حفاظت، اعمال نفوذ و ساختار بازار

We look at a model of lobbying by oligopolistic industry where firms allocate resources between lobbying and internal cost-reducing activities. We ask the following questions: (i) if firms differ with respect to comparative advantage in lobbying, what is the equilibrium allocation of resources between lobbying and cost-reducing activities? (ii) Can lobbying opportunities reverse the profitability ranking among firms? (iii) Under what condition is the conventional wisdom that highly concentrated industries tend to obtain more protection valid? The answers depend on various measures of comparative advantage in lobbying and on the demand curve.

There is a presumption, dating at least to Olson’s (1965) classic analysis of collective action, that small group size is advantageous in influencing endogenous policy. The argument is that, under institutions of representative democracy, governments and candidates for political office have political-support needs that can be better satisfied by ‘cohesive’ coalitions, as these are less prone to defection and free-riding than more ‘diffuse’ coalitions. Hence, under representative democracy, ‘the small exploits the large’. In contrast, under direct democracy where voters determine outcomes, larger group size is more advantageous.1 An implication of the advantage of small group size for collective action is that more concentrated industries should (all else equal) be more successful in securing protection and/or in resisting trade liberalization. Empirical studies have however failed to find an unambiguous relation between industry concentration and policy effectiveness of an industry (see the surveys by Baldwin (1984), Hillman (1989), Potters and Sloof (1996), and also Goldberg and Maggi (1999)). Potters and Sloof (1996, pp. 417–418) summarize the diversity of the extensive empirical evidence as follows: Most scholars indeed find an increased scope for political influence with higher degrees of concentration, but there are many that find no effect or even a negative effect. Equally ambiguous are the results of the use of numbers for the free rider effect. A large number of participants to collective action is usually hypothesized to increase the free riding problem. Sometimes indeed a negative effect of numbers on influence is reported. More often, however, a positive effect is found. Hence there appears to be relatively little direct empirical support for the Olson (1965) influential theoretical study on collective action. One may therefore well wonder what is going on. In this paper we consider theoretical foundations for the source of the empirical ambiguities. There are different possible points of departure. One beginning is Stigler’s (1964) proposal that a theory of oligopoly should start by assuming collectively rational behavior, and then should proceed to investigate the costs of defection from the cooperative equilibrium. Stigler’s perspective on oligopoly provides a reasonable basis for Olson’s collective-action proposition. Smaller group size increases the probability of detection of free-riding behavior and decreases the transactions costs of organizing and monitoring contributions to collective action. More concentrated industries are expected to be more effective in influencing endogenous policy decisions. This is not however the unambivalent picture provided by the empirical evidence. The alternative non-cooperative Cournot–Nash approach adopts as a point of departure individually rational behavior. Policy influence then becomes a case of non-cooperative private provision of a public good. In the latter approach, we have well-established results for the case where consumers choose public good provision (see Cornes and Sandler (1996)). If the public good is a normal good, there are countervailing substitution and income effects on the contribution decisions of other consumers when one consumer increases his or her Nash contribution, so that a larger contribution by one consumer need not decrease the contribution of other consumers. Increasing group size and thereby adding a new prospective contributor to the public good therefore can either increase or decrease total contributions. Also, the total Nash-equilibrium contribution by consumers to provision of a public good is independent of the distribution of income among those consumers who are making positive contributions to provision of the good (Warr, 1983; Kemp, 1984; Bergstrom et al., 1986). The analogy to firms in an industry making contributions in pursuit of a collective policy objective is investigated in Hillman (1991). Consumers are replaced by owners of firms who allocate time and attention between the privately beneficial activity of monitoring their firms’ production activities2 and the public-good benefit of persuading policy makers to implement policies that benefit the entire industry. Firm owners have different comparative advantages in lobbying for protection.3 Results are obtained that are analogous to the consumer outcome: redistribution of aggregate industry profits among a given number of firms in the industry, as implied by a change in the size distribution of firms or industry concentration, need not change the aggregate Nash contribution of resources by firms in the industry to the collective benefit of influencing policy. Prospective neutralities are therefore introduced into the relation between industry concentration and the effectiveness of the collective pursuit of policies beneficial to the industry; that is, changes in concentration as measured by the distribution of profits among a given number of firms can leave unaffected the political influence of the industry as measured by the total resources allocated by the industry to policy persuasion. When group size changes through an increase in the number of firms, the neutralities appear if managerial time and attention available for allocation between productive activity and seeking political influence is an industry-specific input, but not if such inputs are intersectorally mobile. The model in Hillman (1991) views the domestic industry as confronting a competitive world market. International prices of import-competing output are thus exogenously determined, and the domestic price is given by the world price plus the protection provided as a consequence of firms’ contributions to lobbying efforts. The strategic interdependence amongst firms is thus only with respect to contributions to influencing policy, and not with respect to competition in the product market. This permits the industry seeking protection (or resisting liberalization) to be placed within the broader context of a competitive small-country model of international trade. In this paper we consider the relation between industry concentration and policy effectiveness in an internationally oligopolistic industry rather than an internationally competitive industry. As in Hillman (1991), the firms seeking protection are heterogeneous (see also Long and Soubeyran (1996)), and trade policy is endogenously responsive to the total resources contributed by domestic firms to influencing policy. In our model, the resources that an oligopolist deploys for lobbying has an impact on internal cost structure. We consider firms as facing a resource constraint, or as facing a schedule of rising marginal cost of funds that are to be allocated between political activities and internal cost-reducing activities. Because of these factors, as well as the oligopolistic market structure and the consequent endogeneity of domestic price, contributions by firms to influencing policy no longer have the characteristics of contributions to a pure public good. We show how in these circumstances the industry equilibrium is influenced by the properties of the lobbying technology and the domestic demand function, and we establish how an index of concentration is related to effectiveness of collective action of the industry. The specific questions that we address are: (i) with firms differing in comparative advantage in lobbying, what are the characteristics of the equilibrium allocations by firms between privately profitable monitoring and collectively beneficial lobbying activities? (ii) Can the ranking of firms’ profitability be reversed by the introduction of opportunities for lobbying to influence policy decisions? And most basically: (iii) what can be said about the conventional wisdom that more concentrated industries might be expected to obtain more protection? The model is set out in Section 2. We consider the outcomes when lobbying by firms is non-cooperative and cooperative in 3 and 4, respectively. The final section summarizes the conclusions. Before proceeding with the model, we note that our endogenous-policy specification is general in not presupposing any one particular mechanism that translates lobbying inputs into endogenous policy outcomes.4 We simply assume that an increase in the resources available to the industry to influence policy enhances the lobbying effectiveness of the industry. The model is in principle consistent with an underlying political-support function maximized by an incumbent government (Hillman, 1982) or influence over candidates’ trade policy platforms in the context of political competition (Hillman and Ursprung, 1988; Mayer, 1998). In neither type of specification, political support maximized by an incumbent policy maker or political competition among candidates who make policy pronouncement to seek political support, do we find an investigation of the collective-action incentives associated with industry concentration with which we are concerned5 (see however Hillman (1991)). For example, in the micro-foundations for political support proposed by Grossman and Helpman (1994), either an industry has been successful in perfectly internalizing collective action problems to permit collectively optimal political behavior, or otherwise the industry is not at all politically active.6 Hence, in Grossman–Helpman, the issue of the market structure of the industry, and the consequences for collective action in responding to the policy maker’s readiness to ‘sell protection’, do not at all arise. In models where trade policy is endogenously determined as the equilibrium outcome of political competition (as in Hillman and Ursprung (1988)), market structure implicitly affects the competing candidates’ policy platforms, but in a rather simple way because of the homogeneity of firms; the political competition models can in principle address the issue of the relation between concentration and effectiveness of policy influence, but only in the sense of measurement of industry concentration in terms of the number of identical firms composing the industry.7 Since in practice firms in an industry are not identical, a model of collective action to influence policy requires acknowledging differences among individual firms. This we do in the model that follows.

We have shown that in an asymmetric oligopoly where domestic firms allocate entrepreneurial time between lobbying for protection and internal control (monitoring), the availability of lobbying opportunities in general has differential effects on the profits of individual firms. In fact, under non-cooperative lobbying, the ranking of profits is reversed when lobbying opportunities are available, if the monitoring technology exhibits increasing returns, or if the demand curve is locally convex. The reversal is attributable to free riding in a non-cooperative equilibrium. In the cooperative lobbying case, by definition there is no free riding. Our model also lends only limited support to the conventional wisdom that industries with greater concentration tend to obtain more protection. In our model, firms lobby for quantitative import restrictions, such as an aggregate quota. Similar results can also be obtained in a model where firms lobby for tariff protection, see Hillman et al. (2000). In this paper, whether firms cooperate or not is taken as exogenous. Our results on profit reversal point however in the direction of a theory of endogenous coalition formation in the lobbying game. Such a theory would have aspects similar to the theory of endogenous vertical integration.10