The Flow of Urban to Rural Migration During 2017 to 2021 in Iran: Analysis of Labor Force Survey Database

Reverse migration as a demographic phenomenon is influenced by various causes and factors. As this trend continues and large numbers of migrants arrive in rural areas, the government faces the challenge of providing basic rural infrastructure and services to meet the needs of the migrant population. Therefore, managing and formulating supportive and encouraging policies and creating a suitable platform for reverse migration can be one of the most important solutions to address this challenging issue. In this research, for the first time in the country, we evaluate the trend of urban-rural migration and identif its causes using Labor Force Survey Database during 2017-2021. The research findings show that urban-to-rural migrations and their causes are strongly influenced by the socio-economic characteristics of migrants. Therefore, it is necessary to pay attention to different sections of the society and continuously investigate urban-to-rural migrations based on sustainable development indicators in the country's villages. Descriptive analysis and logistic regression model fitting on the data during the five-year period showed that age, sex, activity status and education level are related to urban to rural migration. 

According to the results of the 2011 population and housing census, the amount of urban to rural migrants in Iran is more than the number of rural to urban migrants. It seems that the reverse flow of migration from urban to rural areas, which is called reverse migration, has been started. However, investigating the flow of urban to rural migration and identifying the reasons of this type of migration in Iran, especially in short time intervals, has been less discussed. The issue is due to the lack of a database or a specific framework for investigating and evaluating this type of migration.Until now, the main source of information and data for investigating migration patterns and its reasons has been the latest population and housing census. However, due to the fact that the censuses are carried out with a frequency of five or ten years, it is not possible to continuously monitor internal migrations and their causes. Therefore, it is important to find other data sources to monitor the flow of urban to rural migrations in shorter time intervals. Investigations showed that the only other source other than the census that provides information is the Labor Force Survey (LFS), which is conducted annually by Statistical Center of Iran. The general purpose of this paper is to estimate the number of urban to rural migrants during the years 2017 to 2021 at national level, to investigate its causes, and to identify the socio-economic characteristics of migrants during these years using LFS data. 

The research method is the secondary analysis of the data obtained from the labor force survey. Descriptive statistics are presented and the logistic regression model has also been used to investigate the factors affecting urban to rural migration. According to the definition of the LFS, those household members whose Length of stay (continuously until the survey time) was less than 12 months in that city or village are known as immigrants. The target population in LFS is the members of private settled or collective households. The sampling method is stratified two-stage cluster sampling. The annual sample size in rural areas is about 94 thousand households. Since LFS is not done centrally in a specific time of the year, the duration of people's stay is not measured at a certain point of time and the same for everyone. The causes of migration listed in the LFS questionnaire are similar to those mentioned in the 2011 census and include seeking work, looking for better work, job transfer, education, graduation, beginning or ending of compulsory military service, follow the household and others. 

In this section a summary of results including, the estimation of the number of urban to rural migrants using the LFS data from 2017 to 2021 by reasons, gender and activity status is presented at national level. Table 1- The number of immigrants from the urban to rural areas by gender and the reason for migration during the years 2017 to 2021 According to the Table 1, the highest and the lowest amount of urban to rural migrations among men during the entire period under review was due to begin or end of compulsory military service, education and graduation, respectively. The highest number of urban to rural migrants among women was due to follow the household. In the total period of five years, it can be stated that the total number of migrations was higher for men than for women.According to the results of Table 2, beginning or end of compulsory military service, seeking work, looking for better work, job transfer are main reasons of urban to rural migration among employees during the five-year period. But the reason for migration of inactive people is different from that of employed people, that is follow the household. In the total period of five years, it can be stated that the number of employed immigrants has been more than other people.  Table 2- The number of immigrants from the urban to rural areas by activity status and the reason for migration during the years 2017 to 2021 The effect of various factors on urban to rural migration is investigated using the logistic regression model and based on the available data. In this model, the person's immigration status is considered as the response variable and migration from urban to rural area is considered as a pass event. The independent variables whose information are available from LFS include age, gender, literacy status, activity status, and marital status. Due to the fact that marital status is determined for people aged 10 years and older, model fitting is done for this subgroup of people. The results of fitting the logistic model to the data from 2017 to 2021 are shown in Table 3.As can be seen in Table 3, from 2017 to 2021 urban to rural migration is decreases with increasing age. Also, the odds for women to migrate is less compared to men. Examining the effect of activity status on the odds of migration shows that the chance of migration of unemployed people is higher compared to employed people and the odds of migration of inactive people is lower compared to employed people. But the interaction effect of age and activity status shows that with increasing age, the odds of migration of inactive people is higher compared to working people. This issue can indicate the return of the elderly from the urban to rural areas. Among women, inactive women are more likely to migrate than working women, which can be due to the migration of housewives or female children following the household.
Table 3- Fitting Logistic regression Model on Urban to Rural Migration Data from 2017 to 2021. 

The estimation of the number of urban to rural migrants, its causes, as well as the socio-economic characteristics of migrants at the national level, was presented based on the annual data of LFS. Obviously urban to rural migrations are strongly influenced by the socio-economic characteristics of the migrants. Therefore it is essential to consider different  groups of people based on their characteristics in order to formulating support and encouragement policies and creating a suitable platform to meet their needs. AcknowledgementThis paper is based on the research project: “Feasibility Study of Estimating Urban-to-Rural Migrants, 2016 to 2021”, commissioned by the Vice-president for Rural Development and Deprived Areas of the country and conducted at the Statistical Research and Training Centre of Iran, and we thank them for their support.