Thessaloniki: Profiling of Refugees, Asylum Seekers and Third Country Nationals not registered with the Asylum Service. Potential and Obstacles to Local Integration. 2019

In total, the survey of refugees and asylum seekers covered 1,808 individuals comprising 641 households. The sample was stratified by accommodation type into three strata:

• Those in the urban accommodation scheme who have been provided with apartments
• Those self-accommodated in Thessaloniki, i.e. are either renting an apartment by themselves, or being hosted by friends, relatives or volunteers
• Those who were fully registered residents of the Open Reception Facilities (ORF)
  The sampling frame for refugees and asylum seekers was UNHCR's ProGres database, while for the ORF, a site population list provided by the camp manager was used as a basis to generate a sample.A simple random sample of households was initially drawn for the accommodation scheme strata and the self-accommodated strata shortly before the data collection was due to begin. During data collection, reaching a majority of the sampled households was challenging due to the listed phone numbers being outdated, as persons of concern often change their pre-paid SIM cards. Unannounced home visits were not an option given time and resource constraints. It was therefore decided to aim for full coverage of both these strata, expecting that a high proportion of the persons in the ProGres database for these strata would not be reachable by phone. To assess potential bias introduced by this approach, the demographic profile of the surveyed persons was compared to that of the whole population of refugees and asylum seekers in the UNHCR ProGres database. The age and sex figures of the population were compared to the survey figures. The sample distributions resemble the population distributions quite closely on the basis of these demographic characteristics. As such, the overall impression is that there is little skew in the survey data for these two strata. It is therefore assumed that the survey results are representative and can be applied to the population as a whole. For the strata of the Open Reception Facility (ORF), the most update site registration list was obtained from the Reception and Identification Service (RIS) that manages the site. The enumerators managed to get in touch with at least one representative of each of the registered households living in the site at the time of the data collection. No one declined the request for an interview. It was not relevant to compare the surveyed population to the UNHCR database list to assess representativity, given that the population in the site had changed significantly since the list for that strata had been assembled by the camp manager in the site. Since a full count of the site population was achieved, the results are considered to be representative for the population.
  A different sampling took place for third country nationals not registered with the Asylum Service. The unified registry for persons with police notes (EURODAC II) could not be accessed for the purpose of the profiling study. Although organizations that provide assistance to police note holders hold information about this population group, including UNHCR which provides cash assistance, there is no exhaustive list. Similarly there is no unified registry for undocumented persons. However, through comparing aggregated information from multiple service providers, a population figure of 200 households was estimated as a rough baseline. In the absence of a registry, it was not possible to construct a list from which a random sample could be drawn. Thus, a non-probability sampling strategy was applied, which included convenience sampling approaches. With non-probability approaches it is not possible to establish how well the sample represents the population unless all members of a given target group have been interviewed. Convenience sampling is a type of non-probability sampling method, where the sample is taken from a group of people easy to contact or to reach, e.g. by snowballing techniques where respondents identify other respondents known to them. The enumeration team interviewed 451 persons making up 227 households under the category of third country nationals not registered with the Asylum Service. This number of households interviewed was slightly higher than the number originally foreseen, a possible explanation for this being the aforementioned influx of arrivals to Thessaloniki the same month. The survey results support this theory, as more than half of the survey respondents from this target group had been in Thessaloniki for less than a month at the time of the interview. The high number of recent arrivals made the estimate of the total population more uncertain. In addition, many of the persons who were approached, declined to be interviewed. As a result, it is difficult to assess how representative the interviewees were of the target group.