Voting on Contracts in the Auto Industry

W E welcome the opportunity to discuss 'V Tthe similarities and differences between our paper and the paper by Cappelli and Sterling (CS and second, the monopoly and efficient contract models of union behavior that are popular in the literature are seriously in error because they assume a homogeneity of interests on the part of the rank and file. As both of these papers show, at least in the UAW no such homogeneity of interests exists and, thus, it is impossible to derive a well-defined union objective function as assumed in these models. The papers are also broadly in agreement on two other points. One is that a significant proportion of the union membership must feel that their jobs are threatened before a union will seriously consider a wage concession. This conclusion admittedly is not startling, but, on the other hand, no one has ever provided hard evidence on this subject before. The other important point of agreement between the two papers is that the labor relations climate in the plant has a significant impact on the willingness of union members to accept a wage concession. Not surprisingly, workers are less likely to agree to a wage cut the more adversarial and antagonistic is the relationship between the union and the company. There are also several points on which the two papers disagree. For example, perhaps the key hypothesis advanced by C&S is that the larger the number of layoffs in a plant, the higher will be the Yes vote for a concession. This hypothesis is not so much incorrect as seriously incomplete. As detailed in the Easterbrook article referenced in our paper, in a number of cases unions have let a plant close rather than accept a wage concession. In those cases, C&S's model would predict that the union membership would overwhelmingly vote for a concession (since all workers are threatened with layoff), when in fact the workers voted heavily against it. The reason for this anomaly is that C&S only consider in their model the costs to workers of rejecting a concession and entirely neglect the benefits (if any) to workers from accepting a concession. This omission is a serious one, as is pointed out by both our model and our empirical results. The median voter model developed in our paper predicts that the larger the number of layoffs in a plant, the larger will be the Yes vote for a concession-but only if the workers believe the concession

Adversarial system · Auto industry · Automotive industry · Business · Economics · Industrial organization · Labour economics · Law and economics · Microeconomics · Monopoly · Negotiation · Point (geometry) · Political science · Politics · Rank (graph theory) · Voting · Wage · Engineering · Labor Movements and Unions · Law · Law, Economics, and Judicial Systems · Mathematics · Merger and Competition Analysis · Transport and Economic Policies