distribution of power. They represent diverse interests within the larger society and thus do not share the same power resources. Weaker, less politically legitimate NGOs, for example, those concerned with redistributive policy in the fields of anti-poverty and labour reform, may not be included in the process or their participation may be perfunctory. The inequality characterizing the wider society will simply be reproduced within the open and pluralized new governance policy process as insider policy actors, those with political links to the state and possessing significant power resources, dominate and realize their specific policy goals (Ansell & Gash, 2008; Mol, 2007). To point, some empirical studies of deliberative processes conclude that they often fail to be completely inclusive and minimize or ignore the proposals of policy actors who possess ‘outsider’ status (Eversole, 2010; van der Heijden & ten Heuvelhof, 2012). But this presents a methodological problem if we alternatively accept that we are at a rather early stage in the development of NPG-type arrangements which means that there is still considerable learning to take which will inform their evolution.

Consequently, the core thesis underpinning NPG theory, that government is shedding its pre-eminent role at the centre of the policy process as state-dominated hierarchy gives way to multi-actor plurilateralism (Howlett, Rayner, & Tollefson, 2009), is contestable but only if we accept that the two paradigms can only exist in total isolation from one another. This framing is too pristinely categorical and ignores nuance and subtlety. The counter-narrative contends that what “appears to be a shift away from government may turn out to be a path towards government” (Goetz, 2008, p. 272). But this is not an either/or but rather a both/and situation which in the current moment may be conceived of as transitional leading towards a yet to be determined future state. For example, Capano's study of the shifting governance framework in the higher education systems of four European countries observes in this specific case an illustration of where the state is very much engaged in steering the sector in each case, albeit from a distance and unevenly (Capano, 2011, p. 1639). While Capano interprets this as an example of ongoing state direction, albeit from a distance, we can also understand this as perhaps evidence of an evolving network model incrementally superseding hierarchical direction from the state. Koch's analysis of the integration of four public transport systems in Switzerland is equally categorical in its interpretation of what this case suggests with respect to distributive governance. Koch's conclusion, at least in terms of the cases studied, is that network governance is a step within the process of governance reform moving over time towards a more hierarchically institutionalized form (Koch, 2013, p. 418). In other words, the network governance framework becomes governmentalized. However, while that may well be the case, it is not conclusively demonstrated empirically by Koch and at this point in time, cannot be.

There may be significant continuity with traditional public administration than as networks can simply reproduce the hierarchies of power they supposedly displaced (Davies, 2011, p. 50). However, it is too early and the various cases contesting NPG are too historically bounded at this point to allow anyone to come to definitive conclusions. One alternative conceptualization posits that state–society relations are being transformed through a process of ‘centralized decentralization’. Hoggett wrote this “corresponds to the abandonment of control by hierarchy and its replacement with control by contract” (1991, p. 250). Centralized decentralization is a “paradoxical development through which radical forms of operational decentralisation become combined with the further centralization of strategic command” (p. 249). We see this concretely in the proliferation of special operating agencies, or arms-length agencies, as well as the expansion of the contracting state (Pollitt & Talbot, 2004). Harlan Cleveland understood decentralization as a sub-component of centralization which enable better management of complexity. Perhaps his concept of uncentralization which is based upon a set of agreements respecting standards but beyond this, it is driven by “personal initiative, voluntary cooperation, joint ventures, committee work and networking” (Cleveland, 2000, p. 295). We can re-imagine that centralization/decentralization is actually one end of a continuum and uncentralization occupying the other end. Thus decentralization may be a form of centralization but it also may lead towards much looser systems of co-ordination where a number of policy actors agree on certain principles and objectives and then go about achieving these. Again, we do not have the benefit of sufficient historical perspective to be definitive one way or the other. Thus it is impossible to state that this is an either/or situation anymore than it may well be a both/and one!

3 Policy work in sub-national public administration and NGOs

3.1 Comparing subnational policy work and workers

Traditionally, policymaking was considered to be an “official preserve” of the state where outsiders were able to propose advice on public policies, but only authoritative leaders were entitled to decide and to ‘make policy’ (Colebatch, Hoppe, & Noordegraaf, 2010, p. 19). NPG theory raised many questions about policy work, including (Colebatch et al., 2010, p. 19): the relationships among governmental policy workers; relations between policy workers and non-governmental actors or the importance of non-governmental bodies in the construction of policy regimes. This led to a new interest in the real work of “policy people” outside the centre public administration.

In many countries sub-national governments traditionally control many important areas of policy-making. The importance of sub-national units has increased with the advent of new public management and extensive decentralization/devolution in many developed countries. In multi-level systems, sub-national governments control many important areas of policy-making, especially health, education, social services, local government and land, resources, and the environment (Bache & Flinders, 2004; Hooghe & Marks, 2003). As a consequence sub-national units (such as states, provinces, or regions) require substantial personnel, including so-called policy workers.¹ It is estimated that in federal countries such as the USA, Australia, Brazil or Mexico as many as 50% of national bureaucratic policy analysts work for sub-national states or provincial governments (Howlett, 2009c, p. 2). Similarly, according to Howlett and Newman (2010, p. 125), in Canada approximately half of all government policy analysts are employed at the sub-national level.

Despite the importance of subnational policy work, however, there is a serious gap in our knowledge of what subnational policy workers actually do and how they see they work. This poverty of research first began to change in Canada. In 2006–2007 studies of non-governmental policy analysts (Dobuzinskis, Laycock, & Howlett, 2007) and of regional and central policy analysts employed in the federal civil service were realized (Wellstead, Stedman, Joshi, & Lindquist, 2007; Wellstead et al., 2009). It was followed by series of large N surveys conducted by Michael Howlett and his colleagues during 2008–2009 in Canadian provinces and territories (Bernier & Howlett, 2012; Howlett, 2009a, 2009b, 2009c; Howlett & Newman, 2010; Howlett & Wellstead, 2011). Together with some other evidence (McArthur, 2007; Rasmussen, 1999), Canada is now by far the most researched country on this topic. In sharp contrast, only very limited evidence is available for the other two countries covered in this special issue: Australia (Butcher & Dalton, 2014; Colebatch & Gill, 2006; Hartley, Alford, Hughes, & Yates, 2013; Head, Ferguson, Cherney, & Boreham, 2014) and Czech Republic (Čada & Ptáčková, 2014; Veselý, 2013a; Veselý, Evans, & Wellstead, 2014).

The empirical evidence on subnational policy work can be analysed and interpreted against different “baselines”. First, we can compare federal policy bureaucrats located in the centre with federal policy bureaucrats located in regions. Such a comparison might reveal to what extent the difference between policy work in the centre and regions is influenced by geographic location.² In Canada, such comparison uncovered a different set of practices in regions than those at the national level and revealed the arguably poorer policy capacity of regional organizations (Wellstead et al., 2007, 2009). Regionally based policy analysts working for the federal government, for example, were found to be more commonly engaged in “street-level” advice oriented towards day-to-day firefighting, while the analysts in Ottawa engaged in more long-term strategic planning (Wellstead et al., 2007, 2009). There were differences between the regional and NCR (national capital region) respondents mean scores in all but two of the 12 roles. Not surprisingly, implementing or delivering policies or programmes were the most frequently mentioned activity by regional respondents, while the most frequent activity of NCR respondents was collecting policy-related information. Regional employees were on average older and tended to have had longer tenures within their respective organization. Nearly a third of NCR respondents indicated that they came from an academic background compared to 17% in the regions. Not all differences might be explained solely by geographic location. Institutional factors (position in the political system), do play a role as well. Over three quarters of the NCR respondents indicated they belonged to formal policy units, compared to only 39% of the regional respondents.

For the sake of clarity, it should be emphasized that regionally based policy workers work solely for federal government, and not for provinces. They are thus not subnational policy workers in the proper sense. However, examining differences between capital-based and regionally-based policy workers has theoretical implications relevant for this issue. It suggests that policy work at the regional level might be organized differently than at the central level, not only because of the different location within the overall governance structure, but simply because of geographic location. In other words, it implies that true subnational policy work (for provinces and regions) might have special features because it is embedded in different social, economic and cultural contexts (such as different values and way of life in smaller, less populated areas).

Second, it is possible to analyse the results of subnational units within the country (usually a federal state), by comparing provinces (territories, states, regions, etc.) one against another (Howlett, 2009a). Such an approach can reveal what is more generally valid for the whole jurisdiction (such as Canada) and what is idiosyncratic for the province or territory. When differences for some sub-national units occur, we can analyse factors that are related to such differences.

Third, the autonomous subnational units might be compared with higher levels of public administration within the country, i.e. usually with the central level (Howlett, 2009a, 2009b; Howlett & Newman, 2010; Howlett & Wellstead, 2012; Wellstead et al., 2011). This can help us to disentangle which differences between the central and sub-national level can be attributed to institutional and governance factors. In Canada, subnational and national policy workers were found to differ substantially in their background and job experience. Provincial analysts are predominantly (57.5%) female and relatively young in that more than 70% are under 50 years of age and more than 40% under 40 years old. By comparison, only 61.3% federal policy analysts are under the age of 50 and are majority (51.9%) male (Howlett & Newman, 2010, p. 126). Almost 60% of provincial policy workers had been in their present organizations for less than five years, including 14% for less than one year. This contrasts sharply with the federal situation described by Wellstead et al. (2007) where a sizable number (28.4%) have been in their positions for over 20 years.

According to Howlett (2009a) provincial policy analysts in many ways fit the profile of process-oriented trouble-shooters. They tend to be relatively young and well educated, from a social science background, relatively inexperienced and untrained in formal policy analysis or analytical techniques, work in small policy shops and working on exclusively provincial issues. At the same time they are more often engaged in fire-fighting activities on a day-to-day basis and lack access to academic or professional literature in their subject areas. A survey of Canadian public service senior managers and executives (both provincial and Federal) found the 56% held the view that the policy function could be improved if less time was invested in daily issues management/firefighting (Evans, Lum, & Shields, 2011, p. 14).

To this we can add the need to consider how macro-level political/ideological shifts are expressed in policy work. Several of the contributions to this special issue suggest, directly or more subtly, that such forces are significant. Head et al., 2014, in their attempt to answer the question of how externally produced policy knowledge is transferred and utilized in three of Australia's state governments is a case in point. Despite the in-vogue status of evidence-based policy-making, they find that, at least in the cases they surveyed, this is far from the reality. Astonishingly, only 19% of state policy workers responding to their survey held the view that academic research had ultimately influenced policy changes in their own policy area. If evidence is so little utilized in policy work, then what does inform policy? One is left to conjecture here, as Evans & Shields, 2014 do with respect to the role of NGOs in informing policy processes in the Canadian context, if neoliberal restructuring, including more direct political intervention, explains this apparent dis-interest by the state in research and front-line knowledge.

Fourth, policy work at the subnational level can be compared with other subsystems, especially with the non-governmental sector. How does policy work in public administration and NGOs differ? There have been very few direct comparisons of subnational state and NGO policy workers (Howlett & Oliphant, 2010; Howlett, Tan, Wellstead, Migone, & Evans, 2014; Oliphant & Howlett, 2010; Wellstead & Evans, 2013). The empirical evidence we have so far suggests that they differ profoundly. This stems from different missions and accountabilities, resourcing, and the extent to which there is actually space in the policy process for NGO participation. As noted in the theoretical discussion above, new public governance posits that the policy-making process is an open system which includes a broad range of non-state policy actors. Indeed, various studies speak to a changing policy role for NGOs but this is not precisely the role imagined by NPG theorists. Canadian NGOs do have a voice in policy but this is most frequently found in the agenda setting and implementation phases. There is comparatively little space for NGOs to shape policy formulation and design decisions (Jette & Vaillancourt, 2011; Larner & Craig, 2005). Indeed, in the Czech Republic, NGOs enjoy a much more ambivalent status vis-a-vis the state (Fagan, 2005; Green, 1999). While the case of Australia is similar to Canada where Australia's non-governmental — or non-profit — sector has struggled, to adapt to changing terms of engagement with government. Relations between government and the non-profit sector have sometimes been plagued by hostility and distrust: hindered by a lack of commitment on the part of governments to much needed structural reforms and resistance to modes of engagement preferred by the sector.

Across cases, capacity issues loom large in the NGO sector. A survey of NGOs in three Canadian provinces working in the labour, immigration, environment and health sectors found that 67.2% had no staff exclusively dedicated to policy work (Wellstead & Evans, 2013, p. 71). A lack of capacity to engage effectively in policy is a significant issue in Canada where between 36% and 62% of NGOs report that participating in policy processes is difficult for them (Carter, 2011, p. 433). Only 22% of Canadian NGOs contributed to Federal public policy as part of their mission. The relatively low percentage of NGOs participating in policy processes at the Federal level in Canada has nothing to do with a lack of interest but rather “they do not have the capacity (time or expertise) to engage in public policy” (Carter, 2011, p. 431). And the extent to which governments desire to engage with NGOs generally is a concern. Indeed, this concern is not peculiar to one country or countries sharing a political history. Several of the articles in this issue, notably Evans and Shields, Cada and Ptackova, and Butcher and Dalton, reveal that whether one is examining Canada, Australia or the Czech Republic, common issues emerge respecting NGO policy voice and the potential for their domestication by the neoliberal state. Other studies outside of those collected here echo these conclusions. One study in Australia found that NGOs were deliberately marginalized by government (Maddison & Denniss, 2005). Moreover, their submissions to government were found to have virtually no impact on policy outcomes (Phillips, 2006).

Fifth, we can compare nature of policy work in different contexts, i.e. to compare policy workers in different countries or contrast policy workers in different regions (provinces, territories). This approach seems very promising. Unfortunately it is quite limited due to lack of comparable data. Articles by Veselý, Evans and Wellstead (2014) are a first attempt in this direction. They found that Canadian subnational policy workers differ from their Czech counterparts in all possible aspects — from their background to tasks they carry out. Despite all these differences, however, the structure of policy work displays similar patterns.

Sixth, we can confront the evidence with some normative ideal or assumed standard (e.g. what should be education of PBs, what they should do in they work, etc.). This is very common approach. However, it is also the most problematic one. Setting the results against any normative standards assumes that there are clear and unquestionable baselines, which is hardly the case. Policy workers are for instance often criticized for not using analytical techniques enough. But how is “enough” to be measured? And who is in position to decide which techniques (and in what proportion) should be used? Similarly, many papers attempted to put the empirical results into a policy capacity perspective, usually arguing that in comparison with the national level, there is lack of policy capacity at the subnational level. The problem is, however, that there is no absolute measure — and actually even common meaning — of “policy capacity”. Also, there is no clear normative standard of what policy workers should look like. Sub-national policy workers, for example, are sometimes blamed for being preoccupied with fire-fighting activities at the expense of doing long-term strategic work. But what portion of fire-fighting activities is “normal” or “necessary”, and when does it become a sign of internal incapacity? It might well be that this is simply the best way of conducting policy work in a network society, and that it is the raison d’être of policy bureaucracy (Hawke, 1993).

Last but not least, the results can be interpreted in terms of well-developed theories, which in turn give rise to relevant hypotheses and interesting research questions. This should be, in our opinion, the way forward. However, with only slight exaggeration, we may say that the current policy work literature is divided into two different streams. On the one side, there is an elaborated theoretical stream (Colebatch, 2006a, 2006b; Colebatch et al., 2010), which, however, usually draws only upon anecdotal empirical evidence (often small case studies). On the other side, there are large N surveys leading to rich data sets, but usually rather descriptive analysis and perfunctory conclusions. These two streams rarely interact.

In the next section we attempt to link these two streams together and summarize empirical findings into more general propositions about the nature of policy work at sub-national level.

3.2 Making sense of the data: from surveys to theory

It has been suggested by many authors (Hoppe & Jeliazkova, 2006; Mayer, Bots, & van Daalen, 2004) that actual policy work is far from how it is portrayed in policy analysis textbooks. The empirical evidence unequivocally supports such critique. Policy workers identify themselves with many different roles and undertake multiple functions (e.g. Wellstead et al., 2009, p. 40). Under the rubric of “analysis”, public bureaucrats undertake a much wider range of activities than is usually assumed in the literature (Howlett & Wellstead, 2011, p. 619).

The world of policy workers is extremely heterogeneous, and includes many different tasks. It has been suggested by many authors (e.g. Radin, 2000) that “traditional” policy analysis tasks are just one part among many others such as communication or negotiation. Even when policy analysis tasks are conducted, they are not structured in rational “step-by-step” process, starting with problem identification and ending with solution identification (Colebatch & Gill, 2006). Current policy work cannot be correctly described as neutral, competent and objective performance of tasks associated with the application of technical policy analytical tools (Howlett, 2009b).

But what do policy workers actually do if it is not “textbook” policy analysis? And, even more importantly, why do they do it? It has been argued that the governance environment has changed in many respects with significant consequences for policy work. More diverse sets of actors involved in policy-making and the rise of networks is believed to lead to a shift of styles of policy workers. They are now expected to be engaged in greater consultation, consensus building and public dialogue than previously. Similarly, older technical forms of analysis are believed to be replaced by newer more participatory ones (Wellstead et al., 2011, p. 355).

Empirical evidence gives some support for this claim. Communication and consultation is a very important and rather frequent task of policy workers. It also suggests that regional policy workers are slightly more involved in consultation with the public and stakeholders than their federal counterparts (Wellstead et al., 2009). It was also found that regional respondents thought it was more important to involve non-government organizations in the policy-making process than policy workers working in the capital. If this so also in other jurisdictions, the shift to network governance is a bit more pronounced at the regional level.³

In general, however, government regional policy workers are still quite removed from key non-governmental policy actors (Wellstead et al., 2009). The networks seem to be mostly internal, i.e. regional policy workers interact mostly with other public administration bodies, and not with external and independent actors: “The contacts with internal networks … were notably stronger than with those outside of the government, which, as Meltsner (1976, pp. 177–196) indeed argued, suggests that many analysts are largely agency-cantered or “desk-bound” and prepare, review, or manipulate financial and other kinds of data, interacting most often with other bureaucratic government actors” (Howlett & Wellstead, 2011, p. 622). There also seem to be some differences between network types at the regional and central level. For instance, NCR respondents were more frequently contacted by think tanks than regional respondents. However, we need to learn much more about this aspect of policy work in the future.

The last research topic, where theory and empirical analysis can be linked together, is to what extent is policy work conducted within a short-term or long-term focus. In this respect, Howlett and Wellstead (2012), following McArthur (2007), tested so-called “limited autonomy hypothesis”. McArthur argued that the smaller size of government and the tighter lines of political control found at the sub-national level result in the provincial government agencies and the analysts they employ to work in a context where they enjoy limited autonomy from the demands placed upon them by political masters and by social actors such as trade unions, business associations, think tanks and interest groups. Consequently, sub-national policy workers are supposed to be much more constrained and short-term in their activities and orientations than their national counterparts, and are much more likely to follow political dictates and fashions in government, and much less able to resist pressure-group politics. In other words, because subnational organizations are smaller in scope, the organizational distance between analysts and decision makers is shorter than at the national level. As a consequence, due to their lack of autonomy from ongoing political issues and activities, provincial, territorial and regionally based employees are assumed to deal with more immediate type issues (firefighting) whereas, in contrast, NCR based employees should be more engaged with longer term issues.

As shown above, sub-national policy workers are indeed sometimes found to be more “fire-fighters” than policy workers in comparison to their national level counterparts (Howlett, 2009a). However, this hypothesis was not fully supported by further multivariate analysis (Wellstead et al., 2011). The authors do not suggest any explanation for this finding that contradicts both theory and previous findings. One possible reason could be that, work on short and long term issues is not mutually exclusive, i.e. many policy workers work in different temporal horizons. The question for further investigation is how policy workers are able to “switch” from short to long term issues, and how they cope with this obvious paradox: demand on long-term planning in environment which is in constant flux and in which any attempts to take long-term view is immediately corroded by pressing immediate issues.

4 Overview of the articles in this special issue

The individual contributions to this issue help to illustrate points made above by way of concrete empirical examples in three countries. The first part is devoted to analysis of policy work at subnational autonomous governments (provinces and territories, states, regions). The second part examines non-governmental policy work.

The article by Head et al. explores the use of research and expertise within Australian state governments. It is shown that the capacity to use external sources of expertise in policy work is influenced by broad spectrum of factors. The individual skills and dispositions of policy-related staff are important, but are not sufficient to increase substantially the use of external research expertise by public sector agencies. Meso- and macro factors do play a significant role, too. Among the most important factors are organizational cultures and practices that value expertise and rigorous evidence, and generally supportive research-oriented culture in a given organization.

Vesely, Wellstead and Evans compared Canadian provinces and territories with the Czech Republic regions. They found that in Canada, where policy analysis and policy research is much more institutionalized than in the Czech Republic, policy workers are much more involved in evidence-based work, especially in evaluation and policy research. They also deal more with policy analysis activities such as identification of policy issues and options. In contrast, Czech policy workers are more engaged in consulting with the public and briefing managers and decision-makers.

The three articles presented here concerned with the role of NGOs in the policy process, with one situated in each of the countries examined here, present diverse but still a largely common set of experiences of NGO engagement with the subnational state. Evans and Shields, basing their work on interviews with NGO and government policy workers, find that NGO policy workers view the policy process as one they have little influence upon save for delivery. Similarly, Butcher and Dalton's exploration of the restructuring of state–NGO relations in the Australian states, finds that marketization has reshaped the policy role of non-profit agencies and created a sector characterized by uneven capacities. Cada and Ptackova observe that Czech NGOs draw upon two styles of policy engagement with the local state. The choice of style is an expression of the ideological understanding of power relations within each NGO. All three contributions lead to a general conceptualization of state-NGO policy engagement as fundamentally characterized by ‘insider’ or ‘outsider’ strategies depending on the larger political context the NGOs operate.

Notes

* The work of Arnošt Veselý was supported by the Czech Science Foundation Grant (P404/12/0725). The authors would like to thank two anonymous reviewers for their valuable comments and suggestions.

1 Various authors use different labels for people involved in policy design and policy making. Meltsner (1976) wrote about “bureaucratic policy analysts”, Page and Jenkins (2005) used the term “policy bureaucracy” and “policy bureaucrats”, Howlett sometimes uses term “public sector analysts”, while in more recent articles he follows Colebatch (2006a, 2006b) and call them “policy workers”. With the exception of referring to the particular work of other authors, where we use their labels, in this introductory article we use the most generic “policy workers”. For more discussion, see Kohoutek, Nekola, and NovotnyÌ 2013 and Veselý (2013b).

2 Wellstead, Stedman, and Howlett (2011, p. 357) called this “geographical scope hypothesis” (i.e. work is dominated by the issues arising within boundaries of their jurisdiction), and found strong empirical evidence in its support.

3 But again it must be noted that regional federal policy workers should not be equated with subnational policy workers.