« Back to Results

Environmental Externalities and Agriculture

Paper Session

Saturday, Jan. 5, 2019 10:15 AM - 12:15 PM

Atlanta Marriott Marquis, A708
Hosted By: Association of Environmental and Resource Economists
  • Chair: Taro Mieno, University of Nebraska-Lincoln

Implications of the Federal Crop Insurance Program on Water Sustainability in the United States

Prasenjit N. Ghosh
Auburn University
Ruiqing Miao
Auburn University


Irrigated agriculture is the largest fresh water user in the United States. Growing pressure on
water resources due to climate change and population growth has prompted renewed focus on
irrigation water demand. Currently the U.S. crop insurance program covers more than 80% of
cropland and 130 crops. In its 2017 baseline projections, the Congressional Budget Office has
estimated that the crop insurance program will cost $7.7 billion annually between 2018 and
2027. Government pays, on average, 62% of total premium for yield-based and revenue-based
crop insurance policies, and almost 100% of total premium for catastrophic coverage. There
have long been concerns that crop insurance would have a wide range of distortionary effects
on farmers’ economic decisions. In this article, we particularly focus on how crop insurance
affect farmers’ irrigation decisions. Related studies overlook the impacts of climate and soil
quality while studying the causal relationship between crop insurance and irrigation demand,
therefore, may suffer omitted variable bias. Our study is the first to conduct a comprehensive
analysis of the implications of federal crop insurance on irrigated water use while accounting
for soil quality and climate variation. We first develop a conceptual framework to deduce
testable hypotheses about how insurance may affect water use. Based on the county-level
irrigation and crop insurance data from 1990 to 2010 with a five-year step, our preliminary
results suggest that, ceteris paribus, 1% increase in insured crop acreage, on average, raises
irrigation water uses demand by 0.16%. Albeit the overall deleterious effects of the insurance
policies on water sustainability, preliminary results indicate that revenue insurance policies in
general are associated with relatively less use of water for irrigation over yield insurance
policies. This study provides guidance to policymakers regarding the impacts of federal crop
insurance policies on irrigated water use across the country.

Oil for Food? Oil Spills and Agricultural Productivity

Justice Tei Mensah
Swedish University of Agricultural Sciences


What is the spillover effect of natural resource extraction on the development of the agricultural sector in resource-rich countries? In this paper, I evaluate the effect of oil spills on agricultural productivity using high-resolution georeferenced data on oil spills matched with geocoded plot level panel data from Nigeria. Using an identification strategy that exploits plausibly exogenous variations in the economic value of oil spills, induced by variations in international price of crude oil, I find that exposure oil spills have a significant negative impact on agricultural productivity. A one standard deviation increase in the price of crude oil results in a 1.5 percent decrease in output per hectare of farms exposed to oil spills. Aside the direct effect of pollution, labor reallocation from agricultural to non-agricultural sectors is a likely operative channel.

Adaptation to Environmental Change: Agriculture and the Unexpected Incidence of the Acid Rain Program

Nicholas Sanders
Cornell University
Alan Barreca
University of California-Los Angeles


Burning coal produces sulfur dioxide (SO2), which converts to sulfuric acid and
falls as \acid rain". Environmental damage concerns led to the Acid Rain Program
(ARP) to reduce coal emissions in the Midwest and Eastern United States. Ensuing
SO2 reductions altered the agricultural sulfur cycle, decreasing transmission of sulfur,
a helpful nutrient. Comparing county-level changes in yields by proximity to ARP-
regulated power plants shows the program meant output losses up to 20% for corn and
15% for soybean, the two largest crops by acreage and revenue. We observe limited
adjustment, suggesting producers were slow to react to changing environmental condi-
tions. We present suggestive evidence that centralized information distribution ended
up playing a large role in adaptation, and discuss our results in the context of agricul-
ture's future ability to adjust to climate change and uncertain growing conditions.

Regret minimization, path dependence, and attribute non-attendance in discrete choice experiments

Qi Tian
Michigan State University
Jinhua Zhao
Michigan State University


Cognitive psychologists and behavioral decision theorists have argued that individuals often simplify their decision-making processes when solving complex decision tasks. Discrete choice experiment (DCE), as a popular nonmarket valuation approach, faces challenges of incorporating simplifying strategies in ensuring incentive compatibility and consistent estimation of welfare values. In this paper, we investigate how respondents in DCE use behavioral strategies to make decisions, how such simplification evolves in repeated choice tasks, and how failing to identify these strategies can lead to confounding conclusions.
One behavioral strategy is attribute non-attendance (ANA): respondents strategically or unconsciously ignore some attributes in a choice task in order to reduce cognitive burdens of decision-making. Another behavioral strategy is regret minimization (RM): respondents use reference points (RP) to make choices depending on the net gains/losses from bilateral comparisons between alternatives. The asymmetric weights on gains and losses leave some attributes of certain alternatives appearing to be ignored. This paper, for the first time, discusses the relation between RM and ANA, and investigate how these strategies evolve through repeated choices. We conducted a DCE survey of farmers’ decisions in fertilizer management in the Midwest states of Michigan, Iowa, and Illinois. Our results reveal an adaptive pattern of simplifying strategies: as respondents get more familiar with the choice setting, they are more likely to adopt ANA (under the RUM model). Moreover, we investigate under what situations would either strategy be employed to simplify complex tasks. We find that attributes, that are otherwise estimated to be non-attended, are in fact attended once path dependence behavior is accounted for. Understanding the underlying mechanisms of ANA and RM will facilitate DCE choice modeling and survey design.
Taro Mieno
University of Nebraska-Lincoln
Valerie Mueller
Arizona State University
Wolfram Schlenker
Columbia University
Patrick Lloyd-Smith
University of Saskatchewan
JEL Classifications
  • Q5 - Environmental Economics
  • Q1 - Agriculture