The Stockbridge Dispatch : 2025

Ana Ospina Tobar

Presented, "Nudges in a Simulated Environment and Their Effects on Use of Resilience-Enhancing Technologies"

Abstract: 
For smallholding farmers, learning about new technologies often means risking household resources. This paper reports on efforts to overcome this barrier using a tablet-based game. The game allows farmers to experiment with drought tolerant maize and index insurance in a low-cost environment. In a sample of 235 farmers in Manica Province, Mozambique, we find that experiencing drought in one round of the game is associated with a statistically significantly increase in adoption of the technologies in subsequent rounds. Highlighting the effects of drought on the farmer’s family and community, also increased adoption of the risk management technologies. When evaluating the long-term impact of the game in real-world adoption, we find that farmers who experienced more severe drought seasons during the gameplay, are more likely to purchase the drought-tolerant maize variety.

Rajib Rahman

Presented "Firms' Advertising Strategies and Effectiveness of Traditional Advertising under Sugar Sweetened Beverage (SSB) Tax Policy"

Abstract: I estimate the causal effects of soda tax policy on firms' advertising strategies and its effect on advertising elasticity in Berkeley across major diet and regular soda brands. Using weekly sales and advertising spending data from 2014 to 2015, I analyze the impact of Berkeley, California’s soda tax policy, passed in November 2014, on the advertising effectiveness for diet and regular soft drinks across major brands, as well as its impact on firms' advertising expenditure. While existing literature provides evidence of advertising effectiveness across various products from a marketing perspective, I examine how this effectiveness changes under soda tax policies, a public health tool frequently used to mitigate the health risks of excessive sugar intake from soft drink consumption. I find  significant evidence that long-run own traditional advertising elasticity of regular soda declined by 0.023 under such policies in Berkeley, indicating that advertising became significantly less effective in the taxed category compared to untaxed beverages. I also find significant evidence that firms reduced advertising spending for regular soda by approximately 16.4\% overall and by 16.1\%  and 39.2\% relative to diet soda and light beer, respectively, in the post-policy Berkeley. Additionally, firms adopted spatial marketing strategies, decreasing advertising spending near Berkeley while increasing it in more distant areas. These findings offer new insights into soda tax policies and have important implications for manufacturers' pricing and marketing strategies, as well as for policymakers regarding public health outcomes.

Bijesh Gyawali

presented, "Anti-Bullying Laws and Weight-Based Disparities in Suicidality"

Abstract: Appearance-based bullying is common among teenagers and may inflict substantial psychological harm on its victims. Overweight and obese students are both more likely to be bullied at school and more likely to engage in suicidal behaviors than their healthy-weight counterparts. This study is the first to explore how anti-bullying laws (ABLs) affect disparities in suicidality between overweight and obese U.S. high school students compared to their and healthy-weight counterparts. Using data from the National and State Youth Risk Behavior Surveys and a difference-in-differences approach, we find that ABL adoption is associated with a 6-19 percent reduction in suicidal behaviors among overweight or obese teens; estimates for healthy-weight teens are considerably smaller in magnitude and statistically insignificant. Weight-based disparities in suicidal behaviors are reduced most by ABLs among obese teenage girls. An exploration of mechanisms suggests that improvements in the quality of peer interactions in school – rather than ABL-induced changes in body weight (sample selection) or students’ own-weight perception – generate disparate mental health gains for at-risk youth. We conclude that curbing targeted bullying based on appearance yields important health benefit

Rajib Rahman

Presented "Firms' Advertising Strategies and Effectiveness of Traditional Advertising under Sugar Sweetened Beverage (SSB) Tax Policy"

Abstract: I estimate the causal effects of soda tax policy on firms' advertising strategies and its effect on advertising elasticity in Berkeley across major diet and regular soda brands. Using weekly sales and advertising spending data from 2014 to 2015, I analyze the impact of Berkeley, California’s soda tax policy, passed in November 2014, on the advertising effectiveness for diet and regular soft drinks across major brands, as well as its impact on firms' advertising expenditure. While existing literature provides evidence of advertising effectiveness across various products from a marketing perspective, I examine how this effectiveness changes under soda tax policies, a public health tool frequently used to mitigate the health risks of excessive sugar intake from soft drink consumption. I find  significant evidence that long-run own traditional advertising elasticity of regular soda declined by 0.023 under such policies in Berkeley, indicating that advertising became significantly less effective in the taxed category compared to untaxed beverages. I also find significant evidence that firms reduced advertising spending for regular soda by approximately 16.4\% overall and by 16.1\%  and 39.2\% relative to diet soda and light beer, respectively, in the post-policy Berkeley. Additionally, firms adopted spatial marketing strategies, decreasing advertising spending near Berkeley while increasing it in more distant areas. These findings offer new insights into soda tax policies and have important implications for manufacturers' pricing and marketing strategies, as well as for policymakers regarding public health outcomes.
 

Said Arslan

Presented "Energy Choice, Air Pollution, and Public Health: Evidence from Turkey's Natural Gas Transition"

Abstract: This paper examines the public health impacts of Turkey’s nationwide transition from solid fuels to natural gas. Using province-level data from 2011 to 2019 and a staggered rollout of residential natural gas infrastructure, I investigate whether increased natural gas consumption leads to reductions in air pollution-related health outcomes, including cardiovascular and respiratory (CV-R) mortality, hospital admissions, and inpatient bed-days. Employing dynamic difference-in-differences, event studies, and instrumental variables methods, I find significant health benefits from residential natural gas adoption. Specifically, a one-unit in residential natural gas consumption results in a 1.8% reduction in CV-R mortality and notable declines in hospital utilization. The analysis further reveals that infants benefit substantially, experiencing a 1.3% reduction in mortality. Crucially, the health improvements are driven primarily by residential natural gas use rather than industrial consumption, as residential adoption directly replaces highly polluting solid fuels. These results highlight substantial short- to medium-term public health gains achievable through cleaner transitional energy sources like natural gas in developing countries, offering critical policy insights for similar energy transition strategies elsewhere.

Dan Ai

presented "Heat, Power Outages, and Mortality in the United States"

Abstract: 
We examine the mortality impacts of exposures to high temperatures and power outages using a county-level data set in the United States from 2015 to 2019. We find that each additional hour of power outage leads to a 0.04% increase in the monthly mortality rate. Furthermore, we show that the mortality effects of hot days are exacerbated by the co-occurrence of power outages, with each hour of power outage on a hot day increasing the harm from the hot day by 61%. Widespread and long-lasting power outages during hot days have disproportionately large effects on mortality. We also show heterogeneity across climate regions in the estimated relationships, which is consistent with heavier reliance on technological adaptations to heat such as air conditioning in hotter climate regions. Taken together, our results suggest that the reliability of electricity grids serves as an important means of adaptation to high temperatures and climate change.

Dhiroj Koirala

presented "Assessing the Economic Impacts of the Bhutanese Electricity Subsidy Program"

Abstract: This study examines the economic impacts of Bhutan’s electricity subsidy program, implemented in two phases (2013 and 2020). Using repeated cross-sectional data from five rounds of the Bhutan Living Standard Surveys (2003, 2007, 2012, 2017, and 2022), the study employs a staggered Difference-in-Differences (DiD) framework to exploit variation in subsidy eligibility. Rural households consuming less than 100 kWh/month received subsidies in 2013, with an expansion in 2020 offering up to 200 kWh/month for highland households. This staggered rollout allows for a dynamic evaluation of short- and medium-term treatment effects. The empirical strategy accounts for time-invariant unobserved heterogeneity through a cohortbased pseudo-panel approach, with cluster-robust standard errors at the Geyog level. Parallel trends are validated via pre-treatment placebo tests, and robustness checks include falsification tests, alternative specifications, and sensitivity analyses. Results show that the subsidy reduced monthly firewood expenditures by 30% post-2013 and an additional 15% post-2020, consistent with substitution effects from consumer theory. Electricity consumption increased by 40% initially and by another 20% after the 2020 expansion, contributing to a 65% rise in appliance ownership, especially electric cooking and heating devices. Welfare improvements include a 25% increase in disposable income for non-energy expenditures like education and healthcare and a 20% reduction in respiratory illnesses from decreased indoor air pollution. A distributional analysis using quantile regressions shows that 78% of initial subsidy benefits accrued to the lowest income quintile, with the 2020 expansion covering 85% of low-income highland households. Fiscal incidence analysis reveals that while the subsidy accounts for 2.8% of Bhutan’s GDP, its welfare gains surpass the fiscal cost, supported by cost-benefit analyses and sensitivity checks. This study contributes to energy economics literature by integrating microeconomic theory and econometric analysis to evaluate targeted subsidies. It highlights the effectiveness of phased electricity subsidies in transforming household energy portfolios, improving welfare, and supporting environmental sustainability, offering policy insights for subsidy design and fiscal management in developing countries.

Maryam Feyzollahi

presented "The Short-Term Price Elasticity of Wholesale Electricity Demand: A Double Machine Learning Approach"

Abstract: 
This study estimates the hourly price elasticity of wholesale electricity demand using data from the New York Independent System Operator market spanning January 2021 to June 2022. We address the fundamental simultaneity between electricity prices and quantities through an instrumental variable approach that exploits wind speed variation to generate exogenous price movements via merit order effects. Our analysis employs both traditional two-stage least squares estimation and Double Machine Learning methods that relax parametric assumptions while preserving causal identification. Using 144,144 hourly observations across eleven load zones, we find statistically significant negative price elasticity ranging from -0.031 to -0.085 depending on the econometric approach. While ordinary least squares yields misleading positive correlations due to simultaneity bias, our instrumental variable estimates reveal genuine demand responses that align with economic theory. The results demonstrate that wholesale electricity consumers do respond to price signals within short time horizons, though these responses remain economically small due to technical and behavioral constraints on rapid consumption adjustments. These findings provide the first estimates of hourly price elasticity for a major U.S. wholesale electricity market and offer important insights for electricity market design and renewable energy integration policies.

Nima Rafizadeh

presented "Targeted Interpretability for Causal Accuracy: Evidence from the Price Elasticity of Gasoline Demand"

Abstract: 
This paper introduces targeted interpretability to address the fundamental trade-off between interpretability and accuracy in causal inference. Traditional econometric approaches impose uniform interpretability requirements across all parameters, sacrificing accuracy when relationships are complex. The paper proposes selective preservation of interpretability only where economic theory demands it for policy analysis, while allowing flexibility elsewhere. The approach is implemented through Double/Debiased Machine Learning applied to gasoline demand estimation using a novel weekly state-level dataset spanning 1992-2022. Using rotary rig counts as instruments, the paper reveals that targeted interpretability yields price elasticity estimates substantially different from traditional methods (-0.177 vs -0.512). The findings suggest conventional approaches may mischaracterize consumer responses to price incentives, with significant implications for policy design. The framework demonstrates how economic theory can guide optimal allocation of interpretability requirements, enhancing both statistical precision and policy effectiveness across empirical economics.

Ana Ospina Tobar

Presented "Nudges in a Simulated Environment and Their Effects on Use of Resilience-Enhancing Technologies"

Abstract:
For smallholding farmers, learning about new technologies often means risking household resources. This paper reports on efforts to overcome this barrier using a tablet-based game. The game allows farmers to experiment with drought tolerant maize and index insurance in a low-cost environment. In a sample of 235 farmers in Manica Province, Mozambique, we find that experiencing drought in one round of the game is associated with a statistically significantly increase in adoption of the technologies in subsequent rounds. Highlighting the effects of drought on the farmer’s family and community, also increased adoption of the risk management technologies. When evaluating the long-term impact of the game in real-world adoption, we find that farmers who experienced more severe drought seasons during the gameplay, are more likely to purchase the drought-tolerant maize variety.

Prashikdivya Gajbhiye

Presented "Weather Shocks, Workfare, and the Farmer Suicide Crisis in India"

Abstract: 300,000 Indian farmers have taken their lives in the past two decades, creating a national crisis. This marginalized population is particularly vulnerable to weather-induced crop failures. I examine how weather shocks affect suicide rates, and whether government assistance programs can mitigate these effects, among the more than 120 million people in the Indian state of Maharashtra. I assembled a novel dataset of daily farmer suicides at the village level. Using climate econometrics regressions, I find that hot temperatures have a greater effect on farmer suicides than previously believed. I further find that government assistance can greatly reduce the magnitude of these effects. These findings implicate and emphasize the role of government programs to solve this crisis. The estimated cost per life saved on a single 35°C Day is approximately $183,000, while the VSL is $138.51 million, underscoring the substantial net social benefit of the program.
 

Rui Wang

Presented "Risk Preferences and Household Decision to Adopt Solar Panels"

Abstract:
Individuals making a decision about whether to adopt rooftop solar PV technology face significant financial risk, including technology underperformance, unexpected maintenance and repairs, and changes in policy regarding solar PV incentives. The presence of risk as a factor in the adoption decision suggests that individuals with a higher level of risk tolerance are more likely to adopt the technology than those who are risk averse. In addition, early adopters are also more likely to be risk-tolerant than late adopters. In this paper, we use a lab-in-the-field experiment to elicit individual risk preferences from a subject pool of solar PV adopters and non-adopters, and use this data to examine the effect of risk preference on the decision to adopt solar PV. Our findings confirm our hypothesis that risk preference plays a crucial role in determining solar PV adoption status and the timing of adoption. Our findings suggest that reducing risk in the solar market through policy or through risk-mitigating insurance products can help to broaden solar PV adoption among households.

Dan Ai

Presented "Heat, Power Outages and Mortality in the United States"

Abstract:
We examine the mortality impacts of exposures to high temperatures and power outages using a county-level data set in the United States from 2015 to 2019. We find that each additional hour of power outage leads to a 0.04% increase in the monthly mortality rate. Furthermore, we show that the mortality effects of hot days are exacerbated by the co-occurrence of power outages, with each hour of power outage on a hot day increasing the harm from the hot day by 61%. Widespread and long-lasting power outages during hot days have disproportionately large effects on mortality. We also show heterogeneity across climate regions in the estimated relationships, which is consistent with heavier reliance on technological adaptations to heat such as air conditioning in hotter climate regions. Taken together, our results suggest that the reliability of electricity grids serves as an important means of adaptation to high temperatures and climate change.

Gayan Udugama

presented "Motivating Costly Action in the Face of Climate Change: Hurricane Sandy and Flood Insurance"

Abstract: 
The purchase of flood insurance represents a costly action that individuals can take to mitigate a portion of the financial risks posed by climate change. Understanding what drives such preemptive actions is vital for fostering widespread adaptation and mitigation efforts. In this paper, we examine the extent to which experiencing a major flood event -- Hurricane Sandy in New York City -- led to increased flood insurance uptake, shedding light on what motivates individuals to undertake costly risk-mitigating actions. Using a difference-in-difference framework with address-level flood insurance data,  we assess the effect of varying experiences with Hurricane Sandy on residential insurance uptake. Our findings reveal that properties directly inundated were 7.1 to 10.4 percentage points (34% to 50%) more likely to purchase flood insurance following Sandy. Those that were very near but not directly inundated saw only a 2.6 percentage point increase. This response quickly falls to zero as the distance from direct harm increases. Taken together, these findings suggest that populations may only be motivated to take action after experiencing direct harm. This presents a significant challenge for those advocating preemptive action in the face of climate change.

PhD Student, Ana Ospina Tobar, selected to participate in the 2025 Berkeley Summer School in Environmental and Energy Economics

The goal of the summer school is to provide doctoral students enrolled in a North American Ph.D. program, who have completed the core first year microeconomics and econometrics requirement with an overview of the most important and current topics in various areas of the fields of environmental and energy economics. Previous incarnations of this summer school have helped first year students identify promising dissertation projects. The instructors’ expertise covers a range of important topics. 

Prashikdivya Gajbhiye

Presented "Weather Shocks, Workfare, and the Farmer Suicide Crisis in India"

Abstract: 300,000 Indian farmers have taken their lives in the past two decades, creating a national crisis. This marginalized population is particularly vulnerable to weather-induced crop failures. I examine how weather shocks affect suicide rates, and whether government assistance programs can mitigate these effects, among the more than 120 million people in the Indian state of Maharashtra. I assembled a novel dataset of daily farmer suicides at the village level. Using climate econometrics regressions, I find that hot temperatures have a greater effect on farmer suicides than previously believed. I further find that government assistance can greatly reduce the magnitude of these effects. These findings implicate and emphasize the role of government programs to solve this crisis. The estimated cost per life saved on a single 35°C Day is approximately $183,000, while the VSL is $138.51 million, underscoring the substantial net social benefit of the program.

Sparshi Srivastava

Presented "Effects of Environmental Exposure on Weight-Related Behaviors and Outcomes"

Abstract: We provide evidence on the role of environmental exposures in shaping weight-related goals, behaviors, and outcomes. Using data from the 1991-2010 Behavioral Risk Factor Surveillance System and an identification strategy leveraging temperature deviations from local seasonal norms, we show that adults were less likely to report trying to lose weight, dieting to lose weight, and exercising to lose weight when temperatures fell below 20-25°C (68-77°F). We then show that temperature increases were associated with increases in respondents’ weekly servings of fruits and vegetables and the number of minutes they spent engaged in moderate or vigorous physical activity. Despite these behavioral changes, we find no evidence of a relationship between temperature and BMI"

Dhiroj Koirala

Presented "Evaluating the Impact of the Low-Income Home Energy Assistance Program (LIHEAP) on Energy Consumption and Household Debt Dynamics in the United States"

Abstract: This study examines the relationship between household debt and residential energy consumption in the United States (US), with a particular focus on the moderating role of the Low-Income Home Energy Assistance Program (LIHEAP). Using county-level panel data from 1999 to 2023, I employ a fixed-effects regression framework and an instrumental variable approach to account for endogeneity in household debt. The findings reveal that higher household debt is associated with lower energy consumption, particularly for natural gas expenditures, suggesting that financially constrained households adjust their heating usage more than electricity consumption. Additionally, the analysis indicates that LIHEAP mitigates the adverse effects of household debt on energy expenditures, particularly for natural gas, 1 ensuring greater energy security for low-income households. The study highlights the importance of targeted energy assistance programs and financial stability policies to address energy insecurity and the broader implications of household debt on essential expenditures. The results provide valuable insights for policymakers aiming to improve energy affordability and financial resilience among vulnerable populations.

Onu Das

Presented "Political Representation, Forest Cover and Development in Rural India"

Abstract: This paper examines how political representation for historically marginalized communities shapes local economic activity and environmental stewardship. Leveraging the quasi‑random rotation of chairperson reservations for Scheduled Castes (SCs) and Scheduled Tribes (STs) across 5,000 village councils in Karnataka, India, over five election cycles (1994–2010), I estimate causal effects on satellite‑measured night‑time lights and forest cover. SC/ST-led councils see higher local economic activity but lower forest cover, with larger effects under ST leadership. These patterns reveal trade‑offs between development and conservation and point to the need for policies that enable marginalized leaders to pursue development without depleting the natural resources their communities depend on.