Given the importance of exploration in a firm’s overall innovation program, scholarshave sought to understand organizational factors that give rise to exploration-oriented innovations. We propose theory and empirical evidence that relates firms’ use of financial incentives to their exploratory innovation performance. We expect that a larger proportion of long-term incentives in R&D employee compensation should be positively associated with the creation of exploratory innovation in a firm. In addition, we propose that a higher level of horizontal pay dispersion is negatively associated with the creation of exploratory innovation. We examine innovations reflected in the patents of a unique six-year, unbalanced panel dataset of 94 high-technology firms in the U.S. Empirical results confirm that firms with high level of horizontal pay dispersion have less exploratory patent innovations. However, surprisingly, firms that pay their R&D employees a higher proportion of long-term financial incentives in total compensation have lower level of exploratory innovation. This implies the possibility that popular longterm incentive plans in high-technology sectors (e.g., stock option plans) have failed to achieve their intended goals in practice. We discuss factors that might moderate the negative impact of long-term incentives on exploratory innovation.

Given the importance of exploration in a firm’s overall innovation program, scholars have sought to understand organizational factors that give rise to exploration-oriented innovations. We propose theory and empirical evidence that relates firms’ use of financial incentives to their exploratory innovation performance. We expect that a larger proportion of long-term incentives in R&D employee compensation should be positively associated with the creation of exploratory innovation in a firm. In addition, we propose that a higher level of horizontal pay dispersion is negatively associated with the creation of exploratory innovation. We examine innovations reflected in the patents of a unique six-year, unbalanced panel dataset of 94 high-technology firms in the U.S. Empirical results confirm that firms with high level of horizontal pay dispersion have less exploratory patent innovations. However, surprisingly, firms that pay their R&D employees a higher proportion of long-term financial incentives in total compensation have lower level of exploratory innovation. This implies the possibility that popular longterm incentive plans in high-technology sectors (e.g., stock option plans) have failed to achieve their intended goals in practice. We discuss factors that might moderate the negative impact of long-term incentives on exploratory innovation.

This dissertation examines the relationship between organizational identity and the formation and performance implications of interorganizational alliances. The first study investigates the effect of an organization's identity on its initial alliance portfolio formation, addressing how becoming comprehensible through organizational identity is a fundamental step in order for a new organization to be accepted by the market. Through different categorizations, some new organizations will be more comprehensible and possess clearer identities in the market than others. I develop a theory of how this variation affects the search for alliance partners in terms of the speed of alliance formation and the diversity between the new organization and its partners. The second study investigates how organizational identity affects the impact of alliances on performance outcomes. Alliances that explore and experiment tend to affect organizational outcomes negatively, at least in the short term. Although exploration strategies facilitate learning and adaptation in the long run, they incur costs due to the nature of experimentation. I advance an alternative perspective that organizational identity plays a role in this alliance-performance link. Depending on the strength of an organization's identity in terms of how coherent and taken-for-granted its categorization or social grouping is, the effect on performance may be more or less negative. Overall, this research indicates that organizational identity matters both to an organization's initial alliance portfolio formation and to the impact of this alliance portfolio on performance outcomes. This work contributes to the literature streams of both organizational identity and alliances, and presents the first systematic investigation of the link between them.

The dissertation investigates stock options-related arrangements by individual executives and firms from a sociological point of view. The first study in this dissertation explores the antecedents of stock option exercises by executives in Chinese state-owned firms, behaviors considered deviant from the institutional norms of the Chinese state bureaucracy. This study seeks to answer the following question: When individual beliefs and actions are deeply embedded in their institutional context, as in the case of Chinese executives in overseas-listed firms, who is likely to break with the institutional status quo, and what are their reasons for doing so? Contrary to the existing status-based theory of social deviance, institutional disengagement among Chinese executives often takes place in the middle of an institutional status hierarchy. Characteristics of the institutional environment and the individual biography further interact with individual positions to affect the likelihood that an executive will diverge from the institutional expectation of not exercising stock options. The second study investigates the individual consequences of stock option exercises in Chinese state-owned firms. This study seeks to answer the following questions: When institutional entrepreneurs diverge from the institutional status quo, who is most likely to be punished? Who can bypass the sanction, and for what reasons? The findings suggest that executives were more likely to be punished for divergent behaviors if (a) the executive had low levels of bottom-up power, (b) the prevalence of divergent behaviors among peers was moderate without having reached a threshold that sufficiently legitimized the practice, and (c) the broader audience was concentrated. This study promises to shed new light on the outcomes of institutional entrepreneurship by addressing why some institutional entrepreneurs fail. The third study, explores the spread of stock backdating, an unethical corporate practice about which public information was virtually unavailable until 2005. This "invisible" practice, unlike corporate practices accessible to outsiders, did not diffuse through board interlocks. Rather, stock option backdating spread because of geographic proximity: Firms were more likely to backdate stock options to the extent that other firms located geographically close to them had done so. The effect of geographical proximity was conditional on high levels of local board interlocks, a finding that lends support to the idea of the importance of localized interactions among members of the local business elite. Together these findings suggest that invisible corporate practices follow unique diffusion patterns.

This paper examines gender differences in the participation of university life science faculty in commercial science. Based on theory and field interviews, we develop hypotheses regarding how scientists’ productivity, co-authorship networks, and institutional affiliations have different effects on whether male and female faculty become “academic entrepreneurs”. We then statistically examine this framework in a national sample of 6,000 life scientists whose careers span more than 20 years. We find sharp gender differences in participation in for-profit ventures, which we measure as the likelihood of joining the scientific advisory board (SAB) of a biotechnology firm. Compared to men, women life scientists are much less likely to advise for-profit biotechnology companies. We also identify factors that contour this gender difference, including scientists’ co-authorship network structure and the level of support for commercial science at their universities. Surprisingly, we find that the (conditional) gender gap is largest among faculty members at the highest status institutions.

This paper examines gender differences in the participation of university life science faculty in commercial science. In part based on interviews, we develop hypotheses regarding how scientists’ career achievements—their productivity, co-authorship networks, and institutional affiliations—have different effects on whether male and female faculty will become “academic entrepreneurs”. We then statistically examine this framework in a case cohort dataset containing the career histories of 6,000 life scientists. We find that participation in for-profit ventures, which we measure as the hazard of joining the scientific advisory board (SAB) of a biotechnology firm, is a new arena in which gender differences are sharp: compared to men, women life scientists are far less likely to receive compensated advisory roles at for-profit biotechnology companies. Moreover, the gap in participation rates persists after conditioning on numerous measures of human and social capital. We also find that this gender difference is contoured by a number of factors, such as co-authorship network structure and the level of institutional support for commercial science. Surprisingly, we find that the (conditional) gender gap is largest among scientists employed at high prestige institutions.

This study advances the prior literature concerning the impact of information technology on productivity in academe in two important ways. First, it utilizes a dataset that combines information on the diffusion of two noteworthy and early innovations in IT -- BITNET and the Domain Name System (DNS) -- with career history data on research-active life scientists. This research design allows for proper identification of the availability of access to IT as well as a means to directly identify causal effects. Second, the fine-grained nature of the data set allows for an investigation of three publishing outcomes: counts, quality, and co-authorship. Our analysis of a random sample of 3,771 research-active life scientists from 430 U.S. institutions over a 25-year period supports the hypothesis of a differential return to IT across subgroups of the scientific labor force. Women scientists, early-to-mid-career scientists, and those employed by mid-to-lower-tier institutions benefit from access to IT in terms of overall research output and an increase in the number of new co-authors they work with. Early-career scientists and those in top-tier institutions gain in terms of research quality when IT becomes available at their campuses.