Research
Working Papers
Policy Spillovers to Biodiversity: The Impact of Clean Water Act Investments on Bird Diversity and Abundance [Job Market Paper]
Draft (Coming Soon!) Current Slides
Abstract:
Did the Clean Water Act’s largest spending program- the Title II grants that financed municipal wastewater treatment- produce ecological co-benefits for birds, a category of return that its extensive cost-benefit literature has left almost entirely unquantified? This project tests for them directly, pair ing a century-long, continent-wide citizen-science census- the Christmas Bird Count- with grant records from EPA’s Grants Information and Control System, plant-level engineering attributes from the Clean Watersheds Needs Survey, and the National Hydrography Dataset Plus river network. I study two outcomes. The first is diversity: the richness and evenness of the assemblage conditional on presence, measured with bias-corrected Hill numbers, the scale-invariant “effective number of species” family. The second is abundance: how many birds are counted, modeled directly as a count process by Poisson pseudo-maximum-likelihood, which retains the many zero cells. I leverage the quasi-experimental framework of grant roll-out in a staggered, continuous treatment setting, comparing Christmas Bird Count sites (circles) exposed to upstream grant-funded treatment against not-yet-exposed and never-exposed circles, with a flow-share dose that weights each plant by its discharge relative to the receiving stream. Three diagnostic findings stand out. First, the identifying variation is quasi-exogenous: 80–87% of the cross-circle variationin treated flow-share is set by hydrological network geometry rather than by where grants were targeted, so which circles are treated- and how intensely- is driven by rivers, not by policy selection. Second, the most natural confound is ruled out: observation-effort and extensive-margin placebos show no break at grant commissioning, so grant-driven changes in who counts birds do not masquerade as changes in the birds themselves. Third, a diagnostic split by guild locates the response exactly where the dissolved-oxygen-to-prey-base mechanism predicts and a landbird placebo does not: the abundance of water dependent guilds- waterbirds and waterfowl- rises sharply and significantly in grant-exposed circles while the landbird placebo shows no change. These insights are informative for how water-quality investments are appraised and targeted, and the ecosystem measurements they require are themselves a template for credible program evaluation
Benefits of Conservation Easements to Biodiversity: Evidence from Wetland Birds.
** With Terry Zhang & Wendong Zhang **
Abstract:
By 2019 wetlands covered just 116.4 million acres, roughly 6% of U.S. land. Yet, rates of loss have risen 50% since 2009 with drainage and agricultural conversion. The USDA’s Natural Resources Conservation Service has obligated over $1.8 billion to preserve nearly 800,000 acres of wetland, and the 2018 Farm Bill some $450 million annually to the Agricultural Conservation Easement Program. Despite scale of investments, causal evidence on the ecological returns to easements remains limited. This study thus estimates how wetland easements affect bird relative abundance and species richness, accounting for variation in easement intensity, timing, and hydrological boundaries. We combine eBird observations from 2003 through 2023 with the National Conservation Easement Database for spatial records of CREP, ACEP, and EQIP easements, the National Land Cover Database, PRISM climate data, and CropIndex measures of adjacent agricultural pressure. The empirical strategy proceeds in two steps. First, we construct annual HUC12-level indices of relative abundance and richness from eBird checklists, conditioning on observer effort, expertise, and seasonality and separating breeding from non-breeding seasons to capture habitat sensitivity. Second, we implement a staggered difference-in-differences design comparing treated subwatersheds to not-yet-treated subwatersheds within the same HUC8 watershed. Preliminary results indicate substantial heterogeneity across subwatersheds, with those of higher initial richness tending to maintain or increase relative abundance. Across all species in our sample, greater protection is associated with gains of one to three additional species and with increases in relative abundance relative to less protected wetlands. Analysis of public versus private access and of heterogeneity between resident and migratory populations is ongoing. These timely causal estimates contribute to the economic valuation of wetland ecosystem services and inform efforts to optimize investment in habitat protection for ecological improvements.
Evaluating State Policies Supporting Conservation in Agriculture: Impacts on Water Quality in Minnesota.
** With Catherine L. Kling & M. Mainul Hoque **
Abstract:
Nutrient pollution from agriculture remains one of the most persistent water quality challenges in the United States, where only about a third of rivers and streams meet standards for nitrogen and phosphorus and agricultural runoff is the dominant non-point source. Because conservation policy is largely designed and implemented at the state level, states are central to addressing this pollution, yet rigorous evidence on the effectiveness and cost-efficiency of state-funded programs is scarce. This study develops a framework to evaluate the effectiveness, efficiency, and replicability of state-level conservation policies. We study two major programs in Minnesota: the mandatory Minnesota Buffer Law (MBL) and the voluntary Minnesota Agricultural Water Quality Certification Program (MAWQCP). We exploit the legally binding nature and timing of the MBL to compare changes in surface-water nutrient concentrations across watersheds differentially exposed to the two programs before and after the buffer mandate took effect. Accounting for concurrent conservation efforts (e.g., EQIP and the Wetland Reserve Program) and weather variability, we conduct a cost-benefit analysis, monetizing water quality improvements using social cost of nutrient values from a bottom-up integrated assessment model alongside estimates of practice implementation and maintenance costs. We find that the MBL reduced surface-water nitrogen by an average of 179 pounds (81 kg) per buffered acre annually, operating through both reduced application on formerly cropped land and filtration across the surrounding watershed. Benefits exceed implementation costs by at least one order of magnitude. These results provide strong evidence that where voluntary approaches alone may be insufficient, enforceable, well-supported state policies can cost-effectively reduce agricultural non-point source pollution.
Select Work in Progress
Green Infrastructure: Is It for the Birds?
** With Sheila M. Olmstead, Billie Faircloth, Douglas H. Wrenn, & Daniel A. Brent **
Converging Crises: Understanding the Impact of Climate Change and Violence on Migration in Mexico.
** With Luc M. Esprabens & Dana J. Smith **
