Performing Department
(N/A)
Non Technical Summary
Cattle are the largest source of U.S. farm revenue. With droughts becoming more frequent and severe, cattle producers are making costly decisions to reduce their herd size or stocking rates. While the Federal farm safety net has been focused on row crops, disaster programs and crop insurance, for which payouts based on weather conditions and are linked to the value of forage, are available to cattle producers. The proposed research will study the interlinkage among cattle production, forage production and storage, and the utilization of Federal programs focusing on increasing drought risk. We will first expand on cattle cycle theory to incorporate forage storage and safety net payments that are correlated to local weather shocks. The developed theoretical model will provide the foundation for two empirical studies. The first will use both national data and unique farm-level data to explore how forage stocks provide a buffer against drought. The second empirical study will estimate the complementarities among Federal disasters programs, Federal rainfall insurance for forage payments, and forage storage. We plan to extend the literature that assesses the impacts of drought and market conditions on herd size and stocking rate for cattle production by expanding the focus to the production and management of forage. This research will proprovide a theoretical and empirical foundation for future economic research, extension, and multidisciplinary collaboration on the resilience of the US cattle sector to climate change.
Animal Health Component
0%
Research Effort Categories
Basic
40%
Applied
60%
Developmental
0%
Goals / Objectives
Economic research on forage storage and the interaction of forage-related safety net programs is limited. This project will address these gaps in the literature, to build a foundation for future research in this area, including both economic research and multidisciplinary collaborations and development of bioeconomic models.The first research goal of the project (R01) is to develop a theoretical framework that models the forage and herd size management decisions under the presence of drought risk. The second research goal of the project (R02) is to estimate the supply response of forage production and storageto prices and weather. The second research goal of the project (R03) isto empirically evaluate how various federal risk management policies interact and affect the forage production and management decisions of livestock producers.
Project Methods
InResearch Objective 1, we plan to develop a model that describes the rangeland and cattle management decisions with a particular emphasis on forage storage decisions. We will extend and modify the dynamic programming approaches from the literature (e.g., Kobayashi et al. 2007; Ritten et al. 2010; Kobayashi, Rollins, and Taylor 2014) to simultaneously model herd size, land allocation, and forage inventory. To do so, we start with a dynamic model with four sets of control variables:amounts of age-specific cows, amounts of age-specific cows purchased, amount of forage stored (hay or silage), and amount of feed purchased. The state variables and the parameters include variables related to prices in the cattle and feed markets, and weather variables (e.g., temperature, precipitation, and drought) and the vectoris the stacked equations that include the equations of motion for the state variables and other parameters.The aforementioned dynamic model is an optimal control problem under uncertainty and can be represented as a stochastic dynamic program. The first set of state variables is the inventory of cows at the various ages. The equations of motion are dictated by the sale and the purchase of the cows and time. The second state variable is the vegetation density of the rangeland (i.e., forage yield) as discussed by the rangeland management literature, such as Ritten et al. (2010). We will investigate agronomy, forage science, and animal science literature to further explore alternative specifications of the forage growth and consumption functions. The third state variable is forage inventory, which is one of the key elements of the proposed project.The profit function, the equations of motion, and the constraints can be specified further and parameterized by extending and modifying the previous approaches such as Ritten et al. (2010) and Kobayashi, Rollins, and Taylor (2014). The optimal control problem can be solved by using a Bellman equation and numerical procedure to get the solution of the stochastic dynamic program following Miranda and Fackler (2004). Along with the numerical solutions, we plan to provide some comparative statics with respect to key parameters; cattle prices, feed prices, and weather variables (especially, drought) as done in Jarvis (1974). We also plan to conduct two sets of simulations: a) climate change scenarios and b) scenarios with climate change and risk management policies available. For the first set of simulations, we consider various climate scenarios by differing the frequency and severity of drought occurrences. We solve the optimal control problems under these scenarios and compare the optimal path of herd size and forage inventory. In the second set of simulations, we consider the federal programs such as PRF, LFP and ELAP. We plan to solve the optimal control problems under the climate scenarios of a) with several policy scenarios (e.g., a) insured by PRF, b) insured by PRF and LFP available, and c) insured by PRF and both LFP and ELAP available).InResearch Objective 2,Activity 1, we will usestate-level data from USDA NASS, to estimate how hay production and inventories respond to prices and extreme weather, specifically drought events.In this activity, we plan to utilize a long series of state-level data of hay prices, acres harvested, and on-farm inventories from USDA NASS combined with weather data from the Parameter-elevation regressions on Independent Slopes Model (PRISM) from Schlenker and Roberts (2009). The time coverage of the data we plan to utilize is from 1950 to 2021 (The NASS state-level data for most of the states start from 1950). We start with a linear regression whereacres harvested or on-farm inventories of hayarean outcome variable of interest, with controls that include avector of price and weathervariables for statesin yeart.Identifying the parameters of interest is challenging as there are many unobservable factors that lead to the omitted variable bias (OVB).To mitigate the OVB and identify the parameters of interest, we first employ the Panel two-way fixed effects.However, a drawback for the Panel two-way fixed effects model is that it does not account for time-variant confounders. As an alternative, we consider the estimators that are based on the latent factor model,where are the unobserved common time-specific factors and are unobservable unit-specific factor-loadings. The challenge is to identify the common factors and factor loadings as they are unobservable and hence, it requires certain assumptions and parameter restrictions. Two approaches will be employed - Panel Interactive Fixed Effects of Bai (2009) and Panel Common Correlated Effects of Pesaran (2006). Finally, as our state-level time-series are long, we will also consider Panel Vector Autoregression (Holtz-Eakin, Newey, and Rosen 1988). We plan to mitigate possible spatial or time-series dependency in error terms via cluster-bootstrapped standard errors following Cameron, Gelbach, and Miller (2008).InResearch Objective 2Activity 2,we will estimate the relationship between extreme weather and hay production and inventories, using farm level Kansas Farm Management Association (KFMA) data.To identify the effect of extreme weather on hay production and storage we harness panel data with within-location fixed effects. Hay production and storage the outcome variables of interest, with controls that include time varying economic, weather, and environmental factors.The inclusion of location and year fixed effects means that the identification of the parameters is based on the within location and within year variation in the outcome and covariates and weather variables, which is arguably exogenous. Because many of the outcomes and variables in the analysis are spatially dependent, we plan to account for spatial contemporaneous variation in a few ways. One way is through clustering at the KFMA region-year level and another option is to correct for spatial autocorrelation following the approach developed in Conley (1999) adapted to panel data.In addition to understanding the relationship between weather and forage production and storage, we will alsotest the hypothesis whether storage rates increase after exposure to extreme weather, or if increasing storage is a behavioral or management response to drought. This can be tested by using a ratio of storage and production as the outcome variable and adding weather lagsto the above models.InResearch Objective 3we will empirically evaluate how various federal risk management policies interact and affect the forage production and management decisions of livestock producers, using similar econometric methods as in RO2.Two hypotheses underpin this proposed research. First, experience with receiving LFP or other disaster payments during a drought may limit PRF participation, as PRF requires payment of premiums and enrollment decisions may happen before the onset of drought. If a producer expects to receive a disaster payment, they may decide not to enroll in insurance.However, cattle producers in general have much less experience with government programs than crop producers. This informs our second preliminary hypothesis. Experience with either disaster programs or insurance may make producers more likely to use the other program. While LFP is generally available, it does require producers to file for payments, and still only has partial participation in some areas. Experience with PRF may depend on insurance outcomes and may need to be qualified as "positive" or "negative" experience. Positive experience could be defined as receiving indemnities or positive net indemnities (indemnities less premiums paid).