Why do labor contracts offer guaranteed annual wage increments?

• Author: Bulkley, George

1. Introduction

   An interesting feature of wage contracts for tenured academic staff in United Kingdom universities is that they formally embody the principal that wages should increase with seniority. Individuals are hired at, or near, the bottom of a wage scale that has seventeen different wage levels each separated by approximately 5%. Once hired, lecturers are guaranteed the right to move at least one step up this pay ladder each year.(1) Of course this whole scale is renegotiated annually, resulting in additional wage increases, superimposed on the automatic increments. Many other U.K. public sector labor contracts employ a similar wage ladder, for example those of civil servants, local government employees, school teachers and health service workers. Although employment of these groups is very secure, contracts of university lecturers are particularly interesting because tenure guarantees to the individual, over some range, both increasing wages and future employment. Thus, although the evidence for seniority wages from econometric studies has been mixed [1; 2], we have in these contracts a clear cut case to explain, and one with the additional wrinkle of guaranteed future employment.

   In this paper we adopt an explicitly multi-period framework in order to investigate guaranteed wage increments over time. A critical feature of the model is the assumption that consumption is allocated across periods according to the life-cycle hypothesis. Multi-period contracts are therefore evaluated by the lifetime utility which they yield. It should be emphasized that even for contracts where wage increments are guaranteed, there still remains residual uncertainty about future earnings since the possibility exists that a worker may receive a better outside offer. In section II this uncertainty is motivated by incomplete search and in section III by the fact that ability only becomes known after the first period of employment. These essentially represent different interpretations of the same model. The analysis of section II assumes that all workers are offered a tenure contract and it is shown that seniority wages arise naturally in such a framework. This in followed in section III by a joint analysis of seniority wages and the firm's decision whether to offer a tenure contract. In this case, seniority wages can again arise and, importantly, tenure contracts can actually be more profitable for the firm than contracts with no tenure. Two reasons for tenure emerge in this model. It acts as insurance for workers against the income risk they face and it allows employers to benefit from the tie-in caused by the existence of relocation costs for workers who choose to switch employer.

   There is no pre-existing rationalization of seniority wages that is consistent with the stylized facts we wish to explain. The conventional explanation of increasing wages is human capital theory [3] and this could in principle also explain guaranteed future employment at higher wages if human capital were acquired at a common and predictable rate for all workers in each grade. However, this does not seem a reasonable explanation in the context of university employment, where the vast majority of human capital is acquired early in career and, as far as research output is concerned, most likely declines over time. There are a number of alternative theories of seniority wages that build on the assumption of asymmetric information rather than human capital. These include moral hazard [10], adverse selection [15], and uncertainty on the part of firms about workers' productivity [7]. However, none of these is consistent with a precommitment to both higher future wages and guaranteed employment. For example, one explanation for wage/experience scales that rise faster than productivity is that workers' efforts cannot be directly controlled, or measured and rewarded, so a worker earns less than his marginal product when young, and correspondingly more when old, in order to discourage shirking. This is proposed by Lazear [10] and the idea is developed further in Malcolmson [11]. However in Lazear's model it is essential that shirking workers, if caught, be sacked, and in Malcolmson's that they are not promoted, so this cannot explain the commitment to increasing wages for tenured academics.

   Kuhn [9] develops an interesting model of why trade unions might choose such a pay ladder, based on the idea that seniority wages may be an effective way for a union to extract rents that accrue to the firm's owners, analogously to the use of nonlinear price schedules in product markets. However, seniority is assigned by the union rather than earnt over time and, as the analysis is essentially single period, there is nothing explicit in this model which implies automatic pay rises over time for individual workers and employment is in no way guaranteed. Indeed uncertain employment is at the heart of this model.

   The remainder of the paper is organized as follows. Section II studies the model under uncertainty about future wage offers. The optimality of seniority wages for a firm offering a tenure contract is demonstrated. The issue of whether a firm would ever wish to offer a tenure contract is integrated with the issue of seniority wages in section III. Conclusions are given in section IV.

2. Contracts with Tenure

   In this section we assume a two-period tenure contract is granted to new hires and study the characteristics of the associated wage profile. The offer of tenure contracts has been justified in a number of ways in the literature. The most obvious approach has been to treat it as a device designed to protect academic freedom and allow researchers to pursue ideas that may be in conflict with the beliefs of the employer. An alternative suggestion is that tenure has evolved as a social custom [5] that reflects the collegiate roots of universities. In the economics literature the argument that tenure exists because of the inherent risk involved in following an academic career [6; 8] has found some favor. The basis of the argument is that the training required for academic work is long: from entering as an undergraduate to completing a doctorate takes a minimum of six years. Even after this training the financial rewards are not great but, more importantly, the level of performance in the job, in terms of the rate and quality of publication, is not fully realized until several more years have passed. A final argument for tenure has been provided by Carmichael [4] on the basis that it provides the security needed for those already in post to allow them to appoint new hires that will eventually be competitors with them for the limited funds of the university.

   There are two key features of the models developed here. First, we note that a worker who is offered a tenure contract may nevertheless be exposed to future income uncertainty since he can quit if he receives a better outside wage offer at the beginning of the second period. Secondly, we assume that the lifetime expected income is optimally allocated between the two periods, subject to any capital markets constraints. The implication of this assumption is that the supply of labor in response to any particular wage offer will depend on the discounted value of lifetime expected utility which is implied by the contract.

   The model is studied in the context of two formulations of the labor market. In the first, the firm is assumed to have some monopsony power, facing an upward sloping labor supply schedule (as a function of expected utility). This allows us to examine how Kuhn's idea that seniority wages may be viewed as an optimal price schedule might work in this model. The second assumes the labor market is competitive.

   The Labor Market

   Assume a labor market characterized by a wage dispersion, which is supported by imperfect information and costly search. Persistence of such a wage dispersion in equilibrium may be explained by heterogeneous firm marginal revenue product schedules, see Reinganum [14]. There are many identical workers who live for two periods and they supply inelastically one unit of homogeneous labor per period. Workers in their first period (the "young") are observationally distinct from those in their second period (the "old"). At the beginning of each period, workers have the opportunity to costlessly search an exogeneously determined number of firms, S, after which further search becomes prohibitively expensive. Clearly S must be small, relative to the number of firms, to support...