Based within the 'upper echelons' tradition, the starting premise for this thesis is that
demographic attributes such as age, functional background, educational attainment,
gender, and tenure, influence the decisions made by top management teams (TMTs)
(Pfeffer, 1983; Hambrick & Mason, 1984). Unlike most studies, which use public
archival data, artificial teams, or retrospective interviews with a couple of selected
senior executives, this research design (which is unprecedented in the TMT
literature), investigated the decision making processes, in real time, of 23 authentic
and fully functioning TMTs in the UK manufacturing sector using a state-of-the-art
business simulation.
From a concentrated literature review which focused exclusively on TMTs,
and disentangled the constructs of dissimilarzfy (individual level differences) and
diversity (team level differences), a series of propositions were established. These
hypothesized that demographic variation would lead to cognitive variation, that both
these types of variation would influence team processes, which in turn would affect
decision belief.
Despite the meticulous precision with which the constructs were measured in
this research, and even with the application of sophisticated multi-level modeling
techniques, only limited and sporadic support was observed for these predictions.
Although there were slightly more findings than one would expect by chance alone
(27 from a possible 177), these tended to be isolated and formed no clear pattern.
Moreover, when one went beyond tests of simple statistical significance and
reviewed effect sizes, all 27 results were tiny. The conclusion of this research is that
demographic attributes are not nearly as influential in real TMTs as 'upper echelons'
theory (Hambrick & Mason, 1984) supposes.
It is argued that the lack of convincing results is due to over-riding and
inherent social factors in authentic TMTs, so that individual demographic differences
cease to be novel or important during strategic decision-making discussions. The
practical, theoretical and methodological implications of retaining the global null
hypothesis are discussed in the final chapters.