IAB-BAMF-SOEP Survey of Refugees - 2023: Sampling, nonresponse, and weighting of the sample M9

Steinhauer, Hans Walter; Siegert, Manuel; Süttmann, Felix; Siegers, Rainer; Zinn, Sabine Research Report IAB-BAMF-SOEP Survey of Refugees - 2023: Sampling, nonresponse, and weighting of the sample M9 SOEP Survey Papers, No. 1537 Provided in Cooperation with: German Institute for Economic Research (DIW Berlin) Suggested Citation: Steinhauer, Hans Walter; Siegert, Manuel; Süttmann, Felix; Siegers, Rainer; Zinn, Sabine (2025) : IAB-BAMF-SOEP Survey of Refugees - 2023: Sampling, nonresponse, and weighting of the sample M9, SOEP Survey Papers, No. 1537, Deutsches Institut für Wirtschaftsforschung (DIW), German Socio-Economic Panel (SOEP), Berlin This Version is available at: https://hdl.handle.net/10419/331794 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-sa/4.0/ Hans Walter Steinhauer, Manuel Siegert, Felix Süttmann, Rainer Siegers, Sabine Zinn Series C - Data Documentations (Datendokumentationen) IAB-BAMF-SOEP Survey of Refugees – 2023: Sampling, Nonresponse, and Weighting of the Sample M9 1537 2025 Running since 1984, the German Socio-Economic Panel (SOEP) is a wide-ranging representative longitudinal study of private households, located at the German Institute for Economic Research, DIW Berlin. The aim of the SOEP Survey Papers Series is to thoroughly document the survey’s data collection and data processing. The SOEP Survey Papers is comprised of the following series: Series A – Survey Instruments (Erhebungsinstrumente) Series B – Survey Reports (Methodenberichte) Series C – Data Documentation (Datendokumentationen) Series D – Variable Descriptions and Coding Series E – SOEPmonitors Series F – SOEP Newsletters Series G – General Issues and Teaching Materials The SOEP Survey Papers are available at http://www.diw.de/soepsurveypapers Editors: Dr. Jan Goebel, DIW Berlin Dr. Christian Hunkler, DIW Berlin Prof. Dr. Philipp Lersch, DIW Berlin and Humboldt-Universität zu Berlin Dr. Levent Neyse, DIW Berlin and Berlin Social Science Center (WZB) Prof. Dr. Carsten Schröder, DIW Berlin and Freie Universität Berlin Prof. Dr. Sabine Zinn, DIW Berlin and Humboldt-Universität zu Berlin Please cite this paper as follows: Hans Walter Steinhauer, Manuel Siegert, Felix Süttmann, Rainer Siegers, Sabine Zinn. 2025. IAB-BAMFSOEP Survey of Refugees – 2023: Sampling, Nonresponse, and Weighting of the Sample M9. SOEP Survey Papers 1537. Series C. Berlin: DIW/SOEP This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. © 2025 by SOEP ISSN: 2193-5580 (online) DIW Berlin German Socio-Economic Panel (SOEP) Mohrenstr. 58 10117 Berlin Germany [email protected] IAB-BAMF-SOEP Survey of Refugees – 2023: Sampling, Nonresponse, and Weighting of the Sample M9 Hans Walter Steinhauer1, Manuel Siegert2, Felix Süttmann1, Rainer Siegers1, and Sabine Zinn1,3 1German Institute for Economic Research (DIW Berlin) 2Federal Office for Migration and Refugees - Research Center (BAMF-FZ) 3Humboldt-Universität zu Berlin July 28, 2025 Abstract This paper outlines the sampling and weighting process for the M9 sample, an augmentation sample for the IAB-BAMF-SOEP Survey of Refugees. It augments the samples M3, M4, M5, and M6 with refugees entering Germany between January 1, 2017, and the August 31, 2022. It also adds refugees from Ukraine by including those who entered Germany in the period from June 9, 2022 up to and including August 31, 2022. The sample is drawn from the Central Register of Foreigners (Ausländerzentralregister; AZR) in a stratified two-stage design. At the first stage clusters of districts are sampled and at the second stage refugees are sampled as anchor persons for the household. Design weights are adjusted to correct for nonresponse and coverage. Nonresponse adjustments make use of design information, interviewer characteristics, arealevel and regional information, as well as information from the AZR. The marginal distributions we use for post-stratification are derived from the AZR as well. The M9 sample and its weights are seamlessly integrated into the IAB-BAMF-SOEP Survey of Refugees and the SOEP. 1. Introduction Together with the Institute for Employment Research (IAB) and the Research Center of the Federal Office for Migration and Refugees (BAMF-FZ) the Socio-Economic Panel (SOEP) at the German Institute for Economic Research in Berlin (DIW Berlin) surveys households of refugees in Germany in a longitudinal study. The cooperation conducting the IAB-BAMF-SOEP Survey of Refugees started in 2016 and is designed as a panel study of refugee households, see Brücker et al. (accepted). It is incorporated into the SOEP as subsamples M3, M4, M5, and M6. These subsamples cover mainly refugees from Syria, Iraq, Afghanistan and Africa. In March 2022, the SOEP together with the IAB, the German Family Demography Panel Study (FReDA) at the Federal Institute for Population Research (BiB) and the BAMF set up the IAB-BiB/FReDA-BAMF-SOEP Survey to address the need for data on Ukrainian refugees in Germany; see Steinhauer, Décieux, Siegert, Ette, and Zinn (2024). Ukrainians who entered Germany in the period from February 24, 2022 up to June 8, 2022 were interviewed. Interviews were conducted from August to October 2022 (online and paper questionnaire) and a second wave took place between January and March 2023. After the second wave, half of the participants of the IAB-BiB/FReDA-BAMF-SOEP-Survey became part of the IAB-BAMF-SOEP-Survey of Refugees. The IAB-BAMF-SOEP-Survey of Refugees was augmented in 2023 by two groups: Ukrainian refugees who entered Germany in the period from June 9 to August 31, 2022 (UKR 2023) and refugees from the rest of the world who entered Germany from January 2017 to August 31, 2022 (ROW 2023). Together with the Ukrainian refugees from the IAB-BiB/FReDABAMF-SOEP survey (UKR 2022), the two newly drawn groups form the M9 sample. This paper documents the sampling design for the UKR 2023 and ROW 2023 and the weighting strategy for the 2023 sample M9 of the IAB-BAMF-SOEP Survey of Refugees. Therefore, Section 2 provides details on the population. Sampling is described in Section 3. Section 4 provides information on the fieldwork and its results. Weighting adjustments are presented in Section 5. Finally, Section 6 gives a brief summary. 2. Target Population and Sampling Frame The target population of sample M9 consists of three main groups. The first group, UKR 2022, consists of persons with a Ukrainian Nationality arriving in Germany between February 24, 2022, and June 8, 2022 (cp. Steinhauer et al., 2024, p. 86f). The second group, UKR 2023, is formed by persons with a Ukrainian Nationality arriving in Germany between June 9, 2022, and August 1, 2022. The third group, ROW 2023, covers persons who entered Germany between January 2017 and end of August 2022 and filing an asylum application (excluding the previous groups UKR 2022 and UKR 2023). Note that, Ukrainian refugees arriving in Germany before the Russian invasion are included in the group ROW 2023. Hence, the previous refugee samples M3, M4, M5 and M6 are extended to include persons with entry between January 2017 and August 2022 by the group ROW 2023. The IAB-BAMF-SOEP survey also includes refugees from Ukraine by the two groups UKR 2022 and UKR 2023. To sample from this population, we utilized the Central Register of Foreigners (Ausländerzentralregister, AZR), an official dataset provided by the Federal Office for Migration SOEP Survey Paper 1537 1 v40 and Refugees (Bundesamt für Migration und Flüchtlinge, BAMF). The AZR records all foreign nationals who have resided in Germany for a minimum of three months, thereby encompassing the target refugee population. Further details on the register are provided by Babka von Gostomski and Pupeter (2008). For the registration process of the Ukrainian refugee population and its peculiarities, see Steinhauer et al. (2024). We filtered this register to identify individuals meeting our population criteria. Overall, the AZR included 483,290 individuals who met these conditions, see Table 1. Table 1: Numbers of adult refugees by Nationality and immigration year. Immigration year Nationality 2017 2018 2019 2020 2021 20221Total Syria 21,676 13,287 11,231 11,404 22,441 13,227 93,266 Afghanistan 3,832 3,970 4,768 5,755 14,495 8,972 41,792 Iraq 8,545 6,036 5,262 4,113 9,637 4,168 37,761 Iran 4,565 7,270 5,476 1,992 2,054 1,903 23,260 Ukraine2187 323 320 143 359 117,317 118,649 Turkey 4,573 6,472 6,609 3,730 5,943 6,599 33,926 West-Africa 7,146 7,012 6,067 3,270 3,015 2,708 29,218 East-Africa 8,872 4,900 2,991 2,012 3,407 2,053 24,235 Others 12,860 11,945 12,513 9,083 16,978 17,804 81,183 Total 72,256 61,215 55,237 41,502 78,329 174,751 483,290 Notes: 1Covers only January until August. 2Refugees arriving in Germany before 2022 will be added to the ROW 2023. In total, the population consists of 483,290 refugees. Table 1 presents the number of immigrants from specific nationalities over the years 2017 through the end of August 2022. Overall immigration numbers fluctuated significantly during this period. Starting off with 72,256 refugees in 2017 dropping to 41,502 in 2020. After the corona lockdowns and travel restrictions in 2020 the numbers rose again in 2021. The sharp increase in 2022 reflects only refugees who arrived in Germany until August. It is mostly driven by Ukrainians who arrived in Germany after the war started. Before 2022 only few Ukrainians are registered in the AZR each year. Among the other nationalities, Syria shows consistently high numbers but with a dip around 2018-2020. Afghanistan shows a steady increase until 2021. Other regions and countries like Iraq, Iran, West Africa, and East Africa exhibit varying levels but generally lower than Syria. Of the 483,290 refugees listed above, 87% are registered with regular immigration authorities (Ausländerbehörden, ABH). The remaining part is registered through, for example, police or border patrol. 3. Sampling Design To sample from the desired population, we filter the AZR for refugees registered with regular immigration authorities. Thus, we mainly exclude refugees from initial reception facilities. Further, there are some regional peculiarities. In Saarland there is only one central authority located in Saarlouis. Further, the district Spree-Neiße and the city of Cottbus are joined to a cluster as well as the district and the city of Kassel. Except SOEP Survey Paper 1537 2 v40 for these peculiarities the jurisdiction of the ABH and the district boarders coincide. So in a first step we create primary sampling units (PSU) from the authorities regionally coinciding with cities and districts. The creation of PSU aims to create clusters that are more similar in terms of their size. First, cities and districts with at least 2,000 refugees form a PSU on their own. This mainly concerns city states Berlin, Hamburg, and Bremen, but also Saarlouis (as explained above), Leipzig, Cologne, Frankfurt am Main or Munich. Next we join cities not yet forming a PSU having only one district surrounding it. In subsequent steps we allow for more neighboring districts to be joined yielding the desired size of 2,000 refuges. When these joins become to large in terms of area, we lower the minimum number of refugees per PSU to 1,000. For a detailed description of the clustering see Jacobsen et al. (2019) Note that, refugees are usually distributed according to „Königssteiner Schlüssel“. This holds for the group of ROW 2023. Ukrainian refugees, on the other hand, can choose their place of residence themselves under certain conditions; see Steinhauer et al. (2024). For this reason the clustering is done separately for the two strata. This procedure results in a total of 190 PSUs for both strata (groups). Of these, 90 PSUs are sampled at the first stage with probability proportional to size in each stratum. Within the sampled PSUs we randomly select a fixed number of refugees at the second stage. The number of refugees to be sampled at the second stage is computed in a way that the sampling design in total yields roughly equal inclusion probabilities for refugees at the second stage. This holds for both strata covering the distinct groups of UKR 2023 and ROW 2023. Thus, the design can be summarized as stratified two-stage sampling design. For details of the sampling design of UKR 2022, see Steinhauer et al. (2024). 4. Fieldwork Results and Response Rates The M9 sample covers the two groups of UKR 2023 and ROW 2023 which are surveyed for the first time. The group UKR 2022 was already surveyed before. For details of the field work results and response rates of UKR 2022, see Steinhauer et al. (2024). The BAMF delivered the addresses to infas Institute for Applied Social Science, the field work agency, to be validated. During the fieldwork a total of 33,766 addresses was validated. Of these 1,559 addresses were not used. All valid addresses were sent a pre-contact form via mail. It provided details on the survey, access to a CAWI and PAPI short questionnaire, and asked for interest in participation. All 9,251 households responding to the pre-contact were deployed to field. Of those households from the ROW group who did not respond to the pre-contact a random sample of 5,000 households was drawn and also deployed to the CAPI field. A total of 7,388 pre-contact forms were returned due to false address information. Also, 375 individuals actively refused to participate in the pre-contact phase of the survey. Figure 1 in Appendix A details the processes leading to being interviewed in the survey. The interviews were conducted between June and December 2023. All households started in CAPI mode with the possibility to switch to another mode. SOEP Survey Paper 1537 3 v40 Table 2: Fieldwork results on the household-level according to American Association for Public Opinion Research (2023). Final Disposition Code UKR 2022 ROW 2023 UKR 2023 Total Number Proportion Number Proportion Number Proportion Number 1. Interview (1.10) Complete 1,276 0.271 1,880 0.084 594 0.088 3,750 (1.20) Partial 276 0.059 702 0.031 141 0.021 1,119 2. Eligible, Non-Interview (2.10) Refusal 122 0.026 704 0.032 58 0.009 884 (2.12) Break-off 55 0.012 326 0.015 39 0.006 420 (2.20) Non-contact 171 0.036 902 0.040 63 0.009 1,136 (2.31) Dead 2 0.000 6 0.000 5 0.001 13 (2.32) Physically / mentally unable 16 0.003 57 0.003 11 0.002 84 (2.33) Language problems 29 0.006 286 0.013 15 0.002 330 (2.36) Miscellaneous 662 0.140 3,360 0.150 429 0.064 4,451 3. Unknown eligibility, non-interview (3.11) Not attempted or worked 0 0.000 1,587 0.071 0 0.000 1,587 (3.22) Nothing ever returned 1,363 0.289 8,191 0.367 3,525 0.525 13,079 4. Not Eligible (4.10) Out of sample / screened out 0 0.000 0 0.000 2 0.000 2 (4.20) Moved abroad 62 0.013 13 0.001 25 0.004 100 (4.40) Untraceable 669 0.142 4,158 0.186 1,803 0.268 6,630 (4.50) Non-residential building 12 0.003 159 0.007 10 0.001 181 Total 4,715 1.000 22,331 1.000 6,720 1.000 33,766 Notes: 1Category includes M9-ROW not participating in pre-contact. Proportions might not add up to one due to rounding errors. SOEP Survey Paper 1537 4 v40 Table 2 provides details on the final disposition outcomes for sampled households across three groups, using classification codes defined by the American Association for Public Opinion Research (2023). For each group, the table lists the number of households falling into each AAPOR disposition category: those that resulted in an interview (complete or partial), eligible but not interviewed, households with unknown eligibility, and those who were not eligible. The proportions represent the percentage of contacted households within each specific survey period that resulted in these final categories. In total, there were 4,882 (UKR 2022: 1,552, UKR 2023: 735, ROW 2023: 2,582) complete or partial interviews, resulting in a response rate on the household-level, calculated according to American Association for Public Opinion Research (2023), of RR2 = 4,869 26,853 = 0.181 (UKR 2022: 1,552 3,972 = 0.391, UKR 2023: 735 4,880 = 0.151, ROW 2023: 2,582 18,001 = 0.143). The response rates are higher for Ukrainian refugees (regardless of the group 2022 and 2023) and for ROW 2023 comparable to M6, see Steinhauer, Siegers, Siegert, Jacobsen, and Zinn (2022). The refusal rate is REF 1 = 1,304 26,853 = 0.049 (UKR 2022: 177 3,972 = 0.045, UKR 2023: 97 4,880 = 0.020, ROW 2023: 1,030 18,001 = 0.057) and, thus, is similar to other samples / studies for Ukrainian refugees. For refugees from other countries (ROW 2023) the refusal rate is higher compared to previous refugee samples of the IAB-BAMF-SOEP Survey of Refugees. The proportions for non-contact as well as for untraceable households are notably lower than the earlier samples M3, M4, M5, and M6. 5. Cross-sectional Weighting The computation of survey weights consists of three steps (Brick & Kalton, 1996). In a first step, design weights are derived as inverse of the inclusion probability. Design weights are adjusted to correct for unit nonresponse in a second step. This step is referred to as sample weighting adjustment by Kalton and Kasprzyk (1986). Finally, nonresponse adjusted weights are calibrated so that estimates conform to known population parameters or to meet specific distributions. Kalton and Kasprzyk (1986) refer to this step as population weighting adjustment. For details on the general weighting strategy of the SOEP and the integration of new samples, see Kroh, Siegers, and Kühne (2015). Please note that the AZR lists individuals without providing information about their household context. However, the IAB-BAMF-SOEP Survey of Refugees is a household panel survey in which all adults are interviewed. Consequently, a household with two individuals, for example, has a higher probability of selection compared to a single-person household. To determine a household’s sampling probability, we assign sampling probabilities to all members of existing households, even though these individuals were not initially sampled as anchor persons. This involves accounting for various characteristics used in stratifying and clustering the sample. Once the sampling probabilities for each household member are identified, we calculate the joint probability the household to be part of the sample. For formal documentation of this procedure, see Kroh, Kühne, Jacobsen, Siegert, and Siegers (2017). To address potential selectivity resulting from fieldwork and nonresponse, we employ three models that capture the success of (pre-)contacting a household as well as the decision-making process of households regarding participation. The models incorporate information on: a) all households that have been pre-contacted and did not fall outside the population, SOEP Survey Paper 1537 5 v40 References American Association for Public Opinion Research. (2023). Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys (10th ed.). Retrieved from https://aapor.org/wp-content/uploads/2024/03/Standards -Definitions-10th-edition.pdf Babka von Gostomski, C., & Pupeter, M. (2008). Zufallsbefragung von Ausländern auf Basis des Ausländerzentralregisters: Erfahrungen bei der Repräsentativbefragung ”Ausgewählte Migrantengruppen in Deutschland 2006/2007” (RAM). Methoden, Daten, Analysen,2(2), 149–177. Brick, J. M., & Kalton, G. (1996). Handling missing data in survey research. Statistical methods in medical research,5(3), 215–238. doi: 10.1177/096228029600500302 Brücker, H., Kosyakova, Y., Rother, N., Zinn, S., Liebau, E., Gider, W., … Siegert, M. (accepted). Exploring integration and migration dynamics: The research potentials of a large-scale longitudinal household study of refugees in germany. European Sociological Review, A preprint of the paper is available here: https://doi.org/10.17605/OSF.IO/CQ2KZ. doi: 10.1093/esr/jcaf032 Jacobsen, J., Kroh, M., Kühne, S., Scheible, J. A., Siegers, R., & Siegert, M. (2019). Supplementary of the IAB-BAMF-SOEP Survey of Refugees in Germany (M5) 2017 (SOEP Survey Papers No. 605). Berlin: DIW/SOEP. Retrieved from https:// www.diw.de/documents/publikationen/73/diw_01.c.615884.de/diw_ssp0605.pdf Kalton, G., & Kasprzyk, D. (1986). The treatment of missing survey data. Survey methodology,12(1), 1–16. Kroh, M., Kühne, S., Jacobsen, J., Siegert, M., & Siegers, R. (2017). Sampling, nonresponse, and integrated weighting of the 2016 IAB-BAMF-SOEP Survey of Refugees (M3/M4)—revised version (SOEP Survey Papers No. 477). Berlin: DIW/SOEP. Retrieved from https://www.diw.de/documents/publikationen/73/ diw_01.c.572346.de/diw_ssp0477.pdf Kroh, M., Siegers, R., & Kühne, S. (2015). Gewichtung und Integration von Auffrischungsstichproben am Beispiel des Sozio-oekonomischen Panels (SOEP). In Nonresponse bias (pp. 409–444). Springer. R Core Team. (2025). R: A language and environment for statistical computing [Computer software manual]. Vienna, Austria. Retrieved from https://www.R-project.org/ Steinhauer, H. W., Décieux, J. P., Siegert, M., Ette, A., & Zinn, S. (2024, March). Establishing a probability sample in a crisis context: the example of ukrainian refugees in germany in 2022. AStA Wirtschaftsund Sozialstatistisches Archiv, 18(1), 77–97. doi: 10.1007/s11943-024-00338-0 Steinhauer, H. W., Siegers, R., Siegert, M., Jacobsen, J., & Zinn, S. (2022). Sampling, Nonresponse, and Weighting of the 2020 Refreshment Sample (M6) of the IABBAMF-SOEP Refugee Panel (SOEP Survey Papers No. 1104). Berlin: DIW/SOEP. Torregroza, S., Leschny, K., Ruland, M., Hess, D., Gilberg, R., & von der Burg, K. (2022). IAB-BIB-BAMF-SOEP Ukrainian Refugees in Germany (Wave 1) (Tech. Rep.). Bonn: infas Institute for Applied Social Sciences. SOEP Survey Paper 1537 12 v40 A. Figures Figure 1: Flowchart of the processes leading to panel participation in the M9 samples. UKR 2022 UKR 2023 ROW 2023 Pre-contact Participation Deployment Face to face Sampling for Face to face ROW 2023 only Contact Survey Cooperation Panel Sampled Not surveyed Yes No Yes Yes Yes No No No SOEP Survey Paper 1537 13 v40