Using the IPEA DATA the five macro

Using the IPEA/DATA, the five macro-socioeconomic variables were: the homicide and suicide rate per 100,000 inhabitants, the percentage of poor people, the Gini index as an income inequality measurement, the open unemployment rate and government expenses. These macro-variables were disaggregated for each state for the two years considered for the research. The interaction of these two sets of variables resulted in a total of thirty-seven explicative variables. Chart 1 introduces a description of variables used in the estimation, as well as some statistical information.
Obtained evidence Table 3 introduces estimated results for the ordered multinomial logit model for the probability ratio in the “less happiness” categories with respect to the “more happiness” ones, as show in Eq. (3). However, seeking to analyze the impact of explicative variables on the probability ratio, conclusions shall be based on the “more happiness” categories with regards to the “less happiness” ones. For this purpose, it LY 2157299 cost is enough to consider the value of exp () as a measurement for such impact. Within the statistically significant variables shown for a significance level of 5%, results show that a one decile income growth makes the probability to be happier increase in a 1.1 factor with regards to being less happy. In other words, there is a positive correlation between happiness and income. However, observe Table 2 data, which shows percentages for happiness category for each income level in 2006 and 2014. With regards to the employment status of individuals, the probability to be happier compared to being less happy is 1.2 times higher for those who are employed if compared to other types of occupations (students, housewives and retirees). This result agrees with findings by Guo and Hu (2011), Chuerattanakorn (2007) and Di Tella et al. (2003) for other countries that also verified a significant inverse relation between unemployment and happiness. Within the set of personal explicative variables, the higher impact on the probability ratio of being happier compared to being less happy is explained by the “married” marital status. The probability of being happier is 1.6 times higher for married people if compared to those unmarried. This evidence reinforces results found by Di Tella et al. (2003) for other countries and Corbi and Menezes-Filho (2006) for Brazil. The effect of people’s age on their happiness is represented by a U-shaped curve. Therefore, results suggest that the happiness trajectory is not constant throughout life. Initially, the probability of being happy falls as people age. After certain maturity is reached, happiness probabilities start to grow. This same characteristic was also noticed by Corbi and Menezes-Filho (2006). The geographical area of residence does not seem to have any correlations with the ratio likelihood to be more or less happy. The only state in which stomata relation was statistically significant was Amazonas, if compared to the state of Maranhão (base category). In intertemporal terms, there was no significant difference between the likelihood of being happy in 2006 and 2014. The dummy variable year did not prove significant. The analysis of marginal effects of explicative variables is then made on the probabilities of each happiness category. In other words, how much these likelihoods grow when there is unitary variation in one of these variables and the others are kept constant. Such effects are estimated according to Eq. (5) of Subsection 3.1 and introduced in Table 4. For statistic significance purposes, we considered a level of 5%.
Final considerations The probability ratio of being happier compared to being less happy is 1.2 higher for those who are employed when compared to other types of occupations (students, housewives and retired people). This evidence was also noticed in studies developed by Guo and Hu (2011), Chuerattanakorn (2007) and Di Tella et al. (2003) for other countries.