<p dir="ltr">The Farm Service Agency's Guaranteed Loan Program supports eligible lender's ability to provide credit to farms who would otherwise not qualify for loans by guaranteeing up to 95% of principal and interest if the farmer defaults. The first chapter examines the degree to which bank characteristics influence FSA guaranteed loan rates paid by farmers. We leverage the unique characteristics of a panel of FSA guaranteed loans that include both borrower and lender information. Relative to pooled OLS, our preferred fixed-effects regression specification suggests that both time-varying and invariant lender effects are a significant determinant of FSA guaranteed loan rates. Further, when controlling for lender-effects, the significance of borrower characteristics largely diminish. These findings are consistent with prior studies of broader lending market interaction. This is the first study to examine FSA guaranteed loans which accounts for bank-level variation in lending terms. The findings may be of interest to policymakers, program administrators, lenders, and farmers.</p><p dir="ltr">Bankers’ expectations have been shown to provide reasonable forecasts of land value. In the second chapter, we test the informativeness of bankers’ expectations in predicting FSA guaranteed loan application volumes. Once again, we leverage proprietary administrative data from the FSA and, this time, pair it with survey data from the Federal Reserve Bank of Chicago to evaluate bankers’ forecasts. Results show that bankers’ forecasts are outperformed by naïve models, and including bankers’ expectations does not improve predictive models. Once again, these results will be of interest to FSA program administrators, lenders, and potential borrowers.</p><p dir="ltr">The study of risk is an important thread of farm management research as agriculture is an industry with many sources of risk. In the third chapter, we link broad measures of policy risk in the form of Equity Market Volatility trackers to farmer’s perceptions of risk and uncertainty. We consider disagreement in ex ante sentiment questions to measure farmer risk. Through a series of pairwise VARs, we show which sources of risk matriculate as concerns for farmers measured by uncertainty in the Purdue University-CME Group Ag Economy Barometer. Increases in tax policy, trade policy and infectious disease uncertainty are found to Granger-cause movement in farmer sentiment uncertainty.</p>

<p>     Farmers face significant uncertainty, like weather and prices. Micro-economic theory tells us that when facing uncertainty, an agent, or farmer, makes economic decisions based upon their expectations. This primitive is important for agricultural economics. The “classic” agricultural economic problems: acreage allocation, commodity storage, technology adoption, household labor engagement, etc., are all influenced by the expectations of farmers. Despite expectations pervasive inclusion in economic theory and the decades of attention from agricultural economists, we still know relatively little about how farmers form expectations. This Dissertation is aimed at this opportunity.</p> <p>     The first chapter estimates the degree to which information is incorporated in farmland value expectations. Theoretically, an agent’s expectation should represent all available information. However, there are reasons to believe that an agent may not possess all the pertinent information or they may not be able to interpret the information. Macroeconomists have developed two models to explain the degree to which information may not be incorporated into expectations, The Sticky and Noisy Information Models. I use expectations and actual values of Iowa farmland from 1964 to 2021 to estimate the degree to which new information is not reflected in expectations, or exhibit information rigidities. I find that Iowa farmland market participants do experience information rigidities. From a practical standpoint, farmland is farmers’ most important collateral, the presence of public, simple farmland information may help mitigate lending challenges as a result of farmland value expectations.</p> <p>     The second chapter addresses how commodity price information is incorporated into the financial expectations of farmers. I estimate how unknown or surprise information from a United States Department of Agriculture (USDA) report changes farmers’ attitudes and expectations of their financial conditions. This chapter, synthesizes literature from macroeconomics and commodity price analysis, and uses a unique source of data, the Purdue University/CME Group Ag Economy Barometer. The Ag Economy Barometer reflects the aggregate sentiment of farmers across the US. Like the consumer sentiment index from the University of Michigan, the Ag Economy Barometer can provide a snapshot of sentiment, a measure outside of fundamental economic indicators. Using the corn ending stocks values from the USDA World Agricultural Supply and Demand Estimates (WASDE), I find that</p> <p>farmers’ short– and long–term expectations and attitudes toward large farm investments are increased by information implying a higher corn price. However, this relationship does not exist in the reverse direction and when corn is not actively growing. As a result, if farmers are acting on these changes in expectations, they may be engaging in suboptimal decision making.</p> <p>     The third and final chapter explores the degree to which previous experience is reflected in expectations. The tales of the financial hardship during 1980’s Farm Financial Crisis have been shared across farmers’ dining room tables for decades. The most prominent anecdote relates to the rapid decline in farmland prices. As mentioned in the first chapter, the asset value of farmland is important to farmers. As a result, if experiences like the 1980’s Farm Financial Crisis have created a downward bias toward farmland values, the asset may be undervalued and frictions may exist in the farmland lending market. Macroeconomists show that consumers’ inflation expectations are directly related to their life experiences. I use a panel of farmland market participants in the Purdue Land Value and Cash Rent Survey to estimate the effect of previous experience on farmland value expectations. I find no</p> <p>significant effects. However, my estimates are using variation in cross sectional data. This modeling choice does not rule out the potential of the Farm Financial Crisis effecting all market participants in a similar way, a question outside of my analysis.</p> <p>     Each chapter of this Dissertation addresses how an agent forms their expectations, a necessary first step in my journey as a researcher. I am interested in the link between expectations and economic outcomes. I have built considerable knowledge on expectation formation and will deploy this knowledge exploring the role of expectations in farm outcomes, like acreage allocation, commodity storage, technology adoption, and household labor engagement. In my next step as a researcher, I plan to use the current theoretical advancements in behavioral economics, the explosion in empirical methods and computing, and the availability of data to re-visit the role of expectations in “classic” farm economics problems.</p>