Other research supports the conclusion that above-market wage rates tend to
reduce employees’ turnover intentions or actual turnover (e.g. Guthrie 2001 ; Levine
1993 ; Shaw et al. 1998 ). In one of the few studies to link relative pay rates with
employee attitudes, Levine ( 1993 ) found that higher plant wage rates in their US
and Japanese sample improved employee outcomes on a number of measures,
including job and pay satisfaction levels, intentions to stay with the company, and
reports of willingness to ‘work harder.’ In addition to aVecting turnover, it is
perhaps unsurprising that potential employees are also attracted toWrms paying
higher wages (e.g. Barber and Bretz 2000 ; Cable and Judge 1994 ).
In sum, research suggests that higher relative wages improve the ability to both
attract and retain employees. This is again illustrated by Costco. For example, after
theWrst year of employment, Costco has a 6 percent employee turnover rate, as
compared to 21 percent at Sam’s Club (Holmes and Zellner 2004 ). Does this ability
to retain employees yield economic beneWts? Costco believes that its higher market
wage helps account for its comparative productivity advantage in terms of sales
revenue ($ 795 sales/square foot per employee vs. $ 516 at Sam’s) and operating
proWts ($ 13 , 643 per employee vs. $ 11 , 034 at Sam’s). Because of this productivity
advantage—and despite the fact that their per capita wage rates are higher—
Costco’s labor costs as a percentage of revenue are signiWcantly lower than Sam’s
Club.
The research evidence, however, is a bit more mixed. Supportive evidence for the
eYciency wage argument is found in a study of UK manufacturing plants (Wadh-
wani and Wall 1991 ), where higher relative wages were positively associated with
both output and value added per employee. However, Gerhart and Milkovich
( 1990 ) found that while use of contingency pay for managers positively inXuenced
Wrm performance (ROA), managerial pay-level position did not aVect perform-
ance. And, in Brown et al.’s ( 2003 ) study of pay practices in California hospitals,
while pay level had direct, positive impacts on organizational performance in terms
of both patient care quality and eYciency, there were non-linear, diminishing
returns to increases in market pay position, prompting the authors to assert that
‘there exist limits as to how much selection can be improved, turnover can be
reduced, and motivation can be increased with pay’ (p. 760 ).
Finally, the set of studies on pay eVects in the professional sports arena discussed
earlier is also relevant. While Bloom ( 1999 )Wnds strong negative eVects on baseball
team performance for pay dispersion, heWnds no pay-level eVects (in fact, his
bivariate correlations indicate that relative team payroll has a signiWcant,negative
correlation with teams’Wnishing positions). His results stand in contrast with those
of Frick et al. ( 2003 ) whoWnd strong positive eVects for total team pay in four
diVerent US professional leagues: baseball, basketball, hockey, and football.
These mixed results underscore the need forWrms to consider a host ofWrm-
speciWc variables that may moderate consequences associated with pay-level policy
decisions. This is highlighted by Klaas and McClendon, who applied utility analysis
remuneration: pay effects at work 357