Executive discretion, the latitude for executives strategic decisions, is a powerful moderator of strategic decision making. In spite of its potential contribution to strategic management studies, Hambrick and Finkelstein's (1987) socio-political model of executive discretion has received little empirical research effort. Some of the basic propositions of the model, which incorporates industry, firm and individual characteristics as determinants of discretion have not been empirically tested. The restricted research effort is partly attributable to the lack of quantitative measures for industry-level discretion. This thesis initially uses the correlation between industry-level attentional homogeneity, the similarity in foci of attention of executives in an industry, and industry-level discretion to produce 116 new values for industry-level discretion for 23 U.S. 4-digit SIC coded industries for the years 1990 to1997. Predictive validity for the new values is demonstrated using long-term debt data and annual accounts adjustment data. Theil's (1992b) industry variety measure based on information theory is modified to produce strategic variety measures that permit pan-industry comparisons. Strong support is demonstrated for a positive association between variety in long-term strategic positions and industry-level discretion. Some weak evidence suggesting large firms in low discretion industries may compete using behaviours that impact on current accounts is also identified.

This thesis examines the associations between organisational slack, that pool of actual or potential cushion of resources of an organisation, and executive discretion - the executives’ latitude for strategic action. Bourgeois and Singh (1983), George (2005), Sharfman et al. (1988) and Sharma (2000) have referred to slack as having a discretionary dimension because its ‘ease of recovery’ varies depending on where it is gained from. For the obverse of this association, slack contributes to resource availability in the task environment and therefore executive discretion (Hambrick & Finkelstein, 1987). However until now, this bi-direction association has been largely unexplored empirically. This thesis contributes to both fields by bringing them together to examine and measure aspects of these interactions. These constructs are applied to the annual reports of U.S. firms by measuring industry level discretion using content analysis of presidents’ letters to shareholders and industry average slack using financial ratios. Correlations show that industries with higher levels of slack enjoy greater industry level discretion. However the associations between slack types and industry level discretion are not uniform suggesting that the discretionary dimension of slack is influenced by the task environment and industry context. The present study replicated Keegan and Kabanoff’s (2007) method to examine slack within industries but could not extend their results to available and recoverable slack, which suggest a curvilinear relationship between potential slack and executive discretion. The limited sub-industry results offer opportunity for further research as does the idea of applying the same research question to the organisational and individual level studies of different cohorts of firms and industries. Future efforts should also improve the measurement of the slack construct.