Demanda domiciliar por carnes no Brasil: a questão da separabilidade

• Autor(a) principal: Travassos, Guilherme Fonseca
• Data de Publicação: 2014
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: LOCUS Repositório Institucional da UFV
• Texto Completo: http://locus.ufv.br/handle/123456789/93
• topic: Carne - Consumo; Alimentos - Consumo; Separabilidade fraca; Meat - Consumption; Food - Consumption; Weak separability; CNPQ::CIENCIAS SOCIAIS APLICADAS::ECONOMIA::ECONOMIA DO BEM-ESTAR SOCIAL::ECONOMIA DO CONSUMIDOR

Given the importance of the meat sector in the Brazilian economy, in addition to changes in household demand for meat in Brazil, mainly caused by structural changes such as urbanization, changes in demographics and increased demand for convenience, with greater participation of women in the labor market, this study investigates the demand for meat by Brazilian families based on the household Budget Survey (POF/IBGE) 2008/2009. So it was necessary to know exactly the structure of demand for the product, using the concept of weak separability, a hypothesis often assumed in the analysis of demand, and a condition to represent the consumption in budget stages, allowing the use of only the price and expenditure of the product of interest in the demand estimation. Specifically, we sought to test whether the meat group is weakly separable from other foods in household demand in Brazil and how is the format of the utility tree in the demand for meats. Furthermore, we analyzed the sensitivity of meat consumption in relation to expenditure and price, and the influence of demographic variables, location and composition of households. Finally, the relationship between the demand for processed cuts and the fact that the woman is the head of the household was further verified, assessing the demand for convenience. Results showed that meat is not separable from other foods, and weak separability between meat by animal type or quality was also rejected. Thus, the utility tree chosen as a suitable model was that where all foods are demanded in the same stage of consumption. In order to obtain more reliable results, we sought to test the restrictions of symmetry and homogeneity, which were rejected. Therefore, the demand system used for household demand for meat was the model with no symmetry and homogeneity. Thus, this system was estimated using the QUAIDS model with the Shonkwiller and Yen procedure. Moreover, we also corrected for prices and expenditure endogeneity as a way to obtain more accurate estimates. In the first stage of estimation, we highlight the effects of the educational stock of the head of household, in addition to variables that express household location. In the second stage, we obtained expenditure elasticities, own and cross-price elasticities for each good, which allowed us to analyze the behavior of households in the face of changes in economic variables. Results indicated that the sensitivity of consumers to changes in expenditure is greater than the sensitivity to changes in prices. Thus, policies that help improve income, rather than policies that promote the decline in x prices, are more effective to encourage the consumption of meat in Brazilian households. Also, we noticed a trend of change in intra-group consumption for meat chicken, behavior not observed for pork and beef. Furthermore, trends of increased consumption of poultry meat against other meats were confirmed. Demand for meats also depends on location and household composition. The trend of change in family structure, in which the woman has acted as responsible for household decisions, was confirmed for the demand for meat, not rejecting the hypothesis that the increase in demand for processed items can be due to a greater demand for convenience, being positively related to the fact that the woman is the head of household. Besides, results showed that, when estimating demand systems without considering the hypothesis of weak separability, there were no statistically significant differences in expenditure elasticities and in Marshallian price elasticities . However, there were statistically significant differences in cross-price elasticities and marginal impacts of demographic variables. Thus, the choice of a particular utility tree over the remaining models can influence the interpretation of results, especially in the relations of complementarity and substitutability between goods.