ANALYSIS OF FACTORS AFFECTING HAPPINESS WITH MULTINOMIAL LOGIT MODEL

Abstract

The aim of the study is to determine which factors are statistically significant on the happiness of the individual. In the study, the data for the 2010-2014 period are used in Turkey. The data are obtained from the World Values Survey. A total of 751 data are analyzed by making various adjustments on the data. The factors affecting happiness are determined by estimating the multinominal logit model and the Hausman test is applied to the happiness levels for the validity of this analysis. The Hausman test results show that irrelevant alternatives are independent. The health status, life satisfaction, financial satisfaction and gender variables in the study have significant effects on individual happiness levels. Based on the results of the study, the log likelihood value show that model is generally meaningful, the Pearson and Deviation measures show that there is a good fit in the model, and the Likelihood Ratio test result show that coefficients are statistically significant. According to Multinominal Logit Model results, when the dependent variable categories are considered, it is found that the health status and gender have an increasing effect on the happiness of individuals, and life satisfaction and financial satisfaction to have a reducing effect on the happiness of individuals.