Empirical Bayes shrinkage methods usually maintain a prior independence assumption: The unknown parameters of interest are independent from the known precision of the estimates. This assumption is often theoretically questionable and empirically rejected, and imposing it inappropriately may harm the performance of empirical Bayes methods. We instead model the conditional distribution of the parameter given the standard errors as a location-scale family, leading to a family of methods that we call CLOSE. We establish that (i) CLOSE is rate-optimal for squared error Bayes regret, (ii) squared error regret control is sufficient for an important class of economic decision problems, and (iii) CLOSE is worst-case robust when our location-scale assumption is mis-specified. We use our method to select high-mobility Census tracts targeting a variety of economic mobility measures in the Opportunity Atlas (Chetty et al., 2020; Bergman et al., 2023). Census tracts selected by CLOSE are more mobile on average than those selected by the standard shrinkage method. The gain of CLOSE over the standard shrinkage method is substantial relative to two benchmarks.

Online discourse related to discrimination, including complaints about firm actions, has surged in recent years. While consumer complaints about discriminatory behavior by firms are often rooted in reality, they may at times contain distortions from the truth (e.g., false attributions, exaggerations) regarding experiences of differential treatment based on their membership in certain social categories. Combining evidence from Twitter with five incentive-compatible, online experiments, I investigate consumer motivations behind mentioning discrimination in their complaints within the context of airline customer service. I find that consumers perceive mentions of discrimination to be effective in eliciting a firm’s response, and this view is confirmed by Twitter data on major U.S. airlines: tweets mentioning discrimination-related words elicit faster responses from airlines. This is because consumers consider complaints mentioning discrimination (e.g., “I’ve been discriminated against”) as particularly damaging to the firm’s reputation. As a result, in settings where firms are more concerned about their reputation (e.g., public channels, corresponding track records), consumers are more inclined to strategically mention discrimination in their complaints, even when it is ambiguous whether discrimination actually occurred.

Forthcoming

The prevailing theory of innovation is that breakthroughs emerge from the recombination of technological knowledge elements. However, this narrow conceptualization of the nature of knowledge involved in firms’ inventive searches leaves largely unexplained the question of why similar firms following seemingly equivalent technological search strategies differ in their abilities to develop breakthroughs. In this paper, I argue that innovation is not simply a process of exploring technological spaces, but also application or market spaces. Accordingly, my theory reconceptualizes the breakthrough development process as one involving search across two dimensions of a theoretical knowledge field: technological knowledge and application knowledge. Just as a firm’s search has been traditionally characterized as exploration or exploitation based on its familiarity with technological knowledge, in my model, it is also characterized as more or less exploratory based on a firm’s familiarity with application knowledge. Using three decades of data on firms’ patenting activities and a novel measure of search (Application Focal Proximity), I investigate the nature of application knowledge searches that are linked to breakthroughs. I find that (1) an application search trajectory that beings with a period of exploration and culminates in a period of exploitation and (2) the intensity of focus on a given application space in the final stages of an invention’s development are both positively associated with breakthrough creation. My research contributes to the literature on firm innovation and breakthrough creation.

The idea that well-being interventions can enhance both worker well-being and organizational performance has led many organizations to frame them as a “win-win,” suggesting that there is a “business case” for well-being. Drawing on an ethnography of a company that implemented and later rolled back a 4-day workweek, this paper explores the challenge of sustaining well-being interventions through an investigation of how they are contested. Findings illustrate the key role of how organizations manage tensions between social and financial goals. The company implemented the 4-day workweek with the dual goals of improving well-being (a social goal) and improving productivity (a financial goal). There was consensus that the intervention improved well-being, and administrative data showed neither large increases nor large decreases in productivity. Yet despite strong support from workers and the CEO, a few executives successfully argued to end the 4-day workweek using a two-pronged resistance strategy targeting each of the intervention’s stated dual goals. First, resisters generated concern about productivity by asserting that the 4-day workweek had failed to deliver promised productivity gains, and by arguing that productivity was a trade-off against well-being. Second, they delegitimized well-being improvements by downplaying them and by mobilizing business norms that valorize hard work and financial goals. Ambiguity and subjectivity in how productivity was understood and measured enabled resisters’ strategy and constrained promoters’ advocacy. This study contributes to scholarship on worker well-being in organizations by showing that the use of a win-win framing to motivate well-being interventions can contribute to their fragility.

This paper examines the impact of reduced resource redeployment costs on market entry, entry mode, and organizational structure, using changes in state-level occupational licensing rules for lawyers. My findings suggest that firms are more likely to enter a market when redeployment costs drop, yet less likely to enter using a new physical establishment (greenfield investment). Moreover, the study indicates that firms increase their managerial span of control in response to lower redeployment costs. Relatedly, firms increase the level of worker specialization since expanded market access enables workers to concentrate on a greater depth of specialized knowledge. With these new opportunities for flexible firm expansion, firm performance improves as redeployment costs decrease. Overall, these results imply that firms adopt new entry modes and organizational structures when facing lower redeployment costs to effectively reallocate resources as market conditions evolve

Hiring beyond the founding team is essential to scaling a venture, but labor and search are extremely costly. What hiring strategies do entrepreneurs use, and how do they think about setting compensation for their employees? I conduct a novel survey and experiment with 540 founders of growth-capable U.S. startups that have, on average, 18 employees and $6.8 million of funding. Entrepreneurs initially described hiring and compensation setting at their own firms and then were asked for their best advice about wages for four fictitious job postings. This paper shows that founders report being involved in all aspects of the hiring process, even at the minority of startups that have human resources staff. Startups are less likely than established firms to use paid consultants or wage benchmarking data, relying instead on word-of-mouth advice and free online services to research compensation. Entrepreneurs’ wage advice is highly dispersed and sensitive to information, particularly when the entrepreneur is a first-time founder or lacks experience with a given job. Finally, soliciting a “fair” wage rather than framing wages as a cost induced non-male entrepreneurs to recommend wages that were $11,000 higher, on average. Taken together, these results suggest that startup hiring is a founder-centric process and that wages at startups may be influenced both by available information and by founders’ own utility, beliefs, and preferences.

Even if algorithms make better predictions than humans on average, humans may sometimes have private information which an algorithm does not have access to that can improve performance. How can we help humans effectively use and adjust recommendations made by algorithms in such situations? When deciding whether and how to override an algorithm's recommendations, we hypothesize that people are biased towards following a naïve advice weighting (NAW) heuristic: they take a weighted average between their own prediction and the algorithm's, with a constant weight across prediction instances, regardless of whether they have valuable private information. This leads to humans over-adhering to the algorithm’s predictions when their private information is valuable and under-adhering when it is not. In an online experiment where participants are tasked with making demand predictions for 20 products while having access to an algorithm’s recommendations, we confirm this bias towards NAW and find that it leads to a 20-61% increase in prediction error. In a follow-up experiment, we find that feature transparency – even when the underlying algorithm is a black box – helps users more effectively discriminate when and how to deviate from algorithms, resulting in a 25% reduction in prediction error.

While corporate involvement in modern scientific research is an indisputable fact, the impact of corporate involvement on scientific progress is controversial. Corporate interests can lead to constraints that redirect research activities into applied problems in a way that benefits the company but reduces scientific impact. However, corporations also provide resources such as funding, data sets, collaborators, engineers, and technical problems that researchers may otherwise be unable to access or know about, spurring knowledge creation. This paper empirically assesses the impact of corporate involvement on scientific research by focusing on dual-affiliated artificial intelligence researchers located at the intersection of academia and industry. After controlling for the researcher's quality and topic preferences, I find that corporate involvement leads to up to a 44% increase in field-weighted citations received by a paper. I document evidence that this effect is driven by the resource-constraint tradeoff. Specifically, I show that corporate involvement significantly increases the likelihood of a breakthrough paper and that these effects are magnified by the involvement of firms with greater resources. However, corporate involvement also alters the direction of the dual-affiliate author's research to be more aligned with the firm's commercial interests. This is the first large-scale quantitative study of any field of science to demonstrate a direct positive effect of corporate involvement on science or to describe the underlying mechanism.