Êste trabalho se propôs a analisar alguns aspectos da utilização e da eficiência dos fatôres de produção e do crédito rural em uma região predominantemente agrícola do Estado de São Paulo. Correlativamente, foram perseguidos os seguintes objetivos específicos. a. Descrever e analisar as principais características das propriedades rurais que têm acesso ao mercado de crédito. b. Estimar as produtividades médias e marginais da terra total explorada, do trabalho da família e do capital na forma de “despesas de custeio” e de “animais e maquinaria” nos estabelecimentos que se utilizam do crédito. c. Estimar as produtividades média e marginal do crédito e analisar as principais relações desta variável com a receita total dos agricultores. d. Derivar a demanda a curto prazo do capital na forma de “despesas de custeio” e de “animais e maquinaria” e examinar suas prováveis implicações para a utilização eficiente do crédito. A informação básica analisada na pesquisa foi obtida através de entrevistas pessoais com os agricultores de uma amostra extraída do universo constituído pelos agricultores dos Municípios de Itapetininga e Guareí, localizados na região sul do Estado de São Paulo. A técnica de amostragem empregada foi a dos grupos ao acaso. Da amostra total, que se compunha de 132 propriedades agropecuárias de dez tipos pre-estabelecidos, foram selecionadas aquelas que tiveram acesso ao mercado de crédito no biênio 1964-1965. Nessa condição, encontraram-se 49 propriedades, ou seja, 37% ela amostra total. Embora os dados tenham sido coletados durante o ano agricola 1965-1966, o periodo-base considerado na pesquisa foi o ano de 1965. Preliminarmente, julgou-se necessária a aplicação de um teste de homogeneidade aos dados originais. Assim sendo, o teste do X² (Quiquadrado) foi aplicado para se verificar se o grupo dos agricultores usuários de crédito e a amostra total não diferiam significativamente entre si, no que respeita ao atributo considerado essencial quando da obtenção da amostra, isto é, o tipo de propriedade rural. A hipótese de homogeneidade entre as duas distribuições de frequência foi aceita, o que significa, também, que o grupo dos agricultores. Usuários de crédito é representativo dos tipos de propriedades encontrados na amostra total. As diferenças observadas foram de natureza aleatória. O modêlo econométrico utilizado para a consecução dos principais objetivos propostos anteriormente foi o sugerido por Charles Cobb e Paul Douglas. E duas foram as hipóteses mais gerais de trabalho, a saber: Hipótese Geral 1: Log Y = log a + b₁ log X₁ + b₂ log X₂ + b₃ log X₃ + b₄ log X₄ Hipótese Geral 2: Log Y = log a + b₁ log X₁ + b₂ log X₂ + b₅ log X₅ onde Y = receita agrícola total (NCr$) X₁ = terra total explorada (hectare) X₂ = trabalho da família (equivalente-homem) X₃ = capital na forma de “despesas de custeio” (NCr$) X₄ = capital na forma de “animais e maquinaria” (NCr$) X₅ = crédito para custeio e investimento (NCr$) Visando a estimar as funções de produção que representassem as relações fator-produto e se adaptassem aos objetivos da pesquisa, foram testadas, separadamente, dez equações estimativas, sendo seis pertinentes à Hipótese Geral 1 e quatro relativas à Hipótese Geral 2. A seleção das duas “melhores” equações estimativas baseou-se nos seguintes critérios (a) coerência dos resultados com os princípios econômicos que regem as relações de produção; (b) significância estatística dos coeficientes de regressão parcial e do coeficiente de correlação múltipla; (c) magnitude do coeficiente de determinação múltipla. O ajustamento das equações estimativas foi feito pelo método dos mínimos quadrados é a computação eletrônica das estatísticas originais realizada no Centro de Computação Eletrônica da Escola Superior de Agricultura “Luiz de Queiroz”. As equações estimativas selecionadas foram as seguintes: Para a Hipótese Geral 1 Y = 10,37 X₃ 0,175 X₄ 0,512 R² = 0,61 Para a Hipótese Geral 2 Y = 194,1 X₁ 0,412 X₂ -0,471 X₅ 0,217 R² = 0,54 A partir dessas funções, as produtividades médias e marginais correspondentes às médias geométricas dos valores observados foram as seguintes: (Descrito na Tese) Com base na produtividade marginal da variável X₅, procurou-se também estimar, ainda que grosseiramente, as produtividades marginais do crédito de custeio e do crédito de investimento. Em valores, essas produtividades marginais foram estimadas em NCr$ 2,07 e NCr$ 1,56, respectivamente. Para a derivação das curvas de demanda a curto prazo do capital na forma de “despesas de custeio” e de “animais e maquinaria”, determinaram-se, para cada fator, dois valores de produtividade marginal. Êsses valores, juntamente com aquêles correspondentes às respectivas médias geométricas possibilitaram, então, o ajuste a mão livre dessas curvas. As coordenadas dos três pontos de cada curva de demanda foram as seguintes: Para a Demanda do Fator x₃ (Descrito na Tese) Para a Demanda do Fator x₄ (Descrito na Tese)

This thesis presents an analysis of some aspects of the utilization and efficiency of credit and some other production factors for farms in a predominantly agricultural area in the State of São Paulo. The following are the specific objectives: a. Describe and analyse the principal characteristics of the farms that had access to credit. b. Estimate the average and marginal productivities of total land, family labor and capital in the form of both variable expenses and “fixed” expenses (animals plus equipment), for those farms which utilize credit. c. Estimate the average and marginal productivities of credit and analyse the relationship of credit to total farm receipts. d. Derive the short term demand function of capital in terms of variable expenses and of "fixed" expenses (animals plus equipment), then examine the probable implications for efficient utilization of credit. The data for this research were obtained through personal interviews with farmers in the Municípios (roughly corresponding to counties) of Itapetininga and Guareí, which are located in the southern part of the State. Selection of the sample was by means of random groups. The total sample consisted of 132 farms classified into ten groups according to the type of farming enterprises present. From this sample, 49 farms, which had utilized credit during the 1964-65 period, were selected for this study. The 49 farms utilizing credit during this period constituted 37 percent of the total sample. Although the data were collected during the 1965-66 agricultural year, 1965 was considered as the base period for this research. The Chi Square (X²) test was used to determine whether the distribution of the 49 farmers using credit differed significantly from the total sample with respect to type of farm. The two frequency distributions were accepted as being homogeneous, which indicated that the group of farmers using credit were representative of the types of farms encountered in the total sample. The econometric model utilized to analyze the major objectives previously proposed was the model first suggested by Charles Cobb and Paul Douglas. Two general hypotheses were formulated as follows for the actual analysis: General Hypothesis 1: Log Y = log a + b₁ log X₁ + b₂ log X₂ + b₃ log X₃ + b₄ log X₄ General Hypothesis 2: Log Y = log a + b₁ log X₁ + b₂ log X₂ + b₅ log X₅ where: Y = receita agrícola total (NCr$) X₁ = terra total explorada (hectare) X₂ = trabalho da família (equivalente-homem) X₃ = capital na forma de “despesas de custeio” (NCr$) X₄ = capital na forma de “animais e maquinaria” (NCr$) X₅ = crédito para custeio e investimento (NCr$) To determine the production functions which would best represent the input-output relationships, ten different equations were fitted. Six were relative to General Hypothesis 1 and :four relative to General Hypothesis 2. The equations selected as giving the “best” fit were: For General Hypothesis 1: Y = 10,37 X₃ 0,175 X₄ 0,512 R² = 0,61 For General Hypothesis 2: Y = 194,1 X₁ 0,412 X₂ -0,471 X₅ 0,217 R² = 0,54 From those two functions, the average and marginal productivities corresponding to the geometric means of the observed values were calculated and found to be as follows: (See Thesis) Using the marginal productivity of the variable X₅, an attempt was also made to roughly estimate the marginal productivity of credit used for variable expenses and for “fixed” expenses. These marginal productivities were estimated as NCr$ 2, 07 and NCr$ 1,56, respectively. To derive the short term demand curves for capital, in the form of variable expenses and of “fixed” expenses, two marginal product values for each factor were determined. These two marginal product values selected 50% above and below their respective geometric means, enabled the free-hand fitting of the demand curves. The coordinates of the three points of each demand curve were as follows: Demand of capital for variable expenses (X₃): (See Thesis) Demand of capital for “fixed” expenses (X₄): (See Thesis) The major conclusions of this study are: 1. The analysis indicated that the farmers who used credit are strongly market oriented. They are mostly producers operating large farms with a very low rate of turnover (0.09), on their capital. 2. The structure of total capital investment conforms to the type of agriculture in the area studied which consists of extensive agriculture. Land constitutes about 80 percent of total investment. Next in importance is the investment in machinery and livestock. 3. The amount of credi t used by the farmers is very low. The average loan represents only about 10 percent of the total investment in equipment and livestock 9 which normally· constitutes collateral for obtaining credit. 4. There is evidence indicating that the farmers in the study have a very low average net income. The capital investment, family labor and farm manager appear to be receiving inadequate compensation. This could be a result of the very low productivity of the agricultural resources in the area. 5. The estimated production relationships of capital, in the form of both variable and “fixed” expenses, are considered satisfactory despite the fact that the results for variable expenses should be interpreted with some caution. 6. These two factors of production are strongly associated with the variation in total farm receipts and are both being used in the rational stage of the production function. 7. If all other factors are held constant, a 10 percent increase in average variable expenses or an equal increase in “fixed” expenses could be expected to increase total farm receipts by two percent and five percent, respectively. 8. The larger farms have greater total farm receipts and use more credit, However, family labor was found to be negatively correlated with all factors of production considered and, also, with total farm receipts. 9. The estimated production relationships of land, family labor and credit, used for variable and “fixed” expenses, were found to be satisfactory. 10. With all other factors held constant, an increase of 10 percent in land or an equal increment in credit, used for variable or “fixed” expenses, total farm receipts could be expected to increase by four percent and two percent, respectively. The same increase (10 percent) in family labor tends to reduce total farm receipts by five percent. 11. The marginal value product of land, expressed as a percentage of the average value per hectare in 1965, is 4.7 percent. This tends to indicate that the marginal efficiency of land is very low. 12. The negative marginal productivity of family labor may have extremely important policy implications for agricultural development in the region. However, there is a possibility that the variable overestimates the real contribution of family labor to the production process. 13. The marginal value product estimated for credit, both for variable and “fixed” expenses, indicates that agricultural credit can perform a dynamic and definite role in increasing the level of income and the productive capacity of the farmers in the region. 14. The analysis of the short-run demand curve for variable expenses indicates that farmers are acting irrationally. The data indicate that they should probably reduce variable expenses. 15. The demand curve estimated for “fixed” expenses indicates that the quantity utilized for this factor is close to the equilibrium position. ln this case it was found that “fixed” expense should be increased. 16. The results of this research may have important implications for economic policy decisions. In terms of general agriculture policy, consideration should be given to the problem of the utilization of productive resources. The marginal productivity analysis indicates that large structural modifications are necessary in ·agriculture in the region. Also, this need should be objectively considered before implementation of new credit policies and technical assistance to agriculture in the region