What Determines the Structure of Pensions?
C. What Determines the Structure of Pensions?
In theories of deferred compensation, DB pensions solve a contracting problem between workers and firms Lazear 1986. For example, firms may want to deter
shirking but cannot perfectly monitor workers. Deferred pension accruals induce workers to devote optimal effort so that they do not lose their job and pension, and
perhaps further to retire at an appropriate age. Similar motives for deferred compen- sation arise if firm-specific training or hiring is costly.
While various elements of these theories have found support in explaining the use of DB pensions, they offer little insight about the increasing prevalence of DC pen-
sions. Two sets of explanations have been offered for the shift from DB to DC pensions; neither suggests that it occurred in response to changing retirement motives.
First, regulatory changes since 1974 have tightened DB funding standards, con- strained the structure of DB and DC pensions, and extended tax breaks for DC con-
tributions. Many of these changes were aimed at encouraging and preserving the accumulation of retirement wealth Clark and McDermed 1990; none were explicitly
aimed at influencing retirement ages. Second, Friedberg and Owyang 2002b argued that the value of long-term jobs has declined, thus reducing the usefulness of DB pen-
sions. The focus of their argument on changes in the nature of jobs held by prime-age workers is consistent with the more rapid change in pension structure among younger
workers; a decline in overall job tenure; and the movement of workers from jobs typ- ically covered by DB plans to jobs typically covered by DC plans.
15
If anything, the shift away from DB plans may have increased firms’ use of temporary early retire-
ment inducements aimed at older workers.
16
We recognize nonetheless that pensions and retirement may be endogenously determined. A firm’s choice of pension structure may be influenced by factors corre-
lated with the average age and retirement preferences of workers. However, we do not believe it is feasible to estimate the determinants of pension design.
17
We address con- cerns about endogenous sorting as permitted by our data. We control for observable
characteristics job tenure, firm size, industry, unionization that are related to pen- sion type; none influence the estimated effect of pension characteristics on retirement.
Also, we show that older workers with different pension types are quite similar on other key dimensions like earnings and wealth, along which one might expect observ-
able differences if workers sorted by retirement preferences or related characteristics.