Analysis and estimation of consumer expenditure and budget shares are important for understanding quantitatively the expenditure based behaviour of the people of a country or region. The costs attached with performing consumer expenditure and budget shares survey are significant. These surveys are quite time consuming and even a slight increase in sample size could lead to a significant increase in the cost of the survey under the current structure. This high cost for the survey also reduces its flexibility. Hence it is of paramount importance to be able to provide a statistically sound and relatively cheaper facilitation method of survey to be able to understand the consumption expenditure distribution of a country or a region without much loss of inferential power. In the context of the Indian National Sample Survey, in this paper we perform analysis and estimation of consumer expenditure assuming uniform prices across First Stage Units (FSUs) of a sampling design, and check its feasibility in estimating the consumer expenditure distribution of the country. We also compare it with the existing methodology and infer that there is no significant loss of information in the estimated consumer expenditure distribution and other inferences like the Lorenz curve and the Gini Index, when uniform distribution of prices within any FSU is assumed.
翻译:对消费支出和预算份额进行分析和估计,对于从数量上了解一国或一个区域人民的支出行为十分重要。进行消费支出和预算份额调查所附带的费用是相当可观的。这些调查耗费大量时间,抽样规模即使略有增加,也可能导致调查费用在现有结构下大幅度增加。调查费用高昂,也降低了调查的灵活性。因此,至关重要的是,要能够提供一种统计上合理和相对便宜的便利调查方法,以便能够了解一个国家或一个区域的消费支出分配情况,而不会大大丧失推断力。在印度国家抽样调查中,根据抽样设计第一阶段的统一价格,我们分析和估计消费支出,并检查其可行性,以估计该国的消费支出分配情况。我们还将这一方法与现行方法进行比较,并推断,在假定消费支出估计分配估计数和Lorenz曲线等其他推论和吉尼指数等其他推论中,在假定任何价格统一分布时,不会重大损失。