Publications

Agricultural supply chains play a crucial role in supporting food security in Africa. However, high-resolution supply chain information is often not available, which hinders our ability to determine which interventions in food supply chains would most enhance food security. In this study, we develop a high-resolution supply chain model for essential staple crops in Zambia, aiming to estimate how improvements in transportation infrastructure would impact food security. Specifically, we simulate district-level monthly consumption, trade flows, and storage for maize and cassava in Zambia. We then conduct a counterfactual case study with low transportation costs, discovering that reducing transaction costs leads to higher aggregate net agricultural revenue and aggregate net expenditure. These results indicate that transportation investments are more beneficial to suppliers than to consumers, with implications for household food security in smallholder agriculture. Our study highlights the potential for infrastructure investments to improve food security.

Although it is increasingly recognized that aspirations drive economic behavior and outcomes, it is not fully understood how aspirations are formed (or eroded). However, it has been theoretically established that aspirations are socially constructed and formed under an aspiration window. An aspiration window refers to a cognitive zone of similar individuals based on age, gender, caste, geography, religion, ethnicity, and other social (self-help) groups. In these groups, individuals learn from each other through interaction and experimentation. We examine the relationship between group membership and aspirations. As a proxy for group membership, we use agricultural cooperatives that abound in many developing countries and have been associated with productivity and welfare gains. Given that farmers interact and are exposed to role models in these cooperatives, we also investigate the relationship between exposure to role models and aspirations. We show a positive association between cooperative membership and aspirations as well as between exposure to role models and aspirations. Interacting cooperative membership with exposure to role models, we find a larger association between cooperative membership and aspirations, highlighting the relevance of exposure to role models in these cooperatives. Given the growing evidence regarding the relevance of aspirations in achieving various developmental outcomes, our study highlights some entry points in improving aspirations.

New investments in conservation are needed to halt and reverse the rapid and extensive changes to ecosystems driven by growing human demands for natural resources. A major barrier is matching viable financing solutions to conservation projects. Recent conservation finance studies catalog available financing options but do not provide adequate guidance on which financing pathways are suitable for a particular conservation project. Studies in the natural capital literature identify activities that best serve the conservator's objectives but typically fail to address the question of how to pay for them. We attempt to bridge these literature sources by providing a framework for identifying the specific conditions that must be satisfied by a project in order for an existing financing mechanism to be viable. Notably, our framework quickly reveals financing approaches that can be eliminated. We demonstrate the utility of this approach through conservation case studies on establishment of native forests, coral reef restoration, oyster restoration, and island biosecurity.

Despite a 30-year long history, Renewable Portfolio Standards (RPS) remain controversial and debates continue to surround their efficacy in leading the low-carbon transition in the electricity sector. Contributing to the ongoing debates is the lack of definitive causal evidence on their impact on investments in renewable capacity and generation. This paper provides the most detailed analysis to date of the impact of RPSs on renewable electricity capacity investments and on generation. We use state-level data from 1990–2019 and recent econometric methods designed to address dynamic and heterogeneous treatment effects in a staggered adoption panel data design. We find that, on average, RPS policies increase wind generation capacity by 600–1200 MW, a 44% increase, but have no significant effect on investments in solar capacity. Additionally, we demonstrate that RPSs have slow dynamic effects: most of the capacity additions occur 5 years after RPS implementation. Estimates for wind and solar electricity generation mimic those for capacity investments. We also find similar results using an alternate treatment definition that allows states to meet their RPS requirements with pre-existing renewable generation and renewable generation from nearby states.

Renewable energy forecasting is crucial for integrating variable energy sources into the grid. It allows power systems to address the intermittency of the energy supply at different spatiotemporal scales. To anticipate the future impact of cloud displacements on the energy generated by solar facilities, conventional modeling methods rely on numerical weather prediction or physical models, which have difficulties in assimilating cloud information and learning systematic biases. Augmenting computer vision with machine learning overcomes some of these limitations by fusing real-time cloud cover observations with surface measurements acquired from multiple sources. This Review summarizes recent progress in solar forecasting from multisensor Earth observations with a focus on deep learning, which provides the necessary theoretical framework to develop architectures capable of extracting relevant information from data generated by ground-level sky cameras, satellites, weather stations, and sensor networks. Overall, machine learning has the potential to significantly improve the accuracy and robustness of solar energy meteorology; however, more research is necessary to realize this potential and address its limitations.

1. Calls for using marine protected areas (MPAs) to achieve goals for nature and people are increasing globally. While the conservation and fisheries impacts of MPAs have been comparatively well-studied, impacts on other dimensions of human use have received less attention. Understanding how humans engage with MPAs and identifying traits of MPAs that promote engagement is critical to designing MPA networks that achieve multiple goals effectively, equitably and with minimal environmental impact.
2. In this paper, we characterize human engagement in California's MPA network, the world's largest MPA network scientifically designed to function as a coherent network (124 MPAs spanning 16% of state waters and 1300 km of coastline) and identify traits associated with higher human engagement. We assemble and compare diverse indicators of human engagement that capture recreational, educational and scientific activities across California's MPAs.
3. We find that human engagement is correlated with nearby population density and that site “charisma” can expand human engagement beyond what would be predicted based on population density alone. Charismatic MPAs tend to be located near tourist destinations, have long sandy beaches and be adjacent to state parks and associated amenities. In contrast, underutilized MPAs were often more remote and lacked both sandy beaches and parking lot access.
4. Synthesis and applications: These results suggest that achieving MPA goals associated with human engagement can be promoted by developing land-based amenities that increase access to coastal MPAs or by locating new MPAs near existing amenities during the design phase. Alternatively, human engagement can be limited by locating MPAs in areas far from population centres, coastal amenities or sandy beaches. Furthermore, managers may want to prioritize monitoring, enforcement, education and outreach programmes in MPAs with traits that predict high human engagement. Understanding the extent to which human engagement impacts the conservation performance of MPAs is a critical next step to designing MPAs that minimize tradeoffs among potentially competing objectives.

We estimate spillover effects to local manufacturing plants from the Indonesian palm oil boom, using a stacked difference-in-differences approach. We use new data on the establishment dates and ownership of palm oil mills to identify clean shocks from investments in new plantations. Local plantation booms increased sales and productivity of manufacturing plants, despite increasing blue-collar wages. Using confidential input-output data, we rule out the possibility that this effect is driven by supply chain linkages. Plants increased their share of tradable goods. Local road upgrades point to improved market access as an explanation for this shift.

Climate change is threatening marine systems, and its widespread and dynamic effects are creating challenges for designing and managing marine protected areas (MPAs). The majority of recommendations for climate-resilient MPAs focus on enhancing ecological resilience to disturbance and updating management strategies to respond as changes occur. Here, we assess how existing recommendations for climate resilience are applied in real-world MPA management, using criteria from five key management components: objectives, assessments, design, monitoring, and management. Our review evaluates 172 management plans for 555 MPAs across 52 countries and written in nine languages. We find that MPA management plans contain many underlying scientific and management principles for promoting resilience to climate change, even when “climate change” or related terms are not specifically included: plans include long-term objectives (93.6%), threat-reduction strategies (99.4%), monitoring programs (97.7%), and adaptive management (93%). However, there is substantial variation in the degree to which plans explicitly incorporate climate change into their strategies, from not mentioning it at all (21.5%) to developing detailed climate change-specific action plans (20.9%), with most somewhere in between. In addition to identifying common gaps across management plans, we also provide practical examples of activities MPA managers are undertaking to address climate change.

Marine protected areas (MPAs) have gained attention as a conservation tool for enhancing ecosystem resilience to climate change. However, empirical evidence explicitly linking MPAs to enhanced ecological resilience is limited and mixed. To better understand whether MPAs can buffer climate impacts, we tested the resistance and recovery of marine communities to the 2014–2016 Northeast Pacific heatwave in the largest scientifically designed MPA network in the world off the coast of California, United States. The network consists of 124 MPAs (48 no-take state marine reserves, and 76 partial-take or special regulation conservation areas) implemented at different times, with full implementation completed in 2012. We compared fish, benthic invertebrate, and macroalgal community structure inside and outside of 13 no-take MPAs across rocky intertidal, kelp forest, shallow reef, and deep reef nearshore habitats in California's Central Coast region from 2007 to 2020. We also explored whether MPA features, including age, size, depth, proportion rock, historic fishing pressure, habitat diversity and richness, connectivity, and fish biomass response ratios (proxy for ecological performance), conferred climate resilience for kelp forest and rocky intertidal habitats spanning 28 MPAs across the full network. Ecological communities dramatically shifted due to the marine heatwave across all four nearshore habitats, and MPAs did not facilitate habitat-wide resistance or recovery. Only in protected rocky intertidal habitats did community structure significantly resist marine heatwave impacts. Community shifts were associated with a pronounced decline in the relative proportion of cold water species and an increase in warm water species. MPA features did not explain resistance or recovery to the marine heatwave. Collectively, our findings suggest that MPAs have limited ability to mitigate the impacts of marine heatwaves on community structure. Given that mechanisms of resilience to climate perturbations are complex, there is a clear need to expand assessments of ecosystem-wide consequences resulting from acute climate-driven perturbations, and the potential role of regulatory protection in mitigating community structure changes.

Globalization of fishery products is playing a significant role in shaping the harvesting and use of aquatic foods, but the vigorous debate has focused on whether the trade is a driver of the inequitable distribution of aquatic foods. Here, we develop species-level mass balance and trophic level identification datasets for 174 countries and territories to analyze global aquatic food consumption patterns, trade characteristics, and impacts from 1976 to 2019. We find that per capita consumption of aquatic foods has increased significantly at the global scale, but the human aquatic food trophic level (HATL), i.e., the average trophic level of aquatic food items in the human diet, is declining (from 3.42 to 3.18) because of the considerable increase in low-trophic level aquaculture species output relative to that of capture fisheries since 1976. Moreover, our study finds that trade can improve food security by contributing to increasing the availability and quality of aquatic foods in >60% of the world’s countries. Trade has also reduced geographic differences in the quality of aquatic food consumption among countries over recent decades. We suggest that there are important opportunities to widen the current focus on productivity gains and economic outputs to a more equitable global distribution of aquatic food quantity and quality.

Smallholder agriculture is critical for current and future food security, yet quantifying the sources of smallholder yield variance remains a major challenge. Attributing yield variance to farmer management, as opposed to soil and weather constraints, is an important step to understanding the impact of farmer decision-making, in a context where smallholder farmers use a wide range of management practices and may have limited access to fertilizer. This study used a process-based crop model to simulate smallholder maize (Zea mays) yield at the district-level in Zambia and quantify the percent of yield variance (effect size) attributed to soil, weather, and three management inputs (cultivar, fertilizer, planting date). Effect sizes were calculated via an ANOVA variance decomposition. Further, to better understand the treatment effects of management practices, effect sizes were calculated both for all years combined and for individual years. We found that farmer management decisions explained 27–82 % of total yield variance for different agro-ecological regions in Zambia, primarily due to fertilizer impact. Fertilizer explained 45 % of yield variance for the average district, although its effect was much larger in northern districts of Zambia that typically have higher precipitation, where it explained 72 % of yield variance on average. When fixing a specific fertilizer amount, the “low-cost” management options of varying planting dates and cultivars explained 20–28 % of yield variance, with some regional variation. To better understand why management practices impact yield more in particular years, we performed a correlation analysis comparing yearly management effect sizes with four meteorologically based variables: total growing season precipitation, rainy season onset, extreme heat degree days, and longest dry spell. Results showed that fertilizer’s impact generally increased under favorable weather conditions, and planting date’s impact increased under adverse weather conditions. This study demonstrates how a national yield variance decomposition can be used to understand where specific management interventions would have a greater impact and can provide policymakers with quantification of soil, weather, and management effects. In addition, the variance composition can easily be adapted to a different range of management inputs, such as other cultivars or fertilizer quantities, and can also be used to assess the effect size of management adaptations under climate change.

Air pollution is the fourth-highest risk factor for premature deaths globally and the second-highest risk factor for premature deaths in Ghana. Understanding the socioeconomic correlates of cooking stove/fuel choices and stove stacking patterns is a precursor to designing effective household air pollution interventions. The study applies the multinomial logit model (MNL) to identify socio-economic determinants of household cooking stoves in three unique analyses. In the first analysis, the study examines socio-economic determinants of primary stove choice for LPG, charcoal, and three-stone stoves. Secondly, the study examines determinants of stove choice for primary and secondary stoves when the order is not taken into account: LPG-charcoal, LPG-3-stone, and charcoal-3-stone. The third analysis examines determinants of stove use patterns among stove stackers when the order is taken into consideration: LPG-charcoal, LPG-3-stone, charcoal-3-stone, charcoal-LPG, 3-stone-LPG, and 3-stone-charcoal stove combinations. A cross-sectional design is used to collect data from sampled districts in all 16 regions of Ghana, covering about 7400 household-level respondents in 370 enumeration areas (EAs); across 177 urban EAs and 193 rural EAs.

The results show that primary cooks with higher education, smaller, wealthier, and urban households are more likely to use LPG; more likely to use a stacking mix that includes LPG, and more likely to use a stacking combination where LPG is the primary stove-fuel used. Ghana is currently undertaking a National LPG Promotion Programme (NLPGPP) that seeks to increase LPG access by providing subsidized “starter kits” (stoves, regulators and cylinders) to periurban and rural households. Our findings can aid in targeting households effectively. Our results suggest that households that are female-headed; have less educated household heads; have household heads who double as the primary cook; have larger family sizes, have lower socio-economic status, and are rural, will have an increased probability of adopting and using a cleaner stack of cooking technologies and have an increased probability of using LPG long term, conditional on receiving a starter kit. Such a targeted intervention with a multifaceted approach to shifting households to a cleaner stack of cooking technologies could result in widespread adoption and contribute to significant reductions of CO and PM2.5 emissions in Ghana. A cleaner stack of cooking technologies, given the socio-demographics that are least likely to adopt and sustain use, will have a significant impact on health, the environment, and climate change.

Compared to excise taxes and carbon taxes, setback restrictions on new oil wells have larger health benefits and worker compensation losses, but are more equitable by bringing greater benefits and lower losses to disadvantaged communities in California. For California to meet green gas emissions (GHG) targets, larger setbacks than currently proposed or additional supply-side policies are needed.

To improve the performance of an intra-hour global solar irradiance forecasting algorithm, it is important to detect multiple layers of clouds. Horizontal atmospheric wind shear causes wind velocity fields to have different directions and speeds. In images of clouds acquired using ground-based sky imagers, clouds may be moving in different wind layers. The information provided by a solar forecasting algorithm is necessary to optimize and schedule the solar generation resources and storage devices in a smart grid. This investigation studies the performance of unsupervised learning techniques when detecting the number of cloud layers in infrared sky images. The images are acquired using an innovative infrared sky imager mounted on a solar tracker. Different mixture models are used to infer the distribution of the cloud features. Multiple Bayesian metrics and a sequential hidden Markov model are implemented to find the optimal number of clusters in the mixture models, and their performances are compared. The motion vectors are computed using a probabilistic implementation of the Lucas-Kanade algorithm. The correlations between the cloud motion vectors and temperatures are analyzed to discover the method that leads to the most accurate results. The findings point that a sequential hidden Markov model outperforms the detection accuracy of standard Bayesian model selection metrics.

Pollution concentrations (PM_2.5) in the United States have fallen in recent decades. Despite these improvements, disparities in concentrations between racial/ethnic groups persist. We combine administrative data on the universe of emergency room (ER) admissions across California with satellite information on PM_2.5 concentrations and compare recent trends in racial/ethnic disparities for PM_2.5 and asthma rates. We find that PM_2.5 concentrations fell for the average Black, Hispanic, and White individual. Similarly, disparities in PM_2.5

Like other climate-related disasters, wildfires are intensifying. Property owners can reduce their vulnerability to wildfire losses but are not well informed about the costs and benefits of available self-protection investments. Technological advances mean that insurers are increasingly able to monitor household mitigation behavior. We revisit the problem of self-protection from risk in a setting where households have incomplete insurance and limited information about self-protection investments. Insurer discounts for self-protection generate additional value by informing households about self-protection investments that also reduce uninsured losses. This information provision increases the responsiveness of self-protection to ex post disaster assistance, with implications for optimal government transfers.

Oil supply-side policies—setbacks, excise taxes and carbon taxes—are increasingly considered for decarbonizing the transportation sector. Understanding not only how such policies reduce oil extraction and greenhouse gas (GHG) emissions but also which communities receive the resulting health benefits and labour-market impacts is crucial for designing effective and equitable decarbonization pathways. Here we combine an empirical field-level oil-production model, an air pollution model and an employment model to characterize spatially explicit 2020–2045 decarbonization scenarios from various policies applied to California, a major oil producer with ambitious decarbonization goals. We find setbacks generate the largest avoided mortality benefits from reduced air pollution and the largest lost worker compensation, followed by excise and carbon taxes. Setbacks also yield the highest share of health benefits and the lowest share of lost worker compensation borne by disadvantaged communities. However, currently proposed setbacks may fail to meet California’s GHG targets, requiring either longer setbacks or additional supply-side policies.

Government agencies may be an additional source of unequal representation beyond that stemming from the differential responsiveness of elected officials. We use plausibly exogenous focusing events, which raise public demands for government provision of local public goods, to examine evidence of inequality in agency decision-making. Using the empirical case of wildfire risk management in the western United States, we find that experiencing nearby wildfires raises the salience of wildfire risk and leads agencies to place a greater number of risk reduction projects nearby, even when wildfire risk has already been reduced. This effect predominates among high socioeconomic status communities, especially higher income, more educated, and whiter communities. Empirical evidence is consistent with a formal model in which public agencies perpetuate inequality due to differences in the costs of lobbying across demographically varying communities and differences across communities in the benefits to agencies of responding to their demands.

Understanding the economics of food loss-reduction technologies is crucial to promote their adoption. We offer experimental evidence on returns to technology that studies a wide range of food security outcomes and accounts for the quality, quantity, and economic gains from adoption. Using two randomized experiments on improved storage with farmers in India, we find that food loss-reduction technology delivers large gains on multiple aspects of food security. The largest benefits of loss-reduction technology come from reduced quality and economic losses. Using a novel within-farmer experiment, we offer evidence of significant economic incentives for investing in loss-reduction technology in local rural markets. Our cost-benefit analysis shows that farmers recover the full unsubsidized cost of the reusable improved storage technology in one agricultural season.