We carried out a computerized search of relevant scientific articles and analyzed their main findings. This work also rests on knowledge of the health care field and institutions developed in previous work (Contandriopoulos, 2004
; Contandriopoulos, Denis, & Langley, 2001
; Contandriopoulos, Denis, & Langley, & Valette, 2004b
The first step in the literature review was a computerized search of the major databases of interest, using the combination of lobby, health, and policy as keywords. When search engines accepted wildcards, lobb* was used instead of lobby. We searched the SilverPlatter WebSPIRS databases Canadian Research Index (1982 to present); Econ Lit (1969 to present); International Political Science Abstracts (1989 to present); PAIS International (1972 to present); Philosopher’s Index (1940 to present); Social Sciences Index (1983 to present); Social Work Abstracts (1977 to present); and Sociological Abstracts (1963 to present), as well as the OVID databases MEDLINE (1966 to present); CINAHL (1982 to present); and Current Contents/All Editions (1993 to present). After excluding articles in languages other than English, French, or Spanish and those not of interest, we had a set of more than a thousand articles (n = 1,071). Relevant documents were then selected by reading the abstract, or by relying on the title when no abstract was provided. This allowed us to identify a set of 248 articles on health policy lobbying.
However, given the keywords used, theoretical groundwork outside of health fields would probably not have been identified through this first step. To include such works, we relied on snowball sampling. While analyzing the corpus collected in the first step, we noted the theoretical works on which the authors relied and, when pertinent, added them to our literature review. We thereby identified 216 other documents, for a total review of 464 documents, mostly journal articles and books. This literature review, though centered on health care, is by no means exclusive to it.
This method cannot guarantee exhaustivity, but although we may have missed some potentially helpful articles, we are confident that no influential articles were overlooked. This method gives a broad view of the field of interest group representation and makes it unlikely that any significant strain of ideas was entirely omitted. The most important limitation in such an effort to survey an academic field is that there are no clear, or even logical, boundaries. For example, in the subfield of pharmaceutical policy, there are articles in political science, but also many in the health policy field, that though not directly dealing with lobbying, bring interesting and original evidence from current policy debates—such as Lexchin’s numerous articles on pharmaceutical companies’ influence (Lexchin, 1993
Organized Interests: Definitions, Origins, and Networks
As stated earlier, in health care policy making, there is a vast array of politically active groups, institutions, and organizations. Our use of the terms group
, and institution
is deliberate, because many important actors in the field are institutions such as hospital associations, unions, and regional boards. In fact, many influential “interests” in health care policy making are not groups (Salisbury, 1984
). A major hospital or a pharmaceutical company, for example, cannot be described as a “group” but nevertheless often intervenes efficiently in policy making (Hart, 2004
We consequently rely on a broadly encompassing definition of interest groups that includes both membership-based organizations as defined by Pross (1992
, pp. 3–11) and “memberless” institutions such as hospitals, private companies, and regional boards. We thus support, to some extent, Salisbury’s “imperialist” definition of an interest group as any “active unit, from the isolated individual to the most complex coalition of organizations, … that engages in interest based activity relative to the process of making public policy” (Salisbury, 1994
, p. 17). At least in the health care field, using more restrictive definitions without valid justification would clearly exclude some actors that exert significant pressure on the policy-making process. Although memberless institutions administratively linked to governments will be constrained in their choice of lobbying strategies, this does not impede their capacity to exert political pressure. Moreover, a membership-oriented definition would consider such institutions as hospital associations to be interest groups, whereas a single hospital would not be so considered, whatever its political involvement or power. Hence, the relevance of that hospital’s action would depend on whether it used its association as a political vehicle. Similarly, a good deal of the political action is carried out by individual member organizations rather than by the association itself. Hence, rather than seeing the decision to join an association as triggering a fundamental shift in the ontological political nature of the institution, we suggest it is more fruitfully analyzed as a decision to join, or not, a particular kind of coalition (Hojnacki, 1997
The main modification we would make to Salisbury’s definition is that groups be limited to structured organizations that express themselves through legitimate designated individuals (Bourdieu, 1984
). This limitation excludes latent interests, unorganized individuals, or noninstitutionalized social movements, all of which can try to influence policy on a personal basis. Even though, as Bentley (1967)
noted in The Process of Government
, there is no clear boundary to distinguish between institutionalized and noninstitutionalized groups, the distinction can still be used heuristically to identify groups that are at least institutionalized enough to express themselves in the political arena through a designated spokesperson. The potential self-designation of groups’ spokespersons, as well as their exact representativeness, remains, however, an open question (Ainsworth, 1993
; Contandriopoulos et al., 2004a
; Nelson, Heinz, Laumann, & Salisbury, 1987
). Finally, as do almost all other definitions, we distinguish between interest groups and political parties, using the rationale that, in the words of Berry, interest groups are policy maximizers
, whereas political parties are vote maximizers
Political parties are vote maximizers. To win elections they must dilute policy stands, take purposely ambiguous positions on others, and generally ignore some, so as not to offend segments of the population that they need in their coalition. Interest groups are just the opposite. They are policy maximizers, meaning that they do better in attracting members if their outlook is narrowly focused. (Berry, 1997
, p. 47)
The end point, then, is the definition of interest groups as any institutionalized organization that engages in political activity relative to the process of making or influencing policy without explicitly trying to obtain or exercise the responsibility of government. Strictly speaking, the interest groups we focus on here sometimes are not groups at all, in which case the term interest group
is inaccurate. Similarly, these interests are neither as ontologically “vested” nor as “special” as other denominations would imply. The most obvious and precise term would then be organized interests
. However, our goal not being to add to the stockpile of terms, we will use interest groups
and organized interests
interchangeably. As has been well argued by Hart (2004)
, including memberless groups as hierarchical organizations in the definition necessitates some discussion about group origin and maintenance.
Contrary to Olson’s (1971)
view, we do not consider material incentives to be the main explanatory factor for the existence of organized interests in health care politics. Many influential actors are preexisting organizations such as hospitals, public health agencies, or pharmaceutical companies, whose political capacities are—in Olson’s terms—a by-product of their broader administrative and managerial capacities. For other actors, such as unions and professional associations, it would also prove quite hard to link either their origin or their membership to material incentives. Both institutional factors (Meyer & Rowan, 1991
; Powell & DiMaggio, 1991
) and state regulation (Mawhinney, 2001
; Walker, 1983
) could probably explain their existence, with their interest group status being a by-product. But more fundamentally, in both cases, they are either memberless groups or compulsory membership groups.
Organizations conceptualized as politically active memberless groups could fall into two categories: utilitarian organizations (Clark & Wilson, 1961
) that strive to maximize material rewards (e.g., a pharmaceutical company) and not-only-utilitarian organizations that are concerned about resources but also have nonmaterial goals. In this latter category, we might find hospitals or public health agencies that though interested in obtaining enough resources from their environment to subsist or grow, also pursue other objectives, such as curing disease, relieving suffering, improving public health, and so on. This typology tallies with R. G. Evans’s (1984)
typology of for-profit, not-for-profit, and not-only-for-profit actors in health care.
However, in both cases, the incentives are not targeted at the individual level in the same way as posited by Olson (Salisbury, 1984
). For all these organizations, the political arena is just a part, sometimes a small part, of their struggle to obtain resources from their environment. All are able to draw the majority of their resources either from the market or from stable public funding, as hospitals do. Their political involvement is thus one lever—among others—for obtaining material resources or, quite often, intangible ones. For example, a hospital could fight politically to its last breath to obtain or maintain its university affiliation, even when this affiliation is unrelated to either volume of service or material resources. If one of these organizations proves unable to fulfill its stakeholders’ expectations in regard to its ability to attract resources (whether material or intangible), the executive summit of the organization will probably face sanctions, or at least pressures. However, being memberless groups, the way they redistribute benefits probably resembles more closely the organizational incentive system of Clark and Wilson (1961)
than the group-based logic of Olson (1971)
In the case of compulsory membership groups such as some physicians’ associations or unions, Olson’s subtheory of by-products applies, since, as with the previous category, the group’s existence predates the need or will for political action. However, the distinction is that in Canada, wherever there is a professional association or union, membership is compulsory. Thus, members’ dissatisfaction with redistribution of benefits or incentives would provoke their voicing these rather than their exit (Hirschman, 1970
), with the possible takeover of the union’s or association’s management. The situation for compulsory-membership associations is thus quite similar to the one prevailing in memberless ones.
Finally, in the case of noncompulsory membership groups, such as patients’ organizations, Rx&D,1
and others, restricting incentives to purely material ones remains dubious. Olson’s main objection to the inclusion of intangible benefits is epistemological. A broad conception of costs and gains as both tangible and intangible benefits produces a situation where the existence of politically active groups implies that their subjective perception of gains or costs ought to be high enough; conversely, the absence of politically active groups, notwithstanding objective gains or costs, implies that those gains are perceived as too low. Thus, the existence of politically active groups becomes a tautologically explanatory variable of the existence of their necessary conditions of existence and makes the empirical validation of Olson’s viewpoint more difficult. However, even if this criticism is logically correct, and of practical importance in understanding the necessary conditions of political activity, it is much less pertinent when our interest shifts to understanding the nature, form, or extent of a group’s political activity. Even if it is somewhat tautological to state, as does Bentley (1967)
, that “there is no group without its interest,” it remains empirically true.
In the context of health care, many groups become politically active either to obtain gains that are public goods and as such, are accessible to everybody and not only to those who have mobilized, or even to obtain gains for groups to which they do not belong. For example, groups that lobby for the accessibility of free drugs for social insurance beneficiaries are often constituted of people who are not themselves social insurance beneficiaries. More broadly, the health care sector in Canada, as in other jurisdictions, is somewhat specific, in the sense that people—whether policy makers, lobbyists, or the “public”—need to behave in accordance with the objectified fact that health is not an ordinary good. Health is generally perceived as intrinsically desirable, its accessibility part of citizens’ rights. Thus, even groups seeking additional resources for themselves—for example, blind people asking for more resources in rehabilitation—will, as long as these resources are health care related, often perceive themselves and be perceived as seeking some kind of public good. This “private” public good is public in the sense that anybody could need it and benefit from it one day. Thus, even though we do not provide such an analysis here, we think that Olson’s postulates would need to be reexamined in the context of health care in order to use a typology of groups based on the nature of their goals.
The classical corporatist model of group–state relationships posits that core interest groups will be co-opted by public agencies as insiders in the policy-making process. For example, the Canadian Medical Association (CMA) was repeatedly described as having almost an ideal corporatist relationship with government (Taylor, 1960
). In contrast, the pluralist model posits an open system in which any interest group can organize itself and enter the political arena to express its views. Notwithstanding their divergences, pluralism and corporatism both share at least the conviction that groups do not exist in a social vacuum but rather with specific relationships among themselves and with other elements of society.
There is, for example, consensus on the fact that groups with interests in a similar field will be aware of each other’s positions and strategies, as well as of the relevant government agencies’ positions (Carpenter, 2004
; Heinz, Laumann, Nelson, & Salisbury, 1990
). Unfortunately, the names given to this network of actors are tightly linked to specific theoretical standpoints (policy networks, subsystem, iron triangles, etc.). For a brief description of intergroup relationships, we rely on the definition of policy subsystems or domains by Sabatier and Jenkins-Smith (1999)
, even though, as is usually the case, it does not provide any clear boundary for the relevant domain. For example, should we conceptualize the provincial health care policy-making arena as a specific subsystem with more specific sub-subdomains such as, for example, pharmaceutical regulations? Or is the pharmaceutical regulation arena an international policy subdomain and the specific provincial arena a sub-subdomain? Although there is probably no obvious choice in the conceptualization of the level of analysis for groups’ interactions, this has implications for our capacity to generalize in specific case analyses.
In any case, in their interaction with other elements of society, the groups and public agencies inside a given policy domain can have various relationships ranging from confrontation to collaboration (M. Evans, 2001
; Salisbury, Heinz, Laumann, & Nelson, 1987
). However, there is some imprecision in the exact nature of interrelationships, because most, if not all, collaborative arrangements are indistinctly described as coalitions (Lowi, 1964
, p. 678), even if specific relationships can take many forms. For example, in the health care sector, a frequent specific form of coalition could be called the relay (Burton & Rowell, 2003
; Wysong, 1992
), an example of which could be the relationship between the pharmaceutical industry and patients’ organizations (Herxheimer, 2003
). The pharmaceutical industry can have interests in common with some patients’ organizations, such as the inclusion of a given drug in the government formulary. Rather than openly forming a coalition, these groups will often feel their interests would be better served by another arrangement. A company might help to hire, fund, and train a given patient organization’s staff, but stay away from visibly lobbying for inclusion of its drug on the formulary, which could harm its case more than help it. The patient organization would, however, be there to lobby and would be sufficiently funded and trained to do so. This is thus a relay in the exercise of pressure as well as a reciprocal exchange of commodities, that is, legitimacy for material resources and material resources for legitimacy.
Another level of imprecision can be found in the degree of institutionalization of coalitions. As we have seen, some institutions—for example, hospitals or pharmaceutical companies—are politically active both at the individual organizational level and through membership in organizational associations. The conventional wisdom of group theory is to consider the association as the unit of analysis, thus dodging the question. But if individual organizations or institutions are considered legitimate units of analysis, then associations become institutionalized forms of coalition. However, their institutionalization is not absolute: Organizations sometimes disagree with the association to which they belong and politically oppose it. This conceptualization would add to the fluidity of group population observed by Gray and Lowery (1993)
or Salisbury et al. (1987)