Embodied
Insights into US life expectancy stagnation from birth cohort mortality dynamics
Leah Abrams et al.
Proceedings of the National Academy of Sciences, 17 March 2026
Abstract:
US life expectancy remained essentially flat in the 2010s. This trend has occurred in the presence of advancing innovations in medical care, especially for chronic disease, and a growing economy. While research has identified selected contributing factors, including "deaths of despair" and cardiovascular disease mortality stagnation, a comprehensive understanding of the underlying mortality dynamics has not been provided. Notably missing is a systematic evaluation of birth cohort dynamics of mortality, which can reveal whether specific generations are driving adverse trends. Using Lexis diagrams, a powerful tool for visualizing the demographic landscape, we analyzed changes in mortality from 1979-2023 for all-cause mortality and three major cause groups (cardiovascular disease, cancer, external causes). Data included cohorts born between the 1890s and 1980s. Results reveal that both cohort- and period-based processes have produced stalling life expectancy improvement. The 1950-1959 birth cohort represents a transition cohort, wherein there were general improvements in mortality across cohorts born before and general deterioration in mortality across cohorts born after. Alarmingly, cohorts born after 1970 exhibited deteriorating patterns in all major cause groups at young and middle-adult ages. Layered on these cohort dynamics was a broad mortality deterioration that began around 2010 and was experienced by nearly all living adult cohorts at the time, driven primarily by cardiovascular disease mortality. These patterns reflect the complex, multifaceted nature of stalled life expectancy improvements that cannot be attributed to any single cause or temporal mechanism. It also portends an unprecedented longer-run stagnation, or even sustained decline, in US life expectancy.
Weighing down the future: Long-term effects of childhood obesity on intergenerational mobility
Maoyong Fan, Yanhong Jin & Man Zhang
Journal of Population Economics, February 2026
Abstract:
Amidst declining intergenerational mobility in the United States, this paper identifies a critical and understudied barrier: childhood health. Using data from the National Longitudinal Study of Adolescent to Adult Health (Add Health) linked to neighborhood opportunity measures from the Opportunity Atlas, we estimate the long-term economic consequences of childhood obesity. To address endogeneity, we construct a novel genetic instrumental variable for body mass index (BMI) that removes confounding genetic variation related to intelligence, cognitive function, and educational attainment, thereby mitigating concerns of horizontal pleiotropy. We find that compared to their normal-weight peers, obese children experience substantially less upward income mobility, ending up approximately 20 percentile points lower than their parents in the adult income distribution and being less likely to live in higher-opportunity neighborhoods. This health-based penalty is driven by lower educational attainment, persistent health problems, and adverse occupational sorting. Our results point to an income-health feedback loop that can reinforce poverty across generations, highlighting the policy importance of childhood health interventions to improve both health and economic mobility.
Translocation of bacteria from the gut to the brain in mice
Manoj Thapa et al.
PLoS Biology, March 2026
Abstract:
Recent advances suggest a correlation between gut dysbiosis and neurological diseases, however, relatively little is known about how gut bacteria impact the brain. Here, we reveal that bacteria can translocate directly from the gut to the brain in small numbers when mice are fed an atherogenic, high-fat diet (Paigen diet) that causes alterations in gut microbiome composition and gut barrier permeability. The bacteria were not found in other systemic sites or the blood, but were detected in the vagus nerve. Right cervical vagotomy reduced bacterial burden in the brain, implicating the vagus nerve as a conduit for bacterial translocation from the gut to the brain. Antibiotic treatment perturbed the composition of the gut microbiome and correspondingly changed the bacteria that localized to the brain in the setting of Paigen diet feeding. To further establish the gut as the origin of bacterial translocation to the brain, we gavaged exogenous Enterobacter cloacae into Paigen diet-fed mice, subsequently detecting the E. cloacae in the gut and brain. In addition, we monocolonized germ-free mice with E. cloacae and only cultured the bacteria from the brains of mice fed Paigen diet, but not those fed standard diet. Localization of bacteria to the brain in Paigen diet-fed mice was reversible with return to normal diet. Bacteria were also detected in the brain of murine models of Alzheimer's, Parkinson's, and autism spectrum disorder fed standard diet. These data reveal a bacterial translocation axis from the gut to the brain, impacted by environmental (diet) and genetic factors, and warrant further investigation to determine if this phenomenon also occurs in humans and to elucidate whether it may play a role in diverse neurological conditions.
Smartphones, Online Music Streaming, and Traffic Fatalities
Vishal Patel et al.
NBER Working Paper, February 2026
Abstract:
Modern smartphones present new threats to road safety beyond talking and texting, but the real-world effects are difficult to study. One way to causally assess the impact of smartphones on road safety is to identify arbitrarily timed events during which smartphone-related distraction may exogenously increase -- i.e., a situation that relies not on plausibly random variation in who uses smartphones while driving, but when smartphones are used. We investigated the impact of smartphones on road safety by examining traffic fatalities on days when smartphone use likely surges: the release of major music albums. Using event study analysis, we show that music streaming -- an indicator for smartphone use, where streaming most often occurs -- sharply increases, by nearly 40%, on dates of major music album releases, while U.S. traffic fatalities increase by nearly 15% on those same days. Mobile device use while driving is a known safety issue, but today's smartphones present new and greater opportunities for driver distraction. Our study indicates how features of these phones may have important impacts on distracted driving and traffic fatalities.
When Political Pivots Shift Behaviors but Not Beliefs: Evidence from Trump's Position Reversal over Facemasks during the COVID-19 Crisis
Bartholomew Konechni
American Sociological Review, forthcoming
Abstract:
Political leaders play a potentially important role shaping behaviors and beliefs during crises. In the pandemic, a number of high-status politicians, notably leaders of populist parties, were seen to diminish compliance with institutional recommendations by casting doubt on COVID guidelines. But what happens when such leaders change position and endorse previously discouraged behaviors? Using longitudinal data from the Understanding Coronavirus in America panel with fixed-effects modeling, this article examines how Trump's unexpected endorsement of facemasks in July 2020 affected individuals' likelihood of wearing a facemask and belief in masks' efficacy. I find that Trump's pivot lifted Republicans' use of facemasks, closing 40 percent of the preexisting gap with Democrats and with stronger effects among individuals who were more exposed to the early-summer spike in COVID cases. Additionally, I provide evidence for the unique significance of this moment in the history of the pandemic, showing that at almost no other time did partisan behaviors converge as sharply. However, in contrast to expectations from most dominant theoretical models of behavioral change, especially the health belief model, no corresponding shift in beliefs about facemasks can be detected. These results have important theoretical implications for understanding how pivots can shape behaviors during crises, the validity of existing models in public health, pandemic populism's causes, and directions of future research.
Gambling and Substance Use: Early Evidence From Sports Betting Laws
Kabir Dasgupta & Keshar Ghimire
Health Economics, forthcoming
Abstract:
Previous research documents a strong association between gambling and substance use, suggesting that these seemingly distinct behaviors may share similar environmental, neurobiological, and genetic causes. However, there is a dearth of credible empirical evidence on whether gambling has a causal impact on substance use or vice versa. This paper estimates the impact of gambling on substance use making use of the rapid roll-out of sports betting laws across US states. Using data from the Behavioral Risk Factor Surveillance System and a difference-in-difference (DID) estimation strategy to assess the impact of legalizing sports betting on smoking and drinking behavior among adults, we find that the legalization of online sports betting has increased binge drinking frequency-at the intensive margin among young men by approximately 10 percent, but find no discernible impact on smoking. The results are consistent across traditional two-way fixed effects models as well as more recently developed DID methods designed for staggered treatment adoption.
Drink, but Please Don't Drive: Spillover Effects of Ridesharing on Alcohol Sales and Drunk Driving Arrests
Samantha Galvan, Yongseok Kim & Richard Gretz
Journal of Public Policy & Marketing, April 2026, Pages 115-138
Abstract:
In an effort to provide new insights into the sharing economy, the current research examines its combined spillover effects on both alcohol sales and drunk driving arrests. In an analysis of how ridesharing affects sales of alcohol at restaurants and bars, a generalized difference-in-differences model of monthly sales gathered from alcohol-selling establishments in Texas reveals that Uber availability increases monthly sales of alcohol by 10.1%, compared with cities where Uber is unavailable. A complementary study of the effect of Uber availability on drunk driving arrests helps address public safety concerns related to increased alcohol consumption. In detail, data from San Antonio, Texas, indicate that Uber availability leads to a 5% reduction in daily drunk driving arrests, with greater relative declines among members of minority populations and young people. These findings add to research on the sharing economy by illuminating some meaningful complementary effects, which also have crucial implications for policy makers seeking to craft appropriate regulations and for marketers seeking to communicate the value proposition of sharing economy services.
School Vaping Bans and Youth E-Cigarette Use
Dhaval Dave et al.
NBER Working Paper, February 2026
Abstract:
With the goal of curbing electronic nicotine delivery systems (ENDS) use among youths, 40 states and the District of Columbia have adopted restrictions on nicotine vaping at elementary and secondary (K-12) schools. This study is the first to estimate the effect of K-12 school vaping restrictions (SVRs) on ENDS use among youth. Using data from several nationally representative data sources and a generalized difference-in-differences approach, we find little evidence that school indoor vaping restrictions reduce youth ENDS use. With 95% confidence, we can rule out SVR-induced youth vaping declines of more than 3.5%. Descriptive evidence suggests that students may respond to SVRs by substituting to smaller vaping devices and vaping in school locations that are less easily detectable by school employees (e.g., bathrooms and locker rooms). In contrast, we find that comprehensive vaping restrictions covering school campuses, workplaces, restaurants, and drinking establishments are associated with a 1.5-2.6 percentage-point reduction in teen nicotine vaping. This result is concentrated on the initiation margin of use and is consistent with the hypothesis that broader place-based restrictions increase the costs of youth ENDS use.
The Effect of Paid Sick Leave on Healthcare Expenditures
Reginald Hebert et al.
Yale Working Paper, January 2026
Abstract:
As of 2025, 18 states (including Washington D.C.), 17 cities, and four counties have paid sick leave mandates that allow workers to use paid time off for healthcare utilization. Some groups have argued that paid sick leave mandates could reduce healthcare spending overall by increasing use of preventive care and reducing incidences of infectious disease. We evaluate these claims by studying the effect of paid sick leave mandates on private insurance and Medicaid spending data from the Centers for Medicare and Medicaid Services National Health Expenditure accounts, and household spending data from the Bureau of Labor Statistics. We find evidence that paid sick leave mandates significantly increase per-capita private insurance and Medicaid spending as well as out-of-pocket spending on prescription drugs. We find a 7.5% increase in per-capita private health insurance spending overall and a 13% increase in total Medicaid spending. Additionally, we find a 24% increase in prescription drug spending among households most likely to newly gain paid sick leave. We conclude that paid sick leave mandates are unlikely to reduce healthcare spending, at least in the short term.