‘this lobbying function should be interpreted as derived from given political behaviour and institutions, such as the desire o f politicians to maximize their probability o f reelection’ (p. 247), and later on in footnote 10 they write, ‘the government's desire to maximize social welfare is consistent with its willingness to grant tariff protection, in that the latter can represent its reaction to distributive equity whereas the former corresponds to allocative efficiency’ (p. 257). If the government is prepared to grant protection on equity grounds, then to invoke a lobbying function' and a reference to Brock and Magee (1978) is not justified. The social welfare function used by the government should be sufficient enough to generate the optimal tariff rate. On the other hand, if the government is assumed to maximize political support while granting tariff protection to the injured sector, then to claim that it represents the
government's reaction to distributive equity is not justified. Thus, the position o f Feenstra and Bhagwati is not easily understandable.
There are situations in which the protective instruments are not equivalent. For example, Warr and Parmenter (1986) have shown that in a situation with labour market disequilibrium, protection awarded through government procurement policies may dominate tariffs in terms of maintaining employment elsewhere in the economy, in terms of an increase in employment in the protected sector, or in terms of maintaining the total volume of trade.
How would a self-interested government choose between the tariff and the competitively auctioned quota as a protective instrument, since a relative-price protection to a ‘declining’ industry can be provided in either way? This problem was studied by Cassing and Hillman (1985) in a partial equilibrium setting, but assuming that the industry concerned has a monopoly power in the domestic market.
Cassing and Hillman have shown that the superiority of either instrument depends on the value attached by the government on the tariff revenue. If the
government does not value the tariff revenue, and is concerned only with its political support, then tariffs dominate quotas, whereas if the government also values tariff revenue, then the tariff no longer unambiguously dominates the quota as an optimal instrument.
The irrelevance of the welfare comparison of various instruments, when the level of ‘distortion’ is endogenously determined through the political process, was noted by various authors (for example, Rodrik, 1986). Skirting this limitation, however, Rodrik concluded that tariffs could be welfare superior to production subsidies if the protected sector contains a sufficiently large number of firms. His main argument is that under the tariff regime the free rider problem will lead to ‘under-demand’ of protection than relative to the subsidy regime and therefore the level of distortion under the tariff regime will be lower. A Similar conclusion can also be found in Hall and Nelson (1992).
Lloyd and Falvey (1986) studied the choice of policy instrument when there is uncertainty in the international terms of trade using a political economy approach. They have shown that, for a given distribution of the international terms of trade, the
distribution of the domestic relative price depends upon the nature of the protective instrument employed to provide protection. Given that all interest groups are risk averse the preference orderings of the protective instruments will differ across the interest groups and the choice of a particular instrument therefore determined by the domestic process of conflict resolution. In particular, they argue that such a choice would be determined by the relative political power of the interest groups.
Lloyd and Falvey’s study indicates that any of the numerous protective
instruments could be observed in place as an outcome of the political process depending on which of the conditions that guarantee the equivalence of the protective instruments is violated. In our study, we will ignore this issue by assuming that the conditions of equivalence between the protective instruments are satisfied.
2.5 Implications to the Modelling Strategy
We have reviewed the class of endogenous tariff literature which assumes that government is motivated by public-interest and the class of endogenous tariff literature which assumes that government is motivated by its own self-interest. As far as the behaviour of the tariff rate is concerned the predictions of both approaches are similar. Both approaches predict a compensating nature of tariff changes. However, they imply different strategies in modelling the endogenous process of tariff determination.
If one follows the public-interest approach in modelling government behaviour, then the implication of this choice to our modelling problem is that a social welfare function has to be specified somehow and the condition(s) that will be satisfied at the maximum of the social welfare function has to be included in the system of equations describing the general equilibrium of the economic sphere. A solution of this
augmented system will yield the welfare maximizing tariff rate in general equilibrium. However, there are conceptual problems in adopting this approach which need to be solved before such a model can be made operational. In particular, the existence and uniqueness of a social welfare function is a very real problem.11 Moreover, to guarantee that the social welfare function is maximized through policy choices, it is necessary to assume that the policy makers or the politicians are benevolent and they do not pursue their own self-interest while making policy decisions at the cost of the society. Similar problems arise with the hybrid approach as well. Such problems, however, do not arise if we view the government or the politicians as a class of self- interested agents. In this study, therefore, we will assume that the government is guided by its own self-interest - the interest of remaining in power.
This assumption leads the present study into the class of endogenous tariff literature that adopts a political economy approach to tariff formation. On the basis of the experiences of the previous studies we may draw the following guidelines to model the endogenous process of tariff determination.
The first issue is related to the stylization of political economy. Several ways of stylizing the political and economic spheres exist in the literature. However, this study selects a particular stylization of these two spheres based on the following
consideration.
Most of the previous studies have agreed on the conclusion that domestic tariffs respond to international terms of trade shocks. Since international market condition may change frequently for different exogenous reasons, it is natural to expect that domestic tariff rates are being reviewed continuously. Therefore, this study explicitly assumes that the marginal changes (adjustments) in domestic tariff rates are essentially short-run phenomenon.
Note, however, that this assumption does not necessarily imply a compensating nature of tariff changes. Tariffs may change in either direction in response to a shock. It simply implies that tariff changes do not involve long-run commitments on the part of the policy maker, and because of time inconsistency, the private sector cannot base its optimal decisions involving long-run commitments, including the capacity adjustments in the production sector, on the levels of the existing tariff rates.
This means that the adjustments in the economic markets triggered by tariff changes will be driven by short-run economic interests, and the economic consequences of tariff changes to the private sector are best described by a specific-factor model rather than a model that assumes perfect factor mobility in the economic sphere of the political economy. Because of this reason, in contrast to MBY, this study will employ a specific-factor model to stylize the workings of the economic sphere of the political economy.
On the demand side of the political market this study will allow the players to spend resources in order to obtain favourable tariff policies. In this respect, as in previous studies, it will be assumed that only the owners of the specific factors behave strategically, and the rest behave nonstrategically.12
Whether the country is in a pluralistic or in a dictatorial regime, it will be assumed that the government wants to maximize political support while making policy decisions. The relevant description of the political market will have a ‘certain’, not an
‘expected’ government of two or more parties, in place. In contrast to MBY, who studied the politico-economic system based on an expectational equilibrium, this study will be based on the actual (certain) outcomes of the politico-economic process. Thus,
12 See for example Findlay and Wellisz (1 9 8 2 ,1 9 8 3 ,1 9 8 4 ), Rodrik (1986), Yeldan and Roe