Columbia Business School Research

We document gender and race dynamics in performance evaluations in a multi-national company, examining the impacts of a feature of the performance appraisal process: managers’ knowledge of employees’ self-evaluations. Generally, (White) women were rated higher than men and people of color were rated lower than White employees. Women of color gave themselves the lowest self-ratings. When self-evaluations were unavailable due to a quasi-exogenous shock, manager and self-ratings were less correlated.

Generative Artificial Intelligence (GenAI) tools based on Large Language Models (LLMs) are quickly reshaping how researchers conduct surveys and experiments. From reviewing the literature and designing instruments, to administering studies, coding data, and interpreting results, these tools offer substantial opportunities to improve research productivity and advance methodology. Yet with this potential comes a critical challenge: researchers often use these systems without fully understanding how they work.

With multinational corporations (MNCs) increasingly taking public stances on sociopolitical issues such as immigration, LGBTQ+ rights, and racism, it is imperative that International Business (IB) research keeps pace with normative societal debates. In this paper, we introduce the concept of corporate sociopolitical activism (SPA) to the IB literature and develop theory on why MNCs consistently or inconsistently engage in SPA in response to the same issue in their home country and a host country.

Women and minorities are underrepresented in top leadership roles. Besides “supply-side” explanations that focus on the applicant pool, we offer a novel “demand-side” explanation through perceptual imprinting. Using the reverse correlation method, we found that people’s visual templates of leaders are perceptually imprinted by White male leaders. Across 15 studies (N = 3929), we examine what people expect leaders to look like. As demonstrated by the reverse correlation method, compared to followers, people expected leaders to look more White and male.

Under standard assumptions, individuals and the government are indifferent between traditional tax-deferred retirement accounts and “front-loaded” (Roth) accounts. Adding investment fees to this benchmark, individuals are still indifferent but the government is not. We show that under weak conditions firms charge equal percent fees under both systems, yielding higher dollar fees under Traditional. We estimate that tax deferral increases demand for asset management services by $3.8 trillion, costing the government $23.4 billion in annual fees.

This paper addresses a simple theoretical question of high substantive relevance: What makes a consumption experience special in a consumer’s mind?

How has Wikipedia activity changed for articles with content similar to ChatGPT following its introduction? We estimate the impact using differences-in-differences models, with dissimilar Wikipedia articles as a baseline for comparison, to examine how changes in voluntary knowledge contributions and information-seeking behavior differ by article content. Our analysis reveals that newly created, popular articles whose content overlaps with ChatGPT 3.5 saw a greater decline in editing and viewership after the November 2022 launch of ChatGPT than dissimilar articles did.

In this letter, we summarize our recent work on the welfare impact of recommendation algorithms and propose questions for further study. We model recommendation algorithms as an information structure, which shapes how a third party takes actions that affect the welfare of different individuals in a population. Each recommendation algorithm thus induces a welfare profile, describing the expected payoffs of different individuals when the third party takes actions following the algorithm.

Any judgment about population heterogeneity depends on the definition of the sampling frame (1). In a recent paper, Holzmeister et al. (2) (HJBK hereafter) compare different sources of heterogeneity to population heterogeneity. They find that population heterogeneity is much smaller compared to design and analytic heterogeneity as a source of variation in effect sizes.

Multinational enterprises are at the centre of policy debates in low- and middle-income countries. As some of the most productive and innovative firms in the world, which are at the core of global supply chains, multinational enterprises (MNEs) can accelerate development in the countries hosting them, both directly with their presence, and indirectly through linkages to local economic actors.

Multinational enterprises are at the centre of policy debates in low- and middle-income countries. As some of the most productive and innovative firms in the world, which are at the core of global supply chains, multinational enterprises (MNEs) can accelerate development in the countries hosting them, both directly with their presence, and indirectly through linkages to local economic actors.

Professional house price forecast data are consistent with a rational model where agents must learn about the parameters of the house price growth process and the underlying state of the housing market. Slow learning about the long-run mean generates overreaction to forecast revisions and a modest response of forecasts to lagged realizations. Heterogeneity in signals and priors about the long-run mean helps the model account for cross-sectional dispersion in forecasts. Introducing behavioral biases helps improve the model's predictions for short-horizon overreaction and dispersion.

We aim to stimulate discussion on how innovation research within marketing can use a better world (BW) perspective to help innovation become a driver of positive change in the world. In this "Challenging the Boundaries" series paper, we hope to provide purposeful research opportunities for scholars seeking to bridge innovation research with the BW movement. We frame our discussion with four areas of innovation research in marketing that are particularly relevant to BW objectives.

Research on campaign finance suggests that Americans prefer candidates that are not funded by Political Action Committees (PACs). However, prior research has not examined how perceptions of a candidate who is PAC-funded vs. PAC free might differ for racial minority and female candidates compared to White, male candidates. Using experimental vignettes, we test the causal impact of PAC funding, race, and gender on voter perceptions of the candidate.

One of the most crucial aspects and significant levers that gaming companies possess in designing digital games is setting the level of difficulty, which essentially regulates the user’s ability to progress within the game. This aspect is particularly significant in free-to-play (F2P) games, where the paid version often aims to enhance the player’s experience and to facilitate faster progression.

Language is a uniquely human trait at the core of human interactions. The language people use often reflects their personality, intentions and state of mind. With the integration of the Internet and social media into everyday life, much of human communication is documented as written text. These online forms of communication (for example, blogs, reviews, social media posts and emails) provide a window into human behaviour and therefore present abundant research opportunities for behavioural science.

We develop a framework to measure the welfare impact of macroeconomic shocks throughout the distribution. The first-order impact of a shock is summarized by the induced movements in agents’ feasible sets: their budget constraint and borrowing constraints. We combine estimated impulse response functions with micro-data on household consumption bundles, asset holdings, and labor income for different US households. We find that inflationary oil shocks are regressive, but monetary expansions are progressive, and there is substantial heterogeneity throughout the life cycle.

Significance: Estimating the social cost of carbon (SCC)—the cost of one additional ton of CO2 emitted—is crucial for the analysis of climate change policies. Despite numerous recent studies investigating how fundamental aspects of model structure affect SCC evaluation, findings are scattered, making the relative importance of different modeling elements hard to establish. This paper synthesizes results from the published literature over the last 20 y, revealing a wide range of SCC estimates.

Evaluations play a critical role in the allocation of resources and opportunities.
Although evaluation systems are a cornerstone of organizational and market
processes, they often reinforce social and economic inequalities. The body of
organizational research on inequality and evaluations is extensive, but it is also
fragmented, siloed within specific contexts and types of evaluations (e.g., hiring,
performance). As a result, we currently lack a systemic understanding of the
conditions under which inequalities emerge. This paper provides a unifying

How do venture capital investors react to social movements, especially those that relate to historical underrepresentation in their funding decisions? We use image and name algorithms combined with clerical review to classify race for 150,000 founders and 30,000 investors. Our new data allow us to assess the impact of George Floyd's murder on VC funding of Black entrepreneurs and identify which VCs were most responsive. Although VCs responded swiftly, investment in Black-owned startups reverted to prior levels within two years.

Social movements are catalysts for crucial institutional changes. To succeed, they must coordinate members’ views (consensus building) and actions (mobilization). We study union leaders within Myanmar’s burgeoning labor movement. Union leaders are positively selected on both ability and personality traits that enable them to influence others, yet they earn lower wages. In group discussions about workers’ views on an upcoming national minimum wage negotiation, randomly embedded leaders build consensus around the union’s preferred policy.

We leverage Lachmann’s insight on elite conflict to explain the politics surrounding industry regulation in contemporary America and argue that conflicts between political elites create both constraints on industry players and opportunities for them to shape regulation. The widening urban-rural polarization of American society, in particular, has made urban political elites more liberal than those in state politics. The greater the political polarization of a state,  the more local restrictions the nascent U.S.

Firms with political connections to a regime with an authoritarian history face a dilemma when the regime undergoes a democratic transition. Such connections provide an essential competitive advantage when the regime is in power but become a liability when an institutional transition brings democratic change. This study theorizes that when mass protests expose a regime’s distorted policies favoring elites over others and signal a high probability of regime turnover, firms may hedge against the risks associated with their political connections by engaging in philanthropy.

Advancements in data analytics and increased access to consumer data have revolutionized companies’ price discrimination capabilities. These technological advancements have not only changed how prices are determined but also who determines them, with companies increasingly relying on algorithms rather than humans to set prices. We examine consumers’ fairness perceptions of demographic price discrimination—a prevalent yet controversial practice that can trigger considerable consumer backlash—and find that it depends on who is responsible for setting prices.

The extent of future climate change is largely a policy choice. We illuminate this choice with climate policy curves (CPCs), which link climate policies to subsequent global temperatures. The estimated downward sloping CPCs highlight the key trade-off between initial policy ambition, expressed via an overall effective carbon price, and the subsequent policy burden left for future generations. We also demonstrate how different CPCs can illustrate the range of climate policy paths towards attaining the Paris Agreement temperature goals.

Immigrants are highly entrepreneurial. But, what is the broader relationship between high-skilled immigration and regional entrepreneurship activity beyond the ventures that immigrants establish themselves? Using administrative data on newly awarded H-1B visas in the United States, we document a positive relationship between highskilled immigration and regional entrepreneurship. A doubling of immigrants to a metropolitan statistical area is followed by a 6% increase in entrepreneurship within three years.

Although there have been numerous studies on voluntary departure—i.e., quit behavior—the way race influences voluntary departure is not yet settled. Some studies suggest racial minorities are more apt to voluntarily depart than non-minority employees due to discrimination in the workplace. Other studies suggest racial minorities are more apt to stay due to discrimination in the labor market.

One of the most significant levers available to gaming companies in designing digital games is setting the level of difficulty, which essentially regulates the user’s ability to progress within the game. This aspect is particularly significant in free-to-play (F2P) games, where the paid version often aims to enhance the player’s experience and to facilitate faster progression. In this paper, we leverage a large randomized control trial to assess the effect of dynamically adjusting game difficulty on players’ behavior and game monetization in the context of a popular F2P mobile game.

We propose that social-media users’ own post histories are an underused yet valuable resource for studying fake-news sharing. By extracting textual cues from their prior posts, and contrasting their prevalence against random social-media users and others (e.g., those with similar socio-demographics, political news-sharers, and fact-check sharers), researchers can identify cues that distinguish fake-news sharers, predict those most likely to share fake news, and identify promising constructs to build interventions. Our research includes studies along these lines.

Non-informational cues, such as facial expressions, can significantly influence judgments and interpersonal impressions. While past research has explored how smiling affects business outcomes in offline or in-store contexts, relatively less is known about how smiling influences consumer choice in e-commerce settings even when there is no face-to-face interaction.

Previous research has shown that consumers respond differently to decisions made by humans versus algorithms. Many tasks, however, are not performed by humans anymore but entirely by algorithms. In fact, consumers increasingly encounter algorithm-controlled products, such as robotic vacuum cleaners or smart refrigerators, which are steered by different types of algorithms. Building on insights from computer science and consumer research on algorithm perception, this research investigates how consumers respond to different types of algorithms within these products.

Acquisitions can shift the market structure of a digital platform in ways that affect subsequent entries and hence the platform’s base of complementors. Synergies that complementor acquirers accrue can be entry-deterring. We develop a two-by-two typology of acquisition synergies in a multisided platform based on the two sides of a platform market (user side or complementary-technology side) and two sources of synergies (scale or scope economies).

Organizational secrecy is central to national security, politics, business, technology, healthcare, and law, but its effects are largely unknown. Keeping organizational secrets creates social divides between those who are required to keep the secret and those who are not allowed to know it. We demonstrate that keeping organizational secrets simultaneously evokes feelings of social isolation and status, which have opposing effects on employee well-being.

As concern with climate change increases, people seek to behave and consume sustainably. This requires understanding which behaviours, firms and industries have the greatest impact on emissions. Here we ask if people are knowledgeable enough to make choices that align with growing sustainability intentions.

One measure of the ex ante cost of disasters is the welfare gain from shorten-ing their expected duration. We introduce a stochastic clock into a standard disaster model that summarizes information about progress (positive or negative) toward disaster resolution. We show that the stock market response to duration news is essentially a sufficient statistic to identify the welfare gain to interventions that alter the state.

We develop a financial-economic model for carbon pricing with an explicit representation of decision making under risk and uncertainty that is consistent with the Intergovernmental Panel on Climate Change’s sixth assessment report. We show that risk associated with high damages in the long term leads to stringent mitigation of carbon dioxide emissions in the near term, and find that this approach provides economic support for stringent warming targets across a variety of specifications.

Whether speaking, writing, or thinking, almost everything humans do involves language. But can the semantic structure behind how people express their ideas shed light on their future success? Natural language processing of over 40,000 college application essays finds that students whose writing covers more semantic ground, while moving more slowly (i.e. moving between more semantically similar ideas), end up doing better academically (i.e. have a higher college grade point average). These relationships hold controlling for dozens of other factors (e.g.

In this paper, we develop a computational measure of the firm-level rhetorical nationalism. We first review the literature and develop a four-dimensional theoretical framework of nationalism relevant to firms: national pride, anti-foreign, dominant agenda, and corporate role. We then use machine-learning-based text analysis of over 41,000 annual reports of Chinese public firms from 2000 to 2020 and identify a dictionary of words for each dimension.

Using publicly available data from 299 pre-registered replications from the social sciences, we find that the language used to describe a study can predict its replicability above and beyond a large set of controls related to the paper characteristics, study design and results, author information, and replication effort. To understand why, we analyze the textual differences between replicable and nonreplicable studies.

How should an investor value financial data? The answer is complicated because it depends on the characteristics of all investors. We develop a sufficient statistics approach that uses equilibrium asset return moments to summarize all relevant information
about others’ characteristics. It can value data that is public or private, about one or many assets, relevant for dividends or for sentiment. While different data types, of course, have different valuations, heterogeneous investors also value the same data

We propose a political economy mechanism that explains the presence of fiscal regimes punctuated by crisis periods. Our model focuses on the interaction between successive deficit-biased governments subject to i.i.d. fiscal shocks. We show that the economy transitions between a fiscally responsible regime and a fiscally irresponsible regime, with transitions occurring during crises when fiscal needs are large. Under fiscal responsibility, governments limit their spending to avoid transitioning to fiscal irresponsibility.

Routines shape many aspects of day-to-day consumption. While prior work has established the importance of habits in consumer behavior, little work has been done to understand the implications of routines — which we define as repeated behaviors with recurring, temporal structures — for customer management. One reason for this dearth is the difficulty of measuring routines from transaction data, particularly when routines vary substantially across customers. We propose a new approach for doing so, which we apply in the context of ridesharing.

To investigate general patterns in news information in the United States, we combine a protocol for identifying major political news stories, 11 monthly surveys with 15,000 participants, and a model of news discernment. When confronted with a true and a fake news story, 47 percent of subjects confidently choose the true story, 3 percent confidently choose the fake story, and the remaining half are uncertain. Socioeconomic differences are associated with large variations in the probability of selecting the true news story.

Nearly everyone keeps secrets, but only recently have we begun to learn about the secrets people keep in their everyday lives and the experiences people have with their secrets. Early experimental research into secrecy sought to create secrecy situations in the laboratory, but in trying to observe secrecy in real time, these studies conflated secrecy with the act of concealment. In contrast, a new psychology of secrecy recognizes that secrecy is far more than biting our tongues and dodging others’ questions.

Constituency mobilization is a widely prevalent corporate political strategy, yet we lack systematic evidence on the scope of its effectiveness. One emerging form of constituency mobilization is user mobilization, wherein a company focuses on rallying political support among its users. This approach differs from traditional lobbying, which relies on tightly controlled insider strategies to exert influence over lawmakers. In our study of user mobilization by platform-based companies in the U.S.

Policymakers increasingly rely on behavioral science in response to global challenges, such as climate change or global health crises. But applications of behavioral science face an important problem: Interventions often exert substantially different effects across contexts and individuals. We examine this heterogeneity for different paradigms that underlie many behavioral interventions. We study the paradigms in a series of five preregistered studies across one in-person and 10 online panels, with over 11,000 respondents in total.

Disagreement over divergent viewpoints seems like an ever-present feature of American life—but how common is debate and with whom do debates most often occur? In the present research, we theorize that the landscape of debate is distorted by social media and the salience of negativity present in high-profile spats. To understand the true landscape of debate, we conducted three studies (N = 2985) across online and lab samples.