Abstracts

Session 1A

• Samantha Darnell (Management, Wharton) - When One Door Opens: Effects of Political Polarization and Partisan Reputation on Nonmarket Strategy

Political polarization is a phenomenon that has captured the attention of researchers in many fields, culminating in large bodies of work studying polarization among the political elite, the media, and the general public. Although this topic is of considerable importance to management scholars given the power of increasingly polarized stakeholders within firms’ market and nonmarket environments, we know little about the effects of polarization on firm reputation, strategy, and performance. This paper focuses on the polarization of firm reputation within stakeholder groups, which is particularly relevant in the nonmarket context. The relationship between firm reputation and nonmarket performance has been previously demonstrated, but our understanding of the links between reputation and nonmarket strategy may need to be updated given rising polarization. We develop and introduce two original measures of reputation, partisan reputation and reputation polarization, to represent the heterogeneity of a firm’s reputation among politically diverse sets of stakeholders. Using a hand-collected 10-year panel dataset, we analyze theeffects of firms’ partisan reputation and reputational polarization on key nonmarket outcomes including 1) invited congressional hearings, 2) procurement contracts, and 3) refunded PAC contributions.

• Begum Bilgin (Management, RSM) - Overcoming the Myopia of Volunteering: How Vision Vividness Drives Organizations to Help Distant Outgroups

Although there is enough wealth to meet the basic needs of all people in the world, the fact that it is concentrated in small pockets throughout the world means that more than two billion people remain mired in poverty, unable to meet basic needs related to food, water, shelter, and health care. Given that most of the world’s poorest people live furthest from concentrated pockets of economic wealth, the social impact that most Western organizations can have increased as a function of physical distance. In this vein, one may expect that organizations that devote more resources to not-for-profit help are in a better position to help impoverished reasons. But when the association between physical distance and potential impact is considered in tandem with the theory on intergroup relations and psychological distance, a vexing problem emerges. The literature on intergroup relations and psychological distance has established that, when people have a choice about whom to help, they are strongly inclined to assist those who are physically close to them. People are also more likely to notice problems in their immediate vicinity, even if the suffering far away is more potent and prevalent. Given that people typically think and communicate concretely only about that which is immediately around them in space and time, we contend that wealthy Western organizations that dedicate more resources to outreach activities will only inspire employees to direct these resources toward the world’s global poor when their leaders communicate about their organizations’ broadest aspirations in concrete, rather than abstract, terms. In this way, the concrete and the proximal are inextricably intertwined, so much so that people perceive a concretely described reality to be close by, even if, it is thousands of miles or years away. In this archival study of 519 hospitals, we showed that organizations are likely to have a broader scope of outreach — that is, they are likely to initiate projects in the distant countries and continents that experience the greatest poverty — when their leaders communicate more vivid visions of the future.

Session 1B

• Sergey Sarkisyan (Finance, Wharton) - Central Bank Digital Currency: Will Banks Survive

Will an introduction of CBDC cause disintermediation? I provide empirical and theoretical evidence that CBDC will not necessarily crowd out bank deposits in economies with signicant demand for currency. I estimate the model for the US using data on households' payment choices and find that non-interest-bearing CBDC will lead instead to an inflow of deposits caused by cash substitution. Banks then lower deposit rates and lend more. Similarly, banks will not contract lending if CBDC is intermediated even if they experience an outflow of deposits. Finally, I show that CBDC can lead to disintermediation when it is interest-bearing.

• Chuck Fang (Finance, Wharton) - Liquidity Misallocation on Decentralized Exchanges

There is significant misallocation of liquidity on decentralized exchanges. We arrive at this conclusion by analyzing both returns to liquidity providers (LP returns) and their actions to deposit or withdraw liquidity (LP flows). First, many pools have persistently high (low) LP returns, implying persistently insufficient (excessive) liquidity. This finding is strengthened by comparing actual LP returns against options-implied LP returns, which have risk premia baked in. A strategy that exploits the predictability of persistent LP returns earns a high return of 1.24 percentage points per week. Why do LPs leave money on the table? We show that LPs chase fee revenues -- the part of returns that are prominently displayed as APY -- but ignore impermanent losses -- the part of returns that are never displayed. Pools with higher fee revenues experience higher LP flows, even if they have higher risk of impermanent losses and lower net returns in expectation. A strategy that naively longs (shorts) pools with high (low) fee revenues earn zero returns, whereas a strategy that further sorts on expected impermanent losses earns a high return of 1.02 percentage points per week. We discuss implications of our findings for the long-run sustainability of DeFi.

Session 2A

• Tiancheng Wang (Strategy, INSEAD) - When stars strike: How do award-winning CEOs affect the turnover of non-CEO managers?

Prestigious institutions grant CEOs for their performance and qualifications every year. This paper examines how the award-winning of CEOs affect the turnover of non-CEO managers, including senior executives and middle-level managers. Based on the social learning and negative resource draining perspectives, we expect that senior executives are more likely to leave the firm, whereas middle-level managers are less likely to leave following CEOs win awards. Baseline empirical tests support our main hypotheses. Further, we hypothesize that the ages of CEO and non-CEO managers moderate the effects, and that a higher turnover of both senior executives and middle-level managers following the CEO awarding negatively affects firm performance. Our findings will contribute to the CEO award literature and enrich the understanding of non-CEO managers' turnover.

• Jason Lee (Management, Wharton) - Navigating the uncertainty with entrepreneurial experimentation: the evaluation process of experimentation results

I investigate the process after conducting an entrepreneurial experimentation, examining the challenges in inferring opportunities from the newly acquired information. Studies highlight that entrepreneurial experimentation must be conducted in a “scientific” manner, going through the phases of a proper hypothesis generation – experimentation – evaluation (Camuffo et al., 2020; Levinthal, 2017). Entrepreneurs are advised to iterate the experimentations until they find an optimal market opportunity. With regard to the scientific methodology, how entrepreneurs interpret and evaluate the experimental results have been relatively understudied. The literature has emphasized the benefits of experimentation, namely the lower opportunity cost and generation of alternatives (Camuffo et al., 2020; Gans et al., 2019; Koning et al., 2019). Such studies assume that the optimal option is readily recognized from the alternatives presented by the experimentation. However, evaluating alternatives involves the entrepreneur’s interpretation of the information, which is subject to variation. This study empirically examines the evaluation process in entrepreneurial experimentation, showing that information obtained from experiments can be ambiguous and can induce bias in evaluating experimented results.

Session 2B

• Clara (Chi) Xu (Finance, Wharton) - Scrambling for Collateral

We provide a theory of collateral management whereby a borrower signals her types though costly collateral acquisition. Regulatory constraints stipulate that only high-quality assets qualify as collateral. To acquire high-quality liquid assets (HQLAs), a borrower can either purchase them in a competitive centralized secondary market or enter into "collateral transformation" to swap low-quality collaterals into high-quality ones. In equilibrium, a good borrower is able to signal her type in the lending market by pledging collateral and can always access the over-the-counter (OTC) collateral swap market, conditional on having enough bargaining power. We find that a borrower’s ability to utilize collateral transformation covaries positively with her bargaining power and the degree of asymmetric information, and negatively with her amount of cash on hand.

• Xinwei Li (Finance, INSEAD) - A Real Investment-based Model of Asset Pricing

This paper proposes a pure investment-based approach to recovering the stochastic discount factor (SDF). We first solve for the optimal investment in approximate analytical solutions for a given SDF and profitability process from the investment Euler equation, providing a dissection of determinants of real investment. We then calibrate the model to both match moments of quantities and asset prices and minimize mean squared pricing errors of in-sample test portfolios, in order to discipline free parameters of the SDF. We utilize the calIbrated model to finally recover the SDF from investment and use it to price out-of-sample portfolios. We explicitly characterize the risk-free rate, the equity premium, the term structure of interest rates, and the term structure of equity risk premia.

Session 3A

• Xina Li (Strategy, INSEAD) - Love amid Terror: Terrorist Attacks and Firms’ Employee Treatment

This study examines how firms treat their employees differently in reaction to terrorist attacks. We propose that firms located near the epicenters of terrorist attacks treat employees more favorably after the attacks. This effect is stronger when firms have a greater female board representation and a strong corporate culture of companionate love. Using a sample of S&P 1,500 firms and machine learning to analyze employee reviews, our empirical analysis provides strong support to our propositions. Additionally, we find that firms in high-tech industries have weaker performance in wake of terrorist attacks compared with other firms, but such a performance drop is mitigated by favorable employee treatment. Overall, our study not only contributes to the literatures on stakeholder governance, corporate governance, and corporate culture but also generates valuable practical implications regarding firm responses to extertnal adverse events.

• Khwan Kim (OB, INSEAD) - Exploration and Exploitation in Cultural Tastes from a Core/Periphery Perspective

How does category membership impact consumer behavior? Does a cultural consumer’s openness to novel products depend on what genre she fancies? To tackle these questions, we first conceptualize genre preference as a network interface where individual consumers can be positioned differently depending on the genres they favor. We then theorize the consumer’s position in the genre network as indicative of what type of behavioral constraints to be imposed at what level. This, in turn, influences whether a consumer’s taste structure evolves by exploration or stagnates by exploitation. Using longitudinal data on 44,000 consumers and their listening history from 2018 to 2020 on a global music streaming platform, we extend a core/periphery perspective (Cattani and Ferriani, 2008) and construct a genre network where popular, commercial genres are mostly placed at the core while specialized, niche genres are at the periphery. We show that consumers whose categorical membership occupies the intermediate position in a genre network are more likely to explore new tastes whereas consumers who are at the core or periphery in that genre network are more likely to exploit their existing tastes. We argue that this is because consumers in the core may experience cognitive constraints that engender difficulty decoding product complexity, whereas consumers in the periphery may face normative constraints that hamper deviant consumption.

• Christopher Bruno (Management, Wharton) - Corporate networks, the replacement of the inner circle, and community outcomes

This paper is motivated by two emerging themes: the decline in public companies and board interlocks over the past 20 years and the dearth of research examining social welfare outcomes as a function of changing corporate networks. The latter is especially surprising, given that many of the latest findings around the decline in corporate networks relate to political polarization, short-term orientation, and cohesion amongst firms. The central research question I examine is “how, when, and why do declines in corporate networks affect community outcomes?” To explore answers to this question, I connect two streams of management literature: institutional theory-oriented research on corporate networks and communities and an emerging stream on the impacts of nonprofits on community sustainability. These streams provide foundational building blocks to theorizing that declines in corporate networks relate to declines in nonprofit density, resulting in worsening community outcomes. I plan to empirically test these arguments with a novel, large-scale database of county- and MSA-level community outcomes (e.g., poverty rates, income inequality, renewable energy mix and crime rates) connected to data on corporate networks and interlocks. The findings of this research would have implications for community-oriented theory, anti-trust and nonprofit policy, and companies tackling grand challenges.

• Shun Yiu (Wharton, Management) - Acquisitions and Corporate Purpose

-

Session 3B

• Meghana Yerabati (Finance, INSEAD) - Mind the contagion, Your Honour

Using an unexpected order of the Supreme Court of India that retrospectively cancelled the allocation of coal blocks as an economic setting, we study the working of the bank lending channel in a stable macroeconomic environment. Despite there being no direct shock to the banking sector, lending by banks that are exposed to the firms losing coal blocks reduces significantly for borrowers not impacted by the court order directly, suggesting contagion moving through the bank channel. The borrowers hit by the supply shock seem unable to replace the lost credit and suffer a reduction in real outcomes such as investments, revenues, and profits. We propose to understand the (1) change in overall bank lending technology such as increased risk aversion (2) unavailability of other sources of funds and (3) heterogeneity of impact by bank and borrower type. Importantly, we aim to evaluate the relative potency of the bank lending channel in explaining the economic slowdown when compared to other plausible mechanisms.

• Xinyu Liu (Finance, INSEAD) - Labor shortage, hiring and asset returns

In this paper, I examine how labor shortage a↵ect hiring-return relationship. I construct firm-year level measure of labor shortage using textual analysis of firms’ SEC fillings. I show that the negative relationship between firm’s hiring rate and its future return is only significant for firms who discuss labor shortage. I provide further evidence by documenting the connection of labor shortage with firms’ corporate policy and operation dynamics. Firms of high growth are more likely to discuss labor shortage. Once they do, their current hiring and future investment rate drop, leverage and book to market ratio increase. These results can be rationalized with the interpretation that firm labor shortage reflects firm level labor supply constraint, which results in negative hiring-return relationship according to q theory. Intuitively, high-hiring firms of high labor supply constraint incur the largest adjustment cost. Hence they benefits the most from shocks that lowers adjustment cost, becoming the best hedge of the lowest compensation to risk. Lastly, I show that the findings cannot be explained by small firm bias, can are robust to alternative explanations of labor shortage measure.

• Richard Grice (Marketing, INSEAD) - Heterogeneous Investor Consideration, Mutual Fund Competition, and Fund Fees

This paper investigates the implications for the mutual fund industry of retail investors’ ‘consumer behaviour’. In particular, we study how investors’ heterogeneous consideration determines the structure of competition between funds, and in turn the dispersed fund fees observed in practice. Guided by a micro-theorymodel linking consumer consideration to market structure and outcomes, we measure investors’ consideration sets by prospectus views on the SEC’s EDGAR system, and calculate the networked competition structure induced by investors’ overlapping consideration sets. We validate this structure by using it to calibrate the model, and showing predicted fees are significantly correlated with observed fees. Thus we explain how retail investors’ heterogeneous consideration of investment alternatives grants funds differing degrees of (local) market power, which rationalises a portion of the observed dispersion in fund fees. Therefore our framework explicitly links investors’ incomplete information and constrained choice sets to mutual fund-level outcomes, and generalises the notion of investor search costs.

• Ahmed Guecioueur (Finance, INSEAD) - Money doctors and their prognoses

I illuminate an economically meaningful channel through which fund managers build up trusting relationships with their investor clients. Motivated by the theory of managers as money doctors and the first principles of investors’ portfolio choices, I extract forward-looking information about risk from managers’ communications. Exploiting an institutional setting that enables me to identify causal effects, I find that increased communication about future risks induces existing clients to increase their portfolio holdings in the market portfolio.

Session 4A

• Gui Siqueira (LGST, Wharton) - Dynamic capabilities in corrupt routines

In attempting to shape the institutional environment in which they operate, some firms resort to corruption, a phenomenon that has received increased attention from management scholarship. Little is known, however, about how corrupt practices fit within corporate strategy more broadly. This paper adopts the dynamic capabilities framework to investigate that question. Based on an in-depth case study of Odebrecht, a large construction conglomerate implicated in a large corruption scandal, I suggest that firms that opt to engage in corrupt strategies face a dilemma: either they maintain corrupt routines as lower-order capabilities, in which case they become highly inertial and increase the risk of detection; or they develop higher-order capabilities that allow them to steer corrupt practices according to the changing environment, in which case senior management becomes personally exposed to legal liability. In other words, corrupt strategies present a tradeoff between governance and liability. This finding contributes to our understanding of the prevention of organizational wrongdoing. Whereas the literature tends to focus on measures that target the final behavior (e.g. paying a bribe), this analysis suggests that wrongful behavior can be discouraged by policies that disrupt the governance of corrupt routines, thus rendering them unattractive for the firm.

• Shuping Wu (OB, INSEAD) - OrgTech: Evidence of Organizational Innovations in Patent Data

Organization theorists have long claimed that organizational innovations are nontechnological, in part, because they are unpatentable. The claim rests on the assumption that organizational innovations are abstract ideas embodied in persons and contexts rather than in context-free practical tools. However, over the last three decades organizational knowledge has been increasingly embodied in digital tools which, in principle, can be patented. To provide the first empirical evidence regarding the patentability of organizational innovations, we trained two machine learning algorithms to identify a population of 205,434 patent applications for organizational technologies (OrgTech) and, among them, 141,285 applications that use organizational innovations accumulated over the 20th century. Our event history analysis of the probability of patenting an OrgTech invention shows that ideas from organizational innovations decrease the probability of patent allowance unless they describe a practical tool. We conclude that the present-day digital transformation places organizational innovations in the realm of high tech and turns the debate about organizational technologies into the challenge of designing practical organizational tools that embody big ideas about organizing. We outline an agenda for patent-based research on OrgTech as an emerging phenomenon.

• Nety Wu (Strategy, INSEAD) - Adapting versus Shaping: Competition in the Face of Uncertainty

Firms can respond to uncertainty by adapting their own belief as they learn about the environment, or by shaping the environment to reflect their beliefs. However, the interactions of adapting and shaping and the implications of competition between adapter and shaper are still unclear. I propose to incorporate the multiarmed bandit framework with the Cournot competition to study under what conditions firms benefit from adapting relative to shaping and vice versa, as well as the implications for competition. Initial results show a U-shape relationship between the shaper’s shaping capability and the adapter’s profit. The adapter’s profit would be lowest when the shaper has a moderate level of shaping capability, where the shaper can shape the environment to its benefit. When the shaper’s shaping capability is strong enough, the shaper and the adapter will behave more similarly and take the same action. In this situation, the adapter can earn a higher profit by following the shaper’s lead.

• Emily Ulrich (Management, Wharton) - Far away, but not forgotten: Applying stakeholder justice to internationalization strategy

This paper applies justice theory to predict how social evaluations of a firm’s international stakeholder network affect reputation. Current research on international corporate strategy seldom considers how operating in certain countries affect perceptions of whether a multinational organization is exploitative. Traditional transaction cost theory and knowledge-based views of the firm imply a negative relationship between degree of internationalization and just distribution of value across a firm’s stakeholder network, as firms leverage arbitrage and control opportunities. We contend that perceptions of injustice in a firm’s internationalization strategy can create damaging reputation risks that should also be considered. To do so, we use firm- and country-level wage data to examine international interfirm relationships, creating an aggregate measure of justice based on network-wide equality differences of interfirm dyads. We then create a composite score of activist pressure against a firm and strength of informal institutions in a country, specific to labor concerns, to examine how social institutions moderate the extent to which multinationals create injustice. Ultimately, we expect of our results that these institutions can improve equality across a firm’s stakeholder network, making it more just. In this context, organizational justice theory provides novelty to existing internationalization, stakeholder governance, and social movements literatures.

Session 4B

• Srishti Arora (TOM, INSEAD) - Don’t Fake It If You Can’t Make It: Driver Misconduct in Last Mile Delivery

In the last two decades, last mile delivery (LMD) firms have seen immense growth due to the rise of e-commerce. However, this growth brought new challenges e.g., increased competition amongst LMD firms and heightened customer expectations. Operating on thin margins, LMD firms strive to deliver orders in their first attempt because reattempts cost revenue and reputation. A key reason for delivery failure is the misconduct of delivery field executives (FEs). For instance, a FE can record a claim that ‘the customer was not available’ without even visiting the customer’s address (i.e., a fake remarked delivery). Most attempts at improving LMD performance focus on technology and incentives, assuming workers will follow the designed processes, which may not hold, especially for gig workers. We collaborated with a LMD firm in India to study the impact of FE misconduct on LMD performance. Using instrumental variable regression, we find that fake remarked deliveries cause productivity loss spillovers to the subsequent day and reduce the next day successful deliveries by 1.45%. Interestingly, this decrease is driven by the reduction in first-time-right deliveries, resulting in significant revenue losses because reattempted deliveries are costly. We find that opportunistic circumstances, such as cash deliveries exacerbate this detrimental effect.

• Clara Carrera Moreno (TOM, INSEAD) - Diffusion in Circularity: the Missing Link in Circular System Design

The transition into a circular economy entails non-ownership-based business models (e.g., leasing) and new product designs (e.g., modular product architectures). In this paper, we leverage a new product diffusion approach to formally model the forward and reverse flows in such new systems. In our model, forward flows capture new leases and replacements, and reverse flows capture end-of lease returns (in reusable form or to be disposed of). The formal characterization of these flows allows us to demonstrate that the recycling needs in these systems exhibit uneven multi-peak volumes over time and that remanufacturing needs exhibiting relatively stable flows over a short time span. Motivated by an industry partner’s practice, we further analyze how a firm’s optimal lease duration and new product introduction timing choices influence these flows and disposal needs.

• Alexis Gordon (OID, Wharton) - Gossip and Advisor Selection in the Workplace

While previous research has established the importance of receiving and utilizing advice for improving decision accuracy and has shown that gossip is pervasive, particularly in the workplace, research has failed to investigate how gossip impacts the advice seeking process. I seek to understand how a reputation for gossip (and alternatively a reputation for discretion) impacts willingness to select an advisor when seeking advice in the workplace. More specifically, I provide evidence that individuals prefer advisors who do not have a reputation for gossip when seeking advice about interpersonal work issues and issues related to their ability to do their job. Further, I show that individuals prefer individuals who are discreet when seeking advice about such issues. This suggests that, if individuals want to be sought out for advice, they should refrain from acquiring a reputation of gossip and should instead cultivate a reputation for discretion.

• Lin Chen (TOM, INSEAD) - Fairness Concerns in Heterogeneous Teams: Optimal Team Composition and Contract

Inequality aversion among team members is known to lead to higher team output when agents are homogeneous. However, in heterogeneous teams, an uneven profit distribution among team members can create guilt or envy, potentially hurting output. In fact, a manager may choose to over-use the most efficient team member or, to the contrary, require more even involvement at the cost of lower output. In this paper, we consider a principal multi-agent model in which heterogeneous agents are subject to envy and guilt among each other. We analytically show that when agents' heterogeneity in abilities is sufficiently low and agents are sufficiently inequality-averse, it is optimal for the principal to involve a team instead of only the most efficient agent. In such cases, the optimal team contract turns out to be not only envy- and guilt-free, but also to lead to higher team output. As a consequence, both the principal and the agents may be better off. Our analysis demonstrates that heterogeneity in abilities combined with endogenous fairness concerns does not necessarily create a fairness-efficiency trade-off. Instead, stimulating endogenous inequality aversion may be beneficial for both the principal and the agents in providing a fairer and more efficient solution.

Session 5A

• Yanbo Song (OB, INSEAD) - Forecasting Venture Potential Through Founding Team Mood Heterogeneity

There has been a recent surge in interest in how feelings and other affective qualities that investors observe in entrepreneurs influence evaluations of ventures (e.g., Chen, Yao, and Kotha, 2009; Cardon, Mitteness, and Sudek, 2017; Jiang, Yin, and Liu, 2019). Through a FaceReader analysis of team pitch videos submitted to the five major accelerators in the U.S., our findings suggest that founding teams who display diverse levels of positive emotion (i.e., happy) and diverse levels of negative emotion (i.e., angry) in their pitches are more likely to have superior short-term and long-term performance. As over half of ventures are typically organized by founding teams (Aldrich and Ruef, 2006), this study enhances our understanding of how constellations of teams and observed moods in cofounders influence investor evaluations of the potential of early-stage ventures. It also contributes to the knowledge of innovative processes and research on affective diversity’s benefits for team performance.

• Qinghan Yan (Marketing, INSEAD) - Collaboration for Healthier Lives: A Field Experiment

Collaboration between food producers and retailers can improve the nutritional quality of purchases, the effect of which works better for some food categories and some segments of consumers but not necessarily the others. Using large scale data from a field experiment conducted in a French supermarket chain, we investigate the treatment effect of three marketing interventions used in supermarkets, classified as (1) candy-free checkout lines; (2) stands: end-of-aisle displays and sales promotions with healthier foods; or (3) shopping cart posters nudging healthier food purchase. The 2-year transaction-level data from 21 stores covers 7 food categories, from which we can analyze the quantity of fruits and vegetables purchased, quantity-weighted nutrition quality of consumers' food basket of purchases, and the sales effects of end-of-aisle displays compared to regular promotions. The difference-in-difference estimates (our primary outcome) show distinct impact of interventions on different food categories, controlling for price and promotion. For instance, the transactions and quantities purchased on margarine increased significantly under the nudges; however, this effect is not obvious for yogurt. Nutritional quality is hypothesized to play a role here. Our secondary goal of this study is to evaluate consumer attention (recall of the interventions), evaluations (impact on the reputation of the brand and retailer, interpretation of their motives), and intentions to engage in healthy behaviors, using surveys sent to loyalty card holders in both treated and control stores. Our results will provide unique, conclusive field experimental evidence on the relative efficiency potential of the supermarket intervention for each distinct food category, and to provide insights on promoting healthy shopping behaviors.

• Xinlan Emily Hu (OID, Wharton) - Making the Case for Diversity: How Diversity Narratives Influence Team Performance

At the organizational level, the narrative that diversity improves productivity (the “business case”), as opposed to being a moral imperative (the “moral case”), has been shown to alienate minority recruits (Georgeac 2020, Kaplan 2020). But how do such narratives impact performance after being hired? In Study 1, we recruit women and tell them to imagine that they have been hired because of either the business case, the moral case, or (control) because of their qualifications. We then measure the extent to which participants feel committed to the team and are willing to work hard. In Study 2, we conduct an incentivized behavioral study using a team brainstorming task. We pair a woman with an existing (real) team of men, and we randomize the justification offered for her addition to the team (business case, moral case, or a control justification). We then measure the length, number, and quality of the ideas generated, as well as the diverse member’s willingness to stay with their team. We find that there is no way to sugarcoat being a diversity hire: regardless of whether diversity is justified for business or moral reasons, employees’ willingness to commit to the team plummets. They are less willing to stay with their team and generate worse ideas during brainstorming.

Session 5B

• Daniel Chen (OID, Wharton) - Machine Learning and Prediction Errors in Causal Inference

Machine learning is a growing method for causal inference. In machine learning settings, prediction errors are a commonly overlooked problem that can bias results and lead to arbitrarily incorrect parameter estimates. We consider a two-stage model where (1) machine learning is used to predict variables of interest, and (2) these predictions are used in a regression model for causal inference. Even when the model specication is otherwise correct, traditional metrics such as p-values and rst-stage model accuracy are not good signals of correct second-stage estimates when prediction error exists. We show that these problems are substantial and persist across simulated and empirical data. We propose general methods to identify when prediction errors are biasing estimates and provide consistent corrections for the case where an unbiased subset of the data is available.

• Sundara Natarajan Panchanatham (TOM, INSEAD) - Shared Medical Appointments

Shared medica appointments (SMA) are doctor-patient visits in which groups of patients are seen by one or more healthcare professional in a concurrent session. The setup can potentially cater to one-time or recurrent appointment for the same or a random group of participants while enhancing patient experience, eliminating redundant work, improving medical compliance, and reducing cost for certain treatments. From an operations standpoint, providers face several key challenges in the smooth conduction of SMAs which includes finding the optimal group size, coming up with scheduling mechanisms, figuring out the optimal capacity allocation and nudging more people to opt for SMA. In this work, we mathematically analyze the key operational trade-offs involved in SMAs as a service system and provide normative prescriptions that enable providers to choose the optimal decision variables (capacity/payment/group size) for several system configurations.

• Jiatao Ding (TOM, INSEAD) - Can Predictive Technology Help Improve Acute Care Operations? Investigating the Impact of Virtual Triage Adoption

To choose the appropriate resources for their healthcare needs (primary care (GP) or emergency department (ED)), patients seeking acute care must self-triage based on their own assessments of symptoms and severity. However, as patients typically lack sufficient medical knowledge, self-triage decisions can often be inaccurate. In response, healthcare and technology companies have been developing and deploying virtual triage tools designed to help patients make better self-triage decisions. To date, however, the operational implications of such tools have not been assessed. This paper therefore develops a queueing game model to investigate the impact of virtual triage in the acute care setting and potential policies to maximize its efficacy. We find that, due to its decentralized nature, when virtual triage excessively recommends emergency (primary) care, it could bring about a decrease in ED (GP) visits. Another important finding is that for any arbitrary self-triage accuracy, the adoption of informative virtual triage can worsen system performance. To unlock the potential operational benefits of virtual triage, we characterize the optimal virtual triage accuracy subjective to the receiver operating characteristic (ROC) curve, and how the optimal accuracy changes as acute care system parameters change or as the triage capability of virtual triage improves over time.

Session 6A

• Anna (Anya) Shchetkina (Marketing, Wharton) - Autoregressive differences-in-differences

Difference-in-differences and related methods, such as synthetic control or synthetic difference-in-differences, are a common way of evaluating the effects of marketing campaigns. We show that when these methods are applied to autocorrelated data, the resulting estimates are biased if the history of observations is limited. We show that when there are enough periods of observation before the campaign but not after, the estimates are shrunk toward zero. In other cases, the bias can have any direction, in particular, the chance of a Type I error may be inflated. We develop the autoregressive difference-in-differences (AR DiD), a method that takes into account the autocorrelation in the data and at the same time utilizes the information from units unaffected by a campaign. We demonstrate that the correction of the bias is greatest for larger effect sizes, which occur when the data is strongly autocorrelated. We apply the AR DiD to estimate the effect of a branding campaign on visits to a website and find that the proposed method eliminates bias due to shrinkage to zero compared to standard methods, consistent with analytical results and simulations.

• Gi Kim (Applied Economics, Wharton) - Incentive Structure and Competition among Brokerages in the U.S. Real Estate Market

The sticky commission rates of the real estate brokerage industry in the U.S. has been a puzzle for both researchers and regulators for decades. This paper examines the market forces behind the commission rates and evaluates welfare consequences of potential regulations by estimating a structural model of the intermediated housing market. While the model admits other potential mechanisms, the key force behind the pricing of commission rates is from the current incentive structure of the industry where home sellers are paying the entirety of the commission and a portion the commissions is offered to buying brokerages in order to attract buyers. After estimating the primitives, I conduct a counterfactual to simulate a widely proposed policy of making buyers to pay their own commission, severing the obligatory transfer from listing brokerages to buying brokerages. In equilibrium, the policy would change the way brokerages compete with each other and potentially lower industry-wide commission rates, but may put more financial burden on buyers, especially the first-time home buyers. To evaluate the trade-offs of the policy, I measure the distributional impact of having buyers to pay their own commission rates with varying degrees of cash constraint.

Session 6B

• Ying Huang (Accounting, INSEAD) - Does a US Government Shutdown Have Lasting Effects on SEC Enforcement?

As the SEC’s budget is subject to Congressional approval, recent political tensions and divisiveness in the US increase the probability of government agency shutdowns due to lack of funding. This paper employs the SEC shutdown from 27 Dec 2018 to 25 Jan 2019 to examine long-term governance effects on public company registrants due to the disruptions in regulatory enforcement. We find the number of comment letters issued by SEC falls to zero during the month, but we do not see an abnormal increase in comment letters following the resumption of normal regulatory activities after the shutdown ends. Moreover, insiders earned abnormally high profits from selling their shares during the shutdown period. Our results suggest that there are real costs to government shutdowns in terms of lower levels of SEC oversight, providing implications for the debate on whether the SEC should be self-funded or not.

• Jack Xiaoyong Fu (Finance, Wharton) - Patents: ability or choice?

This paper provides causal evidence on the impact of patenting choices on innovation outcomes. Using a novel database on the court's decisions that randomly change firms' patenting motives, I show that patenting choice alone explains 50% of the variation in patents. I find a typical US public firm strategically chooses not to patent 40% of its inventions, and it can potentially gain 67% more patents even when there is no change in its innovation ability. This quantitative evidence challenges the conventional wisdom of using patents as a measure of innovation ability. Further, I find patenting choice has significant real effect on investment and R&D, and affects patent quality. I conclude by providing some improved measures of innovation ability.