Publications
"Climate Rather than Overgrazing Explains Most Rangeland Primary Productivity Change in Mongolia" (with Tumenkhusel Avirmed, Steven W. Wilcox and Christopher B. Barrett)
Science, (2025) 389 (6766): 1229-1233
Science, (2025) 389 (6766): 1229-1233
|
Understanding variation in rangeland conditions at scale is crucial because rangelands are the dominant land type on Earth, supporting the livelihoods of over two billion people. Using spatially disaggregated, nationwide data from Mongolia, 1984-2024, and seasonal variation in grazing locations we generate quasi-experimental estimates of the causal effects of livestock herd size, weather, and climate change on rangeland biological productivity. At inter-annual frequency, herd size has a significant but modest, negative effect on biological productivity that varies by agro-ecosystems. But these effects are an order of magnitude smaller than those from weather fluctuations. At decadal scale, over which herders can adapt, herd size effects disappear and temperature dominates. In Mongolia, climate change seems to drive long-term change in rangeland biological productivity.
|
Identification Strategy
|
“The Marginal Cost of Traffic Congestion and Road Pricing: Evidence from a Natural Experiment in Beijing” (with Jun Yang and Shanjun Li)
American Economic Journal: Economic Policy, (2020) 12(1): 418-53
American Economic Journal: Economic Policy, (2020) 12(1): 418-53
Marginal External Cost of Congestion
|
Severe traffic congestion is ubiquitous in large urban centers. This paper provides the first causal estimate of the relationship between traffic density and speed and optimal congestion charges using real-time fine-scale traffic data in Beijing. The identification relies on plausibly exogenous variation in traffic density induced by Beijing’s driving restriction policy. Optimal congestion charges range from 5 to 39 cents per km depending on time and location. Road pricing would increase traffic speed by 11 percent within the city center and lead to an annual welfare gain of 1.5 billion Yuan from reduced congestion and revenue of 10.5 billion Yuan.
|
“Does Subway Expansion Improve Air Quality?” (with Shanjun Li, Yanyan Liu, and Lin Yang)
Journal of Environmental Economics and Management, (2019) 96: 213-235
|
Major cities in China and many other fast-growing economies are expanding their subway systems in order to address worsening air pollution and traffic congestion. This paper quantifies the impact of subway expansion on air quality by leveraging fine-scale air quality data and the rapid build-out of 14 new subway lines and 252 stations in Beijing from 2008 to 2016. Our main empirical framework examines how the density of the subway network affects air quality across different locations in the city during this period. To address the potential endogenous location of subway stations, we construct an instrument based on historical subway planning, long before air pollution and traffic congestion were of concern. Our analysis shows that an increase in subway density by one standard deviation improves air quality by two percent and the result is robust to a variety of alternative specifications including the distance-based difference-in-differences method. The total discounted health benefit during a 20-year period from reduced mortality and morbidity as a result of 14 new subway lines amounts to $1.0–3.1 billion, or only 1.4–4.4 percent of the total construction and operating cost.
|
Subway Network Density in Beijing
|
“Leveraging Geospatial Techniques and Publicly Available Datasets to Develop a Cost-Effective, Digitized National Sampling Frame: A Case Study of Armenia” (with Saida Ismailakhunova, Tsenguunjav Byambasuren, and Sarchil Qader)
|
The lack of a reliable national sampling frame poses a major challenge for conducting representative population and household surveys, particularly in developing countries affected by displacement and rapid territorial change. This study addresses this gap by developing Armenia’s first digitized national sampling frame, where accessible survey frames are severely limited. We introduce an innovative pre-EA tool to semi-automatically construct the digital sampling frame using publicly available datasets. Compared with traditional approaches, this method outperforms in several ways: it enables rapid, semi-automated frame construction, minimizes resource requirements, eliminates geometric errors associated with manual digitization, and produces pre-census EAs (pre-EAs) that both nest within administrative boundaries and align with visible ground features. The approach also integrates gridded population data to reflect recent urbanization and migration, generating pre-census EAs and urban–rural classifications suitable for national surveys. The sampling frame was successfully applied in the World Bank’s “Listening to Armenia” survey. Overall, the study demonstrates that automated, data-driven approaches can efficiently produce accurate, scalable, and adaptable national sampling frames, offering potential utility in other countries facing similar constraints.
|
Policy work
"Energy Tariff Reform in Uzbekistan: Efficiency, Distributional Impacts, and Social Concerns" (with William Seitz)
Estimates household and commercial demand elasticities for electricity and gas from administrative utility data and the Listening-to-Citizens monthly panel, measures the welfare loss of cost-recovery tariffs across the income distribution, and evaluates mitigation options; block-tariff design and targeted social assistance emerge as the most effective and most publicly supported.
"Electricity Tariff Reform and Energy Affordability in Georgia: Distributional Impacts and Policy Implications" (with Saida Ismailakhunova, Juan Palacios Mora, and Tsenguunjav Byambasuren)
Despite strong recent economic growth and declines in monetary and electricity poverty, energy affordability remains a concern for a significant share of households in Georgia. This paper examines the distributional and welfare implications of electricity tariff reforms using household survey data, new estimates of demand elasticities, and projections to 2030. Three findings emerge. First, electricity demand is inelastic and heterogeneous, limiting households’ ability to adjust consumption when prices rise. Second, energy and electricity poverty are unevenly distributed, disproportionately affecting lower-income and vulnerable households, while existing subsidies exhibit substantial leakage to higher-income groups. Third, projections indicate that income growth is the primary driver of reductions in energy and electricity poverty across all tariff reform scenarios. While tariff increases may raise affordability concerns, their adverse welfare effects are projected to be modest and offset by sustained economic growth. These results highlight the importance of combining tariff reforms with better-targeted social support, energy efficiency investments, and continued economic growth to mitigate potential welfare losses.
"Elasticity of Electricity Demand in the Kyrgyz Republic"
Estimates residential and non-residential price elasticities of electricity demand from monthly administrative records across all regional distribution networks, using a dynamic model that accounts for seasonality and the lifeline-block tariff structure; provides a reusable elasticity benchmark for tariff analysis and forecasting.
"Working for Yourself or for Your Kids? Childcare Expansion Policy in Uzbekistan" (with Dilnovoz Abdurazzakova and Chiyu Niu)
In developed countries, public childcare programs have increased maternal employment by easing time constraints. However, their impact in lower-middle-income settings with multigenerational households is less understood. In Uzbekistan, for instance, many households include multiple adult women, such as grandmothers and aunts, who traditionally do not work and can provide informal childcare, potentially lowering the demand for public services. This paper examines the effects of a recent preschool expansion policy (2018–22) on women’s labor market participation in this context. To assess the policy’s causal impact, the paper utilizes variations in childcare coverage across districts over time. The results show that the childcare expansion policy led to a 12 percent average increase in female labor supply, with the strongest effects observed for families that value education but face financial constraints. In contrast, the availability of informal caregivers does not decrease the policy’s effect. These findings challenge the idea that time constraints are the primary mechanism linking childcare expansion to women’s employment. Instead, in this context, economic factors—especially the need to afford childcare costs—emerge as the main drivers.
"Dynamically Identifying Community-Level COVID-19 Impact Risks: Uzbekistan" (with William Seitz, Eldor Tulyakov, Obid Khakimov, and Sevilya Muradova)
We build a new database of highly spatially disaggregated indicators related to risk and resilience to the social and economic impacts of the COVID-19 pandemic in Uzbekistan. The outbreak disproportionately affects particular groups – the elderly, the poor, those living in areas under lockdown, and families who rely on remittance income are all examples of groups that are especially vulnerable to effects of the crisis in Uzbekistan. We assemble indicators summarizing concentrations of these and other risk factors at the lowest administrative level in the country, neighborhood-sized units called mahallas. Local official administrative statistics (published for the first time in this study) are combined with monthly panel survey data from the ongoing Listening to the Citizens of Uzbekistan project to produce an overall risk index, which is decomposable by dimension or risk factor to inform targeted and issue-specific responses. We then demonstrate a process for updating key indicators (such as employment or remittance flows) on a monthly basis using linked survey data combined with small area estimation techniques. These neighborhood-level results are intended to improve resource allocation decisions and are particularly relevant in Uzbekistan where local representatives are responsible for implementing key social and economic programs to respond to the outbreak.
Real-time welfare monitoring. I lead the World Bank's Listening-to high-frequency household survey series across Central Asia and Ukraine and serve on the core team of the global High Frequency Monitoring program, producing quarterly poverty updates for Ukraine, food-security and energy-hardship monitoring, and reusable survey and analysis tools used across country teams.
Interactive data stories (Listening to the Philippines). Public, reproducible analyses I build from the Listening-to household and phone surveys:
- Baseline Household Survey, Philippines 2025
- Building a welfare ranking from observable assets
- Monthly phone survey, tracking recovery (Rounds 1–8)
- A real-time data-quality dashboard
Estimates household and commercial demand elasticities for electricity and gas from administrative utility data and the Listening-to-Citizens monthly panel, measures the welfare loss of cost-recovery tariffs across the income distribution, and evaluates mitigation options; block-tariff design and targeted social assistance emerge as the most effective and most publicly supported.
"Electricity Tariff Reform and Energy Affordability in Georgia: Distributional Impacts and Policy Implications" (with Saida Ismailakhunova, Juan Palacios Mora, and Tsenguunjav Byambasuren)
Despite strong recent economic growth and declines in monetary and electricity poverty, energy affordability remains a concern for a significant share of households in Georgia. This paper examines the distributional and welfare implications of electricity tariff reforms using household survey data, new estimates of demand elasticities, and projections to 2030. Three findings emerge. First, electricity demand is inelastic and heterogeneous, limiting households’ ability to adjust consumption when prices rise. Second, energy and electricity poverty are unevenly distributed, disproportionately affecting lower-income and vulnerable households, while existing subsidies exhibit substantial leakage to higher-income groups. Third, projections indicate that income growth is the primary driver of reductions in energy and electricity poverty across all tariff reform scenarios. While tariff increases may raise affordability concerns, their adverse welfare effects are projected to be modest and offset by sustained economic growth. These results highlight the importance of combining tariff reforms with better-targeted social support, energy efficiency investments, and continued economic growth to mitigate potential welfare losses.
"Elasticity of Electricity Demand in the Kyrgyz Republic"
Estimates residential and non-residential price elasticities of electricity demand from monthly administrative records across all regional distribution networks, using a dynamic model that accounts for seasonality and the lifeline-block tariff structure; provides a reusable elasticity benchmark for tariff analysis and forecasting.
"Working for Yourself or for Your Kids? Childcare Expansion Policy in Uzbekistan" (with Dilnovoz Abdurazzakova and Chiyu Niu)
In developed countries, public childcare programs have increased maternal employment by easing time constraints. However, their impact in lower-middle-income settings with multigenerational households is less understood. In Uzbekistan, for instance, many households include multiple adult women, such as grandmothers and aunts, who traditionally do not work and can provide informal childcare, potentially lowering the demand for public services. This paper examines the effects of a recent preschool expansion policy (2018–22) on women’s labor market participation in this context. To assess the policy’s causal impact, the paper utilizes variations in childcare coverage across districts over time. The results show that the childcare expansion policy led to a 12 percent average increase in female labor supply, with the strongest effects observed for families that value education but face financial constraints. In contrast, the availability of informal caregivers does not decrease the policy’s effect. These findings challenge the idea that time constraints are the primary mechanism linking childcare expansion to women’s employment. Instead, in this context, economic factors—especially the need to afford childcare costs—emerge as the main drivers.
"Dynamically Identifying Community-Level COVID-19 Impact Risks: Uzbekistan" (with William Seitz, Eldor Tulyakov, Obid Khakimov, and Sevilya Muradova)
We build a new database of highly spatially disaggregated indicators related to risk and resilience to the social and economic impacts of the COVID-19 pandemic in Uzbekistan. The outbreak disproportionately affects particular groups – the elderly, the poor, those living in areas under lockdown, and families who rely on remittance income are all examples of groups that are especially vulnerable to effects of the crisis in Uzbekistan. We assemble indicators summarizing concentrations of these and other risk factors at the lowest administrative level in the country, neighborhood-sized units called mahallas. Local official administrative statistics (published for the first time in this study) are combined with monthly panel survey data from the ongoing Listening to the Citizens of Uzbekistan project to produce an overall risk index, which is decomposable by dimension or risk factor to inform targeted and issue-specific responses. We then demonstrate a process for updating key indicators (such as employment or remittance flows) on a monthly basis using linked survey data combined with small area estimation techniques. These neighborhood-level results are intended to improve resource allocation decisions and are particularly relevant in Uzbekistan where local representatives are responsible for implementing key social and economic programs to respond to the outbreak.
Real-time welfare monitoring. I lead the World Bank's Listening-to high-frequency household survey series across Central Asia and Ukraine and serve on the core team of the global High Frequency Monitoring program, producing quarterly poverty updates for Ukraine, food-security and energy-hardship monitoring, and reusable survey and analysis tools used across country teams.
Interactive data stories (Listening to the Philippines). Public, reproducible analyses I build from the Listening-to household and phone surveys:
- Baseline Household Survey, Philippines 2025
- Building a welfare ranking from observable assets
- Monthly phone survey, tracking recovery (Rounds 1–8)
- A real-time data-quality dashboard
Working papers
“Roads and Capital Misallocation: Evidence from India’s Infrastructure Boom" (with Anastasia Burya, Martino Pelli, and Jeanne Tschopp). Under review.
|
We study the effects of India's road network expansion on capital allocation across firms and aggregate productivity. Focusing on road improvements between 2011 and 2019, largely driven by the National Highways Development Project, we construct a market access measure using historical OpenStreetMap data that captures changes in travel times between postal codes and major cities. Using a staggered difference-in-differences design, we find that improved market access is associated with reductions in capital misallocation and increases in aggregate productivity in treated postal codes, primarily through better capital allocation rather than within-firm efficiency gains. Firms experience a 25% increase in capital, with ex ante high marginal revenue product of capital (MRPK) firms seeing an additional 25% growth and a 45% decline in MRPK, consistent with reduced capital misallocation. We estimate that the Solow residual rises by 2.7-5%, a gain comparable to the lower-bound effects of India's earlier foreign capital liberalization. We show that these allocative gains are driven primarily by reductions in input wedges rather than by changes in markups, indicating that road infrastructure improves efficiency mainly by alleviating physical constraints on input access.
|
"Which Climate Channel Governs Rangeland Productivity? Channel Attribution across Timescales and Biomes" (with Tumenkhusel Avirmed, Steven W. Wilcox, and Christopher B. Barrett).
|
Climate, not livestock, governs Mongolian rangeland productivity (Purevjav, Avirmed, Wilcox & Barrett 2025), but "climate" in that result is a single composite. We open it into four mechanistically distinct channels: heat exposure, precipitation, wind (aerodynamic/evaporative stress), and snow. We use the same seasonal-grazing instrumental-variable design and a Shapley–Owen variance decomposition, and we do so at two timescales. The central result is that channel dominance is timescale-dependent. Precipitation governs interannual productivity variation (41% of the within-climate signal net of structure; 95% bootstrap CI 34–45%; separable from the lower channels), but heat governs the decadal trend (about 40%; CI 31–47%; separable from precipitation; robust to both 10- and 20-year differencing), and heat's effect is negative in every zone, an evaporative / vapour-pressure-deficit stress, not growth-enabling warmth. This reconciles a standing disagreement in the literature: temperature-dominant and precipitation-dominant findings reflect different horizons. The policy implication is a split mandate: precipitation-deficit management for the short run, and warming-aridification adaptation for the long run.
|
|
"The Queue Behind You: Pricing the Network Externality of Traffic Congestion in Beijing" (with Matías Navarro and Shanjun Li).
When a road jams, the queue spills backward onto the links that feed it, an externality that own-cost tolls do not price. Using two-minute detector data for 1,500 Beijing road segments, we identify this spillback from closure events, sudden flow collapses that shock one segment's density, and fold it into the optimal toll. The estimated transmission varies with the congestion state of the receiving link and is flat in the size of the shock, down to mild flow drops an order of magnitude gentler than the closures. The results suggest the network externality is about a third of the optimal toll, 17 to 37% once kernel, first-stage, and predecessor-set uncertainty are propagated jointly, and as much as two-fifths once the operating-point factor is corrected. It falls mainly on the already-congested cores, and it reprices a tail of roads that own-cost tolls miss, where own congestion is mild but the queue backs onto costly neighbours. Pricing it raises the welfare gain from congestion charges by a sixth to two-fifths.
Traffic congestion is among the most pressing challenges facing large cities in developing countries, and governments respond with a range of policies whose relative effectiveness is rarely measured against the travel demand they are meant to manage. In this paper, I estimate the demand for travel in Beijing using the Household Travel Survey of 2010 and 2014, and I reconstruct each traveler's full choice set, including the travel time and cost of all six modes, by leveraging Baidu and Gaode Maps data and a GIS reconstruction of the subway network. A mixed logit model yields a median value of travel time of about 40 yuan per hour, roughly 60 percent of the market wage. I then embed the demand estimates in an equilibrium simulation built on the speed-density relationship of Yang et al. (2020), and I evaluate five congestion policies under a common government-budget discipline. Distance-based road pricing and a cordon toll are the only self-financing policies, and they deliver the largest welfare gains once public funds carry a premium, while Beijing's driving restriction reduces welfare. The benefits of transit fare subsidies rise with income and the burden of road pricing rises with income, so the equity case usually made against road pricing is not supported by the model.
"One Border, Two Steppes: Comparative Rangeland Quality in Mongolia and China" (with Martino Pelli, Jeanne Tschopp, and Yazhen Gong)
|
We study the impact of Chinese grassland protection policies on grassland quality in Inner Mongolia using the Normalized Difference Vegetation Index (NDVI) and a Regression Discontinuity Design, leveraging the Mongolia-Inner Mongolia border as a source of exogenous variation. Our results suggest these policies improved grassland quality by 11%. Using existing valuations, we estimate the annual benefits of these improvements to range from USD 15.2 billion to USD 24.3 billion, with an additional USD 330 million from avoided sandstorms in Beijing. The Net Present Value of these benefits over 50 years is USD 168-270 billion, far exceeding the estimated USD 9.6 billion cost of conservation programs in Inner Mongolia. These findings highlight the strong economic case for continued investment in grassland protection policies.
|
Work in progress
"When the Road to Russia Closed: Global Shocks and Household Labor Reallocation in Uzbekistan".
Using an 82-round monthly household panel, shows that when COVID-19 closed the Russia migration corridor, households shifted members from migration into local work without a deterioration in food security, while the war-induced wheat-price shock left migration unchanged, identifying the corridor rather than prices as the binding channel.
"Extreme Heat, Male Presence, and Women's Work: Evidence from Uzbekistan" (with Tsenguunjav Byambasuren and William Seitz).
A week with at least one hour above 37°C reduces women's employment probability by 2.6 percentage points with no effect on men; the effect appears only when a working-age man is present in the household, consistent with a male-enforced participation restriction rather than occupational heat exposure.
"Compound Shocks and Household Food Insecurity in Wartime Ukraine, 2023–2026" (with Judy Yang and Obert Pimhidzai).
Tracks food insecurity over three years of war using 30 survey waves and 23,000 FIES responses Rasch-equated to the FAO global standard; prevalence is U-shaped and does not recover after the winter 2024–25 grid attacks, with war damage, income disruption, and energy disruption the largest correlates.
"Food Insecurity and Mental Health in Wartime Ukraine: Evidence from a Household Panel".
Food insecurity and poor mental health are substantially one household caseload: one standard deviation of food insecurity associates with 0.23–0.31 SD worse mental-health and life-satisfaction outcomes, a gradient that is monotone and stable across household types, regions, and frontline proximity.
"Security over Income: Why Ukrainian Workers Prefer Formal Jobs over Simplified-Tax Self-Employment" (with Judy Yang and Obert Pimhidzai).
Two survey experiments show 72% of Ukrainian workers choose a formal job over simplified-tax self-employment at equal pay, and a list experiment finds no evidence of concealed disguised employment, favoring reforms that cut formal hiring costs over worker-focused enforcement.
"Locked in Place? Home Ownership, Displacement, and Labour Mobility in Wartime Ukraine" (with Judy Yang and Obert Pimhidzai).
Near-universal mortgage-free home ownership and a thin rental market brake labor mobility: owners are far less likely than renters to have moved since February 2022, suggesting rental-market de-frictioning is labor-market policy.
"Social Exclusion, Relative Standing, and Subjective Well-Being: Evidence from Six Transition Countries" (with Tauhidur Rahman).
Across six transition countries, multidimensional social exclusion is associated with a 13-percentage-point lower probability of reporting life satisfaction, and within-district consumption rank absorbs the own-consumption effect on well-being.
"Forecasting the 2026 FIFA World Cup: A Market-Calibrated Poisson–Elo Model and the Information Each Result Reveals".
Combines margin-stripped bookmaker odds with Elo-based knockout simulation across 200,000 tournament draws, with all forecasts locked at kickoff and revisions measured match-by-match by Kullback–Leibler divergence.
Research in progress: rangelands and climate
Building on the Science (2025) result that climate rather than herd size drives long-run rangeland productivity change in Mongolia, I am developing a broader research program on coupled climate–pastoral systems: how climate change moves through rangeland productivity to carrying capacity and pastoral livelihoods, how herders adapt across margins, and what governance regimes change these outcomes, from cross-border management comparisons to community pasture institutions.
“Does Transport Infrastructures Create Resilience to Hurricanes? Evidence from India” (with Jeanne Tschopp, and Martino Pelli)
“Understanding Heterogeneity in the Value of Time” (with Ziye Zhang and Shanjun Li)
Using an 82-round monthly household panel, shows that when COVID-19 closed the Russia migration corridor, households shifted members from migration into local work without a deterioration in food security, while the war-induced wheat-price shock left migration unchanged, identifying the corridor rather than prices as the binding channel.
"Extreme Heat, Male Presence, and Women's Work: Evidence from Uzbekistan" (with Tsenguunjav Byambasuren and William Seitz).
A week with at least one hour above 37°C reduces women's employment probability by 2.6 percentage points with no effect on men; the effect appears only when a working-age man is present in the household, consistent with a male-enforced participation restriction rather than occupational heat exposure.
"Compound Shocks and Household Food Insecurity in Wartime Ukraine, 2023–2026" (with Judy Yang and Obert Pimhidzai).
Tracks food insecurity over three years of war using 30 survey waves and 23,000 FIES responses Rasch-equated to the FAO global standard; prevalence is U-shaped and does not recover after the winter 2024–25 grid attacks, with war damage, income disruption, and energy disruption the largest correlates.
"Food Insecurity and Mental Health in Wartime Ukraine: Evidence from a Household Panel".
Food insecurity and poor mental health are substantially one household caseload: one standard deviation of food insecurity associates with 0.23–0.31 SD worse mental-health and life-satisfaction outcomes, a gradient that is monotone and stable across household types, regions, and frontline proximity.
"Security over Income: Why Ukrainian Workers Prefer Formal Jobs over Simplified-Tax Self-Employment" (with Judy Yang and Obert Pimhidzai).
Two survey experiments show 72% of Ukrainian workers choose a formal job over simplified-tax self-employment at equal pay, and a list experiment finds no evidence of concealed disguised employment, favoring reforms that cut formal hiring costs over worker-focused enforcement.
"Locked in Place? Home Ownership, Displacement, and Labour Mobility in Wartime Ukraine" (with Judy Yang and Obert Pimhidzai).
Near-universal mortgage-free home ownership and a thin rental market brake labor mobility: owners are far less likely than renters to have moved since February 2022, suggesting rental-market de-frictioning is labor-market policy.
"Social Exclusion, Relative Standing, and Subjective Well-Being: Evidence from Six Transition Countries" (with Tauhidur Rahman).
Across six transition countries, multidimensional social exclusion is associated with a 13-percentage-point lower probability of reporting life satisfaction, and within-district consumption rank absorbs the own-consumption effect on well-being.
"Forecasting the 2026 FIFA World Cup: A Market-Calibrated Poisson–Elo Model and the Information Each Result Reveals".
Combines margin-stripped bookmaker odds with Elo-based knockout simulation across 200,000 tournament draws, with all forecasts locked at kickoff and revisions measured match-by-match by Kullback–Leibler divergence.
Research in progress: rangelands and climate
Building on the Science (2025) result that climate rather than herd size drives long-run rangeland productivity change in Mongolia, I am developing a broader research program on coupled climate–pastoral systems: how climate change moves through rangeland productivity to carrying capacity and pastoral livelihoods, how herders adapt across margins, and what governance regimes change these outcomes, from cross-border management comparisons to community pasture institutions.
“Does Transport Infrastructures Create Resilience to Hurricanes? Evidence from India” (with Jeanne Tschopp, and Martino Pelli)
“Understanding Heterogeneity in the Value of Time” (with Ziye Zhang and Shanjun Li)