Essays on Farmer Consumption Choices, Beginning Farmers Credit Constraints and Exit

Abstract

The dissertation is divided into three chapters, the first concerning the choice of food safety in Ghana, the second on Beginning farmer credit constraints, and the last on beginning farmer survival. Chapter 1 uses linear and logistic regression to show household composition’s impact on food safety choices for subsistence households in the Ashanti region of Ghana. Three biomarkers act as proxies for the consumption of aflatoxin contaminated groundnuts. OLS regression and a test for endogeneity verify that the impact of consumer-to-worker ratio is robust and unbiased overall. All models show that both the number of children attending school and the number of children under 10 are household composition variables that are significantly associated with contamination levels, and thus food safety decisions. (Key words: aflatoxin, consumer worker ratio, food safety, groundnut safety, household composition, peanut toxin, utility of food safety. JEL Classification: Q1, Q10, Q12, I10, I15). Chapter 2 seeks to answer how capital constraints influence the profitability of beginning farmers and ranchers (BFRs). Using propensity score matching with nearest neighbor and mahalanobis distance algorithms, we confirm that there is a negative treatment effect of between $4,700 and $35,400 for credit constrained beginning farmers and ranchers (BFRs). This represents a -14% to -77% loss with similar per acre losses in the value of production. Constrained BFRs are younger, less liquid, more likely to lease, and concentrated in the South compared to their unconstrained peers. The high variance in the differential value of production shows the breadth of diversity and the challenges in classifying BFR credit needs. (Key words: iii

propensity score matching, beginning farmer rancher, credit constraints, finance, family farm. JEL Classification: Q00, Q14, G02, D13, D14, D24). Chapter 3 uses a 1992-2012 Agricultural Census panel of beginning farmers and ranchers from which we estimate a probit model with marginal effects. We find evidence that climate variation affects farmers, especially seasonal rainfall variation. Farmer scale in terms of sales and assets does decrease exit, but we find no evidence that government payment intensity does the same. We find weak support for the liability of adolescence hypothesis and observe a slowing increase of failure risk when comparing those farming 8 to 10 years with those less than 7. (Key words: beginning farmer rancher, survival, hazard, liability of adolescence, risk, climate. JEL Classification: Q00, Q12, Q14, Q50, D00).