A logit model of farmers' decisions about credit

摘要: Most of the research in agricultural credit tion about personal characteristics farm operator United States is related to large-scale commercial and economic aspects farm-firm household. farming. A number such studies have tried esti- The plan paper as follows. Data used for mate demand (Hesser Schuh) and/or analysis described next section. estimated supply (Melichar). Other major areas re- model presented section three. Empirical results search been (a) function- are discussed four, while last ing rural financial markets, (b) effects national contains a brief summary concluding remarks. monetary policies on investment production agriculture, (c) capital formation agriculture affected by policy, (d) DATA sources-and-uses-of-funds approach financing. review data this study were collected part lated these four given Brake. However, larger management carried out western subject small-farm has Tennessee. Since objective was fobeen little explored. cus limited resource farms, only those farms few which dealt with prob- whose gross sales between $2,500 lems limited-resource (small) basically $20,000 during each previous three years studied their attitudes toward borrowing, without ex- included sample. No hobby farmer inploring validity cluded.' In 1979, randomized block design (McManus; Otto; Snell, Hopkins, Barnett; Spitze select sample 89 farmers who Bevins; Romans; Wise; Woodworth, agreed participate long-term record-keeping Comer, Edwards). general consensus that program. trained enumeremerges from relatively op- ators lived two selected counties. Every seerators small use credit, do lected household visited enumerators amounts. This led some social scientists twice month 1980 order collect reliable believe not want outputs inputs. first borrow. interview, respectively, comprised an opening Conventional methods estimating closing inventory all resources. Land categocredit information rized into cropland, pasture, woodland, improvable, actually neglect waste land groups identify potential borrowed. Such different types land. An exhaustive cannot account farmers' initial decisions machinery equipment taken. Regular whether or borrow; consequently, valuable in- records inputs crop every type wasted. Omitting nonborrowers livestock kept separately. Similarly, separate also distorts properties original sam- sales. pie. Furthermore, considering decision All asked borrow can lead biased estimates they had short-term dur(Heckman; Tobin). Fortunately, we quite de- 1980. Pretesting questionnaire indicated tailed considered amount borrowing very households, matter. Therefore, no effort made reoperators borrowed well cord actual individual predict odds using short- Only yes answers recorded. term conditional upon informa-