The Political Model of Strikes: A New Twist

Abstract

The political model of strikes in which work stoppages are viewed as the result of worker intransigence has occupied an influential position in the strikes literature since its empirical specification by Ashenfelter and Johnson [1]. There are two basic reasons for the popularity of the model. First, it is an eminently accessible treatment in a literature of highly abstract bargaining models, most of which do not have a direct bearing on strikes. Second, until the Ashenfelter-Johnson study, little progress had been made in applying models of behavior. Drawing on Ross's [6] insight that the was essentially a political institution operating in an economic environment, Ashenfelter and Johnson sought to examine the implications for strikes of the divergent goals of the leadership and the rank-and-file membership. The basic idea is that strikes are caused by the unrealistically high wage aspirations of the latter, which are nonetheless processed by the better informed leadership so as to deflect any incipient challenge to its authority and continued tenure in office. The function of a strike is an equilibrating mechanism to square up the membership's wage expectations with what the firm may be prepared to pay. Strikes are attributed exclusively to the union side. The difficulty with the model lies precisely in its assumption that it is the side which bargains badly and hence precipitates strikes. For example, one can construct a model employing the opposite assumption after Eaton [2] and Rabinovitch and Swary [4]. Indeed, the assumption that it is the employer side which bargains badly forms the basis of our revisionist model. Conceptual issues apart, the twist is that our model is applied to the very same data set used by Farber [3] in his re-estimation of the Ashenfelter-Johnson model using data on eighty contract negotiations for ten large manufacturing firms (on which more below).' We show that the two very differently assumptioned models both yield reasonable results using the same data set. The implications of this finding are then discussed.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 7 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.