Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Problematizing partner selection: Collaborative choices and decision-making uncertainty

Problematizing partner selection: Collaborative choices and decision-making uncertainty Although networked collaboration is often linked to positive outcomes, choosing suitable partners for collaboration can be difficult. Actors often only have limited information about the preferences, capabilities, and trustworthiness of prospective collaborators, meaning that considerable uncertainty may underlie collaborative choices. This article analyses the decision-making uncertainty associated with collaborative choices and assesses factors that potentially mitigate such uncertainty. Based on qualitative re- search, it presents a conceptual model that brings together and specifies mitigating factors at the network-, organizational, and individual-levels of analysis. The developed conceptual model provides a clearer theoretical understanding and presentation of the cross-level factors important to consider when studying collaborative behaviour. In addition, through its focus on uncertainty, it provides more consideration of the challenges that individual officials face when choosing collaborators in complex net- worked environments. Keywords Networks, policy networks, decision making, uncertainty, collaborative governance Introduction The collaborative contacts that (public) officials maintain in networked settings are valuable for their home departments and agencies: they provide them with access to relevant information and resources, help them learn about potential policy problems and Corresponding author: *Machiel van der Heijden, Utrecht University School of Governance (USG), Room -1.15b, Bijlhouwerstraat 6, Utrecht 3511 ZC, The Netherlands. Email: m.j.a.vanderheijden@uu.nl 2 Public Policy and Administration 0(0) solutions, and create the potential for forming alliances and coalitions (Alexander et al., 2011). However, choosing suitable partners for collaboration can be difficult (Agranoff, 2006; Feiock, 2013). Individual officials often only have limited information about the preferences, capabilities, and trustworthiness of potential partners (Leifeld and Schneider, 2012; Hamilton and Lubell, 2018). Moreover, they generally lack the time and cognitive capacities to accurately map out their network surroundings (Krackhardt, 1990). As a result, considerable uncertainty may underlie collaborative choices, particularly regarding the value of prospective collaborators. Such uncertainty is important to consider because it challenges an increasing number of public officials operating across organizational and jurisdictional boundaries. Although network behaviour is typically associated with beneficial outcomes (Cristofoli and Markovic, 2016; Meier and O’Toole, 2001), the complex nature of networked envi- ronments means that in practice individual officials will often struggle to successfully engage in such behaviour. Many networking individuals (and their respective organi- zations) will not reap the benefits of networked collaboration, being overwhelmed by the complex settings in which they find themselves. As such, decision-making uncertainty undermines the promise of networked collaboration and problematizes the strategic cues by which many theoretical models assume collaborative choices to occur (see Siciliano et al., 2021). However, the uncertainty underlying collaborative choices can also be mitigated. Policy network scholars emphasize the role that the institutional context of networked collaboration can play in this regard (Leifeld and Schneider, 2012; Scott and Thomas, 2015). In particular, the formal structure of working groups, task forces or committees through which a lot of collaboration occurs, presents policy actors with a clearly defined pool of potential partners with whom they have a chance to become acquainted (Fischer and Sciarini, 2016). This reduces decision-making uncertainty regarding the selection of appropriate network partners, as these institutionalized structures make information about potential partners more readily available (Ostrom, 1998). Beyond this general insight, however, several issues still require further consideration. Firstly, although scholars increasingly find that, on average, joint participation in institutionalized settings and bilateral collaboration are correlated (Leifeld and Schneider, 2012; Lubell et al., 2017), such settings vary considerably in terms of their charac- teristics (see Choi and Robertson, 2014; Fischer and Leifeld, 2015). Some settings have many participants, others only few. Some will meet frequently, others not so much. In that sense, these settings vary in terms of the extent to which they actually provide policy actors with easy access to information about prospective collaborators. As a result, the degree to which institutional settings reduce the decision-making uncertainty underlying collaborative choices can also be expected to vary, depending on their specific characteristics. What characteristics are most relevant in this regard, remains an open question. Secondly, when thinking about the actors actually making the collaborative choices, characteristics of (individual) network participants at both the organizational- and individual-level also seem important to consider. Individuals engaged in networking typically represent organizations that have different structures and resources. Think for van der Heijden 3 instance about the existence of formal boundary-spanning roles or the use of explicit internal coordination structures for network behaviour (see McGuire and Silvia, 2010; Six et al., 2006). Moreover, organizational members themselves have different characteristics, such as their experience with collaborative settings or their expertise on the policy issues discussed therein (Juenke, 2005; Meier and O’Toole, 2010). Both types of characteristics likely influence the time and effort officials require for searching and evaluating col- laborative partners, providing another source of variation in the decision-making un- certainty underlying collaborative choices. Although uncertainty is an important theme for network scholars (Hamilton and Lubell, 2018; Koppenjan and Klijn, 2004), there is little theoretical guidance on what factors affect and mitigate such uncertainty, particularly in relation to col- laborative choices. Therefore, this paper asks the following research question: What factors at network, organizational, and individual levels potentially affect the decision-making uncertainty underlying partner choices in collaborative settings? Answering such a question helps to better understand the (theoretical) relationship between collaborative settings and the informal networks that exist within them (see Emerson et al., 2012; Hileman and Bodin, 2019). Moreover, byalsoshiftingthe analytical focus to the individual officials that operate in these settings and the or- ganizational contexts from which they do so, it provides more consideration of the cross-level factors that may be influential for (the uncertainty of) collaborative choices (see Brass et al., 2004). The resulting conceptual model brings together and specifies factors (at three different levels of analysis) that potentially affect the uncertainty that public offi- cials experience when choosing collaborators. At the network-level, characteristics of the institutional settings through which networked collaboration often occurs are emphasized, noting the frequency of meetings, group size, and decision rules as most important for the uncertainty associated with collaborative choices. At the organizational-level, explicit boundary-spanning units, formal reporting require- ments, and limiting the number of different settings are argued to help officials to more easily make collaborative choices. And at the individual-level, the experience, expertise and available time of individual officials are considered to play an im- portant role in mitigating the uncertainty experienced when making collaborative choices. This conceptual model is developed based on qualitative evidence from a prototypical complex networked environment, that is, international finance policy and regulation. Within this policy field, a wide variety of transnational networks and standard-setting bodies have emerged, around which regulatory and ministry officials from various countries engage in complex patterns of networked interaction (Ahdieh, 2016; Frieden, 2016). This complexity provides a good basis for theorizing on the decision-making uncertainty underlying collaborative choices. Before doing so, however, the next section firstly discusses why decision-making uncertainty is such a (theoretically) relevant concept for better understanding collaborative choices. 4 Public Policy and Administration 0(0) Theoretical framing Collaborative Choices and Decision-Making Uncertainty This article’s theoretical interest is primarily in the uncertainty associated with collab- orative choices and how such uncertainty can be mitigated. In thinking about this question, two aspects of networked collaboration are important to separate (Hamilton and Lubell, 2018; Scott and Thomas, 2017). On the one hand, there often exists a formal (or institutionalized) part of such collaboration, in which representatives of organizations participate in collaborative institutions to solve collective-action problems . On the other, complex webs of informal collaborative linkages exist within and beyond these for- malized settings, in which actors voluntarily contact each other for further information exchange and coalition building. Although these two layers of collaboration influence each other (see Leifeld and Schneider, 2012), it is the latter form of collaboration for which the question of col- laborative choices and decision-making uncertainty is most relevant. This is the level at which (individual) officials make selective choices about with whom to collaborate more closely, typically under conditions of limited information about others’ preferences, capabilities and trustworthiness. Because individual officials do not always have the required time and effort available to gather this information, considerable decision- making uncertainty may underlie (informal) collaborative choices. Particularly the value of prospective collaborators and the predictability of their behaviours often remain unclear. Such uncertainty has been noted by network scholars in several ways. Generally, it refers to the expected behaviours, strategies and intentions of potential collaborative partners (Koppenjan and Klijn, 2004). More specifically, Hamilton and Lubell (2018: 230–231) define a form of political uncertainty to describe a lack of awareness of the policy beliefs and preferences of other actors. And Nohrstedt and Bodin (2020: 1092) define collaborative uncertainty as to indicate the risk that a potential collaborative tie has limited positive outcomes, implying a waste of time and resources. Both describe this uncertainty as a source of transaction costs, requiring actors to learn about the preferences of others and monitor their behaviours. Closely related to the above-provided definitions, this article defines uncertainty as the limited information that actors have about the preferences, capabilities and trustwor- thiness of potential collaborators, making it hard to establish their value as a collaborator. However, rather than describing such uncertainty in terms of transaction costs, the theoretical interest is more in how such uncertainty challenges actors to operate col- laboratively. In practice, high degrees of uncertainty mean that actors are often over- whelmed by the complex networked settings in which they operate. This problematizes network behaviour and existing theoretical frameworks used to explain collaborative choices. Many scholars, for instance, theorize that preference similarity is a main driver of collaborative patterns¸ meaning that actors will establish contacts with like-minded actors in terms of (policy) beliefs or values (Calanni et al., 2015; Henry et al., 2011). However, van der Heijden 5 such explanations assume a well-defined policy space in which preferences of potential partners and that of the choosing actors are known. In reality this is often not the case and, particularly for complex and newly emerging issues, actors might actually enter policy arenas to ‘discover’ their own and others’ preferences (Jones, 2001: 102). In the face of uncertainty, preference similarity (or dissimilarity for that matter) is likely a poor predictor of collaborative patterns. Or take scholars that use resource dependence theory to theorize about collaborative choices. The assumption here is that policy actors use collaborative ties to maximize their access to (political or technical) resources (Park and Rethemeyer, 2012). Actors will thus seek out partners that they perceive as influential or technically competent, due to their control over (or access to) critical resources such as information, technology, personnel or political clout (Henry, 2011; Matti and Sandstrom, 2011). However, assessing the in- fluence or technical competence of others can be difficult, particularly when confronted with a large group of potential collaborators that are relatively unfamiliar (Hamilton and Lubell, 2018). The strategic cues assumed to drive collaborative choices are thus problematic when related to the considerable uncertainty that may exist about the preferences, capabilities and trustworthiness of potential collaborators. Although goal-oriented behaviour is a reasonable assumption for network actors, the complexity of networked environments and the limited time and resources that actors typically have, will often interfere with such behaviour. As a result, public and policy officials will sometimes make poor collaborative choices with collaborators that prove unreliable or incompetent in the long run. In that sense, decision-making uncertainty potentially undermines the promise of networked collaboration. This makes it crucial to think about factors that affect such uncertainty and how it may vary between actors and across settings. Mitigating decision-making uncertainty In thinking about factors that influence the decision-making uncertainty associated with collaborative choices, the mechanisms by which such factors might do so must firstly be spelt out. The previous section established that uncertainty primarily grows out of the limited or inadequate information that actors have about the preferences, capabilities, or trustworthiness of potential collaborators. In addition, given the limited time and ca- pacities that individuals have to gather such information, collaborative choices are often made under conditions of uncertainty. Factors that mitigate the uncertainty of collabo- rative choices, are then those that (1) make it easier to access information on prospective collaborators or (2) provide actors with the time and resources to gather such information. To structure the discussion of potentially relevant mitigating factors, three separate levels of analysis are distinguished: the network-, the organizational- and the individual- level. The network-level refers to characteristics of the institutional settings through which networked collaboration often occurs. The organizational-level refers to characteristics of the organization that network participants typically represent when engaged in network behaviour and collaboration. And the individual-level focuses on characteristics of the actor that actually does the networking on behalf of a particular organization or agency. 6 Public Policy and Administration 0(0) The goal is then to identify factors at all three levels that mitigate decision-making uncertainty associated with collaborative choices. Starting with the network-level, policy network scholars have underlined the im- portance of the institutional settings and structures that often facilitate networked collaboration (Fischer and Sciarini, 2016; Leifeld and Schneider, 2012). Such settings are assumed to promote collaborative behaviour on behalf of participants and allow for the development of trust relationships between them (see Klijn et al., 2010; Vangen and Huxham, 2003). In thinking about collaborative choices, the joint participation in particular working groups, commissions or task forces seems particularly important. These institutionalized structures provide actors with a delineated pool of potentially relevant and suitable partners from which to choose, minimizing search costs. Moreover, the regular meetings of these collaborative institutions facilitate commu- nication between participating actors and gives them the opportunity to learn about each other’s preferences. The link with mitigating uncertainty about collaborative choices is thus clear: by providing easy access to information on potential collabo- rators, institutional settings make it easier for actors to assess the value of prospective collaborators. However, not all institutional settings will provide these benefits in reducing the uncertainty of collaborative choices. Although, on average, joint participation in insti- tutionalized settings and bilateral collaboration are correlated (Leifeld and Schneider, 2012; Lubell et al., 2017), it is important to note that such institutionalized settings may vary considerably. Think for instance about the number of participants that they have (Hertz and Leuffen, 2011), their perceived importance for participating actors (Fischer and Sciarini, 2016), their decision rules (Choi and Robertson, 2014), or the frequency with which scheduled meetings occur within them. These characteristics potentially affect the extent to which institutional settings actually provide actors with relevant information about potential collaborators. By extension, this also affects the degree to which such settings help mitigate decision-making uncertainty regarding further collaborative choices. However, there is currently little guidance on what specific characteristics of these settings are most important to consider. Besides these factors at the network-level, an important consideration is that net- working actors typically represent an organization and operate collaboratively from within a given organizational structure. The ability of actors to form and sustain col- laborative relationships is at least partially a function of the internal dynamics of that organization (McGuire and Silvia, 2010: 281). This makes it relevant to also consider factors at the organizational-level when asking the question of how the uncertainty associated with collaborative choices can be mitigated. For one, the structure of an organization allocates work roles and influences how organizational officials are related to the external environment (Child, 1972). By creating explicit boundary-spanning roles, for instance such structures can determine the time and resource that organizational members have for networking activities. Or internal coordination structures for collaborative efforts can give them access to information on prospective collaborators possessed by colleagues. van der Heijden 7 In addition, the number of different settings in which the organization is involved, likely affects the capacity of its members to operate collaboratively. Although partici- pating in many different collaborative networks provide clear benefits in terms of the access to diverse information (i.e., the strength of weak ties), meaningful collaboration arguably requires extensive networking on behalf of the individual officials representing the organization within these settings (see Hileman and Bodin, 2019). The more different settings (or policy fora) an organization engages in, the less time organiza- tional representatives have for searching and evaluating suitable collaborators within each particular setting. If the networking capacities of the organization are scattered across a large number of settings, many potential collaborators within these settings will remain relatively unfamiliar. This creates additional uncertainty around collaborative choices. And lastly, the individual-level also provides some factors that potentially affect the decision-making uncertainty underlying collaborative choices. Most importantly, actors making collaborative choices likely differ in terms of their experience with networked collaboration and their expertise regarding the (policy) issues discussed therein (Juenke, 2005; Meier and O’Toole, 2010). Regarding expertise, officials with technical knowledge on the policy issues in which they are involved can better assess the capabilities of potential collaborators (Williams, 2002). Moreover, for highly complex and newly emerging policy issues, actors with technical expertise can more easily evaluate their own policy positions, as well as those of others. This gives them more capacity to perform the decision-making task of partner selection, particularly in terms of information-processing and search behaviour (Day and Lord, 1992; Taylor, 1975). For (networking) experience, an important consideration is that experienced officials will have built up more elaborate informal networks through past interactions. This creates a number of benefits for future collaborative choices. For one, experienced actors typically have better mental representations of the policy spaces of the networks in which they operate, as well as a higher sensitivity toward the constellation of other actors and their interests (see Halevy et al., 2019). In addition, network scholars typically con- ceptualize the existing informal network as an information repository to reduce uncer- tainty about the trustworthiness of potential partners and learn about opportunities for new ties (see Gulati and Gargiulo, 1999; Henry et al., 2011). Having a more elaborate informal network (i.e. social capital) allows one to make better use of the information signals encoded in existing network structures and provides more opportunities for making contact, for instance through referrals or introductions via mutual friends (Quintane and Carnabuci, 2016). Overall, mitigating factors at three levels of analysis are thus relevant to consider when studying decision-making on collaborative choices and its underlying uncertainty. The mechanism by which these factors can be expected to mitigate decision-making un- certainty is by providing individual officials with easier access to information on the preferences, capabilities, and trustworthiness of potential collaborative partners. Al- though some specific factors are hinted at in the above-provided discussion, many still require further elaboration. The empirical analysis identifies what concrete factors are important to consider and specifies how they matter for the uncertainty associated with 8 Public Policy and Administration 0(0) collaborative choices. Firstly, however, the data collection and analysis procedures of this study are described. Data collection procedures The main goal of this study is theory elaboration, in which empirical analysis is used to further specify pre-existing conceptual ideas (Fisher and Aguinas, 2017). In particular, the focus is on construct splitting, that is, splitting existing theoretical constructs into specific dimensions based on observed empirical realities (ibid.,: 446). What this means for this article is that the section above has theoretically developed the idea that decision-making uncertainty about collaborative choices can be reduced by factors at different levels of analysis, and the empirical analysis below is used to specify what type of factors are likely to play a role in this regard. Qualitative research is particularly suitable for these purposes, as it enables attention to be given to particular circumstances, whilst its open-ended nature is sufficiently flexible for new insights or themes to emerge (see Piore, 2006). Case Selection and scope conditions To focus the analysis, this article looks at transnational collaboration in the field of fi- nancial regulation and zooms in on the way in which public officials of three Dutch organizations are involved in such collaboration (one ministerial department and two regulatory agencies). This research context is appropriate for the purposes of this research as it provides variation at different levels of analysis, which can be used as a basis for theorizing. For one, national agencies within this sector typically operate in a wide variety of transnational policy settings, both in terms of their level of institutionalization and the scale of collaboration (regional vs. global) (Ahdieh, 2016) . This means that there is a lot of variation in terms of the networked settings in which interviewed respondents operate, allowing us to collect and compare data on the importance of the different characteristics of these settings. In addition, the Dutch context allows for comparison between three different orga- nizations involved in transnational regulation and policy-making. Given the twin peaks model of supervision that exists within the Netherlands, responsibilities for financial sector regulation are split between two separate agencies, with one regulatory agency responsible for securities regulation, and another for banking, insurance, and pensions regulation. Moreover, the ministerial department also does transnational policy work, primarily in the context of the EU but also in global platforms such as the FSB. This allows for comparing officials engaged in (transnational) network activities within three different organizational contexts, whilst the nature of the sector in which they engage is largely similar. Data collection and analysis The presented analysis is based on qualitative research in the form of semi-structured interviews. 16 face-to-face interviews were conducted with Dutch officials involved in van der Heijden 9 international financial policy and regulation (see Appendix A). These respondents oc- cupied different positions in the regulatory policy departments at either the Dutch Ministry of Finance (n = 4), DNB (Dutch Central Bank) (n = 7), or AFM (Securities Regulator) (n = 5). The one common denominator of these respondents is that they were all heavily involved with international networks, at either (or both) the European or global level. The respondents were identified through a combination of snowball and purposeful sampling, in which heads of departments were firstly approached for an interview. These heads of departments were asked to identify individuals in the organization involved in the different networked settings in which the agency or ministry engaged. Maximum variation in terms of networked settings was thus an important criterion for identifying the different respondents. Although this type of sampling creates a potential bias, it does help to recruit respondents that are otherwise difficult to reach. Moreover, interviewing several respondents from the same organization and asking them about similar processes helps to verify the provided accounts and gives a more balanced representation of the phenomenon of interest. Drawing on a topic list, these respondents were interviewed (45 min on average) by the author in a semi-structured fashion. Topics discussed in the interviews were – inter alia – the kinds of international platforms in which they participate, how they prepare for international meetings, who their contacts are, what channels they use to influence the international regulatory process, and how they went about selecting partners and for what reasons (see Appendix B). Grand tour questions were asked about what a typical preparation for international meetings looks like (Leech, 2002), as well as example questions about issues discussed in international meetings. Moreover, explicit probing was carried out for the way in which the respondents identify potential collaborators and what the difficulties are in this regard. The recordings were transcribed and analysed through a process of coding. Three steps were involved in moving from the raw interview transcripts to further theorizing on decision-making uncertainty and partner selection. Firstly, topic coding was used to identify passages relevant to decision-making uncertainty regarding partner selection (Richards, 2020: 110). These passages typically hinted at the time and effort needed for making contact and the way respondents engaged in network behaviour more generally. In addition, passages describing network-, organizational- and individual-level charac- teristics were identified and categorized according to the level at which they belonged. This gave a first selection and initial ordering of interview passages relevant to the theoretical question of interest. As a second step, this collection of passages was then reviewed as to develop analytical categories (ibid.: 110–112). This analytical coding was used to interpret and further select passages categorized under different levels of characteristics, particularly assessing which of these passages were relevant for decision-making uncertainty underlying collaborative choices. At this stage, more specific codes were identified and attached to the passages as to signify specific characteristics (see Fisher and Aguinas 2017), for example ‘group size’ at the network-level or ‘experience’ at the individual level. 10 Public Policy and Administration 0(0) Lastly, the identified categories were related to the existing literature on collaboration and decision-making as to theoretically interpret their meaning, and to formulate specific propositions on how the identified characteristics are important for decision-making uncertainty underlying partner choices. For this last step, the different characteristics were primarily evaluated in terms of their likely effect on the time and effort needed to gather information about potential partners and identify their preferences, capabilities and re- liability (i.e. the core mechanism by which uncertainty is mitigated, as argued in the theoretical framework). Analysis As a precondition to identify which characteristics are relevant for the decision-making uncertainty underlying collaborative choices, a first step is to establish that such un- certainty is actually an issue for the respondents of this study. This is what the next section does. Then separate sections present empirical material pointing to the importance of particular characteristics at three different levels of analysis (i.e. the network-, organi- zational- and individual-levels). This empirical material is interpreted theoretically, on the basis of which several propositions are developed. The problem of partner selection Within the policy field of international finance there exists a large variety of international organizations and standard-setting bodies that can be characterized as (transnational) networks (Ahdieh, 2016: 76). Within the institutionalized settings of these networks, interaction between national officials primarily occurs through the various working groups, commissions, or task forces, which carry out most of its operational work. Being connected in such a way simply means receiving the same group mails or periodically attending the same meetings. Besides these more structured interactions, however, regulators also meet each other informally on a more ad hoc basis. This is the level at which officials voluntarily interact with others to exchange (political or technical) in- formation and engage in coalition building to influence decision-making. For these latter types of networked interactions, the question of collaborative choices and decision- making uncertainty is most relevant. In making these collaborative choices, established drivers of collaboration such as preference similarity and perceived influence play an important role for respondents. They frequently talk about their ‘natural partners’, whom they know think about certain issues the same way. As one regulator stated, ‘if there’s an important issue, for us, coming up, for which we know there will be a lot of difference of opinion, we try to mail, call with similarly-minded countries, to see how we can best go into such a meeting’ [R5]. Similarly, for perceived influence, many respondents expressed a preference for working with the ‘big countries’, that are influential and resourceful. In discussing collaborative partners one respondent noted: ‘you know, these nine countries, why did we choose them.. [because] they are all big, semi-big [countries]. I mean, the small ones... they just don’t have the capacity’ [R6]. van der Heijden 11 However, identifying the preferences and capabilities of actors is not as straightfor- ward as it sounds. Given the wide variety of different topics that are dealt with in financial regulation, the cards are constantly reshuffled for every new topic or issue that national officials have to deal with. Although respondents typically talk about ‘natural partners’ with whom they share similar interests, they are quick to emphasize that ‘your natural partners differ per topic’. A senior regulator noted that ‘for me, it is not really the case that you have a fixed group [...] you really have to search your coalition depending on the topic’ [R3]. Moreover, given the fast developments and innovations in international finance, regulators are also often confronted with topics on which they have not yet formed a position. As one regulatory official noted in preparing an international meeting with an extensive agenda, ‘it is our role, given these 25 points, to ask, how important is this [for us], what is the constellation of power, […] do we have a chance?’ [R5]. As a consequence, the search for suitable partners can be rather complex and uncertain. For each new issue, involved officials will have to find out what the policy positions of other network participants are. Moreover, you have to know which potential partners actually want to collaborate and whether they are capable enough to reciprocate your own efforts. This information about other actors is not always clear, nor are all other network participants equally approachable. As one senior regulator remarked, ‘if you’re in a project group, then you have really active countries, that are just involved. Some countries are not very active in the project group and you have to reach them at a different level’ [R4]. In terms of approaching collaborators, another agency official noted: ‘most big countries have a separate desk, a [network-X] desk that you can contact... I also once sent an email to [a smaller agency], to ask who did the coordination [of network-X].. It turned out to be director himself. So that complicates things’ [R5]. Besides the difficulties in identifying or reaching particular partners, the trustwor- thiness of others is sometimes also difficult to assess. In the interviews, several re- spondents complained about collaborators who are ‘indirect’ in their communications or even ‘unreliable’. In discussing potential partners for collaboration, one senior regulator noted how, ‘with the guys from [country X] I just communicate better, with [regulators from country y] it always stays with niceties, but... what do you really think’ [R3]. Similarly, a ministry official remarked after striking a deal with a foreign counterpart: ‘with them, you’re never completely certain, whether you’re being played with, let me put it that way’ [R16]. In that sense, potential risks for defection remain an issue when choosing potential collaborators. Overall, the interviewed respondents seem to have difficulty estimating the capa- bilities, preferences, and trustworthiness of potential collaborators and devote time and effort to acquire this information. Regardless, they often remain unclear about the value of prospective and current collaborators. This makes decision-making uncertainty, as de- fined in the theoretical section, a relevant consideration for their collaborative choices. The next section looks at the factors that potentially affect this degree of uncertainty. 12 Public Policy and Administration 0(0) Mitigating uncertainty: The network-level The theoretical section noted how institutional settings can help officials to better make collaborative choices (Hamilton and Lubell, 2018; Leifeld and Schneider, 2012). The interviews provide extensive anecdotal evidence for this consideration. As one senior official noted in discussing the way in which he contacts collaborators, ‘[…],usually you speak with the people from your committee, that is your first point of reference. You know them, you experience them in meetings, you sometimes have had dinners with them. Those are the people with whom you have had the most contact.. so you’ll speak with them first’ [R3]. The working groups, committees and task forces in which domestic officials participate thus help them to delineate their choice set of potential partners for further collaboration. Moreover, through the interaction occurring within these groups, they can become acquainted with others, providing a low-cost strategy to identify and select appropriate partners. Notably, institutionalized settings help respondents to identify potential collaborators for both political and technical information exchange. For political-strategic information exchange, one regulatory official described how in looking for potential partners, ‘you look for a coalition with people of whom you know they have similar ideas, and there is only one way to find that out, and that is to make sure you’re in those [working] groups’ [R6]. Similarly, for identifying partners with whom to exchange technical information, a securities regulator noted how in the transnational sessions in which he participates, ‘it becomes more clear what issues are prevalent for different countries. After such a session you can determine, wait.. I have to contact colleagues in Spain or colleagues in Brazil, because they also have problems with their mortgage markets, or whatever’ [R11]. The institutional settings of the networks in which respondents participate thus provide an important context to collaborative choices: informal network patterns often grow out of formalized structures. However, the interviews also demonstrated that these institutional settings vary considerably in terms of their characteristics. Firstly, respondents note how some groups in which they participate only meet two or three times a year, whilst others do so on a more frequent basis. This frequency of meetings is obviously important for how often actors see each other face-to-face and have a chance to become acquainted. In reflecting on his participation in both European and global networks, one senior regulator noted that ‘the frequency and intensity in Europe is much higher. So you meet more often and more intensively within Europe than [....] globally. This means you know each other better, are more familiar with their systems. You know more’. [R2]. Similarly, another regulator remarked about a working group that meets relatively frequently: ‘the advantage of those working groups is, you meet each other multiple times a year. [..] So, if you’re a bit pro- active, within half a year [..] you’ve spoken with everyone in one way or another’ [R7]. Overall, participating in the same working group or policy committee gives officials a chance to meet and become acquainted. This makes it relatively easy for them to acquire information on the preferences and capabilities of potential collaborators, hence reducing the uncertainty about collaborative choices (see also Scott and Thomas, 2017). However, respondents also participate in groups that meet relatively infrequently. This means less van der Heijden 13 opportunity for face-to-face interaction and a smaller chance for these benefits to accrue. In these instances, one can expect that the additional information about the preferences, capabilities and trustworthiness of potential partners that such institutional settings provide is likely to be restricted. The resulting unfamiliarity between actors will make it harder for actors to navigate the networked settings in which they operate and collab- orative choices will remain uncertain (see also Lubell et al., 2017). Given the variation across institutionalized networks in terms of the frequency with which meetings occur, a first expectation is that Proposition 1: The higher the frequency of meetings within institutionalized networks, the lower the decision-making uncertainty underlying collaborative choices. Secondly, respondents also reported a large variety in terms of the number of par- ticipants of the working groups, commissions, boards and task forces in which they participated. Whilst some talked about groups in which only 7 other people participated, others mentioned numbers up to 30 or more. An obvious consequence from this variation in group size is that it determines the time and effort you have to devote to getting to know the others within your group. This leads to selective behaviour on behalf of domestic officials about whom to contact as transnational collaborators. As one regulator men- tioned about identifying partners with similar preferences in a relatively big group, ‘it’s not that I’m going to call all 28, definitely not... You call, with whom you expect you’ll have the biggest chance that it will work’ [R8]. Similarly, a ministry official noted that ‘after a while, you recognize the most important faces. But of course it’s a big group, 28 countries’ [R16]. These reported challenges of networked interaction when groups are larger, make variation in the size of groups that come together within collaborative settings theo- retically interesting to consider. As noted by Hamilton and Lubell (2018), joint partic- ipation in working groups or commissions does not ensure that participants actually interact. Particularly when these groups have a large number of participants, the chances of interaction between two particular members are smaller (see also Fischer and Leifeld, 2015). Network scholars have extensively reported on how with each additional network participant the number of potential connections increases exponentially (see Borgatti et al., 2009), making these institutionalized settings more difficult to navigate. These considerations are important for the uncertainty underlying collaborative choices, as it means that more information is required on a larger number of co-participants. Theo- retically, one can then reason that given the restricted time and effort that officials can put into acquiring such information, a larger group size means that choices about collabo- rative partners will inevitably be characterized by higher degrees of uncertainty. Proposition 2: The larger the size of the groups within institutionalized networks, the higher the decision-making uncertainty underlying collaborative choices. And thirdly, respondents report how the decision rules of the institutionalized settings in which they participate are important for the way in which they make collaborative 14 Public Policy and Administration 0(0) choices. In particular, voting procedures determine the degree to which actors can be selective in collaborative choices, or also need information on all other co-participants. As one ministry official noted in reflecting about his partner selection strategy, ‘some topics go by unanimity, then you get different kinds of negotiations…. Other negotiations go by QMV [Qualified Majority Voting], then you see much stronger, I mean, everyone can count how many votes a country represents, and then you can count […] do I have a blocking minority or not ‘[R15]. Another regulatory official described an instance in which the members of his working group had to reach consensus on a set of recom- mendations, in which ‘it was an intensive process to get everyone on the same line. So it costs quite a lot of time and lot of diplomacy skills and negations to eventually get a version that we could back but also the others[...] It was intensive in the sense that we had different conference calls, write different versions, constantly adjusts things, make a new versions, ask reactions, process reactions, or not, [....] So, there was a whole process beforehand’ [R12]. Based on these descriptions, such decision rules seem to play an important role in shaping and constraining the deliberation and decision processes within institutionalized settings (see also Choi and Robertson, 2014; Fischer and Leifeld, 2015). In this way, they also affect how network participants have to choose collaborators for strategic infor- mation exchange and coalition building. The crucial divide here is between the use of unanimity or majority rules to achieve decision-making. To some extent, majority rules simplify partner selection because public officials can focus their attention on a limited number of actors, whilst others can be ignored. With consensual decision-making, however, also less familiar actors have to be involved. Moreover, actors with more extreme positions have to be facilitated (Miller, 1985). This arguably increases the uncertainties underlying collaborative choices, as more information is needed on a larger number of actors. Moreover, additional time and effort are required in gathering such information from actors that are relatively unfamiliar. Proposition 3: The larger the majority needed for making decisions within institu- tionalized networks, the higher the decision-making uncertainty underlying col- laborative choices. Mitigating uncertainty: The organizational-level Organizational settings also matter for the way in which respondents operate collabo- ratively. Firstly, the way in which the studied organizations internally structure and coordinate external network activities mattered a great deal for the respondents. In particular, the existence of specific units or roles focussing on the strategic aspects and coordination of network activities stands out. Both regulatory agencies had special teams of coordinators overseeing transnational interactions that served as ‘the internal and external point of contact for all matters related to [network X]’. By assisting or advising experts from other units when engaging in transnational activities, they help other officials that arguably have less time to do so. As one regulatory official fulfilling such a co- ordinative role noted, ‘in these [preparatory] meetings, possibly we walk through the van der Heijden 15 agenda of the committee in which someone partakes. Just to see, what’s there, what are the issues... Also, to get a sense of, how is it going, what is the structure, what is the vibe, who are the natural allies’ [R8]. Overall, such structures seem to allow for the pooling of expertise and network capacities within the organization, helping individual officials to more easily gather information on potential collaborators. Theoretically, the above-described coordination structures point to the importance of functionally specialized units and explicit boundary-spanning roles overseeing and co- ordinating network activities. Within organizational theory, such boundary-spanning units are traditionally seen as an important way for organizations to cope with envi- ronmental uncertainty (Aldrich and Herker, 1977; Thompson, 1967). Officials within such roles or units have the time and resources to strategize on networked environments, using this information to identify appropriate partners and advise others within the or- ganization. Moreover, the team-based structures through which international meetings are prepared, lets officials pool their attentional capacities and expertise (6 et al., 2006). In that sense, internal structures help to mitigate decision-making uncertainty regarding partner selection, primarily through a more efficient way of processing information on potential partners and their preferences, interests, and capabilities. Given that organizations can be expected to vary in terms of the degree to which they have such internal structures and formal boundary-spanning roles in place, a reasonable expectation is that Proposition 4: The more explicit boundary-spanning roles that organizations have, the lower the decision-making uncertainty underlying collaborative choices. Secondly, the number of different networked settings in which domestic agencies or ministries participate is important to consider. Respondents of all three organizations carefully reflect on how their organizations have to prioritize in the working groups, task forces, and committees in which they participate, pointing to the ‘scarce resources’ or ‘fte- resctrictions’ [i.e. full time equivalent] of their organizations or units. Different levels of engagement are considered for each working group, task force or committee in which organizational members can potentially participate, ranging from not participating at all or only being an ‘email-member’, to actively ‘writing’ or even trying to become ‘chair’.An underlying rationale for this prioritization is given by a senior banking regulator, who notes that ‘if there are points that you think are important, experience teaches that, if you want to be effective to bring in these points, it takes a lot of time and energy. It’s more than just going to the meetings and making your argument. If you want to be effective, you have to lobby. So, then there’s a lot more to it’ [R3]. In other words, to be influential in networked settings, considerable time and resources are required. However, given that the networking capacities of agencies are inevitably restricted, they have to be carefully distributed across the many different settings and venues in which the agency can potentially participate. The above-provided descriptions hint at the strategic choices that agencies have to make when faced with complex networked environments: participate in many different settings or focus on only a few. An implication of the former choice is that one is confronted with many other potential collaborators, whilst also having less time to spend on establishing and maintaining 16 Public Policy and Administration 0(0) collaborative ties within each separate setting (Hileman and Bodin, 2019). This makes venue shopping and issue prioritization an important consideration for networked col- laboration (see Weible, 2005), as it allows agencies to focus on a restricted number of settings and potential collaborators. This also means less information needs to be acquired on the preferences, capabilities and trustworthiness of prospective collaborators. A re- sulting expectation is that Proposition 5: The larger the number of networked settings in which organizations participate, the higher the decision-making uncertainty underlying collaborative choices. Thirdly, levels of formalization and the standard operating procedures that exist within organizations are important for the way in which domestic officials engage in transna- tional collaboration. Respondents point to the formalized process by which they report on international meetings, and how this has a function in the preparation of future meetings. As one regulatory official described, ‘of all meetings a report is made, and that is shared with the relevant colleagues, and that is usually then, on the one hand the report, and on the other hand the contribution of the experts.... together that is the input for the next rounds. So in that sense, it is an iterative process’ [R7]. Similarly, a ministry official noted how, ‘for meetings that officials go to themselves, they have to write reports. And those are saved and that’s basically the archive. And it is always convenient to know, what has been said, and sometimes you need them again, because you cannot remember it all’ [R16]. Importantly. this stored information can be accessed by organizational members when preparing international meetings, even if they were not involved in previous interactions. In that sense, formal reporting requirements regarding network activities are important to take into account when considering collaborative choices. In particular, the extensive backlogs and reports of network meetings and activities create organizational memory through which (strategic) information regarding previous interactions is stored (Schilke and Cook, 2013). This information can be accessed by organizational members when preparing network meetings, even if they were not involved in previous interactions. These standard operating procedures in reporting about network activities thus encode experiences that help guide organizational behaviour (Moynihan, 2008). In particular, by providing information on the preferences and actions of others in previous meetings, such formalized reports mitigate decision-making uncertainty regarding partner selection when preparing new meetings and thinking about collaborative choices. Proposition 6: The more formalized reporting procedures organizations have about network activities, the lower the decision-making uncertainty underlying collab- orative choices. Mitigating uncertainty: The individual level At the individual-level, a first point to consider is the experience that officials already have with the network settings in which they collaborate. Experienced respondents talk about van der Heijden 17 ‘being used to conducting business internationally’, whilst others talk about having to learn how ‘everything works internationally’. Moreover, respondents note how the contacts they know from one setting can also be important for other settings, particularly when participating in a new working group for the first time. As one regulatory official remarked about establishing contact with foreign counterparts, ‘If you’ve been doing this for a while, you run into international colleagues in these working groups... who you know from other working groups. So that helps’ [R7]. Moreover, such experience is crucial for lowering barriers to cooperate, as it allows you to rely more on the informal networks you have built in the past. As one official noted, ‘if you a have case that crosses borders, and you have to collaborate, it just helps if you already know someone in- formally, once had a dinner, or already collaborated with someone during a meeting’ [R12]. A general observation based on these passages is that experience with collaborative settings seemingly allows officials to resort to the informal ties they have built up in the past. As noted in the theoretical section, such prior ties are an important way in which uncertainty about future interactions can be mitigated (see Gulati, 1995; McEviley et al., 2003). Actors acquire information from their past interactions and resort to this infor- mation when considering potential collaborators. The social capital that experienced actors accumulate over time, benefits them in making new collaborative choices (Henry et al. 2011). In addition, through experience actors seemingly develop networking skills and gain more oversight of the complex institutional environments in which they find themselves (see Meier and O’Toole, 2010; Williams, 2002). Arguably, this makes it easier to acquire and interpret information about prospective collaborative partners. Proposition 7: The more experience actors have with network activities, the lower the decision-making uncertainty underlying collaborative choices. Secondly, the expertise of officials is noteworthy, particularly given the highly technical policy discussions in which interviewed respondents often engaged. As one regulator noted, “all of a sudden you are in these policy discussions and it is really hard to see where you are exactly and what you think [about an issue.]... Over time it gets better” [R7]. Similarly, a ministry official mentioned that “this area is eventually just very complex, and within our department, my division is also the most technical one... and there is a lot of issues... If you’re new in such an area, you really have to invest in it a lot at the beginning” [R14]. Lacking such technical expertise, means that officials have to work harder to find what the relevant issues are and what the main policy discussion are about. For decisions regarding partner selection, this is an important consideration as it likely makes networked environments much harder to navigate. Although these quotes point to the important mitigating effect that experience is likely to have, it seems clear that for highly specialized policy discussions it is potentially harder to evaluate your own (policy) position, as well as that of others. This inevitably com- plicates networked collaboration and the specialized expertise of officials engaging in such collaboration is thus important to consider. Knowledge is the “currency of col- laboration” (Emerson et al., 2012: 16) and actors with specialized expertise will more 18 Public Policy and Administration 0(0) easily integrate and evaluate the knowledge held by different actors (Weber and Khademian, 2008). Relating it to the uncertainty associated with collaborative choices, specialized expertise allows actors to more easily process information regarding the capabilities and policy positions of potential partners. Proposition 8: The more specialized expertise actors have when engaging in network activities, the lower the decision-making uncertainty underlying collaborative choices. Lastly, the interviews pointed to the consideration that respondents vary on the available time they have for transnational network activities in general and preparation of international meetings in particular. Importantly, for many respondents, participating in (transnational) networks is a duty they have besides the other core tasks or functions for which they are responsible. whilst preparing international meetings can take a lot of time. As one ministry official remarked, ‘[..] I already lose one day [a week] just calling the different counterparts. And then you also need another day just reading the underlying documents’ [R13]. Such intensive preparation is problematic for officials that do not have such time available. As one regulatory official noted ‘On average, I try to devote half a day a week, to this work. It would be good to devote much more work into this’ [R5]. Another regulatory official noted that preparing international meetings simply requires ‘a lot of talking on the phone, a lot of conference calls. Ideally you would also meet each other face-to-face, but that isn’t always doable. It just takes too much time’ [R12]. As these quotes point out, the time that officials actually have available for network activities is seemingly important for the way they engage with potential collaborators. Although such available time is partly dependent on the existence of explicit boundary- spanning roles (proposition 4) and the number of different setting in which the orga- nization participates (proposition 5), it is also a noteworthy separate factor to consider given its direct influence on individual-level decision-making. Scholars studying the effects of time availability on decision-making primarily note its importance in terms of search behaviour and the number of alternatives considered (Bluedorn and Denhardt, 1988). More available time means more opportunity to acquire information about other actors in the networked environment. Having the time to get to know the network likely leads to less uncertainty when deciding about with whom to collaborate more closely (Juenke, 2005). Proposition 9: The more time actors have available for network activities, the lower the decision-making uncertainty underlying collaborative choices. Discussion and conclusion Overall, factors at three different levels of analysis may influence the degree to which the decision-making on collaborative contacts is characterized by high or low degrees of uncertainty. Importantly, all of the identified factors do so through the mechanism of providing easier access to information on the preferences capabilities, and trustworthiness van der Heijden 19 Figure 1. Conceptual model: collaborative choices and decision-making uncertainty. of potential partners. This unifying mechanism allows us to specify an overall conceptual model in which the formulated propositions are brought together (see Figure 1). This conceptual model provides a clearer theoretical understanding and presentation of the cross-level factors important to consider when studying collaborative behaviour. The institutional setting at the network-level has an important role as it creates familiarity between actors and helps them minimize search costs (see Hamilton and Lubell, 2018). However, as this article demonstrates, individual- and organizational-level factors should also be considered. At least in part, we can assume the skills and abilities of individual officials and the administrative capacity of an organization to contribute to (the capacity for) collaborative behaviour as well (McGuire and Silvia, 2010). Integrating these factors at different levels of analysis into a single model, helps us think more clearly about the decision-making problems that confront an increasingly large number of public officials that have come to operate outside the boundaries of their organizations, whilst also providing ideas on how such problems may be mitigated. Theoretically, the notion of decision-making uncertainty puts pressure on existing frameworks typically used to hypothesize on partner selection within collaborative networks. Given the complexity of networked environments, models emphasizing ra- tional decision-making and strategic cues potentially require too extensive information- processing capabilities on behalf of individual decision-makers. In practice, it is in- credibly hard to accurately perceive your network surroundings and estimate the char- acteristics of prospective collaborators (Krackhardt, 1990). Individual actors engaged in collaboration will vary in the degree to which strategic network behaviour is realistic to expect. Some will have enough capacities for information-processing allowing them to keep track of what is going on in networked environments and adjust their strategies 20 Public Policy and Administration 0(0) accordingly; other are likely to be overwhelmed. The provided conceptual model provides a first idea on the conditions under which the former or the latter is more likely to apply. Through this latter point, the practical relevance of the developed model also becomes clear: it hints at several strategic choices that agencies can make when engaging in networked collaboration. A primary concern here is that these agencies have to create the conditions through which their officials can cope with the uncertainties emerging from complex networked settings. On the one hand, this means having appropriate (organi- zational) structures in place that allow officials to adequately acquire and process relevant information and focus their attention to relevant aspects of the external environment. On the other, it means making strategic choices given the limited time and resources with which agencies typically operate, such as appropriately prioritizing networked settings and limiting staff turnover as to enable officials to develop network experience and policy expertise. In this way, agencies can help their officials to better cope with the uncertainties that they are likely to encounter when engaging in networked collaboration. Besides these points of relevance, several limitations of the present study should be noted nonetheless. For one, the analysis focuses on one specific policy sector, namely that of international finance. Although the choice for this prototypical complex research context is justified as a basis for theorizing, it is plausible that the gathered evidence potentially emphasizes contingencies particular to this context. Also, regarding the evidence status of the collected qualitative data, the interviews rely on the subjective impressions of respondents in which they provide an ex-post rationalized account of how they collaborate; that is, collaboration is not actually observed. Social desirability may be at work here, in which respondents concerned with impression management portray themselves as more capable and professional than they actually are. Both limitations require that the developed conceptual model is tested in different contexts and through more systematic types of empirical inquiry. In addition, several other theoretical lines of research should be pursued as well. Firstly, the behavioural implications of operating under conditions of uncertainty should be assessed. The developed conceptual model implies that there will often be circum- stances in which actors will be confronted with high levels of uncertainty, whilst making a particular decision is still required. A core insight from behavioural scholars is that such uncertainty typically leads to selective information-processing and the use heuristics on behalf of the decision-maker (see Simon, 1985; Vis, 2019). What these heuristics are for the context of partner selection is an important agenda for future research. In other words, we should open up the analytical possibility that, besides the strategic or rational modes of decision-making that most theoretical models on collaborative choices imply (e.g. Berardo and Scholz, 2010), such decisions are made in different ways, potentially re- flecting an unthinking reliance on past strategies, or even becoming spontaneous with little reference to potential losses or gains (Jones et al., 2006: 44). Secondly, as complex patterns of interaction continue to develop within the public sector, we should think more clearly about how aspects of organizational structure and design allow public officials to better operate in relational modes. By coordinating the activities of many individuals, each with partial and incomplete knowledge, organizations allow decision-makers to overcome many of their individual limitations (Jones, 2001: van der Heijden 21 131). Given the complexity of networked environments, structural design parameters are likely to play such a role for network behaviour as well. In that sense, we should more thoroughly assess kind of organizational structures and routines that create capacity for individual officials to ‘operate collaboratively’ (see McGuire and Silvia, 2010; Thompson, 1967). As complex administrative patterns continue to develop, such or- ganizational mechanisms will become increasingly important and should be explored further. Acknowledgements The author would like to thank Jelmer Schalk, Sandra Groeneveld, Kutsal Yesilkagit, and two anonymous reviewers for helpful comments and suggestions on earlier drafts of this paper. Declaration of conflicting interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article. Funding The author(s) received no financial support for the research, authorship, and/or publication of this article. ORCID iD Machiel van der Heijden  https://orcid.org/0000-0001-5723-3620 Notes 1. Within the literature, these collaborative institutions are labelled in different ways (policy forum, collaborative governance regime, policy venue, etc.) but fulfil a similar purpose: bringing to- gether actors and organizations to collectively tackle complex policy issues and providing them with a space for interaction (see Fisher & Leifeld 2015; Scott & Thomas 2017). 2. Among them are the Basel Committee on Banking Supervision (BCBS), the Financial Action Task Force (FATF), the International Association of Insurance Supervisors (IAIS), and the International Organization of Securities Commissions (IOSCO) at the global level, as well as more institutionalized platforms at the regional level, such as the various ESA’s in the context of the EU. References Agranoff R (2006) Inside collaborative networks: ten lessons for public managers. Public Ad- ministration Review 66(s1): 56–65. Ahdieh RB (2016) Coordination and conflict: the persistent relevance of networks in international financial regulation. Law and Contemporary Problems 78: 75–101. Aldrich H and Herker D (1977) Boundary spanning roles and organization structure. Academy of Management Review 2(2): 217–230. 22 Public Policy and Administration 0(0) Alexander D, Lewis JM and Considine M (2011) How politicians and bureaucrats networks: a comparison across governments. Public Administration 89(4): 1274–1292. Berardo R and Scholz JT (2010) Self-organizing policy networks: risk, partner selection, and cooperation in estuaries. American Journal of Political Science 54(3): 632–649. Bluedorn AC and Denhardt RB (1988) Time and organizations. Journal of Management 14(2): 299–320. Borgatti SP, Mehra A, Brass DJ, et al. (2009) Network analysis in the social sciences. Science 323(5916): 892–895. Brass DJ, Galaskiewicz J, Greve HR, et al. (2004) Taking stock of networks and organizations. Academy of Management Journal 47(6): 795–817. Calanni JC, Siddiki SN, Weible CM, et al. (2015) Explaining coordination in collaborative partnerships and clarifying the scope of the belief homophily hypothesis. Journal of Public Administration Research and Theory 25(3): 901–927. Child J (1972) Organizational structure, environment, and performance: the role of strategic choice. Sociology 6(1): 2–22. Choi Tand Robertson PJ (2014) Deliberation and decision in collaborative governance: a simulation of approaches to mitigate power imbalance. Journal of Public Administration Research and Theory 24(2): 495–518. Cristofoli D and Markovic J (2016) How to make public networks really work: a qualitative comparative analysis. Public Administration 94(1): 89–110. Day DVand Lord RG (1992) Expertise and problem categorization: the role of expert processing in organizational sense-making. Journal of Management Studies 29(1): 35–47. Emerson K, Nabatchi T and Baloch S (2012) An integrative framework for collaborative gover- nance. Journal of Public Administration Research and Theory 22(1): 1–29. Feiock RC (2013) The institutional collective action framework. Policy Studies Journal 41(3): 397–425. Fisher G and Aguinis H (2017) Using theory elaboration to make theoretical advancements. Organizational Research Methods 20(3): 438–464. Fischer M and Leifeld P (2015) Policy forums: why do they exist and what are they used for? Policy Sciences 48(3): 363–382. Fischer M and Sciarini P (2016) Drivers of collaboration in political decision making: a cross sector perspective. The Journal of Politics 78(1): 63–74. Frieden J (2016) The governance of international finance. Annual Review of Political Science 19: 33–48. Gulati R (1995) Does familiarity breed trust? the implications of repeated ties for contractual choice in alliances. Academy of Management Journal 38(1): 85–112. Gulati R and Gargiulo M (1999) Where do interorganizational networks come from? American Journal of Sociology 104(5): 1439–1493. Halevy N, Halali E and Zlatev JJ (2019) Brokerage and brokering: an integrative review and organizing framework for third party influence. Acta Medica Academica 13(1): 215–239. Hamilton M and Lubell M (2018) Collaborative governance of climate change adaptation across spatial and institutional scales. Policy Studies Journal 46(2): 222–247. van der Heijden 23 Henry AD, Lubell M and McCoy M (2011) Belief systems and social capital as drivers of policy network structure: the case of california regional planning. Journal of Public Administration Research and Theory 21(3): 419–444. Henry AD (2011) Ideology, power, and the structure of policy networks. Policy Studies Journal 39(3): 361–383. Hertz R and Leuffen D (2011) Too big to run? analyzing the impact of enlargement on the speed of EU decision-making. European Union Politics 12(2): 193–215. Hileman J and Bodin O (2019) Balancing costs and benefits of collaboration in an ecology of games. Policy Studies Journal 47(1): 138–158. Jones BD (2001) Politics and the Architecture of Choice: Bounded Rationality and Governance. Chicago, USA: University of Chicago Press. Jones BD, Boushey G and Workman S (2006) Behavioral rationality and the policy process: toward a new model of organizational information processing. In: Peters BG and Pierre J (eds), Handbook of Public Policy. London, UK: Sage Publications, pp. 49–74. Juenke EG (2005) Management tenure and network time: how experience affects bureaucratic dynamics. Journal of Public Administration Research and Theory 15: 113–131. Klijn EH, Edelenbos J and Steijn B (2010) Trust in governance networks: its impacts on outcomes. Administration and Society 42(2): 193–221. Koppenjan JFM and Klijn EH (2004) Managing Uncertainties in Networks: A Network Approach to Problem Solving and Decision Making. London, UK: Routledge. Krackhardt D (1990) Assessing the political landscape: structure, cognition, and power in orga- nizations. Administrative Science Quarterly 35: 342–369. Leech BL (2002) Asking questions: techniques for semistructured interviews. PS: Political Science and Politics 35(4): 665–668. Leifeld P and Schneider V (2012) Information exchange in policy networks. American Journal of Political Science 56(3): 731–744. Lubell M, Mewhirter JM, Berardo R, et al. (2017) Transaction costs and the perceived effectiveness of complex institutional systems. Public Administration Review 77(5): 668–680. Matti S and Sandstrom A (2011) The rationale determining advocacy coalitions: examining co- ordination networks and corresponding beliefs. Policy Studies Journal 39(3): 385–410. McEviley B, Perrone Vand Zaheer A (2003) Trust as an organizing principle. Organization Science 14(1): 91–103. McGuire M and Silvia C (2010) The effect of problem severity, managerial and organizational capacity, and agency structure on intergovernmental collaboration: evidence from local emergency management. Public Administration Review 70(2): 279–288. Meier KJ and O’Toole LJ (2001) Managerial strategies and behavior in networks: a model with evidence from US public education. Journal of Public Administration Research and Theory 11(3): 271–294. Meier KJ and O’Toole LJ (2010) Beware of managers not bearing gifts: how management capacity augments the impact of managerial networking. Public Administration 88(4): 1025–1044. Miller CE (1985) Group decision making under majority and unanimity decision rules. Social Psychology Quarterly 48(1): 51–61. Moynihan DP (2008) Learning under uncertainty: networks in crisis management. Public Ad- ministration Review 68(2): 350–365. 24 Public Policy and Administration 0(0) Nohrstedt D and Bodin O (2020) Collective action problem characteristics and partner uncertainty as drivers of social tie formation in collaborative networks. Policy Studies Journal 48(4): 1082–1108. Ostrom E (1998) A behavioral approach to the rational choice theory of collective action: pres- idential address, american political science association, 1997. American Political Science Review 92(1): 1–22. Park HH and Rethemeyer RK (2012) The politics of connections: assessing the determinants of social structure in policy networks. Journal of Public Administration Research and Theory 24(2): 349–379. Piore MJ (2006) Qualitative research: does it fit in economics? European Management Review 3(1): 17–23. Quintane E and Carnabuci G (2016) How do brokers broker? tertius gaudens, tertius iungens, and the temporality of structural holes. Organization Science 27(6): 1343–1360. Richards L (2020) Handling Qualitative Data: A Practical Guide. London, UK: Sage Publications. Six P, Goodwin N, Peck E, et al. (2006) Managing Networks of Twenty-First Century Organi- sations. New York, NY: Palgrave Macmillan. Schilke O and Cook KS (2013) A cross-level process theory of trust development in interorga- nizational relationships. Strategic Organization 11(3): 281–303. Scott TA and Thomas C (2015) Do Collaborative Groups Enhance Interorganizational Networks? Public Performance and Management Review 38(4): 654–683. Scott TA and Thomas C (2017) Winners and losers in the ecology of games: network position, connectivity, and the benefits of collaborative governance regimes. Journal of Public Ad- ministration Research and Theory 27(4): 647–660. Siciliano MD, Wang W and Medina A (2021) Mechanisms of network formation in the public sector: a systematic review of the literature. Perspectives on Public Management and Gov- ernance 4(1): 63–81. Simon HA (1985) Human nature in politics: the dialogue of psychology with political science. American Political Science Review 79(2): 293–304. Taylor RN (1975) Age and experience as determinants of managerial information processing and decision making performance. Academy of Management Journal 18(1): 74–81. Thompson JD (1967) Organizations in Action: Social Science Bases of Administrative Theory. New York, NY: McGraw-Hill. Vangen S and Huxham C (2003) Nurturing collaborative relations: building trust in interorgani- zational collaboration. The Journal of Applied Behavioral Science 39(1): 5–31. Vis B (2019) Heuristics and political elites’ judgment and decision-making. Political Studies Review 17(1): 41–52. Weber EP and Khademian AM (2008) Wicked problems, knowledge challenges, and collaborative capacity builders in network settings. Public Administration Review 68(2): 334–349. Weible CM (2005) Beliefs and perceived influence in a natural resource conflict: an advocacy coalition approach to policy networks. Political Research Quarterly 58(3): 461–475. Williams P (2002) The competent boundary spanner. Public Administration 80(1): 103–124. van der Heijden 25 Appendix A List of respondents Respondent Organization Function R1 DNB Div. Director R2 DNB Div. Director R3 DNB Div. Director R4 DNB Policy advisor R5 DNB Policy advisor R6 DNB Policy advisor R7 DNB Policy advisor R8 AFM Policy advisor R9 AFM Policy advisor R10 AFM Unit head R11 AFM Policy advisor R12 AFM Policy advisor R13 Min. Fin Policy advisor R14 Min. Fin Unit manager R15 Min. Fin Div. Director R16 Min. Fin Unit head DNB = National banking regulator AFM = National securities regulator Min. Fin = Ministry of Finance Appendix B Interview guide Note: The interview guide below was translated from Dutch. Also, in practice, the in- terview guide taken to the actual interview was specified further depending on the particular characteristics of the respondent (i.e. organization, function). Still, the sequence of questions as described below was always the blueprint, as to ensure the basic areas of interest were covered. 1. General Introduction: Explaining the goals of the research and what will be discussed (in general lines) during the interview. Procedures on anonymization and data protection (transcript/recording). 2. Walkthrough of activities and setup of organizational unit: a. Role and relation of the unit to the broader organization. b. Specific tasks of respondent within the unit/organization. 3. Walkthrough international platforms with which the unit is involved: a. Who typically participates on behalf of the unit? b. International activities respondent. Involved in what ways? 26 Public Policy and Administration 0(0) 4. Specific regulatory standards/dossiers currently relevant: a. Specific dossiers with which respondents is involved. b. Identify suitable examples to which to return. 5. Walkthrough of how international meetings are typically prepared: a. Structure of preparation, determining positions, finding coalition partners. b. Coordinating with other units. c. Respondents own role + examples. 6. Walkthrough of reporting on international meetings/activities: a. Reporting to supervisors? In what ways? b. Keeping a log? Communicating activities (to whom)? 7. Coordination with other (national) agencies/ministries: a. Ways of convening, contact. Speaking with one voice? Author Anonymous, b. Examples good or bad. Own involvement? 8. International negotiations/decision-making: ways of influence? a. Different policy instruments/strategies of influence. b. Probing for examples/own experiences. 9. Comparing different types of fora/platforms (European/Global): a. Nature of negotiations? Different strategies required? b. Probing for examples/own experiences. 10. Level of contact with foreign regulators/officials. Role of informal network? a. Meeting the same people? Different per country? b. Determining who is an appropriate partner? c. Role of informal network/social dynamics. 11. Wrap-up and debriefing: a. Additions? Returning to particular questions. b. Debriefing and next steps. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Public Policy and Administration SAGE

Problematizing partner selection: Collaborative choices and decision-making uncertainty

Public Policy and Administration , Volume OnlineFirst: 1 – Jan 1, 2022

Loading next page...
 
/lp/sage/problematizing-partner-selection-collaborative-choices-and-decision-VAMsfGNvGC
Publisher
SAGE
Copyright
© The Author(s) 2022
ISSN
0952-0767
eISSN
1749-4192
DOI
10.1177/09520767221088269
Publisher site
See Article on Publisher Site

Abstract

Although networked collaboration is often linked to positive outcomes, choosing suitable partners for collaboration can be difficult. Actors often only have limited information about the preferences, capabilities, and trustworthiness of prospective collaborators, meaning that considerable uncertainty may underlie collaborative choices. This article analyses the decision-making uncertainty associated with collaborative choices and assesses factors that potentially mitigate such uncertainty. Based on qualitative re- search, it presents a conceptual model that brings together and specifies mitigating factors at the network-, organizational, and individual-levels of analysis. The developed conceptual model provides a clearer theoretical understanding and presentation of the cross-level factors important to consider when studying collaborative behaviour. In addition, through its focus on uncertainty, it provides more consideration of the challenges that individual officials face when choosing collaborators in complex net- worked environments. Keywords Networks, policy networks, decision making, uncertainty, collaborative governance Introduction The collaborative contacts that (public) officials maintain in networked settings are valuable for their home departments and agencies: they provide them with access to relevant information and resources, help them learn about potential policy problems and Corresponding author: *Machiel van der Heijden, Utrecht University School of Governance (USG), Room -1.15b, Bijlhouwerstraat 6, Utrecht 3511 ZC, The Netherlands. Email: m.j.a.vanderheijden@uu.nl 2 Public Policy and Administration 0(0) solutions, and create the potential for forming alliances and coalitions (Alexander et al., 2011). However, choosing suitable partners for collaboration can be difficult (Agranoff, 2006; Feiock, 2013). Individual officials often only have limited information about the preferences, capabilities, and trustworthiness of potential partners (Leifeld and Schneider, 2012; Hamilton and Lubell, 2018). Moreover, they generally lack the time and cognitive capacities to accurately map out their network surroundings (Krackhardt, 1990). As a result, considerable uncertainty may underlie collaborative choices, particularly regarding the value of prospective collaborators. Such uncertainty is important to consider because it challenges an increasing number of public officials operating across organizational and jurisdictional boundaries. Although network behaviour is typically associated with beneficial outcomes (Cristofoli and Markovic, 2016; Meier and O’Toole, 2001), the complex nature of networked envi- ronments means that in practice individual officials will often struggle to successfully engage in such behaviour. Many networking individuals (and their respective organi- zations) will not reap the benefits of networked collaboration, being overwhelmed by the complex settings in which they find themselves. As such, decision-making uncertainty undermines the promise of networked collaboration and problematizes the strategic cues by which many theoretical models assume collaborative choices to occur (see Siciliano et al., 2021). However, the uncertainty underlying collaborative choices can also be mitigated. Policy network scholars emphasize the role that the institutional context of networked collaboration can play in this regard (Leifeld and Schneider, 2012; Scott and Thomas, 2015). In particular, the formal structure of working groups, task forces or committees through which a lot of collaboration occurs, presents policy actors with a clearly defined pool of potential partners with whom they have a chance to become acquainted (Fischer and Sciarini, 2016). This reduces decision-making uncertainty regarding the selection of appropriate network partners, as these institutionalized structures make information about potential partners more readily available (Ostrom, 1998). Beyond this general insight, however, several issues still require further consideration. Firstly, although scholars increasingly find that, on average, joint participation in institutionalized settings and bilateral collaboration are correlated (Leifeld and Schneider, 2012; Lubell et al., 2017), such settings vary considerably in terms of their charac- teristics (see Choi and Robertson, 2014; Fischer and Leifeld, 2015). Some settings have many participants, others only few. Some will meet frequently, others not so much. In that sense, these settings vary in terms of the extent to which they actually provide policy actors with easy access to information about prospective collaborators. As a result, the degree to which institutional settings reduce the decision-making uncertainty underlying collaborative choices can also be expected to vary, depending on their specific characteristics. What characteristics are most relevant in this regard, remains an open question. Secondly, when thinking about the actors actually making the collaborative choices, characteristics of (individual) network participants at both the organizational- and individual-level also seem important to consider. Individuals engaged in networking typically represent organizations that have different structures and resources. Think for van der Heijden 3 instance about the existence of formal boundary-spanning roles or the use of explicit internal coordination structures for network behaviour (see McGuire and Silvia, 2010; Six et al., 2006). Moreover, organizational members themselves have different characteristics, such as their experience with collaborative settings or their expertise on the policy issues discussed therein (Juenke, 2005; Meier and O’Toole, 2010). Both types of characteristics likely influence the time and effort officials require for searching and evaluating col- laborative partners, providing another source of variation in the decision-making un- certainty underlying collaborative choices. Although uncertainty is an important theme for network scholars (Hamilton and Lubell, 2018; Koppenjan and Klijn, 2004), there is little theoretical guidance on what factors affect and mitigate such uncertainty, particularly in relation to col- laborative choices. Therefore, this paper asks the following research question: What factors at network, organizational, and individual levels potentially affect the decision-making uncertainty underlying partner choices in collaborative settings? Answering such a question helps to better understand the (theoretical) relationship between collaborative settings and the informal networks that exist within them (see Emerson et al., 2012; Hileman and Bodin, 2019). Moreover, byalsoshiftingthe analytical focus to the individual officials that operate in these settings and the or- ganizational contexts from which they do so, it provides more consideration of the cross-level factors that may be influential for (the uncertainty of) collaborative choices (see Brass et al., 2004). The resulting conceptual model brings together and specifies factors (at three different levels of analysis) that potentially affect the uncertainty that public offi- cials experience when choosing collaborators. At the network-level, characteristics of the institutional settings through which networked collaboration often occurs are emphasized, noting the frequency of meetings, group size, and decision rules as most important for the uncertainty associated with collaborative choices. At the organizational-level, explicit boundary-spanning units, formal reporting require- ments, and limiting the number of different settings are argued to help officials to more easily make collaborative choices. And at the individual-level, the experience, expertise and available time of individual officials are considered to play an im- portant role in mitigating the uncertainty experienced when making collaborative choices. This conceptual model is developed based on qualitative evidence from a prototypical complex networked environment, that is, international finance policy and regulation. Within this policy field, a wide variety of transnational networks and standard-setting bodies have emerged, around which regulatory and ministry officials from various countries engage in complex patterns of networked interaction (Ahdieh, 2016; Frieden, 2016). This complexity provides a good basis for theorizing on the decision-making uncertainty underlying collaborative choices. Before doing so, however, the next section firstly discusses why decision-making uncertainty is such a (theoretically) relevant concept for better understanding collaborative choices. 4 Public Policy and Administration 0(0) Theoretical framing Collaborative Choices and Decision-Making Uncertainty This article’s theoretical interest is primarily in the uncertainty associated with collab- orative choices and how such uncertainty can be mitigated. In thinking about this question, two aspects of networked collaboration are important to separate (Hamilton and Lubell, 2018; Scott and Thomas, 2017). On the one hand, there often exists a formal (or institutionalized) part of such collaboration, in which representatives of organizations participate in collaborative institutions to solve collective-action problems . On the other, complex webs of informal collaborative linkages exist within and beyond these for- malized settings, in which actors voluntarily contact each other for further information exchange and coalition building. Although these two layers of collaboration influence each other (see Leifeld and Schneider, 2012), it is the latter form of collaboration for which the question of col- laborative choices and decision-making uncertainty is most relevant. This is the level at which (individual) officials make selective choices about with whom to collaborate more closely, typically under conditions of limited information about others’ preferences, capabilities and trustworthiness. Because individual officials do not always have the required time and effort available to gather this information, considerable decision- making uncertainty may underlie (informal) collaborative choices. Particularly the value of prospective collaborators and the predictability of their behaviours often remain unclear. Such uncertainty has been noted by network scholars in several ways. Generally, it refers to the expected behaviours, strategies and intentions of potential collaborative partners (Koppenjan and Klijn, 2004). More specifically, Hamilton and Lubell (2018: 230–231) define a form of political uncertainty to describe a lack of awareness of the policy beliefs and preferences of other actors. And Nohrstedt and Bodin (2020: 1092) define collaborative uncertainty as to indicate the risk that a potential collaborative tie has limited positive outcomes, implying a waste of time and resources. Both describe this uncertainty as a source of transaction costs, requiring actors to learn about the preferences of others and monitor their behaviours. Closely related to the above-provided definitions, this article defines uncertainty as the limited information that actors have about the preferences, capabilities and trustwor- thiness of potential collaborators, making it hard to establish their value as a collaborator. However, rather than describing such uncertainty in terms of transaction costs, the theoretical interest is more in how such uncertainty challenges actors to operate col- laboratively. In practice, high degrees of uncertainty mean that actors are often over- whelmed by the complex networked settings in which they operate. This problematizes network behaviour and existing theoretical frameworks used to explain collaborative choices. Many scholars, for instance, theorize that preference similarity is a main driver of collaborative patterns¸ meaning that actors will establish contacts with like-minded actors in terms of (policy) beliefs or values (Calanni et al., 2015; Henry et al., 2011). However, van der Heijden 5 such explanations assume a well-defined policy space in which preferences of potential partners and that of the choosing actors are known. In reality this is often not the case and, particularly for complex and newly emerging issues, actors might actually enter policy arenas to ‘discover’ their own and others’ preferences (Jones, 2001: 102). In the face of uncertainty, preference similarity (or dissimilarity for that matter) is likely a poor predictor of collaborative patterns. Or take scholars that use resource dependence theory to theorize about collaborative choices. The assumption here is that policy actors use collaborative ties to maximize their access to (political or technical) resources (Park and Rethemeyer, 2012). Actors will thus seek out partners that they perceive as influential or technically competent, due to their control over (or access to) critical resources such as information, technology, personnel or political clout (Henry, 2011; Matti and Sandstrom, 2011). However, assessing the in- fluence or technical competence of others can be difficult, particularly when confronted with a large group of potential collaborators that are relatively unfamiliar (Hamilton and Lubell, 2018). The strategic cues assumed to drive collaborative choices are thus problematic when related to the considerable uncertainty that may exist about the preferences, capabilities and trustworthiness of potential collaborators. Although goal-oriented behaviour is a reasonable assumption for network actors, the complexity of networked environments and the limited time and resources that actors typically have, will often interfere with such behaviour. As a result, public and policy officials will sometimes make poor collaborative choices with collaborators that prove unreliable or incompetent in the long run. In that sense, decision-making uncertainty potentially undermines the promise of networked collaboration. This makes it crucial to think about factors that affect such uncertainty and how it may vary between actors and across settings. Mitigating decision-making uncertainty In thinking about factors that influence the decision-making uncertainty associated with collaborative choices, the mechanisms by which such factors might do so must firstly be spelt out. The previous section established that uncertainty primarily grows out of the limited or inadequate information that actors have about the preferences, capabilities, or trustworthiness of potential collaborators. In addition, given the limited time and ca- pacities that individuals have to gather such information, collaborative choices are often made under conditions of uncertainty. Factors that mitigate the uncertainty of collabo- rative choices, are then those that (1) make it easier to access information on prospective collaborators or (2) provide actors with the time and resources to gather such information. To structure the discussion of potentially relevant mitigating factors, three separate levels of analysis are distinguished: the network-, the organizational- and the individual- level. The network-level refers to characteristics of the institutional settings through which networked collaboration often occurs. The organizational-level refers to characteristics of the organization that network participants typically represent when engaged in network behaviour and collaboration. And the individual-level focuses on characteristics of the actor that actually does the networking on behalf of a particular organization or agency. 6 Public Policy and Administration 0(0) The goal is then to identify factors at all three levels that mitigate decision-making uncertainty associated with collaborative choices. Starting with the network-level, policy network scholars have underlined the im- portance of the institutional settings and structures that often facilitate networked collaboration (Fischer and Sciarini, 2016; Leifeld and Schneider, 2012). Such settings are assumed to promote collaborative behaviour on behalf of participants and allow for the development of trust relationships between them (see Klijn et al., 2010; Vangen and Huxham, 2003). In thinking about collaborative choices, the joint participation in particular working groups, commissions or task forces seems particularly important. These institutionalized structures provide actors with a delineated pool of potentially relevant and suitable partners from which to choose, minimizing search costs. Moreover, the regular meetings of these collaborative institutions facilitate commu- nication between participating actors and gives them the opportunity to learn about each other’s preferences. The link with mitigating uncertainty about collaborative choices is thus clear: by providing easy access to information on potential collabo- rators, institutional settings make it easier for actors to assess the value of prospective collaborators. However, not all institutional settings will provide these benefits in reducing the uncertainty of collaborative choices. Although, on average, joint participation in insti- tutionalized settings and bilateral collaboration are correlated (Leifeld and Schneider, 2012; Lubell et al., 2017), it is important to note that such institutionalized settings may vary considerably. Think for instance about the number of participants that they have (Hertz and Leuffen, 2011), their perceived importance for participating actors (Fischer and Sciarini, 2016), their decision rules (Choi and Robertson, 2014), or the frequency with which scheduled meetings occur within them. These characteristics potentially affect the extent to which institutional settings actually provide actors with relevant information about potential collaborators. By extension, this also affects the degree to which such settings help mitigate decision-making uncertainty regarding further collaborative choices. However, there is currently little guidance on what specific characteristics of these settings are most important to consider. Besides these factors at the network-level, an important consideration is that net- working actors typically represent an organization and operate collaboratively from within a given organizational structure. The ability of actors to form and sustain col- laborative relationships is at least partially a function of the internal dynamics of that organization (McGuire and Silvia, 2010: 281). This makes it relevant to also consider factors at the organizational-level when asking the question of how the uncertainty associated with collaborative choices can be mitigated. For one, the structure of an organization allocates work roles and influences how organizational officials are related to the external environment (Child, 1972). By creating explicit boundary-spanning roles, for instance such structures can determine the time and resource that organizational members have for networking activities. Or internal coordination structures for collaborative efforts can give them access to information on prospective collaborators possessed by colleagues. van der Heijden 7 In addition, the number of different settings in which the organization is involved, likely affects the capacity of its members to operate collaboratively. Although partici- pating in many different collaborative networks provide clear benefits in terms of the access to diverse information (i.e., the strength of weak ties), meaningful collaboration arguably requires extensive networking on behalf of the individual officials representing the organization within these settings (see Hileman and Bodin, 2019). The more different settings (or policy fora) an organization engages in, the less time organiza- tional representatives have for searching and evaluating suitable collaborators within each particular setting. If the networking capacities of the organization are scattered across a large number of settings, many potential collaborators within these settings will remain relatively unfamiliar. This creates additional uncertainty around collaborative choices. And lastly, the individual-level also provides some factors that potentially affect the decision-making uncertainty underlying collaborative choices. Most importantly, actors making collaborative choices likely differ in terms of their experience with networked collaboration and their expertise regarding the (policy) issues discussed therein (Juenke, 2005; Meier and O’Toole, 2010). Regarding expertise, officials with technical knowledge on the policy issues in which they are involved can better assess the capabilities of potential collaborators (Williams, 2002). Moreover, for highly complex and newly emerging policy issues, actors with technical expertise can more easily evaluate their own policy positions, as well as those of others. This gives them more capacity to perform the decision-making task of partner selection, particularly in terms of information-processing and search behaviour (Day and Lord, 1992; Taylor, 1975). For (networking) experience, an important consideration is that experienced officials will have built up more elaborate informal networks through past interactions. This creates a number of benefits for future collaborative choices. For one, experienced actors typically have better mental representations of the policy spaces of the networks in which they operate, as well as a higher sensitivity toward the constellation of other actors and their interests (see Halevy et al., 2019). In addition, network scholars typically con- ceptualize the existing informal network as an information repository to reduce uncer- tainty about the trustworthiness of potential partners and learn about opportunities for new ties (see Gulati and Gargiulo, 1999; Henry et al., 2011). Having a more elaborate informal network (i.e. social capital) allows one to make better use of the information signals encoded in existing network structures and provides more opportunities for making contact, for instance through referrals or introductions via mutual friends (Quintane and Carnabuci, 2016). Overall, mitigating factors at three levels of analysis are thus relevant to consider when studying decision-making on collaborative choices and its underlying uncertainty. The mechanism by which these factors can be expected to mitigate decision-making un- certainty is by providing individual officials with easier access to information on the preferences, capabilities, and trustworthiness of potential collaborative partners. Al- though some specific factors are hinted at in the above-provided discussion, many still require further elaboration. The empirical analysis identifies what concrete factors are important to consider and specifies how they matter for the uncertainty associated with 8 Public Policy and Administration 0(0) collaborative choices. Firstly, however, the data collection and analysis procedures of this study are described. Data collection procedures The main goal of this study is theory elaboration, in which empirical analysis is used to further specify pre-existing conceptual ideas (Fisher and Aguinas, 2017). In particular, the focus is on construct splitting, that is, splitting existing theoretical constructs into specific dimensions based on observed empirical realities (ibid.,: 446). What this means for this article is that the section above has theoretically developed the idea that decision-making uncertainty about collaborative choices can be reduced by factors at different levels of analysis, and the empirical analysis below is used to specify what type of factors are likely to play a role in this regard. Qualitative research is particularly suitable for these purposes, as it enables attention to be given to particular circumstances, whilst its open-ended nature is sufficiently flexible for new insights or themes to emerge (see Piore, 2006). Case Selection and scope conditions To focus the analysis, this article looks at transnational collaboration in the field of fi- nancial regulation and zooms in on the way in which public officials of three Dutch organizations are involved in such collaboration (one ministerial department and two regulatory agencies). This research context is appropriate for the purposes of this research as it provides variation at different levels of analysis, which can be used as a basis for theorizing. For one, national agencies within this sector typically operate in a wide variety of transnational policy settings, both in terms of their level of institutionalization and the scale of collaboration (regional vs. global) (Ahdieh, 2016) . This means that there is a lot of variation in terms of the networked settings in which interviewed respondents operate, allowing us to collect and compare data on the importance of the different characteristics of these settings. In addition, the Dutch context allows for comparison between three different orga- nizations involved in transnational regulation and policy-making. Given the twin peaks model of supervision that exists within the Netherlands, responsibilities for financial sector regulation are split between two separate agencies, with one regulatory agency responsible for securities regulation, and another for banking, insurance, and pensions regulation. Moreover, the ministerial department also does transnational policy work, primarily in the context of the EU but also in global platforms such as the FSB. This allows for comparing officials engaged in (transnational) network activities within three different organizational contexts, whilst the nature of the sector in which they engage is largely similar. Data collection and analysis The presented analysis is based on qualitative research in the form of semi-structured interviews. 16 face-to-face interviews were conducted with Dutch officials involved in van der Heijden 9 international financial policy and regulation (see Appendix A). These respondents oc- cupied different positions in the regulatory policy departments at either the Dutch Ministry of Finance (n = 4), DNB (Dutch Central Bank) (n = 7), or AFM (Securities Regulator) (n = 5). The one common denominator of these respondents is that they were all heavily involved with international networks, at either (or both) the European or global level. The respondents were identified through a combination of snowball and purposeful sampling, in which heads of departments were firstly approached for an interview. These heads of departments were asked to identify individuals in the organization involved in the different networked settings in which the agency or ministry engaged. Maximum variation in terms of networked settings was thus an important criterion for identifying the different respondents. Although this type of sampling creates a potential bias, it does help to recruit respondents that are otherwise difficult to reach. Moreover, interviewing several respondents from the same organization and asking them about similar processes helps to verify the provided accounts and gives a more balanced representation of the phenomenon of interest. Drawing on a topic list, these respondents were interviewed (45 min on average) by the author in a semi-structured fashion. Topics discussed in the interviews were – inter alia – the kinds of international platforms in which they participate, how they prepare for international meetings, who their contacts are, what channels they use to influence the international regulatory process, and how they went about selecting partners and for what reasons (see Appendix B). Grand tour questions were asked about what a typical preparation for international meetings looks like (Leech, 2002), as well as example questions about issues discussed in international meetings. Moreover, explicit probing was carried out for the way in which the respondents identify potential collaborators and what the difficulties are in this regard. The recordings were transcribed and analysed through a process of coding. Three steps were involved in moving from the raw interview transcripts to further theorizing on decision-making uncertainty and partner selection. Firstly, topic coding was used to identify passages relevant to decision-making uncertainty regarding partner selection (Richards, 2020: 110). These passages typically hinted at the time and effort needed for making contact and the way respondents engaged in network behaviour more generally. In addition, passages describing network-, organizational- and individual-level charac- teristics were identified and categorized according to the level at which they belonged. This gave a first selection and initial ordering of interview passages relevant to the theoretical question of interest. As a second step, this collection of passages was then reviewed as to develop analytical categories (ibid.: 110–112). This analytical coding was used to interpret and further select passages categorized under different levels of characteristics, particularly assessing which of these passages were relevant for decision-making uncertainty underlying collaborative choices. At this stage, more specific codes were identified and attached to the passages as to signify specific characteristics (see Fisher and Aguinas 2017), for example ‘group size’ at the network-level or ‘experience’ at the individual level. 10 Public Policy and Administration 0(0) Lastly, the identified categories were related to the existing literature on collaboration and decision-making as to theoretically interpret their meaning, and to formulate specific propositions on how the identified characteristics are important for decision-making uncertainty underlying partner choices. For this last step, the different characteristics were primarily evaluated in terms of their likely effect on the time and effort needed to gather information about potential partners and identify their preferences, capabilities and re- liability (i.e. the core mechanism by which uncertainty is mitigated, as argued in the theoretical framework). Analysis As a precondition to identify which characteristics are relevant for the decision-making uncertainty underlying collaborative choices, a first step is to establish that such un- certainty is actually an issue for the respondents of this study. This is what the next section does. Then separate sections present empirical material pointing to the importance of particular characteristics at three different levels of analysis (i.e. the network-, organi- zational- and individual-levels). This empirical material is interpreted theoretically, on the basis of which several propositions are developed. The problem of partner selection Within the policy field of international finance there exists a large variety of international organizations and standard-setting bodies that can be characterized as (transnational) networks (Ahdieh, 2016: 76). Within the institutionalized settings of these networks, interaction between national officials primarily occurs through the various working groups, commissions, or task forces, which carry out most of its operational work. Being connected in such a way simply means receiving the same group mails or periodically attending the same meetings. Besides these more structured interactions, however, regulators also meet each other informally on a more ad hoc basis. This is the level at which officials voluntarily interact with others to exchange (political or technical) in- formation and engage in coalition building to influence decision-making. For these latter types of networked interactions, the question of collaborative choices and decision- making uncertainty is most relevant. In making these collaborative choices, established drivers of collaboration such as preference similarity and perceived influence play an important role for respondents. They frequently talk about their ‘natural partners’, whom they know think about certain issues the same way. As one regulator stated, ‘if there’s an important issue, for us, coming up, for which we know there will be a lot of difference of opinion, we try to mail, call with similarly-minded countries, to see how we can best go into such a meeting’ [R5]. Similarly, for perceived influence, many respondents expressed a preference for working with the ‘big countries’, that are influential and resourceful. In discussing collaborative partners one respondent noted: ‘you know, these nine countries, why did we choose them.. [because] they are all big, semi-big [countries]. I mean, the small ones... they just don’t have the capacity’ [R6]. van der Heijden 11 However, identifying the preferences and capabilities of actors is not as straightfor- ward as it sounds. Given the wide variety of different topics that are dealt with in financial regulation, the cards are constantly reshuffled for every new topic or issue that national officials have to deal with. Although respondents typically talk about ‘natural partners’ with whom they share similar interests, they are quick to emphasize that ‘your natural partners differ per topic’. A senior regulator noted that ‘for me, it is not really the case that you have a fixed group [...] you really have to search your coalition depending on the topic’ [R3]. Moreover, given the fast developments and innovations in international finance, regulators are also often confronted with topics on which they have not yet formed a position. As one regulatory official noted in preparing an international meeting with an extensive agenda, ‘it is our role, given these 25 points, to ask, how important is this [for us], what is the constellation of power, […] do we have a chance?’ [R5]. As a consequence, the search for suitable partners can be rather complex and uncertain. For each new issue, involved officials will have to find out what the policy positions of other network participants are. Moreover, you have to know which potential partners actually want to collaborate and whether they are capable enough to reciprocate your own efforts. This information about other actors is not always clear, nor are all other network participants equally approachable. As one senior regulator remarked, ‘if you’re in a project group, then you have really active countries, that are just involved. Some countries are not very active in the project group and you have to reach them at a different level’ [R4]. In terms of approaching collaborators, another agency official noted: ‘most big countries have a separate desk, a [network-X] desk that you can contact... I also once sent an email to [a smaller agency], to ask who did the coordination [of network-X].. It turned out to be director himself. So that complicates things’ [R5]. Besides the difficulties in identifying or reaching particular partners, the trustwor- thiness of others is sometimes also difficult to assess. In the interviews, several re- spondents complained about collaborators who are ‘indirect’ in their communications or even ‘unreliable’. In discussing potential partners for collaboration, one senior regulator noted how, ‘with the guys from [country X] I just communicate better, with [regulators from country y] it always stays with niceties, but... what do you really think’ [R3]. Similarly, a ministry official remarked after striking a deal with a foreign counterpart: ‘with them, you’re never completely certain, whether you’re being played with, let me put it that way’ [R16]. In that sense, potential risks for defection remain an issue when choosing potential collaborators. Overall, the interviewed respondents seem to have difficulty estimating the capa- bilities, preferences, and trustworthiness of potential collaborators and devote time and effort to acquire this information. Regardless, they often remain unclear about the value of prospective and current collaborators. This makes decision-making uncertainty, as de- fined in the theoretical section, a relevant consideration for their collaborative choices. The next section looks at the factors that potentially affect this degree of uncertainty. 12 Public Policy and Administration 0(0) Mitigating uncertainty: The network-level The theoretical section noted how institutional settings can help officials to better make collaborative choices (Hamilton and Lubell, 2018; Leifeld and Schneider, 2012). The interviews provide extensive anecdotal evidence for this consideration. As one senior official noted in discussing the way in which he contacts collaborators, ‘[…],usually you speak with the people from your committee, that is your first point of reference. You know them, you experience them in meetings, you sometimes have had dinners with them. Those are the people with whom you have had the most contact.. so you’ll speak with them first’ [R3]. The working groups, committees and task forces in which domestic officials participate thus help them to delineate their choice set of potential partners for further collaboration. Moreover, through the interaction occurring within these groups, they can become acquainted with others, providing a low-cost strategy to identify and select appropriate partners. Notably, institutionalized settings help respondents to identify potential collaborators for both political and technical information exchange. For political-strategic information exchange, one regulatory official described how in looking for potential partners, ‘you look for a coalition with people of whom you know they have similar ideas, and there is only one way to find that out, and that is to make sure you’re in those [working] groups’ [R6]. Similarly, for identifying partners with whom to exchange technical information, a securities regulator noted how in the transnational sessions in which he participates, ‘it becomes more clear what issues are prevalent for different countries. After such a session you can determine, wait.. I have to contact colleagues in Spain or colleagues in Brazil, because they also have problems with their mortgage markets, or whatever’ [R11]. The institutional settings of the networks in which respondents participate thus provide an important context to collaborative choices: informal network patterns often grow out of formalized structures. However, the interviews also demonstrated that these institutional settings vary considerably in terms of their characteristics. Firstly, respondents note how some groups in which they participate only meet two or three times a year, whilst others do so on a more frequent basis. This frequency of meetings is obviously important for how often actors see each other face-to-face and have a chance to become acquainted. In reflecting on his participation in both European and global networks, one senior regulator noted that ‘the frequency and intensity in Europe is much higher. So you meet more often and more intensively within Europe than [....] globally. This means you know each other better, are more familiar with their systems. You know more’. [R2]. Similarly, another regulator remarked about a working group that meets relatively frequently: ‘the advantage of those working groups is, you meet each other multiple times a year. [..] So, if you’re a bit pro- active, within half a year [..] you’ve spoken with everyone in one way or another’ [R7]. Overall, participating in the same working group or policy committee gives officials a chance to meet and become acquainted. This makes it relatively easy for them to acquire information on the preferences and capabilities of potential collaborators, hence reducing the uncertainty about collaborative choices (see also Scott and Thomas, 2017). However, respondents also participate in groups that meet relatively infrequently. This means less van der Heijden 13 opportunity for face-to-face interaction and a smaller chance for these benefits to accrue. In these instances, one can expect that the additional information about the preferences, capabilities and trustworthiness of potential partners that such institutional settings provide is likely to be restricted. The resulting unfamiliarity between actors will make it harder for actors to navigate the networked settings in which they operate and collab- orative choices will remain uncertain (see also Lubell et al., 2017). Given the variation across institutionalized networks in terms of the frequency with which meetings occur, a first expectation is that Proposition 1: The higher the frequency of meetings within institutionalized networks, the lower the decision-making uncertainty underlying collaborative choices. Secondly, respondents also reported a large variety in terms of the number of par- ticipants of the working groups, commissions, boards and task forces in which they participated. Whilst some talked about groups in which only 7 other people participated, others mentioned numbers up to 30 or more. An obvious consequence from this variation in group size is that it determines the time and effort you have to devote to getting to know the others within your group. This leads to selective behaviour on behalf of domestic officials about whom to contact as transnational collaborators. As one regulator men- tioned about identifying partners with similar preferences in a relatively big group, ‘it’s not that I’m going to call all 28, definitely not... You call, with whom you expect you’ll have the biggest chance that it will work’ [R8]. Similarly, a ministry official noted that ‘after a while, you recognize the most important faces. But of course it’s a big group, 28 countries’ [R16]. These reported challenges of networked interaction when groups are larger, make variation in the size of groups that come together within collaborative settings theo- retically interesting to consider. As noted by Hamilton and Lubell (2018), joint partic- ipation in working groups or commissions does not ensure that participants actually interact. Particularly when these groups have a large number of participants, the chances of interaction between two particular members are smaller (see also Fischer and Leifeld, 2015). Network scholars have extensively reported on how with each additional network participant the number of potential connections increases exponentially (see Borgatti et al., 2009), making these institutionalized settings more difficult to navigate. These considerations are important for the uncertainty underlying collaborative choices, as it means that more information is required on a larger number of co-participants. Theo- retically, one can then reason that given the restricted time and effort that officials can put into acquiring such information, a larger group size means that choices about collabo- rative partners will inevitably be characterized by higher degrees of uncertainty. Proposition 2: The larger the size of the groups within institutionalized networks, the higher the decision-making uncertainty underlying collaborative choices. And thirdly, respondents report how the decision rules of the institutionalized settings in which they participate are important for the way in which they make collaborative 14 Public Policy and Administration 0(0) choices. In particular, voting procedures determine the degree to which actors can be selective in collaborative choices, or also need information on all other co-participants. As one ministry official noted in reflecting about his partner selection strategy, ‘some topics go by unanimity, then you get different kinds of negotiations…. Other negotiations go by QMV [Qualified Majority Voting], then you see much stronger, I mean, everyone can count how many votes a country represents, and then you can count […] do I have a blocking minority or not ‘[R15]. Another regulatory official described an instance in which the members of his working group had to reach consensus on a set of recom- mendations, in which ‘it was an intensive process to get everyone on the same line. So it costs quite a lot of time and lot of diplomacy skills and negations to eventually get a version that we could back but also the others[...] It was intensive in the sense that we had different conference calls, write different versions, constantly adjusts things, make a new versions, ask reactions, process reactions, or not, [....] So, there was a whole process beforehand’ [R12]. Based on these descriptions, such decision rules seem to play an important role in shaping and constraining the deliberation and decision processes within institutionalized settings (see also Choi and Robertson, 2014; Fischer and Leifeld, 2015). In this way, they also affect how network participants have to choose collaborators for strategic infor- mation exchange and coalition building. The crucial divide here is between the use of unanimity or majority rules to achieve decision-making. To some extent, majority rules simplify partner selection because public officials can focus their attention on a limited number of actors, whilst others can be ignored. With consensual decision-making, however, also less familiar actors have to be involved. Moreover, actors with more extreme positions have to be facilitated (Miller, 1985). This arguably increases the uncertainties underlying collaborative choices, as more information is needed on a larger number of actors. Moreover, additional time and effort are required in gathering such information from actors that are relatively unfamiliar. Proposition 3: The larger the majority needed for making decisions within institu- tionalized networks, the higher the decision-making uncertainty underlying col- laborative choices. Mitigating uncertainty: The organizational-level Organizational settings also matter for the way in which respondents operate collabo- ratively. Firstly, the way in which the studied organizations internally structure and coordinate external network activities mattered a great deal for the respondents. In particular, the existence of specific units or roles focussing on the strategic aspects and coordination of network activities stands out. Both regulatory agencies had special teams of coordinators overseeing transnational interactions that served as ‘the internal and external point of contact for all matters related to [network X]’. By assisting or advising experts from other units when engaging in transnational activities, they help other officials that arguably have less time to do so. As one regulatory official fulfilling such a co- ordinative role noted, ‘in these [preparatory] meetings, possibly we walk through the van der Heijden 15 agenda of the committee in which someone partakes. Just to see, what’s there, what are the issues... Also, to get a sense of, how is it going, what is the structure, what is the vibe, who are the natural allies’ [R8]. Overall, such structures seem to allow for the pooling of expertise and network capacities within the organization, helping individual officials to more easily gather information on potential collaborators. Theoretically, the above-described coordination structures point to the importance of functionally specialized units and explicit boundary-spanning roles overseeing and co- ordinating network activities. Within organizational theory, such boundary-spanning units are traditionally seen as an important way for organizations to cope with envi- ronmental uncertainty (Aldrich and Herker, 1977; Thompson, 1967). Officials within such roles or units have the time and resources to strategize on networked environments, using this information to identify appropriate partners and advise others within the or- ganization. Moreover, the team-based structures through which international meetings are prepared, lets officials pool their attentional capacities and expertise (6 et al., 2006). In that sense, internal structures help to mitigate decision-making uncertainty regarding partner selection, primarily through a more efficient way of processing information on potential partners and their preferences, interests, and capabilities. Given that organizations can be expected to vary in terms of the degree to which they have such internal structures and formal boundary-spanning roles in place, a reasonable expectation is that Proposition 4: The more explicit boundary-spanning roles that organizations have, the lower the decision-making uncertainty underlying collaborative choices. Secondly, the number of different networked settings in which domestic agencies or ministries participate is important to consider. Respondents of all three organizations carefully reflect on how their organizations have to prioritize in the working groups, task forces, and committees in which they participate, pointing to the ‘scarce resources’ or ‘fte- resctrictions’ [i.e. full time equivalent] of their organizations or units. Different levels of engagement are considered for each working group, task force or committee in which organizational members can potentially participate, ranging from not participating at all or only being an ‘email-member’, to actively ‘writing’ or even trying to become ‘chair’.An underlying rationale for this prioritization is given by a senior banking regulator, who notes that ‘if there are points that you think are important, experience teaches that, if you want to be effective to bring in these points, it takes a lot of time and energy. It’s more than just going to the meetings and making your argument. If you want to be effective, you have to lobby. So, then there’s a lot more to it’ [R3]. In other words, to be influential in networked settings, considerable time and resources are required. However, given that the networking capacities of agencies are inevitably restricted, they have to be carefully distributed across the many different settings and venues in which the agency can potentially participate. The above-provided descriptions hint at the strategic choices that agencies have to make when faced with complex networked environments: participate in many different settings or focus on only a few. An implication of the former choice is that one is confronted with many other potential collaborators, whilst also having less time to spend on establishing and maintaining 16 Public Policy and Administration 0(0) collaborative ties within each separate setting (Hileman and Bodin, 2019). This makes venue shopping and issue prioritization an important consideration for networked col- laboration (see Weible, 2005), as it allows agencies to focus on a restricted number of settings and potential collaborators. This also means less information needs to be acquired on the preferences, capabilities and trustworthiness of prospective collaborators. A re- sulting expectation is that Proposition 5: The larger the number of networked settings in which organizations participate, the higher the decision-making uncertainty underlying collaborative choices. Thirdly, levels of formalization and the standard operating procedures that exist within organizations are important for the way in which domestic officials engage in transna- tional collaboration. Respondents point to the formalized process by which they report on international meetings, and how this has a function in the preparation of future meetings. As one regulatory official described, ‘of all meetings a report is made, and that is shared with the relevant colleagues, and that is usually then, on the one hand the report, and on the other hand the contribution of the experts.... together that is the input for the next rounds. So in that sense, it is an iterative process’ [R7]. Similarly, a ministry official noted how, ‘for meetings that officials go to themselves, they have to write reports. And those are saved and that’s basically the archive. And it is always convenient to know, what has been said, and sometimes you need them again, because you cannot remember it all’ [R16]. Importantly. this stored information can be accessed by organizational members when preparing international meetings, even if they were not involved in previous interactions. In that sense, formal reporting requirements regarding network activities are important to take into account when considering collaborative choices. In particular, the extensive backlogs and reports of network meetings and activities create organizational memory through which (strategic) information regarding previous interactions is stored (Schilke and Cook, 2013). This information can be accessed by organizational members when preparing network meetings, even if they were not involved in previous interactions. These standard operating procedures in reporting about network activities thus encode experiences that help guide organizational behaviour (Moynihan, 2008). In particular, by providing information on the preferences and actions of others in previous meetings, such formalized reports mitigate decision-making uncertainty regarding partner selection when preparing new meetings and thinking about collaborative choices. Proposition 6: The more formalized reporting procedures organizations have about network activities, the lower the decision-making uncertainty underlying collab- orative choices. Mitigating uncertainty: The individual level At the individual-level, a first point to consider is the experience that officials already have with the network settings in which they collaborate. Experienced respondents talk about van der Heijden 17 ‘being used to conducting business internationally’, whilst others talk about having to learn how ‘everything works internationally’. Moreover, respondents note how the contacts they know from one setting can also be important for other settings, particularly when participating in a new working group for the first time. As one regulatory official remarked about establishing contact with foreign counterparts, ‘If you’ve been doing this for a while, you run into international colleagues in these working groups... who you know from other working groups. So that helps’ [R7]. Moreover, such experience is crucial for lowering barriers to cooperate, as it allows you to rely more on the informal networks you have built in the past. As one official noted, ‘if you a have case that crosses borders, and you have to collaborate, it just helps if you already know someone in- formally, once had a dinner, or already collaborated with someone during a meeting’ [R12]. A general observation based on these passages is that experience with collaborative settings seemingly allows officials to resort to the informal ties they have built up in the past. As noted in the theoretical section, such prior ties are an important way in which uncertainty about future interactions can be mitigated (see Gulati, 1995; McEviley et al., 2003). Actors acquire information from their past interactions and resort to this infor- mation when considering potential collaborators. The social capital that experienced actors accumulate over time, benefits them in making new collaborative choices (Henry et al. 2011). In addition, through experience actors seemingly develop networking skills and gain more oversight of the complex institutional environments in which they find themselves (see Meier and O’Toole, 2010; Williams, 2002). Arguably, this makes it easier to acquire and interpret information about prospective collaborative partners. Proposition 7: The more experience actors have with network activities, the lower the decision-making uncertainty underlying collaborative choices. Secondly, the expertise of officials is noteworthy, particularly given the highly technical policy discussions in which interviewed respondents often engaged. As one regulator noted, “all of a sudden you are in these policy discussions and it is really hard to see where you are exactly and what you think [about an issue.]... Over time it gets better” [R7]. Similarly, a ministry official mentioned that “this area is eventually just very complex, and within our department, my division is also the most technical one... and there is a lot of issues... If you’re new in such an area, you really have to invest in it a lot at the beginning” [R14]. Lacking such technical expertise, means that officials have to work harder to find what the relevant issues are and what the main policy discussion are about. For decisions regarding partner selection, this is an important consideration as it likely makes networked environments much harder to navigate. Although these quotes point to the important mitigating effect that experience is likely to have, it seems clear that for highly specialized policy discussions it is potentially harder to evaluate your own (policy) position, as well as that of others. This inevitably com- plicates networked collaboration and the specialized expertise of officials engaging in such collaboration is thus important to consider. Knowledge is the “currency of col- laboration” (Emerson et al., 2012: 16) and actors with specialized expertise will more 18 Public Policy and Administration 0(0) easily integrate and evaluate the knowledge held by different actors (Weber and Khademian, 2008). Relating it to the uncertainty associated with collaborative choices, specialized expertise allows actors to more easily process information regarding the capabilities and policy positions of potential partners. Proposition 8: The more specialized expertise actors have when engaging in network activities, the lower the decision-making uncertainty underlying collaborative choices. Lastly, the interviews pointed to the consideration that respondents vary on the available time they have for transnational network activities in general and preparation of international meetings in particular. Importantly, for many respondents, participating in (transnational) networks is a duty they have besides the other core tasks or functions for which they are responsible. whilst preparing international meetings can take a lot of time. As one ministry official remarked, ‘[..] I already lose one day [a week] just calling the different counterparts. And then you also need another day just reading the underlying documents’ [R13]. Such intensive preparation is problematic for officials that do not have such time available. As one regulatory official noted ‘On average, I try to devote half a day a week, to this work. It would be good to devote much more work into this’ [R5]. Another regulatory official noted that preparing international meetings simply requires ‘a lot of talking on the phone, a lot of conference calls. Ideally you would also meet each other face-to-face, but that isn’t always doable. It just takes too much time’ [R12]. As these quotes point out, the time that officials actually have available for network activities is seemingly important for the way they engage with potential collaborators. Although such available time is partly dependent on the existence of explicit boundary- spanning roles (proposition 4) and the number of different setting in which the orga- nization participates (proposition 5), it is also a noteworthy separate factor to consider given its direct influence on individual-level decision-making. Scholars studying the effects of time availability on decision-making primarily note its importance in terms of search behaviour and the number of alternatives considered (Bluedorn and Denhardt, 1988). More available time means more opportunity to acquire information about other actors in the networked environment. Having the time to get to know the network likely leads to less uncertainty when deciding about with whom to collaborate more closely (Juenke, 2005). Proposition 9: The more time actors have available for network activities, the lower the decision-making uncertainty underlying collaborative choices. Discussion and conclusion Overall, factors at three different levels of analysis may influence the degree to which the decision-making on collaborative contacts is characterized by high or low degrees of uncertainty. Importantly, all of the identified factors do so through the mechanism of providing easier access to information on the preferences capabilities, and trustworthiness van der Heijden 19 Figure 1. Conceptual model: collaborative choices and decision-making uncertainty. of potential partners. This unifying mechanism allows us to specify an overall conceptual model in which the formulated propositions are brought together (see Figure 1). This conceptual model provides a clearer theoretical understanding and presentation of the cross-level factors important to consider when studying collaborative behaviour. The institutional setting at the network-level has an important role as it creates familiarity between actors and helps them minimize search costs (see Hamilton and Lubell, 2018). However, as this article demonstrates, individual- and organizational-level factors should also be considered. At least in part, we can assume the skills and abilities of individual officials and the administrative capacity of an organization to contribute to (the capacity for) collaborative behaviour as well (McGuire and Silvia, 2010). Integrating these factors at different levels of analysis into a single model, helps us think more clearly about the decision-making problems that confront an increasingly large number of public officials that have come to operate outside the boundaries of their organizations, whilst also providing ideas on how such problems may be mitigated. Theoretically, the notion of decision-making uncertainty puts pressure on existing frameworks typically used to hypothesize on partner selection within collaborative networks. Given the complexity of networked environments, models emphasizing ra- tional decision-making and strategic cues potentially require too extensive information- processing capabilities on behalf of individual decision-makers. In practice, it is in- credibly hard to accurately perceive your network surroundings and estimate the char- acteristics of prospective collaborators (Krackhardt, 1990). Individual actors engaged in collaboration will vary in the degree to which strategic network behaviour is realistic to expect. Some will have enough capacities for information-processing allowing them to keep track of what is going on in networked environments and adjust their strategies 20 Public Policy and Administration 0(0) accordingly; other are likely to be overwhelmed. The provided conceptual model provides a first idea on the conditions under which the former or the latter is more likely to apply. Through this latter point, the practical relevance of the developed model also becomes clear: it hints at several strategic choices that agencies can make when engaging in networked collaboration. A primary concern here is that these agencies have to create the conditions through which their officials can cope with the uncertainties emerging from complex networked settings. On the one hand, this means having appropriate (organi- zational) structures in place that allow officials to adequately acquire and process relevant information and focus their attention to relevant aspects of the external environment. On the other, it means making strategic choices given the limited time and resources with which agencies typically operate, such as appropriately prioritizing networked settings and limiting staff turnover as to enable officials to develop network experience and policy expertise. In this way, agencies can help their officials to better cope with the uncertainties that they are likely to encounter when engaging in networked collaboration. Besides these points of relevance, several limitations of the present study should be noted nonetheless. For one, the analysis focuses on one specific policy sector, namely that of international finance. Although the choice for this prototypical complex research context is justified as a basis for theorizing, it is plausible that the gathered evidence potentially emphasizes contingencies particular to this context. Also, regarding the evidence status of the collected qualitative data, the interviews rely on the subjective impressions of respondents in which they provide an ex-post rationalized account of how they collaborate; that is, collaboration is not actually observed. Social desirability may be at work here, in which respondents concerned with impression management portray themselves as more capable and professional than they actually are. Both limitations require that the developed conceptual model is tested in different contexts and through more systematic types of empirical inquiry. In addition, several other theoretical lines of research should be pursued as well. Firstly, the behavioural implications of operating under conditions of uncertainty should be assessed. The developed conceptual model implies that there will often be circum- stances in which actors will be confronted with high levels of uncertainty, whilst making a particular decision is still required. A core insight from behavioural scholars is that such uncertainty typically leads to selective information-processing and the use heuristics on behalf of the decision-maker (see Simon, 1985; Vis, 2019). What these heuristics are for the context of partner selection is an important agenda for future research. In other words, we should open up the analytical possibility that, besides the strategic or rational modes of decision-making that most theoretical models on collaborative choices imply (e.g. Berardo and Scholz, 2010), such decisions are made in different ways, potentially re- flecting an unthinking reliance on past strategies, or even becoming spontaneous with little reference to potential losses or gains (Jones et al., 2006: 44). Secondly, as complex patterns of interaction continue to develop within the public sector, we should think more clearly about how aspects of organizational structure and design allow public officials to better operate in relational modes. By coordinating the activities of many individuals, each with partial and incomplete knowledge, organizations allow decision-makers to overcome many of their individual limitations (Jones, 2001: van der Heijden 21 131). Given the complexity of networked environments, structural design parameters are likely to play such a role for network behaviour as well. In that sense, we should more thoroughly assess kind of organizational structures and routines that create capacity for individual officials to ‘operate collaboratively’ (see McGuire and Silvia, 2010; Thompson, 1967). As complex administrative patterns continue to develop, such or- ganizational mechanisms will become increasingly important and should be explored further. Acknowledgements The author would like to thank Jelmer Schalk, Sandra Groeneveld, Kutsal Yesilkagit, and two anonymous reviewers for helpful comments and suggestions on earlier drafts of this paper. Declaration of conflicting interests The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article. Funding The author(s) received no financial support for the research, authorship, and/or publication of this article. ORCID iD Machiel van der Heijden  https://orcid.org/0000-0001-5723-3620 Notes 1. Within the literature, these collaborative institutions are labelled in different ways (policy forum, collaborative governance regime, policy venue, etc.) but fulfil a similar purpose: bringing to- gether actors and organizations to collectively tackle complex policy issues and providing them with a space for interaction (see Fisher & Leifeld 2015; Scott & Thomas 2017). 2. Among them are the Basel Committee on Banking Supervision (BCBS), the Financial Action Task Force (FATF), the International Association of Insurance Supervisors (IAIS), and the International Organization of Securities Commissions (IOSCO) at the global level, as well as more institutionalized platforms at the regional level, such as the various ESA’s in the context of the EU. References Agranoff R (2006) Inside collaborative networks: ten lessons for public managers. Public Ad- ministration Review 66(s1): 56–65. Ahdieh RB (2016) Coordination and conflict: the persistent relevance of networks in international financial regulation. Law and Contemporary Problems 78: 75–101. Aldrich H and Herker D (1977) Boundary spanning roles and organization structure. Academy of Management Review 2(2): 217–230. 22 Public Policy and Administration 0(0) Alexander D, Lewis JM and Considine M (2011) How politicians and bureaucrats networks: a comparison across governments. Public Administration 89(4): 1274–1292. Berardo R and Scholz JT (2010) Self-organizing policy networks: risk, partner selection, and cooperation in estuaries. American Journal of Political Science 54(3): 632–649. Bluedorn AC and Denhardt RB (1988) Time and organizations. Journal of Management 14(2): 299–320. Borgatti SP, Mehra A, Brass DJ, et al. (2009) Network analysis in the social sciences. Science 323(5916): 892–895. Brass DJ, Galaskiewicz J, Greve HR, et al. (2004) Taking stock of networks and organizations. Academy of Management Journal 47(6): 795–817. Calanni JC, Siddiki SN, Weible CM, et al. (2015) Explaining coordination in collaborative partnerships and clarifying the scope of the belief homophily hypothesis. Journal of Public Administration Research and Theory 25(3): 901–927. Child J (1972) Organizational structure, environment, and performance: the role of strategic choice. Sociology 6(1): 2–22. Choi Tand Robertson PJ (2014) Deliberation and decision in collaborative governance: a simulation of approaches to mitigate power imbalance. Journal of Public Administration Research and Theory 24(2): 495–518. Cristofoli D and Markovic J (2016) How to make public networks really work: a qualitative comparative analysis. Public Administration 94(1): 89–110. Day DVand Lord RG (1992) Expertise and problem categorization: the role of expert processing in organizational sense-making. Journal of Management Studies 29(1): 35–47. Emerson K, Nabatchi T and Baloch S (2012) An integrative framework for collaborative gover- nance. Journal of Public Administration Research and Theory 22(1): 1–29. Feiock RC (2013) The institutional collective action framework. Policy Studies Journal 41(3): 397–425. Fisher G and Aguinis H (2017) Using theory elaboration to make theoretical advancements. Organizational Research Methods 20(3): 438–464. Fischer M and Leifeld P (2015) Policy forums: why do they exist and what are they used for? Policy Sciences 48(3): 363–382. Fischer M and Sciarini P (2016) Drivers of collaboration in political decision making: a cross sector perspective. The Journal of Politics 78(1): 63–74. Frieden J (2016) The governance of international finance. Annual Review of Political Science 19: 33–48. Gulati R (1995) Does familiarity breed trust? the implications of repeated ties for contractual choice in alliances. Academy of Management Journal 38(1): 85–112. Gulati R and Gargiulo M (1999) Where do interorganizational networks come from? American Journal of Sociology 104(5): 1439–1493. Halevy N, Halali E and Zlatev JJ (2019) Brokerage and brokering: an integrative review and organizing framework for third party influence. Acta Medica Academica 13(1): 215–239. Hamilton M and Lubell M (2018) Collaborative governance of climate change adaptation across spatial and institutional scales. Policy Studies Journal 46(2): 222–247. van der Heijden 23 Henry AD, Lubell M and McCoy M (2011) Belief systems and social capital as drivers of policy network structure: the case of california regional planning. Journal of Public Administration Research and Theory 21(3): 419–444. Henry AD (2011) Ideology, power, and the structure of policy networks. Policy Studies Journal 39(3): 361–383. Hertz R and Leuffen D (2011) Too big to run? analyzing the impact of enlargement on the speed of EU decision-making. European Union Politics 12(2): 193–215. Hileman J and Bodin O (2019) Balancing costs and benefits of collaboration in an ecology of games. Policy Studies Journal 47(1): 138–158. Jones BD (2001) Politics and the Architecture of Choice: Bounded Rationality and Governance. Chicago, USA: University of Chicago Press. Jones BD, Boushey G and Workman S (2006) Behavioral rationality and the policy process: toward a new model of organizational information processing. In: Peters BG and Pierre J (eds), Handbook of Public Policy. London, UK: Sage Publications, pp. 49–74. Juenke EG (2005) Management tenure and network time: how experience affects bureaucratic dynamics. Journal of Public Administration Research and Theory 15: 113–131. Klijn EH, Edelenbos J and Steijn B (2010) Trust in governance networks: its impacts on outcomes. Administration and Society 42(2): 193–221. Koppenjan JFM and Klijn EH (2004) Managing Uncertainties in Networks: A Network Approach to Problem Solving and Decision Making. London, UK: Routledge. Krackhardt D (1990) Assessing the political landscape: structure, cognition, and power in orga- nizations. Administrative Science Quarterly 35: 342–369. Leech BL (2002) Asking questions: techniques for semistructured interviews. PS: Political Science and Politics 35(4): 665–668. Leifeld P and Schneider V (2012) Information exchange in policy networks. American Journal of Political Science 56(3): 731–744. Lubell M, Mewhirter JM, Berardo R, et al. (2017) Transaction costs and the perceived effectiveness of complex institutional systems. Public Administration Review 77(5): 668–680. Matti S and Sandstrom A (2011) The rationale determining advocacy coalitions: examining co- ordination networks and corresponding beliefs. Policy Studies Journal 39(3): 385–410. McEviley B, Perrone Vand Zaheer A (2003) Trust as an organizing principle. Organization Science 14(1): 91–103. McGuire M and Silvia C (2010) The effect of problem severity, managerial and organizational capacity, and agency structure on intergovernmental collaboration: evidence from local emergency management. Public Administration Review 70(2): 279–288. Meier KJ and O’Toole LJ (2001) Managerial strategies and behavior in networks: a model with evidence from US public education. Journal of Public Administration Research and Theory 11(3): 271–294. Meier KJ and O’Toole LJ (2010) Beware of managers not bearing gifts: how management capacity augments the impact of managerial networking. Public Administration 88(4): 1025–1044. Miller CE (1985) Group decision making under majority and unanimity decision rules. Social Psychology Quarterly 48(1): 51–61. Moynihan DP (2008) Learning under uncertainty: networks in crisis management. Public Ad- ministration Review 68(2): 350–365. 24 Public Policy and Administration 0(0) Nohrstedt D and Bodin O (2020) Collective action problem characteristics and partner uncertainty as drivers of social tie formation in collaborative networks. Policy Studies Journal 48(4): 1082–1108. Ostrom E (1998) A behavioral approach to the rational choice theory of collective action: pres- idential address, american political science association, 1997. American Political Science Review 92(1): 1–22. Park HH and Rethemeyer RK (2012) The politics of connections: assessing the determinants of social structure in policy networks. Journal of Public Administration Research and Theory 24(2): 349–379. Piore MJ (2006) Qualitative research: does it fit in economics? European Management Review 3(1): 17–23. Quintane E and Carnabuci G (2016) How do brokers broker? tertius gaudens, tertius iungens, and the temporality of structural holes. Organization Science 27(6): 1343–1360. Richards L (2020) Handling Qualitative Data: A Practical Guide. London, UK: Sage Publications. Six P, Goodwin N, Peck E, et al. (2006) Managing Networks of Twenty-First Century Organi- sations. New York, NY: Palgrave Macmillan. Schilke O and Cook KS (2013) A cross-level process theory of trust development in interorga- nizational relationships. Strategic Organization 11(3): 281–303. Scott TA and Thomas C (2015) Do Collaborative Groups Enhance Interorganizational Networks? Public Performance and Management Review 38(4): 654–683. Scott TA and Thomas C (2017) Winners and losers in the ecology of games: network position, connectivity, and the benefits of collaborative governance regimes. Journal of Public Ad- ministration Research and Theory 27(4): 647–660. Siciliano MD, Wang W and Medina A (2021) Mechanisms of network formation in the public sector: a systematic review of the literature. Perspectives on Public Management and Gov- ernance 4(1): 63–81. Simon HA (1985) Human nature in politics: the dialogue of psychology with political science. American Political Science Review 79(2): 293–304. Taylor RN (1975) Age and experience as determinants of managerial information processing and decision making performance. Academy of Management Journal 18(1): 74–81. Thompson JD (1967) Organizations in Action: Social Science Bases of Administrative Theory. New York, NY: McGraw-Hill. Vangen S and Huxham C (2003) Nurturing collaborative relations: building trust in interorgani- zational collaboration. The Journal of Applied Behavioral Science 39(1): 5–31. Vis B (2019) Heuristics and political elites’ judgment and decision-making. Political Studies Review 17(1): 41–52. Weber EP and Khademian AM (2008) Wicked problems, knowledge challenges, and collaborative capacity builders in network settings. Public Administration Review 68(2): 334–349. Weible CM (2005) Beliefs and perceived influence in a natural resource conflict: an advocacy coalition approach to policy networks. Political Research Quarterly 58(3): 461–475. Williams P (2002) The competent boundary spanner. Public Administration 80(1): 103–124. van der Heijden 25 Appendix A List of respondents Respondent Organization Function R1 DNB Div. Director R2 DNB Div. Director R3 DNB Div. Director R4 DNB Policy advisor R5 DNB Policy advisor R6 DNB Policy advisor R7 DNB Policy advisor R8 AFM Policy advisor R9 AFM Policy advisor R10 AFM Unit head R11 AFM Policy advisor R12 AFM Policy advisor R13 Min. Fin Policy advisor R14 Min. Fin Unit manager R15 Min. Fin Div. Director R16 Min. Fin Unit head DNB = National banking regulator AFM = National securities regulator Min. Fin = Ministry of Finance Appendix B Interview guide Note: The interview guide below was translated from Dutch. Also, in practice, the in- terview guide taken to the actual interview was specified further depending on the particular characteristics of the respondent (i.e. organization, function). Still, the sequence of questions as described below was always the blueprint, as to ensure the basic areas of interest were covered. 1. General Introduction: Explaining the goals of the research and what will be discussed (in general lines) during the interview. Procedures on anonymization and data protection (transcript/recording). 2. Walkthrough of activities and setup of organizational unit: a. Role and relation of the unit to the broader organization. b. Specific tasks of respondent within the unit/organization. 3. Walkthrough international platforms with which the unit is involved: a. Who typically participates on behalf of the unit? b. International activities respondent. Involved in what ways? 26 Public Policy and Administration 0(0) 4. Specific regulatory standards/dossiers currently relevant: a. Specific dossiers with which respondents is involved. b. Identify suitable examples to which to return. 5. Walkthrough of how international meetings are typically prepared: a. Structure of preparation, determining positions, finding coalition partners. b. Coordinating with other units. c. Respondents own role + examples. 6. Walkthrough of reporting on international meetings/activities: a. Reporting to supervisors? In what ways? b. Keeping a log? Communicating activities (to whom)? 7. Coordination with other (national) agencies/ministries: a. Ways of convening, contact. Speaking with one voice? Author Anonymous, b. Examples good or bad. Own involvement? 8. International negotiations/decision-making: ways of influence? a. Different policy instruments/strategies of influence. b. Probing for examples/own experiences. 9. Comparing different types of fora/platforms (European/Global): a. Nature of negotiations? Different strategies required? b. Probing for examples/own experiences. 10. Level of contact with foreign regulators/officials. Role of informal network? a. Meeting the same people? Different per country? b. Determining who is an appropriate partner? c. Role of informal network/social dynamics. 11. Wrap-up and debriefing: a. Additions? Returning to particular questions. b. Debriefing and next steps.

Journal

Public Policy and AdministrationSAGE

Published: Jan 1, 2022

There are no references for this article.