UvA-DARE (Digital Academic Repository) Why interest organizations do what they do: assessing the explanatory potential of ‘exchange’ approaches

Summary Important parts of the literature on strategies of interest organizations consider the political activities of interest organizations as serving to build and maintain exchange relationships between organizational representatives and constituents, policymakers and the news media. The incentive for organizational survival produces activities that links social and political domains. This review integrates various strands of existing thinking on interest groups under an ‘ exchange ’ perspective. It produces a view on interest organizations as being strategically strongly constrained through different demands made on the organization when engaging in relationships with supporters, policymakers and journalists. While challenging, research designs should account for the interrelated nature of these relationships rather than treating organizations as a strategically autonomous actors.

The online version of this article is available Open Access.
Why do interest organizations do what they do? A number of 'classic' and recent studies evaluate these political activities as exchange relationships of interest organizations with different parts of the organizational environment (for example, Salisbury, 1969;Gamson and Wolfsfeld, 1993;Schmitter and Streeck, 1999;Bouwen, 2004). In this review, I seek to connect the existing 'exchange-inspired' thinking on the mobilization of members, the policy-related activities and outside-oriented strategies. I structure these studies into a triangular scheme of exchanges by the interest organization with three types of actors: (i) their supporters, (ii) political institutional actors and (iii) the news media and other actors related to public opinion. These exchanges are governed by what I label the logic of support, logic of influence and the logic of reputation. 1 The contribution of this review is that it explicitly incorporates the news media in the model, further specifies the supporters and institutional exchange relationships, and presents the logics in an interrelated manner. First, the exchange relationship of interest organizations with the news media matters as important channel of interest group influence on public opinion, and it matters especially in view of recent controversy over 'publicity as a weapon for the weak' (for example, Thrall, 2006). Second, it also matters indirectly as it affects the relationships of interest organization with their constituents and public authorities. By incorporating recent literature on outside strategies, these relationships are further specified to account for the effect of public opinion and news media reporting. Third, exchange relationships with distinct fields of actors (supporters, politics, media) tend to be studied independent of each other. There are small subfields on legislative lobbying, the internal (member) organization of groups or 'outside-oriented' strategies. These distinct types of organizational behavior are strongly interdependent, which makes specialized studies on any one of them vulnerable to underspecification. That is, for instance, opportunities for legislative lobbying largely depend on public opinion. Parliamentarians are probably more likely to talk to group representatives when relevant sections of public opinion are supportive of the cause of the group. Several of such interdependencies are explicated in the following review and could help in the design of more specified models of patterns of behavior of interest organizations.
Furthermore, group scholars have been relatively unsuccessful in designing country-comparative research (but see Dür and Mateo, 2013). According to Lowery et al (2008Lowery et al ( , p. 1234, this is largely due to a lack of theory 'that is capable of encompassing both multiple levels of analysis and variety in contexts in which phenomena of interest occur'. Although the outline presented here is surely not such a theory, it may help constructing a meso-level theory that is capable of connecting country-or issue-level frameworks with organizational level assumptions by explicitly labeling contextual constraints and opportunities for certain political activities. In the following, I first provide a behavioral definition of interest organizations and differentiate them from political parties and social movements. Then I describe neocorporatist thinking on exchange theory and discuss how exchange relationships shape interest organizations. Third, I assess the literatures bearing on the three logics of exchange. I conclude with some modest suggestions for improving the study of the behavior of interest organizations. action or loose networks of actors, commonly included in studies of social movements (for example, Snow et al, 2004, p. 7; but also note organizational definitions of social movements, see, for example, Burstein and Linton, 2002;Diani, 2012). Second, Beyers et al (2008Beyers et al ( , p. 1106 point out that interest organizations should be interested in influencing public policies. This may entail changing existing policy, keeping the status quo policy, reducing government action or seeking government recognition of a problem. Organizations may seek influence via direct interaction with officials or by other means. These other means, however, should not include directly seeking public office through elections. This differentiates interest organizations from political parties. The conceptual focus on lobbying behavior in this socalled functional or behavioral definition of interest organizations makes it, as highlighted by Jordan et al (2004), impossible to distinguish between interest organizations and politically active universities, companies or other institutions (see also Salisbury, 1992, p. 43). Such a differentiation is needed in order to conceptually account for the phenomena associated with the aggregation and representation of interests or causes that are shared beyond the organization itself. However, being organized or representative need not have the form of 'collective action' as traditionally conceived and organizations such as companies or schools also aggregate interests. That is, we know that such institutions have to deal with important internal politics that may be similar to those faced by membership organizations. Heinz et al (1993, p. 384) note that 'the government affairs officers of corporations, the executives of trade associations and the heads of citizen-government groups must justify the cost of their operations to their respective organizational constituencieswhether the constituency is the CEO, industry members or contributors' (for similar argument see Hart, 2004;Hillman et al, 2004;Wilts, 2006). This latter conceptualization of internal politics of firms leads to the empirical inclusion of the same organizations as in functional or behavioral definitions, but adds a conceptual interest in intra-institution relationships as a special case of 'logic of support'.
This review takes the perspective of the interest organization as a building block for theory construction on differences in political activities between organizations. This is not self-evident because others have taken issues (for example, Kingdon, 1984;Baumgartner et al, 2009), countries (for example, Eising, 2009) or individuals (for example, Olson, 1965;Salisbury, 1969) as conceptual building blocks. Such perspectives have somewhat different research aims, and their focus is useful for explaining differences in policy outcomes, examining institutional effects or collective action, respectively. When evaluating activities of organizations, countrylevel theories only provide very general hypotheses-that is, bluntly formulated 'French unions are more likely to call a strike', 'US business associations tend to choose inside lobbying' (for example, Balme and Chabanet, 2008, p. 36;Woll, 2008). Such propositions could easily be incorporated in an organizational-level approach, although it is more difficult to meaningfully disaggregate country-level hypotheses to the organizational level. In other words, the meso-level organizational perspective could help linking macro-level theories to more concrete, testable hypotheses by subdividing the comparative research tasks into meaningful segments (for related argument see Lowery et al, 2008).

How Exchange Relations Shape Political Activities
In the following, I review sociological thinking on exchange relationships, and discuss neo-corporatist thinking about exchange relations and propose several adaptations.
As a concept in (early) sociological theories explaining the workings of power, exchange is defined as a type of behavioral interaction between actors that is voluntary, reciprocal and mutually beneficial (Emerson, 1962(Emerson, , 1976Blau, 1964;Levine and White, 1961). Exchange theoretical work is closely associated resource dependency (Jacobs, 1974;Hannan and Freeman, 1977;Pfeffer and Salancik, 1978;Aldrich, 1979) and resource mobilization theory (McCarthy and Zald, 1977). Exchange relationships involve the reciprocal flow of valued behavior, or, formulated differently, the exchange of resources between two actors. Emerson (1976, p. 347) specifies that a 'resource is an ability, possession, or other attribute of an actor giving him the capacity to reward (or punish) another specified actor'. Resources are clearly not restricted to financial resources and are behavioral and relational: they are 'not possessions or attributes of individual actors, but rather they are attributes of the relationship between actors' (Emerson, 1976, p. 348). They may be used in exchange relationships but also in various other types of interaction such as regulation, use of force or competition (Blau, 1964, p. 124). Consequently, 'social exchange is limited to actions that are contingent on rewarding reactions from others and that cease when these expected reactions are not forthcoming ' (1964, p. 6). This bilateral exchange is, even by relatively powerful actors, usually favored over any other type of interaction because the mutually rewarding character makes it the most cost-effective way to generate power (for example, Jordan and Richardson, 1982, p. 95). It is therefore a useful perspective also in cases where one actor depends more strongly on the other than vice versa, so-called asymmetry, as is commonly the case for relationships of interest organizations with policymakers, supporters or journalists. Exchange relationships between different domains and associated actor types are assumed to be more valuable than relationships among interest organizations as dissimilar resources can be put into exchange. Such exchange relationships are central in this review and they commonly require intermediaries organizing the connections between domains. Interest organizations are in the business of managing exchange relationships between different domains.
In interest group literature, we find the most explicit sociological exchangetheoretical references in Salisbury (1969), neo-corporatist industrial relationships literature (for example, Molina and Rhodes, 2002) and neo-corporatist organizational thinking. I take the latter as a starting point for this review. On the first, Salisbury (1969) solely focuses on the process of group formation. His main interest is in the relationship between group leaders/entrepreneurs and (potential) members, prioritizing that relationship relative to political or public relationships. The review presented here does not contradict his point but more explicitly specifies the additional organizational constraints arising from different types of exchange relationships. Second, political-economic studies of interest organizations tend to use a definition of exchange that is firmly embedded in industrial relations (for example, Molina and Rhodes, 2002;Molina, 2006: 645-646;Baccaro and Simoni, 2008). For instance, Molina and Rhodes (2002, p. 322) follow Marin's (1990, p. 40) definition of 'political exchange as forms of mutually contingent, macropolitical and noneconomic transaction between autonomous, organized, collective actors with divergent/competitive/antagonistic but functionally interdependent interests, the binding character of which cannot be based on law and contract' (also see Pizzorno, 1978). This definition of political exchange is specifically aimed at understanding agreements between sometimes clearly antagonistic interests employers and unions, and only partially applies to the various exchange relationships of a broader range of interest organizations with actors that need not have interests that are antagonistic to theirs.
We find the most explicit and elaborate use of the above-mentioned sociological exchange tradition in neo-corporatist studies of interest organization. The term neocorportatism is here used to denote a research tradition within studies of interest representation, not studies of state formation, normative studies or otherwise. In a neo-corporatist perspective, interest organizations are first and foremost intermediaries between government and their constituents. Interest organizations offer their supporters collective interest representation before government. The combination of interest articulation (government representation) and interest aggregation (act collectively) put them in an intermediate position between the state on the one hand and society or the economy on the other (Streeck and Kenworthy, 2005, p. 451). These two environments operate under two logics of exchange that produces organizational tensions (Schmitter and Streeck, 1999, pp. 19-30). The explication of such organizational tensions and associated 'logics' (also see Offe and Wiesenthal, 1980) is the main contribution to above-mentioned sociological exchange theories and is further elaborated in the next section.
In this perspective, organizational leaders who successfully mobilize a large proportion of potential members are rewarded with policy access, while 'narrow' and partially mobilized interests will receive only limited access. This produces non-competitive (corporatist) environments, where there are only very few policynetworked interest organization per sector, and where members, presumably interested in policy access, do not find other or new interest organizations to choose from (Schmitter, 1974(Schmitter, , 1979. The interests of members are consequently compromised through their association in relatively broad organizations. Leaders may be further inclined to compromise members' preferences and interests through their close involvement in the policy process and presumed concomitant distance from their 'base'. The incentives governing these two exchange relationships produce a tradeoff between a more diverse constituency or broader membership base and more compromised policy demands but a stronger negotiation position vis-à-vis policymakers (Streeck and Kenworthy, 2005, p. 451).
This traditional neo-corporatist tension with its focus on exchange logics has in one way or another become part of most studies on political strategies of interest organizations. At the same time, research interest in the effect of public opinion and the news media has also been growing (for example, Kollman, 1998;Thrall, 2006;Kriesi et al, 2007;Binderkrantz, 2012), associated with debates about the 'mediatisation' of contemporary politics (for example, Mazzoleni and Schulz, 1999;Kriesi, 2004). In order to reflect this, I include a third exchange relationshipthat between the news media and the interest organizationwhich I label 'logic of reputation'. This relationship is shaped through outside-oriented and public strategies, such as the transformation of policy information to newsworthy events, voicing constituents' concerns in public debate and organizing members' participation in public action. The logic of reputation is distinct from the logic of influence and support, as it is governed by different 'media institutional' principles (Schudson, 2002), such as the news values held by journalists. It also affects the opportunities and constraints for other relationships, therefore I also explicitly refer to studies that assess the impact of the news media on the relationship with supporters and political institutions, producing a more elaborate specification of the logics of influence and support than in neo-corporatist views outlined above.
This has a number of implications on our thinking on tensions with which interest organizations have to deal. First, by including public opinion and news media considerations of leaders, the strategic tensions become stronger and choices available to the organization narrower. For example, 'going public' may be beneficial in terms of the reputation of the interest group among a broader public, but at the same time have mixed effects on supporters and negative effects in terms of the logic of influence. As Lipsky (1968Lipsky ( , p. 1054) argues, 'people in power do not like to sit down with rogues' and 'few [advocates] can be convincing as advocate and arbiter at the same time'. Second, it broadens the applicability of the neo-corporatist exchange scheme toward social movement organizations. The scheme is, as intended, biased toward activities typical to business associations such as inside lobbying. Other groups, such as social movement organizations, are traditionally expected to have a tactical preference for outside or public strategies. Third, one of the key differences between political systems is related to the relative importance of outside opportunities and strategies. Balme and Chabanet (2008, p. 34) point to 'protest systems', such as France, where acquiring increased publicity is strategically dominant, where 'unfavorable policy and institutional opportunities [for pluralist lobbying or corporatist consultation] may be partially compensated for by the opening of windows of opportunity for the mobilization of interests using media activity' (for similar arguments on political opportunities see Kriesi et al, 1995). The assessment of such country differences also requires the explicit incorporation of media relationships in organization-level theories.

Constraints in Gaining Support, Influence and Reputation
To repeat, interest organizations organize exchange relationships between members and political institutions, political institutions and the news media, and the news media and members. Through these reciprocal flows of valued behavior, they connect political institutions, supporters and the news media. For the sake of the clarity of this review, I only focus on those exchange relationships that not only connect interest organizations with other actors but also help interest organizations to forge exchange relationship in other domains. It is only in these cases that interest organizations potentially function as intermediaries. Each exchange relationship is understood in association with relationships under other logics. The range of activities described below is narrowed down by its exclusive focus on exchange relationships (as opposed to among others formal regulatory relations), on inter-domain relationships (as opposed to contacts among interest organizations) and on relationships that potentially facilitate relationships with other domains (as opposed to, for instance, 'purely' membership maintenance activities).
Each logic of exchange consists of four types of resources, and several environmental factors affecting the value of these resources. Each of the connected arrows in Figure 1 is discussed below and the discussion in this section is summarized in Table 1 and the numbers in the text refer to this table.

The logic of support
The logic of support assumes that activities are driven by the need for organizational survival. As Lowery (2007, p. 46) notes, 'the most fundamental goals of organizations must be to survive as organizations'. Organizational survival and maintenance primarily depends on the recruitment of members and the mobilization of an existing supporters' base. Salisbury (1969, p. 2) classically argued that 'interest group origins, growth, death and associated lobbying activity may all be better explained if we regard them as exchange relationships between entrepreneurs/organizers, who invest capital in a set of benefits, which they offer to prospective members at a pricemembership'. Besides membership dues, these exchange relationships concern activities related to (i) constituents' policy compliance, (ii) members' public action, (iii) members' control over leaders and (iv) the public visibility of a political issue. The nature of these activities is shaped by the type of supporters (citizens, companies), their interests (diffuse, concentrated) and the interest community of which the organization is part (level of competition).
First, members or supporters provide organizations with (future) compliance with policies, (potential) political support for governing parties and newsworthy participation in public political events. For instance, business interest associations could facilitate strategic adaptations of their members' organizations to new policies (Schmitter and Streeck, 1985). Members can offer such cooperation or not. Further, as highlighted in studies of social movements, supporters socialize among each other and participate in the (public) activities of interest organizations such as conferences, campaign events and demonstrations. Such membership socialization (Beyers, 2008(Beyers, , pp. 1202(Beyers, -1204 and participation forms the basis of exchange relationships in the institutional (loyalty fosters compliance) and the news media environment (participation could create newsworthy 'events'). As Lipsky (1968Lipsky ( , p. 1149) notes, and is further emphasized below, 'leaders' ability to control protest constituents and guarantee their behavior represents a bargaining strength' in negotiations with policymakers. Second, in return for their compliance or public support, leaders supply members with control mechanisms and with favorable news media contributions. This control varies between active, representative 'voice' mechanisms, such as leadership elections, to more passive, administrative 'exit' mechanisms under which leaders follow (through polls and so on) the preferences of their constituents so as to prevent them from leaving (Hirschman, 1970, p. 21). This distinction between administrative and representative means is commonly seen as mutually exclusive (Child et al, 1973;Kriesi, 1996, p. 153;Schmitter and Streeck, 1999, pp. 19-20) and is closely related to the much-discussed professionalization of membership groups (for example, Verba et al, 1995;Skocpol, 2003;Binderkrantz, 2009;Walker et al, 2011;Maloney, 2012). Administrative means are associated with the efficient operation of the organization, such as relying on professional management as opposed to volunteers, using task specialization and relying on donations instead of membership dues. Representative means, in contrast, refer to 'widespread membership involvement' via participation and volunteering, and sometimes even independently voicing policy positions (for example, Gamson, 2004, p. 253). This may create committed or 'disciplined' membership base that may be offered as future policy compliance (or not). Both the provision of opportunities for participation and the 'threat' of members leaving is a liability for organizational leaders as members gain some control over the organization (for example, Jordan and Maloney, 2007, pp. 145-170), and sometimes provide 'biased' signals on their own policy preferences (Lohmann, 1993). Further, organizational entrepreneurs seek to provide members with attention to the organization in the news media and other public venues (Gais and Walker, 1991, p. 103;Jordan and Maloney, 1997). The public visibility of the organization and its cause is, for instance, often emphasized via internal organizational media (Website, newsletters, magazines). Members want that their concerns and interests receive recognition in the public debate; in Salisbury's (1969, p. 16) terms these are 'expressive benefits'.
On the basis of this, we may hypothesize that we should expect leaders to seek to socialize or discipline their supporters while restricting membership control over leaders.
The extent to which both partially contradictory goals can be achieved depends on contextual factors related to social and organizational context: the types of members, supporters and interests represented, and the density of the interest community. First, the magnitude of collective action problems varies between diffuse interests on the one hand and specialized or concentrated interests on the other hand (Olson, 1965). Through their attention to certain interest organizations, the news media help diffusely interested constituents alleviate collective action problems. It seems that citizens are organized on diffuse interests, and companies (or other organizations) seek collective action on specialized interests. Exchange relations in the news media environment are consequently more important for citizen groups with diffuse interests than for other interest organizations. As pointed out by Gais and Walker (1991, p. 106), they rely on 'an outside strategy of public persuasion and political mobilization' (also see Binderkrantz, 2008). Offe and Wiesenthal (1980, p. 79) present this argument in a more abstract manner when they argue that 'the relatively powerless', such as most citizen groups, require public discourse to 'simultaneously define and express interests', whereas the 'powerful' only need the media to (strategically) express their interests. In other words, diffuse interests rely on public action to help 'overcoming the comparatively higher costs of collective action by changing the standards according to which these costs are subjectively estimated within their own collectivity' (Offe and Wiesenthal, 1980, p. 78). Furthermore, the nature of the interest of the constituents (diffuse, specialized) probably correlates with the distribution of their interests across policy sectors. As Beyers (2008Beyers ( , p. 1201) notes, 'the larger the scope of an organization, the more policy sectors and issues it needs to cover, the larger and more heterogeneous its constituency'. As we know that political activities vary by policy sector, organizations that are active in a variety of sectors are also likely to require a broad range of influence tactics and membership-related activities. Organizations that represent diffuse interests thus, necessarily, engage in a broader range of both institutional and public activities than organizations that represent narrower interests.
This leads to the hypothesis that we should expect leaders of organizations representing diffuse interests to have more limited control over their supporters than leaders of organizations representing specific interests.
Second, interest organization population density reduces the value of the supporters-related resources in exchanges with policymakers. In neo-corporatist thinking, competitive pressures from similar organizations (density) have a straightforward detrimental effect on the capacity of interest organizations to strike policy deals. As Streeck and Kenworthy (2005, p. 451) note, 'the logic of influence tends to place a premium on interest organizations being broadly based and representing more general instead of highly special interests'. Policymakers are not interested in a variety of exchanges with specialized actors in a fragmented interest community, but seek to exchange a representational monopoly with a single interest organization that aggregates various interests within a sector. This is impossible in dense communities. Similarly, but more nuanced, population ecology studies show that organizations in dense interest communities and with similar political positions are likely to specialize by tactic (Gray and Lowery, 1998). In terms of exchange relationships, this suggests that to avoid conflict with fellow organizations and to cope with the relatively low value of their supporters-related resources they divide the interest community along lines of institutional interaction, news media contacts and supporters base (with the latter taking priority Gray and Lowery, 1996). Density in the interest organization population is consequently likely to strongly reduce the range of strategic options, as is further reflected in the construction of interest group identities in specific niches (Heaney, 2004). Third, studies of social movement industries similarly find that tactical and positional specialization is related to interest community density (McCarthy and Zald, 1977;Downey and Rohlinger, 2008). This effect operates together with, in the words of King (2008, p. 1598), the 'influence of a broader environmental resource base' within society that favors unspecialized, more general organizations. Competitive pressures, they suggest, lead to a community that is typical of concentrated markets: a couple of (older) generalists on the one hand, and various specialists emerging and flowering in the fringes of the sector on the other (for similar findings see Browne, 1990).
In terms of the logic of support, the following hypothesis may be formulated: we should expect leaders of organizations in dense environments to have more limited control over their supporters than leaders of organizations operating in monopoly environments.

The logic of influence
The interaction of interest organizations with political institutions has frequently been conceptualized as an exchange relationship. This is the case in pluralist (Bauer et al, 1963, pp. 154-178), economic (for example, McChesney, 1997), resource-dependency (Bouwen, 2004, p. 338;Beyers and Kerremans, 2007;Poppelaars, 2009), policy-oriented (for example, Jordan and Richardson, 1982) and corporatist approaches to interest representation. However, in each of these traditions, the exchange relationship is largely considered independent of interactions of interest organizations with their supporters or in the media. In order to influence public policy, interest organizations engage in activities related to: (v) supporters' policy discipline and political/technical information, (vi) policy frames in the news media that are 'acceptable' to policymakers, (vii) the maintenance of organizational privileges and favorable policies and (viii) monitoring institutional activity. Whether and how these activities produce successful exchange relationships depend on the type of political control potentially exerted upon the actor or venue lobbied (parliament, minister, bureaucratic venue) and on the system of structured interest group access in general.
First, organizational leaders offer policymakers the compliance (or not) of their constituents. As discussed earlier, leaders have varying modes and levels of control over their supporters. Leaders may have some formal powers but also have means such as informal persuasion via internal media, appeals to supporters' loyalty and socialization via membership participation. In situations of limited constituent control, they can still offer political actors information on political support. This is what according to Hall and Deardorff (2006) must be understood as a legislative subsidy by interest organizations to members of parliament. Such information on political support or possible non-compliance or resistance could help to increase the effectiveness of policies (Poppelaars, 2009, pp. 4-6).
Furthermore, in relation to public debate, interest organizations can attempt to transform public claims into policy-relevant material such as policy statements and reports. These are valuable for policymakers in order to respond to public challengers. Whereas organizations provide input for the news media (see below), the point here is about the ways in which news media attention is translated into policies (or not) or may provide ammunition for bureaucrats in competition with other parts of the state apparatus, as emphasized by Jordan and Richardson (1982, p. 85). Interest organizations contribute to the transformation of the 'popular rhetoric' of the news media into the policy speak of political institutions, or, in contrast, strategically use news media frames to challenge existing policy frames in the policy sector. Thus, interest organizations offer to or withhold from politicians the instruments or, again in Hall and Deardorff's (2006) words 'legislative subsidies', they need to respond to news media signals.
Second, state officials can give interest organizations access to the policy process. This could range from a formal constitutional role, including legally binding decision powers or representational monopolies, to informal irregular contact with a parliamentarian. Access is beneficial for the interest organizations because it has indirect effects on the membership environment and may lead to favorable policies. Formal access could indirectly have structuring effects on the membership environment by making certain organizations more attractive to potential members than other organizations. Access is also the first step for policymakers toward making substantive policy concessions favorable to the constituents of the interest organization (Bouwen and McCown, 2007).
In order to maintain or construct a positive reputation of the interest represented, interest organizations monitor policies and assess upcoming legislation to transform this information for the news media (and through internal media, to inform their constituents). They rely on policymakers to supply them with policy-relevant or politically strategic information. Among others, they need this expertise for the provision of 'information subsidies' in the news media environment (Gandy, 1982;Hamilton, 2004). For instance, as Keck and Sikkink (1999, pp. 97-98) point out, strategies such as 'accountability politics' require thorough policy research, attending to both policy circles and the broader news media.
Under the logic of influence, the following hypothesis should hold: Interest organizations are provided with policy information and policy access when they offer various forms of legislative subsidies to policymakers.
The institutional position of specific policy venues and the broader (national) institutional context affects the logic of influence. First, two aspects of the specific position of the institutional counterpart matter: the level of democratic control of political actors and the opportunities for venue shopping. Political actors who are under relatively direct electoral control, such as parliamentarians, are likely to favor news media resources. In contrast, venues or actors that are at arm's length of public scrutiny are likely to be interested in the exchange of constituency-related resources, such as information about future compliance. For instance, Bouwen (2004) found that the European Parliament was more interested in information on broader political support and the European Commission more interested in specific (technical) information about the sector and members of interest organizations. Further, the authority to make decisions could be concentrated in a specific institution, a policy venue or, as seems to be increasingly the case, be spread across a system of actors as tends to be the case in federal or multi-level governance systems (for example, Hooghe and Marks, 2003). In the latter situation, interest organizations could 'shop' for receptive venues in order to find the best offer for their resources (Jones andBaumgartner, 1991, p. 1050;Mazey and Richardson, 2001).
Formulated as hypothesis, we should find that interest organizations have to offer different types of legislative subsidies to different institutional venues, and that they have to offer fewer subsidies (per venue) in case multiple venues are available.
The broader (national) political context, secondly, structures the influence environment. This context guarantees (or not) the reciprocity of exchange relations. It is assumed that in corporatist influence environments the credibility of the commitments of the exchange partners is strengthened via such things as social norms, political culture and aspects of the party system. In corporatist systems, it is more likely that interest organizations build lasting and important exchange relations with government actors than in countries with a pluralist tradition. In terms of the logic of influence, we may thus hypothesize that interest organizations in corporatist countries have to offer fewer 'subsidies' than interest organizations in other systems.

The logic of reputation
A third rationale for the existence of interest organizations is their plausible presentation of arguments, their support in public opinion and their recognition by the news media. The news media or public environment consists of journalists and other relevant news intermediaries (blogs, conferences, public debates). The interaction between journalists and interest organizations has always been of particular interest in social movement research (for example, Gamson, 2004;Andrews and Caren, 2010), and within this field it has occasionally been explicitly theorized as an exchange relationship (for example, Wolfsfeld, 1984Wolfsfeld, , 1991Terkildsen et al, 1998;Koopmans, 2004;Binderkrantz and Christiansen, 2012).
Recent studies do not find supporting evidence for two commonly held assumptions about this relationship: the first being that challengers of public policy outside or excluded from established institutional arenas are expected to necessarily equip themselves with symbolic, media-and public-oriented action repertoires, and the second being that the news media is an ally to resonate their voices (Lasswell, 1950, pp. 235-236;Dahl, 1956, p. 145;Schattschneider, 1960, p. 40;Lipsky, 1968Lipsky, , p. 1144Wilson, 1974, pp. 284-289;1961, p. 291). As regards the strategies of interest organizations, recent empirical evidence suggests that interest organizations with strong institutional presence 'persist' their influence and also secure media presence (Beyers, 2004;Kriesi et al, 2007, pp. 53-55). Interest organizations, for various reasons, do not or cannot 'compensate' their institutional absence with strong media presence. As regards the receptiveness of the news media to interest organizations, recent research is concerned with the relative presence of interest organizations in the media compared with other actors or with the difference among interest organizations in media attention (Danielian and Page, 1994;Thrall, 2006;Binderkrantz, 2012;van Dalen, 2012). Both 'biases' are sometimes labeled 'media bias' and tend to be supported by empirical evidence. This is again in contradiction to the earlier expectation that in the 'open' media arena we should find lower levels of bias than in the 'closed' institutional arena.
In the following section, it is discussed that interest organizations offer: (ix) newsworthy 'events and drama', and (x) policy-relevant newsworthy information, in exchange for (xi) latent public opinion support, and (xii) validation of their organization and issue expansion or containment of issues of interest to the organization. Actual exchanges are conditioned by the media system of a country and by the actual public opinion on specific policy issues.
First, interest organizations could provide journalists with information or events that fit with criteria of newsworthiness (Galtung and Ruge, 1965). Organizations ranging from social movement organizations to business interest associations produce various types of attention-grabbing media events. When it comes to social movement organizations, Gamson and Wolfsfeld (1993, pp. 116-117) point out that despite media selection bias (for example, McCarthy et al, 1996;Oliver and Maney, 2000;Earl et al, 2004) 'social movements often make good copy for the media. They provide drama, conflict and action; colorful copy; and photo opportunities'. Business interests also seek favorable coverage; recent studies highlight the direct public engagement on the part of corporate public affairs departments. They provide the media directly with stories and make statements related to public policy, especially when challenged (Wilts and Skippari, 2007, p. 132;de Bakker and den Hond, 2008;Dahan, 2009). Further, through their involvement in the policy process, interest organizations have expertise on public policy and on the ways their constituents are affected by policies. Such knowledge about and statements on the political game could be valuable news input. This is especially so when it is produced in a format that news media can easily process. These are so-called 'information subsidies' such as press statements, pre-fabricated items or contact details of members willing to provide stories to journalists. Gandy (1982, p. 14) criticizes such Public Affairs activities and suggests that they undermine the autonomy of the news media: 'often the value of an information subsidy for any source is increased to the extent that the source can disguise the promotional, partisan and self-interested quality of the information' (see also Negrine, 1996, pp. 11, 27). However, despite 'colorful copy' and information subsidies, various recent studies point to the disadvantaged position of interest organizations relative to other policy actors in getting media attention (Thrall, 2006;Binderkrantz, 2012;van Dalen, 2012). Such studies find support for Bennett's (1990) 'idea that news is "indexed" implicitly to the range and dynamics of government debate'; in other words, interest organizations rank 'low' on the index of sources of journalists, with official, government sources being 'high' on their index (for similar argument see Cook, 1998). 'Official sources' are prioritized relative to other sources of political information (Bennett et al, 2007, p. 36), setting in motion media mechanisms 'cumulative inequality', again to the disadvantage of most interest organizations (Gamson and Wolfsfeld, 1993, p. 117). Interest organizations consequently will have to invest substantial resources in order to get journalists to report on their cause and position. It seems that only those interest organizations that are successful under the logic of support (that is, mobilize a large number of people) or under the logic of influence (that is, have extensive access to government officials) are able to 'persist' their success when engaging in exchange relationships with journalists (Beyers, 2004;Kriesi et al, 2007;Dür and Mateo, 2013).
Second, news media coverage provides latent public support for the goals of the organization. News media activate and mobilize audiences and increase public support for the causes of interest organizations and social movements. This strengthens membership recruitment. As Rucht (2004, p. 211) writes: 'Positive [public] reactions range from increased sympathy to occasional acts of support to continuous and full commitment [in the form of membership of the organization]'. In their survey of campaign organizations, Jordan and Maloney (2007, p. 96) found that 96 per cent of the organizations report that new members are attracted because individuals directly contact the groups at their own initiative. Public reputation largely determines whether individuals consider contacting an organization. For the recruitment of new members, organizations depend in large part on the 'free publicity' of news media coverage. As Gamson and Wolfsfeld (1993, p. 116) note, 'movements must reach their constituents through some form of public discourse'. Similarly, but formulated as a 'second level' of mobilization, Offe and Wiesenthal (1980, p. 96) emphasize that for a broad range of causes the first task of interest organizations is to convince citizens through public discourse to 'see better what it really is' that they want, before potential members support certain interest organizations.
Furthermore, interest organizations seek organizational validation and standing to be taken seriously by supporters and policymakers. In the words of Gamson and Wolfsfeld (1993, p. 116), 'the media spotlight validates the fact that the movement is an important player'. Validation matters for both new and more established interest organizations; considering their 'informal' character, interest organizations require validation per political issue and need to continuously reaffirm their validity as relevant actors. For new organizations that have no policy access at all, it does not matter how their activities make it to the news media; in all cases they accrue the policy or supporters benefits of organizational validationas Gamson (2004, p. 252) writes: 'No news is bad news', or as Lipsky (1968Lipsky ( , p. 1151) writes on protest organizations: 'like a tree falling unheard in a forest, there is no protest unless protest is perceived and projected' (also see Koopmans, 2004, p. 373).
Journalists could set the political agenda or expand the policy conflict. Depending on circumstances, this may either spoil the policy game for interest organizations or provide badly needed government recognition. On the one hand, as Lowery (2007, p. 37) summarises the literature: 'the influence of organized interests seems to be negatively associated with the scope of lobbying battles as measured by the number of organizations involved, the intensity of their lobbying, and how attentive the public is'. In general, interest organizations are likely to lose policy influence when news media attend to 'their interests'. Thus, organized interests would act against their own policy interests (but not necessarily against their supporters preferences) when they seek media attention. On the other hand, media attention is needed in cases where interest organizations would like to change the range of actors involved or seek government recognition of a social problem as an policy issue. The first is the case for 'losers' of status quo policies, as Schattschneider (1960, p. 40) points out that it is the 'weak who want to socialize conflict, that is, involve more and more people in the conflict until the balance of forces is changed'. The latter is the case for 'new' issues where the media spotlight serves to recognize a situation as a public problem and potential political issue requiring government attention (for example, Walgrave and van Aelst, 2006).
Under the logic of reputation, it is hypothesized that (all kinds of) interest organizations should extensively attempt to control (but not necessarily seek) attention in the press but that journalists require 'newsworthiness' that can only be provided by a selected few interest organizations.
The character of exchange relations between interest organizations and news media depends on the public opinion regarding an issue and the media system in a country. First, as regards public opinion, in line with the conditional nature of outside strategies mentioned above, Kollman (1998, pp. 155-164) argues that the selective application and timing of outside-oriented strategies by organized interests depends on the combination of the popularity and the salience of the policy positions of the interest organizationthat is, in situations in which groups perceive broad public support (popularity) on policies their members care about (salience), organizations will seek media contacts (Kollman, 1998, pp. 82-88). Kollman (1998, pp. 155-164) labels this 'comforting politics' and introduces a typology to understand three other states of public opinion that structures the likely strategies. This is shown in Figure 2.
In each situation, leaders of organizations attach varying relative values to exchanges in the news media environment and in the influence environment. In a situation of 'elitist politics', first, when policies are not supported by a broad public and considered unimportant by the supportive part of the public, groups will not seek exchanges in the news media environment. In that case, the involvement of additional political actors through public attention will change the existing balance of power in a way that is unfavorable to the organization. In the influence environment, however, organizations will try to suggest the policymakers that the issue is salient, when in reality it is not (astroturf ). Second, when policies favorable to certain groups are not supported by a broad public and when supportive publics (constituents) consider the policy important, groups will not signal their position to the news media. The organization nurtures its exchanges in the supporters environment, but 'will take care to avoid expanding the conflict too much to spark the opposition' (Kollman, 1998, p. 160). This is 'classical interest group politics'. Third, broad public support but low salience leads to 'latent support politics'. Interest organizations will want to exploit the favorable public opinion by increasing the salience of the issue. They need the news media to convince the potential intensely supportive public that the issue is worth acting upon.
Second, media systems vary in the extent to which the media system parallels the conflict lines of the (party) political system. This is what Hallin and Mancini (2004) label 'political parallelism'. In systems with a high level of parallelism, we find that newspapers and the news media in general are organized along political lines. This could, for instance, be via direct ownership of newspapers by political parties or via party political ties to the public broadcasting system. From the perspective of interest organizations, this matters because in systems with a high political parallelism it is less likely that the news media provide an alternative, indirect venue to political institutions. The structures of opportunities and access are very similar. In cases of low political parallelism, the media system could provide a fruitful, additional arena in which to address conflict dimensions that have not (yet) been institutionalized in the political system. In other words, the so-called 'compensation hypothesis' is more likely to be valid in systems with low levels of political parallelism, whereas the 'persistence hypothesis' should find stronger support in other systems. The country differences are in line with the distinction of 'protest regimes' and other systems of interest representation mentioned earlier. These country differences in media opportunities for interest organizations have received limited scholarly attention (perhaps with the exception of some research on social movement organizations, Kriesi et al, 1995).

Conclusion
The key conceptual benefit of the presented review is that it brings together aspects of both interest organizations (as actors) and their environment (as structures). It highlights the constraints or tensions that arise from the interaction of interest  (Kollman, 1998, pp. 155-164). organizations with other types of actors. I hope to have clarified that despite the seemingly broad range of activities and strategic opportunities of interest organizations they are nevertheless constrained in interacting with journalists, supporters and politicians. These constraints partially arise from contradictory demands from each of these actors and from factors related to, among others, organizational populations, the media system and institutional venues.
Scholars of interest organizations seem to have greater difficulty incorporating such constraints in their theories than researchers on strategies of political parties. Scholars of political parties regularly point to contradictory demands placed on the party organization in seeking votes, public office or policies (for example, Strom, 1990). So, why is this not more explicitly incorporated in research designs on interest organizations? This may be because of focus on only the public policy-related activities of organizations, consequently turning a blind eye on supporters and public opinion. Such constraints may also remain unobserved because research designs tend to allow very little (or no) variation between countries, issues or group types. How to organize the research field and individual research designs in such a way that they remain specific enough to be falsified and simultaneously reflect the complex and sometimes contradictory factors related to activities of interest organizations? I have three modest suggestions on how to deal with this dilemma.
First, as should be clear by now, the explanation of the range of activities associated with exchange relationships requires the inclusion of a lot of potential explanatory factors. Therefore, researchers should not aim to explain even more than that in a single research design. That is, related questions regarding the interest community (Why is the interest organization population as it is?) and the questions related to the exercise of influence (Why do some organizations seem successful and other not?) should be considered separately. Although separate, these aspects of interest representation or stages of the influence production process can still be related at a general level (Lowery and Brasher, 2003, pp. 16-20;Lowery and Gray, 2004). As described, the interest populations' phase provides potential explanations for differences in activities, and the nature of the activities, in turn, may help to explain variation in success of interest organizations. Following this logic, Mahoney (2008) fruitfully divides research tasks into even smaller parts of the process of influence production: for instance, she also distinguishes an argumentation and a framing stage.
Second, the relational nature of political activities should be included in research designs. It takes two to tango. In our studies, we should simultaneously include interest organizations and exchange partners such as members, journalists and policymakers. For instance, in the examination of media-oriented activities, we should not only interview interest organizations but also journalists. Only focusing on the public affairs activities of interest organizations will probably leave the factors (public opinion, newsworthiness) underlying media attention resulting from initiatives of journalists unspecified. Such a one-sided approach will produce only partial explanations of media attention to issues of interest organizations. Thus, although a natural way to narrow down the research task, one cannot limit oneself to only a single side of an exchange relationship. This also means that researchers should not fully specialize in any of the logics of exchange. For instance, consider the evaluation of lobbying practices in parliaments. Here we would specialize on the relationship between parliamentarians and representatives of interest organizations. On the basis of the usual needs of parliamentarians, we can probably produce a broad specification of the types of exchanges (political support, policy information) that are likely to occur. However, we would then study these in relative isolation from factors related to, among other things, membership or public opinion. At the minimum, when choosing such a research focus, one needs a broad idea about the relationship with news media/ public opinion (for example, Does the interest organization lobby on a 'popular' issue?) and about the nature of the membership environment (Who are members? Does the organization have competitors?). That is, as highlighted, lobbying interactions with parliament on 'popular' issues will emphasize broad public support (latent or comforting politics). Legislative lobbying by interest organizations from a competitive environment is likely to focus on policy expertise, as their political support cannot be credibly offered due to the fragmented nature of the support base. A more complete specification of lobby interactions thus requires including a broader range of factors than only those related to the venue lobbied. This is also the case for the activities related to the media and supporters. By specializing in a single type of exchange relationship, we will never fully explain why interest organizations do what they do.
Several general hypotheses regarding activities of interest organizations have been presented. Elsewhere, I have evaluated them more empirically (Berkhout, 2010;Braun et al, 2011). The review presented provides a starting point for the examination of various explanatory factors for the activities of interest organizationsthat are, in turn, crucial to understand the political influence of such activities.