over time receive fewer and fewer job offers and may, finally, even be regarded as
“unemployable”. The resulting detachment from the labour market implies that the long-term
unemployed may exert little or no downward pressure on wage increases. Moreover, a growing
number of unsuccessful job seekers in the pool of the unemployed may reduce the speed by
which vacant jobs can be filled by suitable candidates.
When specific skills are an important aspect of the employment relation, involuntary separation
from a job may imply long waiting periods for re-employment; and when the loss of specific
skills and the associated wage premium eventually has to be accepted, specific capital no longer
provides a buffer between productivity and the value of employment elsewhere or non-
employment, so turnover from new jobs, probably associated with recurrent unemployment, may
be rapid.
Another strand of reasoning emphasises the wage-bargaining behaviour of the employed insiders
and the role of adjustment costs. For example, when unions bargain mainly on behalf of the
incumbent workforce, a temporary adverse shock to employment will tend to perpetuate itself,
because real wage demand is adapted to the now smaller number of employed insiders.
Generally speaking, shifts in the employment composition in favour of groups facing little risk of
unemployment may affect the overall bargaining stance of unions and thus reduce the wage-
moderating impact of a given rate of unemployment.
For insider effects to persist, the employed insiders must command some degree of market
power. This could stem from several sources such as training costs or statutory seniority
systems, but also from various forms of job security legislation. While the resulting reduction in
turnover may well be in the interest of both the firm and the workers, the crucial point with
regard to the persistence issue is that turnover costs render it difficult for outsiders to effectively
compete for jobs.