Working Papers
"Levees and levies: Local financing of climate infrastructure maintenance and housing market dynamics"
Local financing of climate infrastructure structurally connects infrastructure quality with housing value through property tax revenue. I study how this connection shapes the long-term dynamics of neighborhood housing value and climate risk. Using novel data on U.S. flood protection levees, I show that lower-income neighborhoods are disproportionately behind levees of maintenance deficiencies, leading to higher levee failure risk. Exploiting exogenous shocks to local economic conditions and levee failures, I find that weaker housing markets degrade levee maintenance, while levee failures in turn reduce housing values and local revenues. These bi-directional dynamics reinforce a feedback loop of fiscal stress and climate vulnerability, contributing to underinvestment in infrastructure with positive net benefits.
"Working out locked in: Determinants and interventions for physical activity during COVID-19" with Yuchen Chai, Fábio Duarte, Stephanie Jegelka, Siqi Zheng
Public health hazards, such as extreme temperatures, air pollution, and pathogens, pose challenges for physical activity (PA) adherence due to voluntary or mandatory self-protection measures. We collected exercise records for 30 million individuals across more than 60 countries from Strava, a PA self-tracking app, and investigated the drivers of PA adherence during the COVID-19 pandemic. We use neural network methods to automate the characterize of PA adherence prior and during the pandemic based on personal exercise habits and social network interactions, and use an explainable neural network approach with econometric analysis to reveal the impact of city-level policies, socio-demographics, and built environment factors on PA adherence.
"Unveiling dynamics in climate disaster information transmission across regions: A variance decomposition network approach" with Yilan Xu, Siqi Zheng, Zhentao Shi, Yi Huang, Sebastien Box-Couillard
We introduce an innovative machine-learning enhanced variance decomposition method to uncover the spatial network structure of climate information transmission using high-frequency social media data. Our findings reveal the structure and determinants of climate disaster information networks and pinpoint locations with high disaster risk yet limited information connectivity. Our approach is adaptable to different geographic scales and its capacity to capture real-time information dynamics can facilitate timely interventions and resource allocation during public crises.
"Observing the unobservable: connecting persistent emotional state to behavior in risky contexts" with Juan Palacios, Jude Bayham, Eli Fenichel, Siqi Zheng
Economists rely on experiments and surveys to measure typically unobserved subjective attributes, e.g., emotional state, which often are important in understanding behavior. Dependence on these data collection modes limits the ability to connect subjective traits with behavior. We use social media posts and natural language processing to infer individuals' persistent emotional state and link this attribute to the behavior of 500,000 individuals during the COVID-19 pandemic. We find that persistent fear propensity has significant moderating effects on social distancing behaviors and can be used to target interventions at scale.
Selected Publications
"Unequal impacts of rising temperatures on global human sentiment" with Jianghao Wang, Nicolas Guetta-Jeanrenaud, Juan Palacios, Devika Kakkar, Nick Obradovich, Siqi Zheng
One Earth, 2025
Estimating the potential impact of climate change on human emotional well-being is a critical component of understanding the manners in which future climatic changes may undermine human mental health. We analyze over 1.2 billion social media posts from 157 countries to reveal how rising temperatures affect human sentiment worldwide and project future impacts under climate scenarios. We find that temperatures above 35°C negatively impact emotional well-being globally, with effects three times greater in low- and middle-income countries (25.0% decline in sentiment) than in high-income countries (8.1%). Even accounting for climate adaptation through income growth, global sentiment will be significantly lower by the end of this century.
"Social cost of lifestyle adaptation: air pollution and outdoor physical exercise"
Journal of Environmental Economics and Management, 2024
The social cost of environmental hazards depends on the well-being impacts of both exposure and adaptation. While the monetary expenditure of adaptation is increasingly considered, little research assesses the social cost and inequality associated with non-market lifestyle adaptation. I present novel evidence regarding how urbanites adjust their outdoor physical exercise in response to air pollution and related information policies using 27 million exercise records from a smartphone exercise app. I quantify the health cost associated with foregone exercise, analyze the distributional consequences, and present implications for environmental and health policies.
"Intraday adaptation to extreme temperatures in outdoor activity" with Jianghao Wang, Nick Obradovich, Siqi Zheng
Scientific Reports, 2023
Linkages between climate and human activity are often calibrated at daily or monthly resolutions, which lacks the granularity to observe intraday adaptation behaviors. Ignoring this adaptation margin could mischaracterize the health consequences of future climate change. We construct an hourly outdoor leisure activity database using billions of cell phone location requests in 10,499 parks in 2017 all over China. We then use climate econometrics to investigate the within-day outdoor activity rhythm in response to extreme temperatures and project the impacts of future climate change.
"Global evidence of expressed sentiment alterations during the COVID-19 pandemic" with Jianghao Wang, Juan Palacios, Yuchen Chai, Nicolas Guetta-Jeanrenaud, Nick Obradovich, Chenghu Zhou, Siqi Zheng
Nature Human Behavior, 2022
The COVID-19 pandemic has created unprecedented burdens on people’s physical health and subjective well-being. While countries worldwide have developed platforms to track the evolution of COVID-19 infections and deaths, frequent global measurements of affective states to gauge the emotional impacts of pandemic and related policy interventions remain scarce. Using 654 million geotagged social media posts in over 100 countries, coupled with natural language processing, we develop a global expressed sentiment dataset and identify the affective effects of COVID-19 and lockdown policies.
"Encouraging the resumption of economic activity after COVID-19: evidence from a large scale-field experiment in China" with Juan Palacios, Erez Yoeli, Jianghao Wang, Yuchen Chai, Weizeng Sun, David G Rand, Siqi Zheng
Proceedings of the National Academy of Science (PNAS), 2022
As the COVID-19 pandemic comes to an end, governments find themselves facing a new challenge: motivating citizens to resume economic activity. What is an effective way to do so? We investigate this question using a field experiment in the city of Zhengzhou, China, immediately following the end of the city’s COVID-19 lockdown. We assessed the effect of a descriptive norms intervention providing information about the proportion of participants’ neighbors who have resumed economic activity.
"Health perception and commuting choice: a survey experiment measuring behavioral trade-offs between physical activity benefits and pollution exposure risks" with Juan Palacios, Mariana Arcaya, Rachel Luo, Siqi Zheng
Environmental Research Letters, 2021
Previous literature suggests that active commuting has substantial health benefits. Yet, in polluted regions, it can also cause additional health risks by increasing riders' pollution exposure and raising their inhalation rate. We examine the effect of perceived air pollution on stated commuting choices using an on-site survey experiment for 2,285 commuters in Zhengzhou, a heavily polluted city in central China. The results from our randomized controlled trial show that pollution avoidance tendency significantly attenuates the effect of policies encouraging active commuting, suggesting intertwined relationships between different public health targets.
"Dockless bike sharing alleviates road congestion by complementing subway travel: Evidence from Beijing" with Siqi Zheng
Cities, 2020
Dockless bike sharing provides a flexible transportation alternative, enriching the potential of both a feeder mode for subway access and a direct substitute for subway trips. We use 3.2 million geo-coded bike-sharing trips to evaluate the interaction between dockless shared bikes and the existing subway system in Beijing. Employing a generalized Difference-in-Differences identification strategy, we find that subway lines with high bike-sharing intensity showed an 8% larger growth rate in subway ridership, and the rush hour road congestion around those subway stations drops by 4%. We demonstrate that the synergy between dockless bike sharing and the subway system outweighs substituting effects and can help achieve a greener and healthier city.
"Low-carbon innovation induced by emissions trading in China" with Junming Zhu, Xinghua Deng, Lan Xue
Nature Communications, 2019
Emissions trading scheme (ETS) has been adopted by an increasing number of countries and regions for carbon mitigation, but its actual effect depends on specific program design and institutional context. We provide firm-level evidence of the innovation effect directly from China’s pilot emissions trading, based on latest patenting information and a quasi-experimental design. We show that China’s pilots increase low-carbon innovation of ETS firms by 5–10% without crowding out their other technology innovation, and the effect is driven by mass-based allowance allocation.