PEMODELAN BAYESIAN KONSUMSI RUMAH TANGGA AGREGAT MENGGUNAKAN PRIOR ZELLNER

Abstract

In the development of statistics, there are two views of parameters, namely frequentist and Bayesian. In Bayesian, the parameter is a random variable, not a constant like a frequentist view. The research aims to estimate the function or model of household consumption agrees using the Bayesian method. The data used in this study are GDP (y) and household consumption (x) at constant prices (2000) for the 1983Q1 - 2016Q4 period sourced from the Statistics-Indonesia. This study results that the Bayesian regression modeling of the household consumption function agrees with Zellner's previous use. The income coefficient in this model is significant and gets a marginal propensity to consume the value of 0.5702. This implies that more than half of people's income is used for consumption purposes.