Search

Studies have documented that individuals distrust and reject algorithm advice (i.e., algorithm aversion). One seemingly natural way to reduce algorithm aversion is to provide post hoc algorithm justifications for algorithm predictions. By increasing the interpretability of the algorithm, decision-makers would presumably be better persuaded and be less likely to reject algorithm advice. This assumption, however, overlooks the fact that providing more information can undermine judgment quality by lowering accuracy and increasing confidence. Building on this "more is less" view of decision-making, we argue that algorithm justifications can increase human rejection of algorithm advice. Two experimental studies demonstrate that providing algorithm advice with justifications (i.e., positive and negative factors driving the recommendation) lowered the acceptance of the algorithm's advice, and that this effect was driven by lower credibility assigned to the algorithm, and an enhanced (false) sense of expertise. We further examine the interaction between expert identity and justifications, and show that the more-is-less bias is more evident when individuals are primed to see themselves as experts. Our findings highlight the importance of social psychological factors in understanding human-algorithm collaboration. [ABSTRACT FROM AUTHOR]

In this article, the authors discusses the differential regulatory outcomes for firms focusing on firm-government ties, while controlling for a number of other firm characteristics, as well as for various ways of extracting political rents and the movement of employees between the private and the public sectors. The author studied ties between General Electric's crop producing firms and the regulatory body U.S. Department of Agriculture, they concludes the ties significantly improve regulatory speed and regulatory processes.

I argue that the historical evolution of the FDA as an arbiter of science results in faster regulatory approval for drugs sponsored by high status firms. Using review times for 1,303 New Drug Applications approved by the FDA, I find that firms with higher knowledge status receive faster approval for their drugs. [ABSTRACT FROM AUTHOR]

The Food and Drug Administration (FDA) exerts significant influence on firm outcomes and strategies through the regulatory approval process. Prior research suggests that the agency's desire to protect its bureaucratic reputation from audiences such as consumers, the media, and politicians determine the duration of a drug review. I argue that the FDA's historical evolution as an arbiter of science has infused the agency with values that focus the agency's attention on specific types of signals. In particular, I contend that firm status in the knowledge domain becomes salient as the FDA attempts to resolve the uncertainty in determining drug quality. Furthermore, I propose that innovation strategies and external shocks generating uncertainty enact the FDA's scientific identity to a greater degree, rendering a firm's status in the knowledge domain even more valuable. Using the review times for 1,303 New Drug Applications (NDAs) approved by the FDA between 1990 and 2004, I find that firms with higher knowledge status receive faster approval of their drugs, with roughly a one month advantage for a top 13% firm over an average firm. The benefits of being a high status actor were further enhanced when firms attempted to do radical innovation, but status had limited effects when the regulator suffered from a legitimacy crisis. The paper concludes with a discussion of the contributions to organizational theory and the practical implications for innovative strategies of pharmaceutical and biotechnology firms. ..PAT.-Unpublished Manuscript [ABSTRACT FROM AUTHOR]

The speed of product development is crucial to creating and sustaining competitive advantage. This paper examines how a development project's relation to the developing firm and its technological knowledge base influences the rate at which products succeed or fail. More specifically, I introduce two dimensions along which ¡°distance¡± between a firm and project can be created: the locus of initial development and the technological relevance to a firm's knowledge base. Integrating previously conflicting views on the success rates of alliances, I propose that projects originating from outside the firm will have a higher rate of successful development, but the effect is enhanced when the collaborative project is technologically relevant to a firm's knowledge base allowing the firm to screen out potential ¡°lemons¡±. Furthermore, I test the idea that projects will have a slower rate of success when projects are either highly and lowly relevant to a firm's existing technology base. Finally, I examine the rate of development for failed projects and propose that factors that promote faster development will paradoxically slow the rate of failure for these projects. Using project data of 2,114 drug development projects from 1980 to 2003, I find that the hazard of succeeding is higher when drugs are developed in an alliance between firms and especially so when these in-licensed projects are closer to its technological knowledge base. However, alliances are also shown to extend the development period for failed drugs, implying that strategies of buying promising drug candidates can result in higher costs. [ABSTRACT FROM AUTHOR]

Alliances with established organizations provide young firms with access to resources necessary for survival. Previous organizational research has focused on prior alliance activity as a driver of alliance formation. Yet, in the context of young firms which have limited track records, such explanations do not suffice. This paper proposes an alternative mechanism of alliance formation that focuses on the individuals that form the upper echelon of a firm. We theorize that the career affiliations of top managers facilitate the formation of alliances by acting as a signal of legitimacy, status, and substance. Utilizing a typology of upper echelon experience that distinguishes between upstream, horizontal, and downstream employment-based affiliations, we propose that the prominence of upper echelon affiliations signal a firm's quality and legitimacy, which in turn promotes alliance formation. We further unpeel this general proposition to examine the specific ways in which these signals transpire, suggesting that status dynamics and homophily tendencies determine the formation of prominent and specific types of alliances. We test our ideas with an event-history dataset of 3,200 career histories of top managers who took biotechnology firms public between 1979 and 1996, and find that firms possessing more upper echelon affiliations with prominent downstream and horizontal organizations have a higher rate of forming alliances. Further, we find that the better the match between the type of employment affiliation and the type of alliance partner, the greater the rate of alliance formation. We conclude with a discussion of the implications for organizational research on young entrepreneurial firms. [ABSTRACT FROM AUTHOR]

This paper examines the effect that external third-party rankings have on collaboration networks within organizations. Using detailed patient-level treatment data for over 10 million in-patient stays drawn from a broad national sample of hospitals in the US between 2002 and 2011, the paper finds that a rise in a specialty's ranking in the US News & World Report's "Best Hospitals in the US" list is associated with a decrease in the likelihood of collaboration through deference (i.e., ego effect) and an increase in the likelihood of other specialties deferring to the specialty (i.e., halo effect). This effect was stronger when the focal specialty was of high prestige (e.g., cancer or cardiovascular), but weaker when ranked in the top half of hospital specialties. [ABSTRACT FROM AUTHOR]