John Cawley

Estimates of the impact of body mass index and obesity on health and labor market outcomes often use instrumental variables estimation (IV) to mitigate bias due to endogeneity. When these studies rely on survey data that include self‐ or proxy‐reported height and weight, there is non‐classical measurement error due to the tendency of individuals to under‐report their own weight. Mean reverting errors in weight do not cause IV to be asymptotically biased per se, but may result in bias if instruments are correlated with additive error in weight. We demonstrate the conditions under which IV is biased when there is non‐classical measurement error and derive bounds for this bias conditional on instrument strength and the severity of mean‐reverting error. We show that improvements in instrument relevance alone cannot eliminate IV bias, but reducing the correlation between weight and reporting error mitigates the bias. A solution we consider is regression calibration (RC) of endogenous variables with external validation data. In simulations, we find IV estimation paired with RC can produce consistent estimates when correctly specified. Even when RC fails to match the covariance structure of reporting error, there is still a reduction in asymptotic bias.

This chapter summarizes the research on the economic costs of obesity. Specifically, it reports estimates of the causal effect of obesity on two major categories of costs: 1) direct medical care costs; and 2) indirect job absenteeism costs. Recent estimates indicate that obesity raises medical care costs by US$2,781 per individual with obesity, doubling medical care costs relative to those of normal weight. The aggregate medical care costs of obesity in adults in the United States in 2016 were US$289 billion. Obesity also raises a worker's number of missed workdays by an average of three per year, more than doubling the number of workdays missed due to illness. Estimates of the aggregate productivity loss due to obesity-related job absenteeism in the United States range from a lower bound of US$14.9 billion to an upper bound of US$29.8 billion. These costs are useful for estimating the negative externalities associated with obesity, for calculating the cost-effectiveness of policies to prevent and reduce obesity, and for making the business case for a private firm to take steps such as adopting a wellness program or including weight loss drugs on their health insurance formulary.

Economics has long studied how consumers respond to the disclosure of information about firms. We study a case in which the disclosed information is unrelated to the product or firm leadership, but which could still potentially affect consumer patronage through the mechanism of repugnance, as described in Roth (2007). The information in this case concerns the arrest of Jared Fogle, the advertising pitchman for the Subway sandwich franchise, who was arrested in 2015 on charges of sex with a minor and child pornography. We study how the disclosure of this information, which was widely covered in the media, affected patronage of Subway. We estimate synthetic control models using data from a large nationwide survey of consumers regarding the restaurants they patronize. Despite the close and long-standing association of Jared Fogle with Subway, and heavy publicity of his crimes, we consistently fail to detect any effect of the Jared Fogle scandal on the probability of visiting a Subway restaurant. These results contrast with past studies of negative information disclosure, which tend to find negative impacts on sales, revenue, or stock price of the relevant companies. The absence of an effect in this case suggests that repugnance did not drive demand, and that consumers largely separated the offenses of a symbol of the firm from the products of the firm. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

A subset of harm reduction strategies encourages individuals to switch from a harmful addictive good to a less harmful addictive good; examples include e-cigarettes (substitutes for combustible cigarettes) and methadone and buprenorphine (substitutes for opioids). Such harm reduction methods have proven to be controversial. Advocates argue that people struggling with addiction benefit because they can switch to a less harmful substance, but opponents argue that this could encourage abstainers to begin using the harm reduction method or even the original addictive good. This paper builds on theories of addiction to model the introduction of a harm reduction method, and it demonstrates the conditions under which each side is correct. The three key factors determining whether the introduction of a harm reduction method reduces or worsens health harms are: 1) the enjoyableness of the harm reduction method, 2) the addictiveness of the harm reduction method, and 3) the substitutability of the harm reduction method with the original addictive good. Knowledge of these conditions can help inform regulation of harm reduction methods. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

Taxes on sugar‐sweetened beverages (SSBs) are relatively new and there is limited evidence about their impact on SSB consumption or body mass index (BMI) (as opposed to prices, purchases, or sales), their impact on youth (as opposed to adults), or their impact in non‐Western nations. This paper adds to the evidence across all these dimensions by estimating the effect of an SSB tax on SSB consumption and the BMI of youth in Mauritius, an island nation in the Indian Ocean, which we compare to Maldives, another island nation which did not implement an SSB tax during the time of our data. Results of difference‐in‐differences models indicate that the tax in Mauritius had no detectable impact on the consumption of SSBs or the BMI of the pooled sample of boys and girls. However, models estimated separately by sex indicate that the probability that boys consumed SSBs fell by 9.4 percentage points (11%). These are among the first estimates of the effect of SSB taxes on youth consumption and contribute to the limited evidence on the impact of SSB taxes on weight, and in non‐Western countries.

Objective: The aim of this study was to evaluate the effectiveness of the Children’s Food and Beverage Advertising Initiative (CFBAI) in reducing children’s exposure to ads for candy and sweetened beverages.Methods: Survey data were used to determine the television programs that children watch and the time slots during which they watch television. Advertisement placement data were used to count the number of candy and sweetened beverage (SB) ads appearing on programs and during those time slots. Advertisement placement data and children’s exposure measures were examined for 2003 to 2013.Results: There was a dramatic decline in children’s exposure to ads for candy and SBs. The declines occurred before CFBAI implementation and occurred for each demographic group. There was no evidence that advertisers moved ads to programs watched by both children and teens/adults, i.e., programs not likely governed by the CFBAI.Conclusion: There was a striking decline in ad placements and children’s exposure to ads for candy and SBs, much of which occurred when the CFBAI was being negotiated. Voluntary agreements have the potential to be successful, and some gains may occur even as firms and government negotiate the agreement.

Introduction Efficacy tests of physical activity interventions indicate that many have limited or short-term efficacy, principally because they do not sufficiently build on theory-based processes that determine behaviour. The current study aims to address this limitation. Methods and analysis The efficacy of the 8-week intervention will be tested using a three-condition randomised controlled trial delivered through an app, in women with a prior hypertensive pregnancy disorder. The intervention is based on the integrated behaviour change model, which outlines the motivational, volitional and automatic processes that lead to physical activity. The mechanisms by which the behaviour change techniques lead to physical activity will be tested. Following stratification on baseline factors, participants will be randomly allocated in-app to one of three conditions (1:1:1). The information condition will receive information, replicating usual care. Additionally to what the information condition receives, the motivation condition will receive content targeting motivational processes. Additionally to what the motivation condition receives, the action condition will receive content targeting volitional and automatic processes. The primary outcome is weekly minutes of moderate-to-vigorous physical activity, as measured by an activity tracker (Fitbit Inspire 2). Secondary outcomes include weekly average of Fitbit-measured daily resting heart rate, and self-reported body mass index, waist-hip ratio, cardiorespiratory fitness and subjective well-being. Tertiary outcomes include self-reported variables representing motivational, volitional, and automatic processes. Outcome measures will be assessed at baseline, immediately post-intervention, and at 3 and 12 months post-intervention. Physical activity will also be investigated at intervention midpoint. Efficacy will be determined by available case analysis. A process evaluation will be performed based on programme fidelity and acceptability measures. Ethics and dissemination The Medical Ethics Committee of the Erasmus MC has approved this study (MEC-2020-0981). Results will be published in peer reviewed scientific journals and presented at scientific conferences. Trial registration number Netherlands trial register, NL9329.