Miller, T. and Del Carmen Triana, M. (2009) explore a very significant research on boards of directors, they examine how board Racial and Gender diversity relates to firm performance and its impact on performance, through the two intervening variables: Innovation and Reputation, which are considered to be key predictors of firm performance. The writers also discuss why board diversity should influence innovation, reputation and hence firm performance.
Theoretical contribution has been made to corporate governance and diversity literature "â€¦by analyzing board diversity within the framework of two major theories: behavioural theory of the firm and signalling theory" (ibid.:756) The said literature has been extended by the discovery of the two intervening variables that aim to eliminate the complexity of the board diversity - firm performance relationship.
Behavioural theory suggests that board's decision will be innovative depending on the level of comprehensiveness and evaluation of the available information during the decision making process. This theory of the firm has been applied to show the relationship between board diversity and innovation while Signalling theory suggests that firm's reputation in public is obtained by the use of visible signals. Miller and Triana (2009) build up on the previous research by that uses this theory to describe how board's characteristics influence firm's reputation.
Quantitative research method has been used with strong positivist foundation as it contains an element of both Deductivism and Inductivism (Bryman and Bell, 2007). This is evident where the writers, after deducing the hypotheses resulting from the two theories and their mediating effects, find inconsistency that lead to further research (feed-back loop) For example "...the inconsistency in findings may be attributed to a missing moderator [variable] which represents how much influence the female board members actually have on the board." (Miller and Triana (2009) - emphasis added)
The research design exhibits characteristics of a cross-sectional design (Bryman and Bell, 2007). This is apparent as more than one variable was examined, pattern of association showed the relationship between variables was examined but there was no certainty denoting a causal relationship hence, the internal validity was weak. (ibid.:55-56); demographic data was obtained simultaneously by a stratified probability sampling method based on a number of criteria from Fortune 500 firms (Miller and Triana, 2009), this indicates replication in a difference setting is possible.
Miller and Triana (2009) translated concepts into some reliable measures. Board Diversity was measured by degree of heterogeneity using Blau's Index and Proportional measure of women and racial minorities within each board; Innovation was measured by Research & Development (R&D) Intensity; Residual reputation was a measure for reputation and performance was measured using Return on Investment (ROI) and Return on Sales (ROS). (ibid.)
SUMMARY OF RESULTS
Based on the Behavioural theory, two hypotheses were tested predicting a positive relationship between board gender and racial diversity and Innovation; and two hypotheses from the Signalling theory predicting a positive relationship between gender and racial diversity and Reputation.
Multiple regression analysis was done whereby the independent variables Innovation and Reputation were regressed on the control variables then on board diversity using Ordinary Least Squares (OLS) regression method. In order to test the mediating effect "procedure outlined by Baron and Kenny (1986) and Judd and Kenny (1981)" (ibid.:769) was used.
Both racial diversity measures were positively related and statistically significant to firm performance. Innovation and reputation were statistically significant with both racial diversity measures but not with gender diversity measures, However, gender diversity was related to Innovation but not to reputation, to explain this, the writers believe that female directors may not hold powerful position on boards compared to males, thus portraying a weaker signal and they are rarely involved in activities that improve firm's reputation. Also because of globalization of the business, more emphasis has been placed on racial diversity compared to gender diversity. (ibid.:769-776).
CRITIQUE AND CONCLUSSION
Miller and Triana (2009) have certainly done a praiseworthy effort by being the first in probing the 'mediators' of board diversity - firm performance relationship which is a very interesting topic at present especially for women and racial minorities who encounter the 'glass ceiling'. The results of this study were unsuccessful in answering the questions due to a number of limitations, leading to further research.
Firstly, obtaining firm's data only from Fortune 500 implies external validity was weak as their findings cannot be generalized to companies worldwide. Using Forbes Global 2000 or Fortune Global 500 would be appropriate.
As far as gender diversity statistics is concerned, Igbaria (1995) draws attention on the importance of considering information from different occupational groups and industry-wise data.
Additionally, the measure of Innovation (R&D intensity) may not capture innovation outcomes for single-industry studies hence, the author suggests more fine grained measures in future studies.
This article can be criticized for considering a very generic definition of Innovation as "strategies that provide new opportunities for firm to create products or services" (ibid.:756).
Crossan and Apaydin (2010) in their recent research define Innovation as 'process' and 'outcome'.
"Lack of clarity in separation of these two facets of innovation may be intrinsically problematic. This problem is compounded when innovation outcomes are confused with [processes like] market performance..." (Sood and Tellis (2005) cited in ibid.)
Another limitation was the use of archival sources data which is an unobtrusive measure. No primary data was collected and hence, the ecological validity was weak. To overcome this, triangulation approach by including qualitative method such as a follow-up Interview and questionnaire could have been used to cross check the findings.
The meditational model by Baron and Kenny (1986) has been criticized too, as meeting all four steps does not conclusively establish that mediation has occurred. (Kenny, 2009)
Moreover, only accounting-based performance (in terms of ROI and ROS) was measured in this study, this challenges the term 'firm-performance' that also applies to new product/market performance or service level performance. Having said that, several innovation researchers cited by Crossan and Apaydin (2010) have used different measures of performance such as, Earnings before interest, taxes, depreciation and amortization (EBITDA), Return on Assets and market share being among the few.
Ultimately, this study has raised some interesting findings for management researchers such as Investigation of other mediating processes like 'ability to provide financial resources' and finding 'proxies for social capital such as committee memberships' (Miller and Trianna, 2009) In my opinion other possible mediators could also be explored, such as firm's culture and Leadership style, knowledge management, firm's network structure and tie intensity, in order to understand board diversity - firm performance relationship.