Knowledge that Transforms

In a tax—public goods reciprocity framework between citizens and the state, managers view taxes as a payment to the government in exchange for public goods, and hence they adjust their willingness to pay taxes as public good quality changes. We show that corporate tax planning intensity increases with ground‐level ozone pollution. Revisions in ozone pollution regulations cause counties that failed the revised and more stringent standards to reduce ozone pollution. Consequently, firms headquartered in these counties reduced corporate tax planning intensity relative to firms in other counties. The ozone‐tax link varies in the predicted directions with public attention to pollution, potential welfare loss due to ozone, managers’ stakeholder orientation, taxpayers’ polluting status, political preferences, and civic norms. We also find consistent results for Superfund cleanups of hazardous waste sites. Our research sheds light on reciprocity as a potential mechanism influencing corporate tax compliance.

The controllability principle suggests that employees should not be held accountable for factors outside their control. This study develops novel theory to challenge that thinking. According to the theory, holding employees accountable for uncontrollable factors like peer performance can lead to improved decision‐making by increasing how much employees learn from those uncontrollable factors. I expect this effect to occur because goal‐focused employees only consider information that seems goal‐relevant, and uncontrollable factors only seem goal relevant when employees are held accountable for them. Results from a decision‐making experiment support the theory. In particular, paying participants based on uncontrollable factors improves their decision‐making despite providing them with weaker economic incentives. This positive effect is more pronounced when the uncontrollable factors are more informative and when individuals are more goal‐focused. These findings reveal a previously unexplored benefit of disregarding the controllability principle that can help explain why broad, uncontrollable metrics are so prevalent and successful in practice.

We use machine learning for relative valuation and peer firm selection. In out‐of‐sample tests, our machine learning models substantially outperform traditional models in valuation accuracy. This outperformance persists over time and holds across different types of firms. The valuations produced by machine learning models behave like fundamental values. Overvalued stocks decrease in price and undervalued stocks increase in price in the following month. Determinants of valuation multiples identified by machine learning models are consistent with theoretical predictions derived from a discounted cash flow approach. Profitability ratios, growth measures, and efficiency ratios are the most important value drivers throughout our sample period. We derive a novel method to express valuation multiples predicted by our machine learning models as weighted averages of peer firm multiples. These weights are a measure of peer–firm comparability and can be used for selecting peer‐groups.

We analyze the standard error bias associated with the use of generated regressors—independent variables generated from first‐step regressions—in accounting research settings. Under general conditions, generated regressors do not affect the consistency of coefficient estimates. However, commonly used generated regressors can cause standard errors to be understated. Problematic generated regressors include predicted values, coefficient estimates, and measures derived from these estimates. Widely used generated regressors in accounting include measures of earnings persistence, normal accruals, litigation risk, and conditional conservatism. Using simple regression models and simulation, we demonstrate how generated regressors can produce understated standard errors in accounting research settings. We also demonstrate how the magnitude of the standard error bias is inversely related to the precision of the generated regressor. Finally, we discuss bootstrapping as a correction for the bias and demonstrate the pairs cluster bootstrap as a tool to improve inferences in common accounting settings involving generated regressors.

This study examines whether taxpayers intentionally avoid Internal Revenue Service (IRS) third‐party reports. In 2017 an IRS amendment created a quasi‐exogenous shock that reduced third‐party tax reporting of pari‐mutuel gambling winnings from certain types of wagers. I consider the effect that this rule change had on taxpayer behavior. Using a difference‐in‐differences research design comparing thoroughbred racing in the United States to Canada, I find a 27% increase in gambler's investment into wager‐types that became less likely to trigger third‐party reports. Further, I provide evidence that this effect was because of third‐party reporting, not withholding, and was stronger in more informed gambling populations. These findings suggest that taxpayers knowingly avoid third‐party reports, enabling underreporting of income to the IRS. This has important policy implications because underreported individual income is the largest driver of the $496 billion annual gap between legal tax liability and actual tax collections in the United States.

A firm that must decide whether to retain or terminate a manager can rely on several sources of information to assess managerial ability. When it relies on a performance signal and monitoring, we show that a more informative signal can surprisingly increase the value of monitoring. Then, signal precision and monitoring are complements. This happens if a more precise information system makes some signals more negative indicators of managerial ability that still do not trigger termination. When the turnover cost is high enough and the manager is more entrenched after a positive performance, an increase in signal precision increases expected monitoring. In firms with a high turnover cost, a less informative signal is compounded by worse monitoring after a disappointing performance. This “bad corporate governance trap” makes it hard for these firms to eventually improve performance.

Beginning in 2018, U.S. public firms were required to report the ratio of the chief executive officer's (CEO) compensation to their median employee's compensation in the annual proxy statement. Exploiting the staggered reporting of pay ratios, we find little evidence that total CEO compensation changes in response to pay ratio disclosure reform. However, we do find that boards significantly adjust the mix of compensation awarded by reducing the sensitivity of CEO pay to equity price changes, particularly when the CEO is likely to garner media scrutiny, and by reducing reliance on stock‐based and other compensation components that are most susceptible to media coverage surrounding the pay ratio disclosure. Firms ultimately disclosing higher pay ratios garner more media coverage around the filing of their proxy statement, and more negative‐toned coverage in the subsequent month. Finally, we find evidence that greater pay disparity is associated with greater selling activity by retail investors and more negative say‐on‐pay votes following pay ratio reform, consistent with a broad set of investors responding to public scrutiny resulting from pay ratio disclosures.

We examine the economic consequences of a rule designed to improve consumers' understanding of mortgage information. The 2015 TILA‐RESPA Integrated Disclosures rule (TRID) simplifies the mortgage disclosures provided to consumers. As a consequence, TRID‐affected mortgages become a less attractive investment opportunity to banks. Our main results document that mortgage applications affected by TRID are less likely to be approved following the rule's effective date. We find evidence consistent with both a decrease in consumers' information processing costs and an increase in banks' secondary market frictions, providing insight into the potential channels through which this reduction in mortgage credit operates. We also find that banks partially compensate for reduced mortgage lending by increasing small business lending, and that fintechs absorb mortgage demand in areas with reduced mortgage lending by banks. Our study documents real actions that firms take in response to disclosure transparency regulation and contributes to the literature on the economic consequences of such regulation.

Nearest neighbor (NN) matching is widely used in observational studies for causal effects. Abadie and Imbens (2006) provided the first large‐sample analysis of NN matching. Their theory focuses on the case with the number of NNs, M fixed. We reveal something new out of their study and show that once allowing M to diverge with the sample size an intrinsic statistic in their analysis constitutes a consistent estimator of the density ratio with regard to covariates across the treated and control groups. Consequently, with a diverging M , the NN matching with Abadie and Imbens' (2011) bias correction yields a doubly robust estimator of the average treatment effect and is semiparametrically efficient if the density functions are sufficiently smooth and the outcome model is consistently estimated. It can thus be viewed as a precursor of the double machine learning estimators.

This paper describes the response of the economy to large shocks in a nonlinear production network. A sector's tail centrality measures how a large negative shock transmits to GDP, that is, the systemic risk of the sector. Tail centrality is theoretically and empirically very different from local centrality measures such as sales share—in a benchmark case, it is measured as a sector's average downstream closeness to final production. It also measures how large differences in sector productivity can generate cross‐country income differences. The paper also uses the results to analyze the determinants of total tail risk in the economy. Increases in interconnectedness can simultaneously reduce the sensitivity of the economy to small shocks while increasing the sensitivity to large shocks. Tail risk is related to conditional granularity , where some sectors become highly influential following negative shocks.