being relatively Xat or egalitarian, with limited diVerences across levels and
positions, to those that are steep or hierarchical, displaying large diVerences across
levels and positions. With respect to pay basis, I review a limited number of studies
that have examined the use of person-based, as opposed to job-based, pay systems.
Pay dispersion. A starting point for this research stream is tournament theory,
which seeks to both explain and inform pay structures. Developed by Lazear and
Rosen ( 1981 ), tournament theory suggests that it may be eYcient (i.e. it makes
economic sense) to pay those at the top of a pay structure a wage that exceeds their
marginal contributions. Steep diVerentials will beneWt organizations if these pay
structures serve to attract a talented pool of employees and entice them to perform
at high levels in a contest for the top ‘prizes.’ A number of studies have examined
tournament theory in the context of sports, with supportive evidence provided by
studies such as Ehrenberg and Bognanno’s ( 1990 ) study of golf tournaments and
Becker and Huselid’s ( 1992 ) study of auto racing.
In a study of professional baseball teams, Bloom suggests that more egalitarian
pay structures ‘can be beneWcial forgroupperformance because they may inculcate
feelings of fairness and common purpose, foster cooperative, team-oriented
behavior, and support common goal orientations’ ( 1999 : 26 ). Bloom examines
the impact of pay dispersion on individual and team performance andWnds that
more egalitarian pay structures are associated with higher levels of individual and
team performance. Bloom’s results for baseball are consistent with several other
studies of this sport (e.g. DeBrock et al. 2004 ; Jewell and Molina 2004 ).
What about the ‘real’ world? A number of studies have looked at pay dispersion
eVects at the executive level. When compared to other nations, US companies have
often been described as (and criticized for) having pay structures that are ‘overly’
hierarchical. This suggests that tournament structures may be an American phe-
nomenon, but this is belied by international evidence (e.g. Conyon et al. 2001 ;
Eriksson 1999 ). Other tournament features that have been supported using execu-
tive pay data include greater pay dispersion associated with fewer promotions
(Leonard 1990 ), more dynamic industry environments (Bloom and Michel 2002 ;
Leonard 1990 ), greater numbers of tournament ‘participants’ (e.g. Conyon et al.
2001 ; Leonard 1990 ; Main et al. 1993 ), and higher executive turnover (Bloom and
Michel 2002 ).
Consistent with the conclusion of Gerhart and Rynes ( 2003 ), it does appear that
the bulk of the evidence from studies at this level has established the existence of
tournament-like structures within the executive ranks. While evidence is generally
supportive that these structures exist at this level, are tournament structures
‘eYcient’? Do more dispersed executive pay structures positively impactWrm
performance? The evidence in this regard is mixed (cf. Conyon et al. 2001 ; Eriksson
1999 ; Leonard 1990 ; Main et al. 1993 ).
Researchers have also examined performance outcomes associated with pay dis-
persion among workers outside of the executive ranks. These studies are generally
352 james p. guthrie