Investigating Factors affecting Income Inequality in Iran using a Combined Cohort Analysis

Objective: Income inequality, discusses how income is distributed across a given population. Income inequality is often associated with wealth inequality. Populations can be divided in different ways to show different levels and forms of income inequality, such as income inequality based on gender or race. Inequality in income distribution is one of the most important issues in the field of macroeconomics, and it has occupied a large part of studies in this field. The reasons for inequality in income distribution are different. Some studies find such differences to be the result of some factors, including differences in economic systems (such as Alvardo et al., 2018; Duke et al., 2019 & Fletcher, 2017), differences in knowledge and skills among individuals (Basmayes and Iverson, 2014). Differences in wage levels in sectors of economic activity, differences in payments based on education level, age, gender and race (Ridejoy, 2011). In any country, corruption, monopolies, shadow economy and other negative conditions have a negative impact on income distribution. According to financial policies in any country, it can have a negative or positive effect on income inequality. The potential for negative or positive effects is related to specific economic outcomes of fiscal policy (IMF, 2017). Although the goal of fiscal policy is to reduce income inequality, fiscal policy goals may not be achieved; Therefore, it is necessary to evaluate the results in the implementation of any financial policy. The causes of income inequality are not the same for all countries. The causes of inequality are different depending on the level of economic, demographic, and cultural development, and these factors have different effects on the inequality of income distribution in different countries. The impact of factors affecting income inequality is also different. National income distribution and inequality depend on many economic factors. Determining which are the main factors is largely dependent on empirical study (Cornia and Kiiski, 2001). 
Method: In order to understand the influence channels of factors affecting income inequality, the basic model used in the studies of Paweenawat and McNown (2014) has been used. Also, the parameters (0β) to (4β) respectively represent the sensitivity of each of the explanatory variables to the variance variable of the logarithm of household income.
Also, (t) represents the time, which in this research is the study period from 1992 to 2022. All the data required for this study were extracted and categorized from the website of Central Bank of Iran and Iran Statistics Center. The sample size of household gross income-expenditure data was variable in the studied years and was between 15,000 and 21,000 households, and the average of each year was used for the studied variables. Because it is not possible to use different samples in different years, and due to the random nature of the sample, there may not be a constant person in all the samples. The classification of the examined consumer items was also based on the COICOP code.
 Results: The results of the Mana test show that all research variables except the variance of the number of people in the households (Hous) were at the level of the Mana.  Also, the results have shown that non-significant variables have been signified by one-time differentiation. The ARDL method can be analyzed by interpreting the short-term, long-term and error-correction equations. In the short-run equation, the dependent variable is displayed with an interval on the right side. Choosing the optimal interval can be done using Akaike, Hanan-Queen and Schwartz-Baysin criteria.  In the present study, the Schwartz-Baysin criterion was used due to the reduction of the degree of freedom and the fact that it saves time in choosing the interval. According to the short-term model estimation results, the variables have a significant effect on the dependent variable at the 90 and 95% confidence level.
The determination coefficient of the model is 71% and the F statistic is significant at the level of 99%. Considering that Band's test statistic is higher than the critical values presented in the table above, the hypothesis of no long-term relationship between model variables is rejected. In fact, based on the obtained results, the long-term relationship between model variables is confirmed at the 95% confidence level. Pesaran et al. (1999) modeling approach and unconstrained error correction model (UECM) are used to estimate the long-term relationship between model variables. In the UECM model, the long-term coefficients are obtained by dividing the coefficients of the explanatory variables by the interval of the first level of the dependent variable (with a negative sign).
 Conclusion: The results showed that there is a U-shaped relationship between income inequality and income. In other words, increasing household income does not always work in the direction of reducing inequality. In other words, from a certain income level, an increase in income causes an increase in the inequality of income distribution. In order to reduce inequality among the members of a household, it is suggested that officials and economic planners pay special attention to the category of employment creation and use appropriate policy packages. According to the estimated results, in order to reduce the inequality in income distribution among the members of a household, an alternative policy such as increasing job opportunities was used. Potential opportunities to acquire education and skills and access to financial capital through appropriate and efficient markets in order to improve income distribution is another solution in this field. Another solution is empowering young people, graduates and other job seekers for employment and earning. According to the estimated results, in order to reduce the inequality in income distribution among the members of a household, an alternative policy such as increasing job opportunities was used. Providing the minimum livelihood for low-income groups and vulnerable members of the society, with emphasis on benefiting from public health, enjoying public education, having minimum acceptable shelter and food security, developing entrepreneurship and empowering the poor and low-income members, expanding support Social media can also help further improve income distribution. Other results showed that the average logarithm of the number of employees and the average logarithm of household income had a negative and significant effect on income inequality. Also, the variance of the number of household members and the logarithm of the number of household members also have a positive and significant effect on inequality.