This paper analyzes a matching model in which labor market participants use temporary employment as a waiting station between searches. Searchers entering the market see all available options. The best match, however, may not be particularly productive. Since all currently available traders are known, immediate search is not worthwhile. Over time, turnover replenishes the stock of potential traders and poor matches eventually find it profitable to search again. Searchers therefore take the best available match and simultaneously formulate if and when to look again. The best matches become indefinite; lower quality matches coexist as temporary employment. This duration increases with match quality and declines as matching improves.
Many employment arrangements are temporary.1 Temporary employment spells can, of
course, occur for a variety of reasons,2 but recent evidence from the US and from the UK establishes that such arrangements play an important role in the job matching process.3 Many workers as well as many firms use short-term employment as a stepping stone on the way to finding better, more stable arrangements, primarily with other partners. (Polivka, 1996; Houseman 1999, 2001;
Booth et al., 2002; Addison and Surfield, 2006.)
Although temporary employment is closely linked with on-the-job search, conventional
matching models are ill-equipped to account for such interim arrangements. It appears that
employers and firms prefer to stay temporarily matched while waiting for conditions to improve.
After some time has passed, they return to the market seeking a more suitable match.
The standard random search set-up cannot capture this phenomenon since steady-state matching
conditions do not meaningfully change over time. Moreover, as discussed below, the standard
model is also unable to account for some empirical regularities associated with job search and
temporary relationships. To overcome these limitations, this paper adopts an alternative but plausible approach to matching frictions which leads to temporary arrangements. Suppose that when either a firm or a worker goes on the market in search of a partner, the searcher immediately becomes fully informed about the current stock of differentiated opportunities available. Employment agencies along with other networks provide details of all existing potential partners.4 If a highly desirable match is found in the stock of potential partners, further search offers little prospect of improving matters and the relationship becomes indefinite. Searchers, however, may not especially like what they find among the available stock. If a poor pool of currently available candidates exists, searchers might reasonably expect to find preferred partners in the future as new, potentially better
options flow into the market. Given that it takes time for turnover to replenish the market and
given that visiting the market is a costly activity, searchers will not return to the market straightaway. Instead, traders can make the most of the situation by temporarily matching with the best candidate at hand. After some duration that depends on the quality of the current match, they return to the market to see if anything better has turned up. Firms as well as workers can use employment as a holding station between search activity.
Many employment arrangements are temporary.1 Temporary employment spells can, of
course, occur for a variety of reasons,2 but recent evidence from the US and from the UK establishes that such arrangements play an important role in the job matching process.3 Many workersas well as many firms use short-term employment as a stepping stone on the way to finding better, more stable arrangements, primarily with other partners. (Polivka, 1996; Houseman 1999, 2001;
Booth et al., 2002; Addison and Surfield, 2006.)
Although temporary employment is closely linked with on-the-job search, conventional
matching models are ill-equipped to account for such interim arrangements. It appears that
employers and firms prefer to stay temporarily matched while waiting for conditions to improve.
After some time has passed, they return to the market seeking a more suitable match.
The standard random search set-up cannot capture this phenomenon since steady-state matching conditions do not meaningfully change over time. Moreover, as discussed below, the standard model is also unable to account for some empirical regularities associated with job search and temporary relationships. To overcome these limitations, this paper adopts an alternative but plausible approach to matching frictions which leads to temporary arrangements. Suppose that when either a firm ora worker goes on the market in search of a partner, the searcher immediately becomes fully informed about the current stock of differentiated opportunities available. Employment agencies along with other networks provide details of all existing potential partners.4 If a highly desirable
match is found in the stock of potential partners, further search offers little prospect of improving
matters and the relationship becomes indefinite. Searchers, however, may not especially like
what they find among the available stock. If a poor pool of currently available candidates exists,
searchers might reasonably expect to find preferred partners in the future as new, potentially better
options flow into the market. Given that it takes time for turnover to replenish the market and
given that visiting the market is a costly activity, searchers will not return to the market straightaway.
Instead, traders can make the most of the situation by temporarily matching with the best
candidate at hand. After some duration that depends on the quality of the current match, they
return to the market to see if anything better has turned up. Firms as well as workers can use
employment as a holding station between search activity.This paper presents a model in which the firm confronts a world with constant marketplace
flows yet due to worker heterogeneity and the mechanics of the job center, decision making is
duration dependent. Matching frictions cause firms to in effect adopt an s-S inventory strategy
by limiting job durations with less productive workers during which job candidates accumulate
in the marketplace.
A firm entering the labor market with a job opportunity observes the entire stock of current job
searchers in the labor market and selects the best match among qualified workers. At the same
time, the firm also formulates, conditional on the quality of its worker, if and when it will search
again, that is the duration of employment. After seeing all of the currently available workers,
immediate search is not worthwhile. Over time, however, new workers gradually enter the labor
market and replenish the stock of potential employees.
As this process proceeds and the unseen inventory of potential workers in the job queue grows,
the opportunity for more productive worker-firm pairings improves, thereby increasing the return
to search. Eventually firms with workers below a reservation productivity (temporary or limitedduration employees) find it profitable to re-advertise in the labor market. If the firm finds a bettern atch, the buyer terminates the existing partnership and begins with the improved arrangement. If the trader does not find a better partner, it returns to the current partner and starts waiting for regeneration again. Short-te xist with workers in more stable position. The better the worker, the longer the duration before search recurs. Eventually, a threshold quality is reached. Above this threshold employment becomes stable without the prospect of further search. Turnover likewise
becomes deterministic rather than being primarily caused by exogenous, random shocks. Over
time, a particular job will experience longer and more profitable pairings. In other words, the
duration of employment increases with worker quality.
By relabeling firms as workers and workers as firms, the analysis here can be adapted to
address issues of worker search.Workers may accept low wage jobs intending to search for better offers at some point in the future. Unlike stationary on-the-job search in which search intensity is constant, workers do not immediately start looking around. The accepted offer is currently the best available opportunity so an instant return to the market is not worthwhile. However, given new and potentially better jobs appear while engaged at work, the worker plans a future return to the job market expecting a better opportunity the next time search occurs. This logic implies that individual earnings rise while turnover slows over an individual’s career, a well established empirical regularity.