What Drives Ukrainian Internal Migrants?

The war in the East of Ukraine has been ongoing for five years. According to some estimates, about 3 million people lived in the occupied territories prior to the conflict. Over one million of them relocated to the government-controlled area of Ukraine – mostly to the regions close to the occupied area (see map on Figure 1). Donetsk and Luhansk oblasts are two biggest hosts for internally displaced persons (IDPs). Thus we focus our attention on these particular regions.

Migration theory suggests that primary motives of people are economic. While there is a lot of support to that for the labour migration, the situation differs when the relocation is involuntary. Non-monetary factors begin to play higher role in the forced migration. Violence, survivorship aspect, social networks and political institutions are factors that matter in people’s decision–making process during the displacement (Balcilar and Nugent, 2018; Ibáñez and Vélez, 2008; Adhikari and Prakash, 2013;Verme, 2017).

In order to find a link between migration motives and household characteristics, we use the survey data for 2016-2018 provided by the non-governmental organization REACH.

These surveys aim at understanding trends and changes in  humanitarian needs of people. They were conducted in summer or early fall (mostly in August/September) of 2016, 2017 and 2018. Households were reached using computerized random point selection (random stratified clustering) within each region based on population density within each stratum. Population data was obtained from State Statistics Service of Ukraine.  

Every year a different number of households was surveyed depending on the available  resources and purposes of IDP support programs. Each household was surveyed only once. Surveys were held as face-to-face interviews with a person who represents the entire household. We used the pooled data from these surveys.

The data includes a block of questions regarding vulnerabilities, employment status, household structure (vulnerabilities and employment of household members) protection, economic and food security, housing, access to health services and education, income, savings and humanitarian assistance. The dataset includes Donetsk and Luhansk regions (5 km, 20 km and more than 20 km from the contact line).

In 2016, 2632 households were surveyed, of them 1149 displaced, in 2017 – 546 households, of them 25 displaced, in 2018 – 2565 households, of them 161 displaced. We analyzed only the answers of displaced households. 

Our data include 1335 IDPs’ households from Donetsk and Luhansk oblasts. Figure 2 shows the motives for relocation indicated by those households (households could select any number of motives and they were not asked to rank them).

Figure 2 shows that network connections are the most popular motives for location choice. Safety and security is in the third place, while other motives are chosen by less than 20% of respondents. Only 14% of IDPs take into consideration job opportunities when deciding where to relocate. 

Table 1 presents the correlation matrix of displacement motives. As we can see, family and friends connections are slightly negatively correlated with other motives, while motive of accessing water and electricity, healthcare, safety reasons and free/cheap accommodation are positively correlated with a half of other motives. Safety as a reason for relocation is positively correlated with basic services. And the only correlation higher than 0.5 is for access to basic services and healthcare services. Perhaps these reasons are selected together by households which relocated from severely damaged areas. 

To understand people’s motives we linked them to the following parameters: age, gender, and education of the household head, number of adults and children in a relocated household, presence of vulnerable household members, humanitarian assistance, savings, as well as official IDP status and the region of origin (Donetsk or Luhansk oblast). We estimated the model using seemingly unrelated estimations with one equation for each motive. Right-hand side variables in all equations were the same. This model specification takes into account that people could be driven by several motives.

Job opportunities

We start the discussion of results with the economic motive – presence of job opportunities. Households where a head is male, under 30 and more educated are more likely to be driven by employment opportunities. Thus, for households whose head has a university education the probability of selecting this motive rises by 17% compared to other levels of education.

Basic communal and healthcare services  

These motives matter more for families with higher level of education (the likelihood of selecting this motive increases by 6% for university and by 4% for vocational education). Households with savings are also more likely to select a new place for living because of better access to basic communal and health services. Family composition matters — namely, the presence of a vulnerable household member  increases the probability of choosing this motive by nearly 7%.

Social networks (family / friends)

Social networks proved to be crucial motives for the majority of households. Family connections are slightly more likely to matter for larger families (each additional household member increases the probability of selecting this motive by 3%). And  for households with children the probability to move to friends increases by 4%. 

Households from Luhansk oblast are 11% less likely to be driven by this motive than households from Donetsk oblast. We don’t have a plausible explanation for this result which is statistically significant. 

People who have a damaged house in the area of origin are 6% less likely to move due to this motive but the probability that they select safety and security as a reason for relocation rises by 9%. Older people (53+) are more likely to rely on their friends. 

Free/cheap accommodation and proximity to home

In this case, the only supportive factor turns out to be the presence of a vulnerable household member (in this case the probability of displacement due to this motive increases by 5%). Perhaps, it is harder to travel with a vulnerable household member.  If a household has a damaged house in the area of origin, the likelihood to migrate due to the motive of proximity to home declines.

Policy implications

We found that IDPs’ choices of place for relocation are driven by quite distinct factors. Knowing these factors for each particular household would allow to provide more targeted assistance to them.

For example IDPs during the process of registering for the official status can be directly asked about their motives. Depending on the motives they can be provided with additional assistance. 

Thus households that are driven by the motive of accessing basic and healthcare services are the most in need. They would appreciate the information about health facilities and the registration/admission process in those. These IDPs can be provided additional in-kind support (e.g. drugs or accommodation) along with regular monetary assistance for a limited period of time.

Households that are driven by the safety factor need psychological support, especially if they include vulnerable members such as children, elderly and people with special needs. 

Equally important is the factor of adaptation of IDPs who are driven by these motives. Host communities are important in this process. They may assist in creating common initiatives for IDPs and locals. Adaptation of households driven primarily by safety reasons may take longer.

People who are driven mainly by job motive could appreciate legal advice, psychological support and assistance in finding housing.

The adaptation process for households who are driven by family and friends connections will be easier and thus assistance should concentrate on helping them to find a job. Information about work opportunities and retraining programs will help these people to adapt in the new place.

Finally, it is crucial to continue gathering panel data on IDPs in order to investigate the patterns and detect the trends. Data give us useful insight and it will guide policymakers in the development of policies regarding IDPs.

References