Subjective Well-Being and Migration: Empirical Insights from Hungary and Austria

Subjective Well-Being and Migration: Empirical Insights from Hungary and Austria Research Report Dávid Sümeghy, Borbála Göncz, György Lengyel, Ádám Németh, Zsolt Németh, Lilla Tóth 2024 FWF–NKFIH Joint Project 1 This working paper was developed in the context of the project MIGWELL – Well-being and Migration: the Hungary – Austria Migration Nexus Work Package 3: ‘Additional Surveys’ MIGWELL is an FWF–NKFIH International Joint Project. In Austria the project is funded by the Austrian Science Fund. In Hungary the project has been implemented with the support provided by the National Research, Development and Innovation Fund (Ministry of Innovation and Technology of Hungary), financed under the ANN funding scheme. Project code: I 5616 (in Austria), 139465 (in Hungary) Project duration: 01.02.2022 – 31.07.2025 (in Austria), 01.12.2021 – 31.05.2025 (in Hungary) Project Partners: Institute for Urban and Regional Research at the Austrian Academy of Sciences Centre for Empirical Social Research at the Corvinus University of Budapest Department of Finno-Ugrian Studies, University of Vienna Scientific Advisory Board: Department of Sociology, University of Vienna Hungarian Demographic Research Institute Hungarian Central Statistical Office MIGWELL’s website: LINK MIGWELL’s outputs will be uploaded to: LINK Inquiries can be directed to: Institute for Urban and Regional Research at the Austrian Academy of Sciences, [email protected].at or Centre for Empirical Social Research at the Corvinus University of Budapest, [email protected] Suggested citation: Sümeghy, D., Göncz, B., Lengyel, Gy., Németh, Á., Németh, Zs. and Tóth, L. (2024), Subjective WellBeing and Migration: Empirical Insights from Hungary and Austria. MIGWELL Working Papers, No. 3.2, 218 p. FWF–NKFIH Joint Project 2 Table of contents MIGWELL at a glance ............................................................................................................................................ 4 1. Introduction ............................................................................................................................................... 5 2. Sources and methods ................................................................................................................................. 6 2.1. Secondary sources ..................................................................................................................................... 6 2.2. Narrative interviews .................................................................................................................................. 7 2.3. Focus groups ............................................................................................................................................. 9 2.4. Cognitive interviews ............................................................................................................................... 10 2.5. Surveys .................................................................................................................................................... 11 2.5.1. MIGWELL survey in Hungary ............................................................................................................... 11 2.5.2. MIGWELL survey in Austria ................................................................................................................. 12 2.6. Methods of quantitative data analysis ..................................................................................................... 15 2.6.1. Models for satisfaction domains and satisfaction with life ..................................................................... 15 2.6.2. Causal model for determining the effect of migration on subjective well-being .................................... 16 2.7. Challenges and limitations ...................................................................................................................... 18 3. Migration plans and experiences among Hungarians living in Hungary and Austria ............................. 21 3.1. Migration potential in Hungary ............................................................................................................... 21 3.2. The attitudes toward emigrants ............................................................................................................... 23 3.3. Migration plans in Hungary .................................................................................................................... 25 3.4. The Hungarian-born population in Austria: an overview ........................................................................ 29 3.5. Motivations of migration to Austria ........................................................................................................ 32 4. The objective factors of well-being among Hungarians living in Hungary and Austria ......................... 34 4.1. Demographic and spatial characteristics ................................................................................................. 35 4.2. Financial situation ................................................................................................................................... 38 4.3. Economic status and job ......................................................................................................................... 46 4.4. Housing conditions ................................................................................................................................. 52 4.5. Household composition ........................................................................................................................... 56 4.6. Social connections................................................................................................................................... 60 4.7. Health status ............................................................................................................................................ 65 4.8. Work-life balance .................................................................................................................................... 67 4.9. External factors ....................................................................................................................................... 69 4.10. Transnational activity .............................................................................................................................. 72 5. Subjective well-being patterns among Hungarians living in Hungary and Austria ................................ 77 5.1. Overall life satisfaction ........................................................................................................................... 77 5.2. Domain satisfaction................................................................................................................................. 78 5.3. Affective well-being ............................................................................................................................... 84 5.4. Eudaimonic well-being ........................................................................................................................... 88 6. The variables associated with domain satisfaction among Hungarians living in Hungary and Austria .. 91 6.1. Satisfaction with job................................................................................................................................ 91 6.1.1. Model for the whole sample .................................................................................................................... 91 6.1.2. Extended model for Austria .................................................................................................................... 97 6.2. Satisfaction with income ....................................................................................................................... 100 6.2.1. Model for the whole sample .................................................................................................................. 100 6.2.2. Extended model for Austria .................................................................................................................. 107 6.3. Satisfaction with housing conditions..................................................................................................... 108 6.3.1. Model for the whole sample .................................................................................................................. 108 FWF–NKFIH Joint Project 3 6.3.2. Extended model for Austria .................................................................................................................. 115 6.4. Satisfaction with social relationships .................................................................................................... 116 6.4.1. Model for the whole sample .................................................................................................................. 116 6.4.2. Extended model for Austria .................................................................................................................. 125 6.5. Satisfaction with health ......................................................................................................................... 127 6.5.1. Model for the whole sample .................................................................................................................. 127 6.5.2. Extended model for Austria .................................................................................................................. 134 6.6. Satisfaction with work-life balance ....................................................................................................... 135 6.6.1. Model for the whole sample .................................................................................................................. 135 6.6.2. Extended model for Austria .................................................................................................................. 143 6.7. Satisfaction with public services ........................................................................................................... 143 6.7.1. Model for the whole sample .................................................................................................................. 143 6.7.2. Extended model for Austria .................................................................................................................. 148 6.8. Satisfaction with green areas ................................................................................................................. 149 6.8.1. Model for the whole sample .................................................................................................................. 149 6.8.2. Extended model for Austria .................................................................................................................. 152 6.9. Summary of findings ............................................................................................................................. 153 6.9.1. Satisfaction with job.............................................................................................................................. 153 6.9.2. Satisfaction with income ....................................................................................................................... 153 6.9.3. Satisfaction with housing conditions..................................................................................................... 154 6.9.4. Satisfaction with social relationships .................................................................................................... 154 6.9.5. Satisfaction with health ......................................................................................................................... 155 6.9.6. Satisfaction with work-life balance ....................................................................................................... 155 6.9.7. Satisfaction with public services ........................................................................................................... 156 6.9.8. Satisfaction with green areas ................................................................................................................. 156 6.10. Satisfaction inequalities ........................................................................................................................ 157 6.11. Correlations between domains .............................................................................................................. 160 6.12. Correlation with overall satisfaction with life ....................................................................................... 163 7. The variables associated with life satisfaction among Hungarians living in Hungary and Austria....... 165 7.1.1. Model for the whole sample .................................................................................................................. 165 7.1.2. Extended model for Austria .................................................................................................................. 175 8. The drivers of emigration intentions in Hungary .................................................................................. 176 8.1. Microcensus vs the MIGWELL survey ........................................................................................................ 176 8.2. The role of economic, cultural and political factors, the assessment of the situation in the country, and social connectedness ....................................................................................................................................... 178 8.3. The differential effect of subjective well-being ........................................................................................... 180 8.4. Choosing Austria as destination ................................................................................................................... 181 9. How does migration affect subjective well-being in Austria? .............................................................. 183 9.1. Changes by subjective well-being dimensions ...................................................................................... 183 9.2. Changes in affective and eudaimonic variables .................................................................................... 192 9.3. Causal model ......................................................................................................................................... 203 10. Return migration potential .................................................................................................................... 205 11. Literature ............................................................................................................................................... 211 12. Appendix ............................................................................................................................................... 215 FWF–NKFIH Joint Project 4 MIGWELL at a glance The MIGWELL project focuses on the nexus of migration and well-being in Hungary and Austria. Using quantitative and qualitative research methods, it seeks to explore the impacts of migration on subjective well-being in the case of Hungarian immigrants in Austria as well as the effects of subjective well-being differences on emigration potential in Hungary. The approach of this project is innovative not only because it links the concepts of ‘well-being’ and ‘migration’, but also because it interprets their two-way causal relationship within one research framework. Since the Covid-19 pandemic might have a profound impact on both pillars, MIGWELL will also reflect on the rapidly changing socio-economic and well-being related issues that have emerged due to the epidemic throughout the life cycle of the project. The theoretical expansion of these concepts and the empirical findings of the project may contribute to more effective policies in both countries. FWF–NKFIH Joint Project 5 1. Introduction Our first project report provided an overview of the key definitions and some typologies of migration processes in general, followed by a literature-based review of the main theories on migration and well-being (Németh et al., 2022). It was followed by the presentation of the accumulated knowledge of the migration processes in and between Hungary and Austria (Németh et al., 2023a). The third report investigated the subjective well-being patterns in the two countries and the key factors that are responsible for the SWB differences (Németh et al. 2023b). The main goal of Work Package 3 was to implement the narrative and cognitive interviews, set up the surveys and analyse the newly collected primary data. While the first two qualitative methods functioned as exploratory, pre-test works, the surveys provided new information about the SWB profiles of the concrete target groups. Therefore, this fourth project report aims to provide an in-depth analysis of our primary data, integrated into the previously built MIGWELL database. The first report set out in detail the logical structure of the MIGWELL project (Figure 1). In the following, we will use this structure as a starting point and follow this logic to explore the connections between SWB and migration. However, it’s important to note that certain aspects of the comparative analysis will be given more emphasis than others. This is because we aim to focus on the key issues that, based on our understanding of the literature, extensive secondary analyses, and the available methodological tools, offer the most significant insights. Figure 1. Logical structure of the project. Source: Németh et. al., 2022: 43. FWF–NKFIH Joint Project 6 2. Sources and methods The MIGWELL project applies a mixed-methods approach including secondary data analyses and the combination of quantitative and qualitative methods: focus groups, interviews, and surveys in both countries. This chapter provides an overview of the diverse methodological toolkit we applied in Work Package 3. The empirical fieldwork consisted of successive steps of different types of data collection, with each step following the sequence listed below. Each stage was designed with a clearly defined objective and sought to broaden the horizon of the research, building always on the experience of the previous stage: secondary data analysis (WP2)  narrative interviews with actual and potential migrants or stayers  focus groups with stakeholders and experts  cognitive interviews with actual and potential migrants or stayers  surveys in Hungary and Austria  primary data analysis 2.1. Secondary sources As the previous project report has already highlighted, several potential data sources could have been suitable for the secondary analysis (Németh et al., 2023b: 7-10). Considering all advantages and disadvantages, we finally decided to use the microdata set from the Statistics on Income and Living Conditions (EU-SILC), a representative sample survey of private households in the European Union. Indicators related to subjective well-being were collected in the 2013, 2018 and 2022 ad-hoc modules. Our other important secondary source was the 2016 Hungarian Microcensus with a distinct subjective well-being module. As a result of our database building process, in Work Package 2 we created an integrated and weighted MIGWELL dataset. (About a detailed description of the process see Németh et al., 2023b: 11-15, 141-167). At the time the previous project report was written, the 2022 EU-SILC data were not yet available. Therefore, our first task in Work Package 3 was to update the FWF–NKFIH Joint Project 7 database with the new EU-SILC data for Hungary and Austria, which were published in late 2023. As a result, the dataset expanded to 133,557 observations (i.e., survey respondents) and 96 variables (their answers to selected questions). This updated dataset served as the foundation for integrating our primary survey data (Chapter 2.5). 2.2. Narrative interviews The secondary sources represent a top-down approach because they define the components of well-being from a particular scientific position. Such lists are integral parts of the multidimensional frameworks for comparative reasons (McGregor et al., 2015:3). However, in order to scrutinise the migration-SWB nexus from the people’s perspective, we had to understand the way in which potential migrants or stayers in Hungary and Hungarian immigrants in Austria construct their well-being, i.e., the areas of life to which they attach the greatest importance and the extent to which they are satisfied with achieved well-being outcomes. Moreover, we wanted to explore the motivations and their intentions of staying or moving to another country (including the option to return to Hungary) as well as their preand post-migration experiences through the prism of subjective well-being. In general, the narrative interview technique is based on the idea that oral life stories have an identity-giving function. It aims to capture the way in which people interpret, construct and narrate their life events (Riessman 1993; Rosenthal 2004). This method is especially useful in analysing migration stories and related biographies: understanding the way in which migration creates narrative identity and the kinds of discourses that are employed (Kovács & Melegh 2001). Therefore, 10-10 narrative interviews with immigrants in Austria and potential emigrants/stayers or return migrants in Hungary were conducted to gather information about people's preferences and values in permissive, non-threatening environments (Table 1-2). The typically two-hours-long narrative interviews followed the standardised guidelines, designed specifically for the three main types of our target population (Tóth et al., 2023). 1 Additionally, the semi-structured interviews provided the opportunity to go into more depth on certain topics where there was a need to explore more complex issues. The translated transcriptions proved particularly valuable for the subsequent research activities, first and foremost the focus groups with stakeholders, as they allowed us to address topics considered highly relevant to SWB. Indirectly, insights gained from the narrative interviews 1 The narrative interview outlines are available under the project website: https://www.oeaw.ac.at/fileadmin/Institute/ISR/pdf/Forschung/MIGWELL_WP3_Narrative_interview _outlines.pdf FWF–NKFIH Joint Project 8 facilitated the development of additional survey questions as well. For example, it has become evident that the questionnaires should reflect on overqualification or transnational household structures. In this project report, there is no separate deliverable dedicated to the qualitative analysis of the interviews. Instead, the forthcoming book by Göncz et al. (in print) will incorporate excerpts from these primary sources, addressing specific topics such as life domains and migration plans or experiences. These qualitative insights are invaluable, as they reveal the perspectives "behind" the statistical analysis. By drawing on real-life experiences, they contextualize the quantitative data and contribute to a deeper understanding and explanation of certain social phenomena. Gender Age Vocation / profession Address in Austria (NUTS2) Past or still present address in Hungary (NUTS2) 1 female 34 social worker Styria Western Transdanubia 2 male 54 catering (waiter) Upper Austria Northern Hungary 3 male 38 researcher Styria Southern Great Plain 4 male 55 house caretaker (hotel) Tyrol Southern Great Plain 5 female 36 childcare educator in school Lower Austria Western Transdanubia 6 male 53 house painter Tyrol Central Hungary 7 female 47 hotel, elderly care Tyrol Western Transdanubia 8 male 43 shop assistant, fast food worker Vienna Budapest 9 female 60 cleaning woman, babysitter Vienna Southern Transdanubia 10 female 50 kindergarten teacher, dance group leader Vienna Western Transdanubia Table 1. Basic data on the narrative interview partners in Austria. Source: own table Gender Age Vocation / profession Present status 1 female 50 economist potencial migrant 2 male 56 physician potencial migrant / stayer 3 female 45 teacher return migrant 4 female 46 sport marketing expert return migrant / potencial migrant 5 female ? teacher potencial migrant 6 female 32 philosopher - aesthete potencial migrant 7 female 62 ex-bank branch manager, present elderly carer commuter in fortnightly or monhly shift 8 male 65 painter, constraction contractor return migrant 9 male 45 painter return migrant 10 male 33 electrician potencial migrant Table 2. Basic data on the narrative interview partners in Hungary. Source: own table FWF–NKFIH Joint Project 15 2.6. Methods of quantitative data analysis 2.6.1. Models for satisfaction domains and satisfaction with life Satisfaction was interpreted as a scale variable, and therefore modelled using linear regression. This choice was based on the experience of the study by Ferrer-i Carbonell and Frijters (2004), which analysed the methodology of satisfaction modelling. Satisfaction dimensions and life satisfaction were first analysed on the full sample. This unweighted sample consisted of a total of 1688 (the two surveys in Hungary and the survey of Hungarians in Austria). The formulae of the models and a summary of their fit can be found in the Appendix. In general, the modelling process examined how individual variables (e.g., gender, education, number of friends) are related to satisfaction in the overall sample. However, we also introduced moderating effects in the models, using interactions. The moderating effect is interpreted in terms of the variable called „Hungarian group”. This variable can take three values: Hungarians who want to stay in Hungary, potential migrants from Hungary and Hungarians living in Austria. This was necessary because it is possible that some of the variables may be inversely related to satisfaction for example in the group of Hungarians who want to stay in Hungary and Hungarians in Austria. Separate models were also run for Hungarians in Austria for each dimension. In these models we included variables specific to this group, such as year of migration, reasons for migration, transnational activity. In these models we also included a moderating effect, the satisfaction before migration. It is possible that, for example, the gender variable is associated with satisfaction differently when the respondent was dissatisfied before migration than when she/he was satisfied before migration. The interpretation of the models was carried out following the guidance of Arel-Brundock and colleagues (2024), using the marginaleffects package. We typically begin our discussion of the relationship of satisfaction with each variable by presenting the model estimates. For example, for women and men, we calculate a mean satisfaction estimates with the associated confidence intervals. However, this is not the basis to find out the relationship between the variable and satisfaction. It is possible that the difference in the mean estimate is caused by one or more other variables. To establish the relationship, we use the average marginal effects and their corresponding confidence intervals. These are obtained in a counterfactual analysis. Let us stick to the gender variable and during this method two values are calculated for each respondent. In one case we convert everyone's gender value to female and in the other to male. Of course, all other variables are controlled, left as they were in the original responses. We then average the differences between the two values to get the average marginal effect. To establish the moderating effect, we examine these average marginal effects separately by Hungarian groups. In the interpretation, the level of significance is set at p<0.05, but in some cases relationships below p<0.1 are also indicated. FWF–NKFIH Joint Project 16 2.6.2. Causal model for determining the effect of migration on subjective well-being Causal analyses have become increasingly popular in the social sciences thanks to the proliferation of various causal inference techniques. To investigate the impact of migration on subjective well-being, the perfect method would be a randomised control trial, of which the only example so far in this field is the study by Stillman and colleagues (2015). Migration experiments raise ethical issues in addition to the obvious financial and material difficulties of implementation. In the absence of RCTs, researchers rely on observational data, longitudinal surveys if available, and cross-sectional surveys as a last resort. Studies looking for causal effects rely mainly on propensity score matching (PSM) in addition to fixed effects and instrumental variables approaches. One of the most important assumptions (exchangeability) for valid causal procedures is that the treated (migrated) and control (stay-at-home) samples are as similar as possible for certain variables. These are the so-called confounder variables, which in this case affect both the intention to migrate and subjective well-being. Such variables could include for example year of birth, gender or education. Pairing on the basis of time-independent variables is preferred, and the inclusion of a variable that reflects post-migration conditions introduces a bias. In this respect, education is an issue, as it could have changed after migration, but since most of the relevant literature includes this variable in the PSM procedure, we have also added it. In addition to the three variables described above (year of birth, gender, education), Erlinghagen (2021) includes retrospective variables that examine marital status and labour market situation before emigration (and matches it with their recently measured counterparts among the stayers). These variables are not available in our own survey, but a retrospective assessment of satisfaction with these objective indicators is. By including them, we argue that unmeasurable objective variables become measurable through these subjective domains. An important assumption is that there should be a relationship between objective variables and domain satisfactions, and therefore we conducted some tests on a sample of Hungarians living in Austria. We found significant relationships between current objective variables and current domain satisfaction, so there is reason to believe that these relationships existed in the past, between pre-migration objective factors and pre-migration subjective domain satisfaction. However, measurement error can also occur for these retrospective subjective variables, as we mentioned in the previous section on the drawbacks of retrospective questions 3 . Our formulation of the cause-and-effect relationship is illustrated in the following directed acyclic graph (Figure 2). 3 The problem of measurement error may also arise for retrospective objective variables FWF–NKFIH Joint Project 17 Figure 2. Causal pathways between “exposure” (green), outcome (blue), measured confounders (grey) and unmeasured confounders (white). In the case of the “treated” migrants some variables refer to their premigration measurements. The PSM procedure and the evaluation of the results were carried out in R using the packages MatchIt 4 , cobalt 5 and marginaleffects 6 . For this part of the analysis, we used data from the Hungarian representative and the Austrian snowball survey. We thus tried to match the „treated” individuals in the Austrian survey with control individuals from the Hungarian survey, who does not possess migration intension. The Austrian survey was restricted to those who emigrated after 2017 in the causal modelling, in order to reduce time-induced bias and achieve more robust results. In the PSM procedure, the „treatment” migration variable was modelled using a logit model, and then paired with the propensity scores of each individual. In the process, individuals who were very similar in terms of variables affecting both migration and subjective well-being were matched into subgroups. Matching was performed using the nearest-neighbours-method, proceeding in random order. Five controls were assigned to a treated person, and control persons could belong to more than one treated person at a time. Pairing could only be performed if the distance between matched individuals was below 0.37 7 (caliper). Persons excluded from joint support were removed from the analysis. The final data table contained 413 observations, of which 277 were unique observations. 4 Ho D, Imai K, King G, Stuart E (2011). “MatchIt: Nonparametric Preprocessing for Parametric Causal Inference.” Journal of Statistical Software, 42(8), 1–28. doi:10.18637/jss.v042.i08 5 https://cran.r-project.org/web/packages/cobalt/index.html 6 https://marginaleffects.com 7 0.2 of the standard deviation of the logit of the PS (Austin, 2011) FWF–NKFIH Joint Project 18 As Figure 3 shows, the balancing was successful, the variables were previously unbalanced in the original data table, but in the remaining data table the Standardized Mean Differences for all variables except satisfaction with job (0.1012) are all below the conservative limit of 0.1. Using the PSM method, with different matching rules, the average treatment effect (ATE), the average treatment effect in the treated (ATT) and the average treatment effect in the control (ATC) can be calculated. The majority of researchers report ATT in their studies, but due to the removal of observations because of the caliper and common support, only the average treatment effect in the remaining matched sample (ATM) can be calculated in the present case (Greifer, 2022). The ATM effect was modelled using g-computation with cluster-robust (subclass and id based) standard errors. As usual for causal analyses, a sensitivity test was carried out to estimate the risk from unmeasured confounders. For the test we used the sensemakr 8 package. Figure 3. Love plot of absolute standardized mean differences before and after matching. 2.7. Challenges and limitations Our method to gather information for the pre-migration well-being is based on the respondent's retrospective perception and compares the respondent's values in 2024 with his/her well-being in the year/weeks before emigration. The advantage of this method is that it is time-saving, inexpensive and simple to measure in questionnaire surveys, and does not require, for example, longitudinal follow-up. However, it has the disadvantages of the embellishing effect of nostalgia, the bias due to justification of migration (lower values may be given for the period before migration) and the fact that, depending on individual characteristics, positive 8 https://cran.r-project.org/web/packages/sensemakr/index.html Satisfaction with relationships Satisfaction with income Satisfaction with health Satisfaction with job Education: Tertiary Education: Secondary Education: Primary or less Gender Year of birth 0.0 0.2 0.4 0.6 0.8 Absolute Standardized Mean Differences Sample Unadjusted Adjusted FWF–NKFIH Joint Project 19 experiences in the past may be over-represented. For example, extroverts may remember more positive influences from the past (Barrett, 1997), while older respondents may describe earlier events in their lives more positively than they actually were (Comblain et al., 2005). The latter may be related to the general „rosy view” bias (Mitchell et al., 1997), which also calls for a more positive perception of past events. Conversely, negative experiences are more quickly forgotten and selective forgetting takes place (DePrince et al., 2012). A challenge to the reliability of retrospective data is the retrieval strategy used by the respondent to recall and judge past memories. Sometimes, even motivated respondents’ efforts do not allow for successful retrieval (Blair & Burton, 1987). In addition, the success of retrieval is also influenced by individual variables such as age or education (Berntsen & Rubin, 2002). However, the accuracy of retrieval in the case of migration is helped by the fact that this process can be considered as a kind of anchor point (Shmotkin, 2005), which occurs at a particularly significant stage in the life course. According to Oberauer and colleagues (2006), people remember events more accurately when they are rare, have serious consequences and are highly emotional and cognitively affective. Migration is certainly such an event. Further distortion may be caused by the fluctuation of subjective well-being prior to migration and the phenomenon that SWB typically declines in the year before emigration. Fredrickson (2000) and Kahnemann and colleagues (1993) also point out that people tend to judge certain past periods by their most intense point of discomfort or their end point. In this case, the retrospective assessment of subjective well-being before migration would be lower than the true one, since the respondent would judge only the last, most negative period. In summary, retrospective SWB assessment can be subject to both positive and negative biases and therefore the limitation of measurement error should be taken into account. Overall, however, retrospective evaluation, although often criticised in SWB studies, is a method frequently used in the absence of other data (e.g., Erlinghagen, 2021; Lundholm & Malmberg, 2006). Highlighting the importance of pre-migration experiences, Tosi and Impicciatore (2022) advocate the inclusion of retrospective questions in cross-sectional surveys. Although this method is far from being an accurate recollection of past experiences, it reflects these experiences well (Tov, 2012). Furthermore, the degree of happiness of an individual in the past and in the present is related. Bottan and Truglia (2011) were the first to describe this relationship, which they termed the „general habituation” channel. According to the model, a greater sense of happiness in the past increases the likelihood of a greater sense of happiness in the present. In addition to the potential biases posed by retrospective measurement, two limitations should be noted that compromise the validity of the causal analysis on the total Hungarian population living in Austria. One is due to the fact that only those Hungarians in Austria who had lived in Austria for at least one year could be included in the survey. Since we restricted the sample to them in the analysis, the phenomenon of selection bias arises. Another potential bias is nonresponse bias due to the convenience sample. On the one hand, due to the snowball sampling method, we did not have the same chance of reaching each potential respondent. For example, FWF–NKFIH Joint Project 20 those with no Hungarian acquaintances were almost certainly left out. Furthermore, many did not agree to participate in the survey because they were distrustful of the research. As we know that social relationships and trust are also related to life satisfaction, censoring on those could lead to biased estimates. FWF–NKFIH Joint Project 21 3. Migration plans and experiences among Hungarians living in Hungary and Austria 3.1. Migration potential in Hungary Migration potential measures the share of those with the intention to migrate in a society: an indicator of the share of those planning to leave their country for a shorter or longer period (Sik, 2003). It does not account for the actual act of migration but refers to plans, it cannot be used as a direct estimator for migration. Migration can be considered as an outcome of a decision-making process and grasped through its graduality, starting with the formulation of the intention, through the concrete steps to the realization of migration. Migration intentions represent the early phases of this process where plans might change and only parts of intentions will be realized. Analysing the early phases of this process allows to grasp the main dimensions of selection: the socio-demographic characteristics of these population and their differences in terms of objective and subjective well-being that seem to be an important driver of both intentions and migration itself (GödriFeleky, 2017). Studies linking migration intentions and the actual act of migration and looking at how much previous intentions were realized are scarce. Nevertheless, a previous study found that 17% of migration plans were fulfilled over a 3-year follow-up period (Gödri-Feleky, 2017), and that previous migration plans were statistically significant predictor of actual migration. As such, the extent of migration potential is largely dependent on the way it is measured: whether plans, intentions, or actual actions taken are considered. Migration potential is measured in Hungary from the early 90s by TÁRKI with questions about planning to go abroad to work for a few weeks, for a few months/years or indeterminately. These results show that the migration potential of the Hungarian population in terms of shortand long-term employment increased during the 2000s compared to the 1990s, it peaked in 2012 (19% expressed any kind of intentions), and it came down to 13% by 2016 (Figure 4). As for who would like to migrate, results from previous research show that it is higher among younger people, with more material and social resources who are also less satisfied, pessimistic, and feel discriminated (Sik & Simonovits, 2002; Sik & Örkény, 2003). FWF–NKFIH Joint Project 22 Figure 4. Migration potential in Hungary between 1993 and 2016 (%). In our project we measured migration intentions in a somewhat different way, following, and aiming to be comparable with the results of the 2016 Hungarian Microcensus. This Microcensus measured migration potential through the question: “Do you plan to move abroad in the next 2 years for work, study or other reasons?” to which respondents could answer “no” (86,5%), “I don’t know” (5,7%) or “yes” (6,3%). In order to nuance and better grasp the graduality of intentions to move, in our study we asked additional questions about whether one has ever thought about moving or whether one would like to live abroad. The migration potential in Hungary in 2016 was 6,3%, if we narrow the population to 64 years or younger, taking out older generations less likely to migrate, 8,4% of the population are planning to move abroad. This indicator is very much in line with what was measured in our representative survey in 2023: 8,3% (Figure 5). Migration intentions in a wider sense, i.e. migration inclinations or positive attitudes towards migration concern 23-25% of the population. In order to increase our sample size and be able to analyse people with migration intentions we relied on a boost sample as well complementing our representative sample. People recruited in this part of the survey have either ever thought about moving or expressed a like towards moving abroad. 62% of them plan to move within 2 years. FWF–NKFIH Joint Project 23 Figure 5. Migration intentions in Hungary in 2016 and 2023 (%). Source of data: Hungarian Microcensus 2016, MIGWELL surveys. Own figure. As far as the sociodemographic characteristics of those who intend to migrate previous studies have already confirmed that migration potential is higher among more educated younger men from urban areas (e.g., Sik, 2018). Our results indeed confirmed these findings – a detailed overview will be presented in the upcoming chapters. 3.2. The attitudes toward emigrants Attitudes towards immigrants has amply dealt with in previous research, however, attitudes towards emigrants or emigration in general might also be an indicator of the context, being supportive or hostile, and has an impact on plans to migrate. It has already by demonstrated that the support of the immediate relations of a person has an influence on migration plans, nevertheless, general attitudes of the overall population has not been considered so far, according to our knowledge. When asked to rate the seriousness of different social problems, those planning to migrate and those not are in understanding in terms of the order of these problems: inflation, corruption, poverty, and the state of the health care system are considered as the most important problems Hungary faces, followed by the state of the education system, while emigration and integration of immigrants are viewed as being second order problems. Those planning to migrate themselves, however, consider these problems as being more important than their counterparts not willing to move (Figure 6). FWF–NKFIH Joint Project 24 Figure 6. Perception of the seriousness of different social problems (mean values on a 0-10 scale, where 0 means not at all serious and 10 means very serious). Source of data: MIGWELL surveys. Own figure. Mirroring the research tradition that measures attitudes towards immigrants we formulated and tested a few statements regarding the perception of emigration and emigrants. Not surprisingly, those planning to migrate have a more positive attitude towards emigration, however, the general attitudes are rather positive as well. Nearly all with migration intentions agree (9899%), at least to some extent, that every person has the right to choose the country where he or she gets to thrive, that the right to move freely within the EU is a good thing, and that a stay abroad offers many new opportunities for children. The majority of the general public was also positive about these statements (85-88%). People were slightly less sure about whether moving abroad bring money and experience back home, but still very positive in this respect 75% of the general population and 91% of those planning to migrate agreeing with the statement. Negative symbolic messages that one might encounter in the media such as people emigrating would let the country down was rejected by the majority, 71% of not planning to migrate and 99% of those with migration intentions disagreeing with the statement (Figure 7). FWF–NKFIH Joint Project 31 Figure 14. The share of Hungarian citizens in the Austrian municipalities by 2011 and 2022. Source: Statistics Austria FWF–NKFIH Joint Project 32 Figure 15. The Hungarian-born population according to the regions of Austria, 2002-2024Source: Statistics Austria. Own figure. 3.5. Motivations of migration to Austria The leading reason for migrating to Austria is related to material well-being, with 57.5% of respondents citing financial motives (Figure 16). The second most popular answer is a more predictable future (39.4%), while the third is better working conditions (37.4%). Reasons that also exceeded 10% include education (25.2%), social relationships (19.4%), and helping their future back in Hungary (planned return) (11.6%). Political reason was mentioned by 8.1% of respondents. No statistically significant relationship at the p<0.1 level can be shown between the six most common reasons for migration and gender. The largest differences appear for financial reasons (69.2% among men, 53.3% among women) and better working conditions (42.1% among men, 32.9% among women). However, an individual’s current level of education shows a relationship with financial reasons (χ²: 12.43, p=0.006), reasons related to education (Fisher’s exact test, p<0.001), and reasons connected to helping their future back in Hungary (Fisher’s exact test, p=0.014). Financial motives are most common among those with vocational qualifications (72.5%), but also high among those with only primary education (68.8%) (compared to 57.9% for secondary school graduates and 47% for university degree holders). The same trend is characteristic of the motive of helping their future back in Hungary at a later stage. This reason is most common among those with primary education (25%) and vocational qualifications (18.8%). For education, however, the trend is reversed: higher values appear among secondary school graduates (35.5%) and university degree holders (28.2%). 12.2 13.2 8.6 2.1 2.3 3.4 19.4 18.7 15.4 8.6 8.7 15.4 3.7 3.6 7.2 8.4 10.9 11.6 3.3 4.4 7.4 1.8 1.6 3.0 40.4 36.4 25.9 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 Burgenland Carinthia Lower Austria Upper Austria Salzburg Styria Tyrol Vorarlberg Vienna FWF–NKFIH Joint Project 33 The presence of financial motives also shows a relationship with age (W = 9866, p = 0.017), as do better working conditions (W = 12878, p = 0.03) and reasons related to education (W = 3873, p < 0.001). When financial motives are present, the median age is 40, compared to 37 when they are absent. Better working conditions are likewise more typical among older respondents (median 39 years vs. 37 years when absent), whereas education-related reasons are characteristic of younger individuals (median 25 years vs. 42 years when absent). The education-related reason (W = 11494, p < 0.001) also shows a relationship with the year of migration, as does the relationship-related reason (W = 9337.5, p = 0.003). When educationrelated motives are present, migration is more recent (median 2018 vs. 2014 when absent); in contrast, relationship-driven migration is more typical of Hungarians who moved abroad earlier (median 2010 vs. 2015 when absent). Figure 16. Migration motivations of Hungarians residing in Austria. Source: Own figure. %0 %10 %20 %30 %40 %50 %60 %70 %80 %90 %100 FinancialPredictable future Work standards Education Relations Help future in Hungary Public service Political Safety ther Professional Reason for migration PERC Not marked Marked FWF–NKFIH Joint Project 34 4. The objective factors of well-being among Hungarians living in Hungary and Austria This chapter presents the general demographic and socio-economic data of our target groups – the Hungarians living in Hungary and Austria – based on the MIGWELL survey conducted in 2024. For Austria, we also show the national averages of the EU-SILC 2022 and, for reference, the data of the native-born population and the immigrant arriving form the “new” EU countries. The aim of this chapter is to collect the most important objective factors that may have an impact on subjective well-being. These objective factors will be compared across different subsamples of the database. The comparative analysis will cover the brief description of the following relations (Figure 17):  Differences between potential stayers and potential leavers in Hungary.  Differences compared to the Hungarian average.  Differences between Hungarians in Austria and Hungarians in Hungary.  Differences between the Hungarians and the native-born population in Austria, supplemented by the Austrian national averages and the scores for immigrants from the new EU countries. The in-depth analysis of Relation 1 was the main purpose of the previous MIGWELL Research Report (Németh et. al., 2023). About Relation 5 we will write in more details in Chapter 9. Figure 17. Overview chart of the comparative analysis based on the logical structure presented in Figure 1. Source: Own figure. FWF–NKFIH Joint Project 35 It is important to emphasise that the results of the Hungarian MIGWELL survey are representative of the total adult population in Hungary. However, for potential migrants, an additional survey was needed to achieve an adequate sample size. Therefore, the subgroup of potential migrants is not a representative sample. In Austria, the nationally aggregated data from EU-SILC 2022 are representative of the total adult population. For methodological reasons, statistical representativeness cannot be ensured after the breakdown by country of birth. For more details, see chapter 2.5. 4.1. Demographic and spatial characteristics In Hungary, the sex ratio of potential stayers is exactly the same as the national average (47% male, 53% female), but the opposite is true for potential emigrants, where 53.2% are male. Of all the subgroups studied, the highest proportion of men is found here. Since the gender breakdown and the age structure of the Hungarian-born population was available in the database of Statistics Austria, we could take this information into consideration during the survey sampling. The 54% female surplus of Hungarians in Austria is smaller than for immigrants from the new EU countries, but higher than the Austrian average and also compared to the native-born population (51-51% in both cases) (Figure 18). Figure 18. The share of people by gender in Hungary and Austria by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own figure. Based on our survey, Hungarians in Austria have a younger age composition in general than the other subgroups in the EU-SILC 2022 database (Figure 19). The average age of the respondents in the MIGWELL survey was 38 years, compared to the average age in Austria of 49 years. Figure 19 shows only one younger subgroup: those planning to emigrate from FWF–NKFIH Joint Project 36 Hungary are on average 35 years old, which is significantly lower than both those planning to stay (49 years) and the national average (48 years). This is not surprising given that international migration participants are typically young adults in their 20s and 30s (Fassmann et al., 2017; Tálas, 2020: 73). Figure 19. Age distribution of people in Hungary and Austria by analytical subgroups. Source of data: EUSILC 2022, MIGWELL surveys 2024. Own figure. The MIGWELL survey in Hungary was representative of the geographical distribution too; however, spatial representativeness is not true in the case of potential migrants after the additional data collection phase. In Hungary the proportion of potential stayers is almost exactly the same as the national proportion. Respondents living in Pest County and the Southern Greater Plain region are slightly over-represented among the potential emigrants. In Austria, the EU-SILC data are representative of the total adult in general but statistical representativeness cannot be ensured on subnational level, particularly after the breakdown by country of birth. In comparison with the spatial distribution of the Hungarian population in Austria (Chapter 3.4), Vienna and Styria are slightly overrepresented in the MIGWELL sample, while other regions are slightly underrepresented. (However, the difference is typically smaller than one percentage point) (Figure 20). FWF–NKFIH Joint Project 37 Figure 20. The share of people by NUTS2 regions in Hungary and Austria by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own figure. In terms of the degree of urbanisation, there is a significant difference between potential migrants and those who intend to stay. 37% of those who plan to leave Hungary would migrate from low and 33% from high population density settlements, i.e., typically from villages or large cities. These proportions are only around 30% and 17% for those intending to stay. A fifth of Hungarians in Austria did not give a specific settlement name, so no data are available for this variable. However, it can still be concluded that they are basically urban dwellers, similar to immigrants from the new EU countries. While EU-SILC data show that 45% of the nativeborn population live in low-density settlements, only 19% of Hungarian respondents do (Figure 21). From the above we can also conclude that a significant proportion of Hungarian migrants not only cross an international border but also start a new life in a big city, coming from a rural environment. This change in lifestyle alone can have an impact on subjective well-being. Figure 21. The share of people by the degree of urbanisation of their place of residence in Hungary and Austria by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 38 4.2. Financial situation In order to outline the context, we need to draw attention to changes in the national averages of certain macro-level indicators. According to the 2013 EU-SILC survey, the personal incomes in Austria were three times higher on average than in Hungary (ca. 27,300 EUR vs. 7,600 EUR). Regarding the median values, the difference was similar: ca. 22,000 EUR in Austria and 6,600 EUR in Hungary. This income gap has narrowed in the 2018 and 2022 EU-SILC wave; the mean and median values then displayed “only” a 2.3-fold difference between the two countries (ca. 27,600 and 22,800 EUR vs. 12,200 and 10,600 EUR) (Table 5). In terms of equivalised disposable household income, one can observe similar trends: the average and median values of the income level showed a 2.3-fold and 2.4-fold difference in favour of Austria, respectively. Austria Hungary 2013 2018 2022 2013 2018 2022 Mean 27,326 28,603 27,599 7,627 9,827 12,216 Median 22,070 23,671 22,780 6,600 8,520 10,644 Table 5. Average total personal income in Austria and Hungary (yearly, gross, EUR/per capita), considering all types of incomes and benefits, adjusted for PPP and inflation, and standardized to 2022 local currency units using the corresponding official HICP indexes. Source of data: Eurostat microdata. Own table. The at-risk-of-poverty rate is a relative measure indicating the proportion of individuals whose equivalised disposable income (after social transfers) falls below the at-risk-of-poverty threshold, defined as 60% of the national median equivalised disposable income after social transfers. This methodological approach explains why the proportion of people living below the at-risk-of-poverty threshold is similar in the two countries. In Austria, approximately 13% of the population falls into this category, compared to 12% in Hungary, with these rates remaining remarkably stable over time. In 2013, the share of materially deprived individuals was relatively low in Austria, at just 4%, and dropped further to 2% by 2022. In contrast, over a quarter of Hungary’s population (26%) faced material deprivation in 2013. The subsequent decline to 9% highlights a significant improvement in living standards in Hungary. Low work intensity, which measures the proportion of individuals aged 0–59 living in households where adults worked less than 20% of their total potential working time in the previous year, remained relatively stable in Austria at around 5%. In Hungary, this indicator fell significantly from approximately 9% in 2013 to 4% in 2018 but experienced a slight uptick by 2022. FWF–NKFIH Joint Project 39 Finally, we examined the proportion of individuals for whom debt repayment – excluding mortgage or main dwelling-related loans – constitutes a financial burden. In 2022, those without such debts represented around 79% of the population in Austria and 80% in Hungary. However, debt repayment challenges persist for about 3% of Austrians and 5% of Hungarians. Based on the experience of the cognitive interviews, it has become clear to us that when asking about personal and household income, it can be problematic to get the specific amount. The sensitivity of the topic is only one of the main reasons. It is a difficult task for respondents to add up the income of different members of the household, especially if some members are paid in HUF and others in EUR. This took an unrealistically long time during the cognitive interviews and clearly increased the risk of non-response. We therefore decided to treat the data on income as a categorical variable in both countries. (For this reason, of course, we had to reclassify the EU-SILC variables too.) Even so, 17.5% and 35.1% of respondents in Hungary did not answer our questions on personal and household income respectively. For Hungarians in Austria, the refusal rates were much lower, at 8.7% and 4.8% respectively. The income distributions of potential emigrants in Hungary are basically the same as the national average (Figure 22). However, the (non-representative) sample of potential emigrants shows some interesting differences. If we exclude those who refused to answer, the share of people with a net income of HUF 200,000 or less is much lower among potential emigrants (18.8%) than among those who want to stay (33.5%) and the Hungarian national average (34.9%). Medium income earners (201-500 thousand HUF) account for 70.3% of those planning to emigrate, compared to 58% for potential stayers and 55% for the national representative sample. At the same time, the share of high-income earners (above HUF 500,000) varies within a very small range (between 4.1% and 4.8%). In summary, it can be said that it is not the poorest and not the wealthiest who typically want to emigrate from Hungary, which is consistent with the general "migration hump" theory (Zelinsky, 1971). FWF–NKFIH Joint Project 40 Figure 22. Personal incomes in Hungary by analytical subgroups (monthly, net, HUF). Source of data: MIGWELL survey. Own figure. Household data also show a similar pattern. While one in three households with no intention to migrate is low-income (33.6%, maximum 400,000 HUF), this proportion among potential migrants is only 16%. The proportion of households with an income above 1 million HUF is very low in both subgroups (3-5%). 4 out of 5 households planning to emigrate can be considered as medium income, i.e., between 400,000 and 1 million HUF (Figure 23). Figure 23. Household incomes in Hungary by analytical subgroups (monthly, net, HUF). Source of data: MIGWELL survey. Own figure. HUNGARY AUSTRIA 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Personal income Does not have income now 100 thousand Ft under 101 – 200 thousand Ft 201 – 300 thousand Ft 301 – 400 thousand Ft 401 – 500 thousand Ft 501 – 750 thousand Ft 751 – 1 million Ft 1 million – 2 million Ft 2 million – 4 million Ft Unknown HUNGARY AUSTRIA 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Personal income Does not have income now 100 thousand Ft under 101 – 200 thousand Ft 201 – 300 thousand Ft 301 – 400 thousand Ft 401 – 500 thousand Ft 501 – 750 thousand Ft 751 – 1 million Ft 1 million – 2 million Ft 2 million – 4 million Ft Unknown FWF–NKFIH Joint Project 47 Figure 28. The share of people by the highest level of education by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. In Hungary, the distribution of those planning to emigrate by economic status is radically different from the distribution of the representative national sample (Figure 29). 84.5% of potential emigrants are employees, compared to 54.9% in the national sample and 57.1% for potential stayers. There are very few retired people among them and the self-employed rate is also lower than the national average. In contrast, the proportion of unemployed persons and students is higher. In Austria, the self-employed category is missing in the EU-SILC database. Nevertheless, it can be seen that the share of employed persons is roughly the same as in Hungary, but with a relatively higher share of students and pensioners. For immigrants from the new EU Member States, the size of the economically inactive population is much smaller than the Austrian average (27% and 35% respectively). In the non-representative Hungarian sample obtained by the Austrian MIGWELL survey, students were over-represented, accounting for 13% of the respondents compared to the national average of 6%. This is to some extent a consequence of snowball sampling. The high proportion of employees is not surprising, however, given the generally young age composition of Hungarians in Austria, many of whom are target earners (see the MIGWELL Research Report Nr. 2.1: Németh et al., 2022). FWF–NKFIH Joint Project 48 Figure 29. The share of people by current economic status by analytical subgroups. Source of data: EUSILC 2022, MIGWELL surveys. Own figure. There are no significant differences between the Hungarian subgroups compared to the national sample. However, the proportion of both the highest prestige occupations (managers, professionals, technicians and associate professionals: 16.7% in total compared to 19.6% in the national sample) and low prestige occupations requiring no qualifications (6% compared to 13.5% in the national sample) is slightly lower among those planning to emigrate. The other ISCO categories – with the exception of skilled agricultural, forestry and fishery workers – are over-represented among potential emigrants. In particular, clerical support workers and service and sales workers have a strong migration intention: they account for 43% of all potential emigrants. The share of people having the highest-prestige occupations (managers, professionals, technicians and associate professionals) is the highest among the Austrian-born analytical subgroup (43.7% after excluding the non-respondents) (Figure 30). For immigrants from the new EU Member States and the Hungarians it is 26.6% and 31.5% after excluding the nonrespondents. This is much higher than what Horváth found (2022: 120) on the basis of the 2016 Hungarian Microcensus (17%). By contrast, the MIGWELL survey slightly underestimated the categories of "craft and related trades workers" and "plant and machine operators, and assemblers" (15.4% compared to 23% in the Microcensus), and this is particularly true for elementary occupations (9% compared to 20% in the Microcensus). The largest group size is service and sales workers: about one third of respondents belong to this category, compared to only one in five nationally. This figure is consistent with Horváth's findings, who calculated that 31% of Hungarians in Austria could be classified in this category in 2016. In summary, the FWF–NKFIH Joint Project 49 non-representative MIGWELL survey in Austria presumably reached a higher proportion of higher-skilled Hungarians than lower-skilled ones. Figure 30. The share of people by the International Standard Classification of Occupations (ISCO 08) by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. A comparison of the distributions by educational attainment and current occupational structures reveals a significant presence of overeducated Hungarians in Austria – a phenomenon that is not new. Drawing on data from the 2016 Hungarian Microcensus, Horváth (2022, 118) found that approximately one-third of Hungarians in the Austrian labour market were employed in positions below their level of education. This issue was particularly acute among mothers with children left behind in Hungary, 85% of whom were working in roles that did not match their qualifications (Horváth 2022, 136–137). Similar findings emerged from the 2024 MIGWELL survey: excluding respondents who refused to answer, 30% of Hungarians in Austria identified themselves as overqualified for their jobs (Figure 31). %0 %10 %20 %30 %40 %50 %60 %70 %80 %90 %100 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup ISC 08 Managers Professionals Technicians and Associate Professionals Clerical Support Workers Service and Sales Workers Skilled Agricultural, Forestry and Fishery Workers Craft and Related Trades Workers Plant and Machine perators, and Assemblers Elementary ccupations Armed Forces ccupations Currently I don t have job, and I did not have previously. Unknown FWF–NKFIH Joint Project 50 Figure 31. Overqualification. (“Compared to your level of education (primary, secondary or tertiary), what kind of work are you doing in Austria?”). Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. Figure 32. Number of working hours per week by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. There is little difference between subgroups in terms of the duration of the employment contract (Figure 33). The proportion of Hungarian respondents in Austria with a fixed-term employment contract is slightly higher (14%) than in Hungary (7-8%). About 5% of respondents stated that they are currently working without an official work contract. In reality, this is probably higher, but on the one hand not everyone dares to admit this in survey, and on the other hand, as seen above, the MIGWELL survey could reach relatively few people with elementary occupations. HUNGARY AUSTRIA 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Qualification compared to current job Adequate for education level Below education level Above education level Unknown FWF–NKFIH Joint Project 51 (According to the empirical experience, this situation is most common in agriculture or construction). Figure 33. Duration of employment contract by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. Excluding non-respondents, there is relatively little difference between sub-groups in Hungary regarding their perceived opportunities to achieve a higher position at work (Figure 34). Among potential migrants, 29.5% believe that advancing in their job is at least somewhat realistic, compared to 35% of potential stayers. In contrast, the figure is significantly higher among Hungarians in Austria, with 56.6% expressing a similar outlook. Figure 34. Opportunity to move up to a higher position or to have more autonomy at the workplace by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. HUNGARY AUSTRIA 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Type of contract Fixed term, less than 6 months Fixed term, more than 6 months Permanent No formal contract Unknown HUNGARY AUSTRIA 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Opportunity for higher position Not at all Rather no Rather yes Yes, absolutely Unknown FWF–NKFIH Joint Project 52 4.4. Housing conditions In the ECD How’s Life framework, housing conditions were captured by three headline indicators: the number of rooms per person, housing costs, and the number of dwellings lacking basic facilities (OECD 2013). Since the lack of basic facilities does not pose a widespread problem in Hungarian and Austrian households, we focus on the reported housing problems which directly affect people’s satisfaction with their housing situation. As the previous MIGWELL Research Report pointed out (Németh et al., 2023: 40-41), respondents in Hungary more frequently reported different types of internal housing problems. According to the EU-SILC, leaking roof, rot in window frames and floor, damp walls/floors or foundation (22% vs. 10%), and dark rooms (8.5% vs. 5.5%) were almost twice as frequent in Hungary as in Austria. On the other hand, the external negative factors, such as noise from the streets, crime and vandalism in the area, pollution, filth, and other environmental problems seem to be stronger in Austria (Figure 35). Considering the analytical subgroups of this Research Report, the housing problems are rated slightly worse by potential emigrants in all the aspects examined. In particular, there is a significant difference in the level of noise from neighbours and from the street. However, this can be explained by the fact that the non-representative sample of potential emigrants includes a higher proportion of people living in densely populated settlements (see Chapter 4.1) and are therefore more likely to encounter this problem. It is also interesting to note that, compared to the Hungarian population, some problems are less frequent in the neighbourhoods where Hungarian households in Austria live (e.g., pollution, filth, or other environmental problems), while others are more frequent (e.g., dark rooms, noise from neighbours or from the street). Regarding crime, violence or vandalism, there are no differences between the perceptions of the analysed subgroups; only 4-5% reported that this is a problem in the area they live. FWF–NKFIH Joint Project 53 Leaking roof, damp walls/floors/foundation, or rot in window frames or floor: Too dark, not enough light: Noise from neighbours or from the street: Pollution, filth, or other environmental problems: Crime, violence or vandalism in the area: Figure 35. Percentage of people living in accommodation with the following housing problem by analytical subgroups. Source of data: MIGWELL surveys. Own figure. FWF–NKFIH Joint Project 54 Figure 36. Total number of rooms in the dwelling the household lives by analytical subgroups (With the living room but without the kitchen and bathroom. The half-room counts as a room, and the "American kitchen" - living room and kitchen function combined - also counts as a room). Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. While the representative sample in Hungary found that 86% of people live in own properties (or live with a close relative of the owner, i.e., partner, spouse or child), only 24% of Hungarians in Austria do so (Figure 37). This is not a surprising discrepancy, since, as the expert interviews showed, property prices are much higher in Austria and it is not a realistic objective for most immigrants to accumulate the necessary wealth. Also not surprisingly, those who do not plan to leave Hungary are slightly more likely to be homeowners than potential emigrants (87% and 81% respectively). On the one hand, buying a house or flat is one of the main drivers for Hungarians wanting to work in Austria. On the other hand, in general, people who rent are characterised by greater flexibility and more active spatial mobility. The relative majority of Hungarians in Austria rent the whole dwelling (42%), but one in five of those interviewed rent only part of the dwelling. People living in commercial accommodation are typically those who work in a hotel and receive accommodation in that building from their employer (2% in the sample). FWF–NKFIH Joint Project 55 Figure 37. Legal relationship to the property by analytical subgroups. (“I am the owner of the dwelling, or its spouse, partner or relative.). Source of data: MIGWELL surveys. Own figure. The previous three sub-chapters dealt with the material components of people's well-being. The figure below can in some ways be seen as a summary of these, but rather from the perspective of the respondents, as it was intended to operationalise their perceived social position. We asked the respondents to „imagine a ladder with the upper level representing the upper part of society and the lower level representing the lower part. Where would you currently place yourself on the ladder in Hungary / Austria"? In Hungary, compared to the national average, the perceived social position of potential stayers was slightly higher (5.4) and that of potential emigrants slightly lower (4.8). This is interesting because the previous chapters have repeatedly shown that the objective financial position of those planning to emigrate is generally not worse than that of those who imagine their future in Hungary. In fact, for some income or housing indicators we even see better values for potential emigrants. Hungarians in Austria placed themselves on average at 6.6 on this imaginary 0-10 social ladder (Figure 38). Here, we specifically asked them to think about their current position within the Austrian society. (Their previous perceived position before the move, on average, was 6.1 in Hungary). This high score is surprising, because based on the literature we expected that after migration a significant proportion of people would be in a lower social status in their new home country, and this would be reflected in their perceived position as well. Further research is needed to explore the reasons, but their transnational living situation may also have an impact. Many of the Hungarians in Austria are more or less transnational migrants who spend a lot of time in Hungary and very often maintain two addresses or households there. It is possible that this phenomenon obscures – and in a sense makes unnecessary – a realistic assessment of their position within the Austrian society, and that unconsciously people perceive and interpret their FWF–NKFIH Joint Project 56 improved living standards in the context of the Hungarian society they also belong to. (For more on migrant transnationalism, see chapter 4.10). Figure 38. Perceived social position by analytical subgroups. (“Imagine a ladder with the upper level representing the upper part of society and the lower level representing the lower part. Where would you currently place yourself on the ladder in Hungary / Austria?”). Source of data: MIGWELL surveys. Own figure. 4.5. Household composition The theory called “new economics and sociology of labour migration” (NELM) underlines that migration decisions are usually made by families and households, rather than by isolated individuals (Németh et al., 2023: 34-35). It was important to ask survey respondents to provide information on the size and composition of their households. In the case of the Hungarian sample, this was done without any problems, but in Austria, there were several complicating factors. We have therefore included a brief clarification (see below) of what the household concept is in general and how it should be interpreted in practice. „What is a household? A household is made up of persons who live together in a dwelling or part of a dwelling and share at least part of the household's expenses (e.g. food, daily expenses). Dependent persons are also members of the household. Many Hungarians living in Austria are members of two households: one in Hungary (often with other members of their family living there) and one in Austria. If this applies to you, please answer the question below accordingly! Note: If you live with tenants and only pay rent together, this is a one-person household. It is also a one-person household if you live alone, and get accommodation from your employer e.g. in a hotel.” FWF–NKFIH Joint Project 63 Figure 44. Distribution of trust in other people: on a 0-10 scale. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. Figure 45. Distribution of trust in people in one’s neighbourhood: on a 0-10 scale. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. In line with having someone to discuss personal matters, despite the differences in the frequency of contact with friends and family, a similarly high proportion of respondents in Austria (84%) and Hungary (89%) report that they can ask for help from relatives, friends, or neighbours, which is even higher among potential migrants in Hungary (94%) and Hungarians living in Austria (91%) (Figure 46). Looking at material and non-material help separately it seems that asking for financial help is a more sensitive issue: a somewhat lower share of respondents would be able to ask for it. In the case of Hungarian migrants living in Austria, nevertheless, asking 0.0 5.2 0.5 5.7 0.10 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 0.0 5.2 0.5 5.7 0.10 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Migration intention Trust in people in your neighbourhood FWF–NKFIH Joint Project 64 for non-material help would be the issue as only 58% could ask such thing from relatives, friends, or neighbours (Table 7). It is indeed true for this population that they have their support network both in Hungary and in Austria, and this seems to pose a problem when in need of nonmaterial help, probably necessitating the helper being present. Figure 46. Percentage of people being able to ask for help from relatives, friends, or neighbours. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 HU in AT 2024 Material help 78.8 79.5 87.1 84.5 Non-material help 85.7 86.6 92.3 58.1 Any kind of help 89.0 89.3 94.0 91.3 Table 7. Percentage of people being able to ask for material or non-material help from relatives, friends, or neighbours. Source of data: EU-SILC 2022, MIGWELL surveys. Own calculation. Beyond people's personal connections, their eventual memberships in associations, creating some kind of embeddedness and network, also affects their quality of life and well-being. In this respect the share of respondents not being a member of any association is an interesting indicator: 75% of Hungarians do not have any kind of such membership, while this lack of engagement is even higher among those intending to migrate (82%), while it is lower among Hungarians living in Austria (68%) (Table 8). %0 %10 %20 %30 %40 %50 %60 %70 %80 %90 %100 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Help No es Unknown FWF–NKFIH Joint Project 65 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 HU in AT 2024 Social, charitable organization 3.3 3.1 0.9 Church, religious organization 6.6 6.4 4.3 Party, political organization 1.5 1.4 1.3 Trade union, employee representation 8.9 9.4 9.0 Sport, recreation or other cultural association 9.7 8.3 7.3 Cultural organisation 4.8 Other voluntary organisation 0.8 0.6 0.4 24.8 No membership in any kind of organisation 75.4 76.9 82.0 68.1 Table 8. Membership in voluntary organisations (%). Source of data: EU-SILC 2022, MIGWELL surveys. Own calculation. 4.7. Health status One of the most significant factors influencing well-being is health, which also has an impact on how well people are able to engage in daily activities that enhance their well-being. Here, we will rely on respondents' self-assessment of their health status and whether they feel that their activities are limited due to health issues in the absence of objective measures of their health condition (information about diseases and conditions causing poor health or disability). By its very nature, self-perceived health measurement is subjective. It is supposed to cover the various aspects of health, such as mental and physical, and focuses on the overall condition of health rather than the specific circumstance. As already presented in the previous (WP2) report based on 2013 and 2018 EU-SILC data, perceived health is better in Austria than in Hungary. Around 70% of Austrians evaluated their health condition as very good or good, while only about 60% of Hungarians declared the same. Our more recent data confirm these tendencies with 62.6% of the Hungarian sample (in 2024) and 69.8% of the Austrian sample (in 2022) declaring to be in a very good or good health condition. Hungarians who are planning to migrate, however, perceive their health condition much better than the rest (90.6% in a very good or good shape) and 98.3% at least in a fair condition (Figure 47). This seems to be in line with previous empirical results that migrants, and those with migration plans are in better health condition than others also due to the fact that overall, they are younger, or, in a reversed causal logic, people with better health are those that are prone to migrate or consider it. Interestingly, however, this difference is slightly less notable in the case of Hungarian people who have already migrated to Austria. Nevertheless, their perception of their own health condition remains more positive (79.1%) than either the native FWF–NKFIH Joint Project 66 Austrian population (70.4%) or immigrants from the new EU Member States living in Austria (65.1%) as measured in the 2022 EU-SILC. Figure 47. Perceived health. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. According to previous, 2013 and 2018 EU-SILC data no notable difference existed between the two countries in terms of the share of people declaring a severe limitation in their activities due to their health condition (around 7-10%), while the share of people declaring to be somewhat limited was slightly less important in Hungary (18-19%) than in Austria (25%). More recent data confirm these tendencies with one third of Hungarians (33.3% in 2024) declaring to be strongly or otherwise restricted due to their health conditions and 28.8% of Austrians in 2022) (Figure 48). In line with previous results, the share of Hungarians with health restrictions among those planning to migrate is much lower, only 7.7%. On the other hand, those Hungarians that have migrated to Austria are rather similar in this respect (26,7% declaring to be severely or otherwise restricted) to either their Austrian counterparts or other immigrants from new EU Member States (28.5%) with slightly less of them being severely restricted (3.5%). It has to be noted, however, that the differences in the perception of limitations due to health condition might be partly shaped by different cultural contexts, regulations, and institutional and policy approaches in the two countries to the diagnostics and formal acknowledgement of such conditions. %0 %10 %20 %30 %40 %50 %60 %70 %80 %90 %100 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Health Very bad Bad Fair Good Very good Unknown FWF–NKFIH Joint Project 67 Figure 48. Percentage of people declaring to be strongly or otherwise limited in their everyday activities due to their health condition. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. 4.8. Work-life balance Another aspect of the quality of life is work-life balance, in terms of the time one can devote to family life and leisure activities, and how much free time one has beyond working and commuting. In our research it is measured through the working hours, participation in leisure activities and being able to meet with friends. The landscape of working hours changed somewhat since 2018 in Austria where part-time work became more widespread, while the main tendencies remained similar in Hungary with a dominance of full-time jobs. According to 2022 EU-SILC data the average working hours in Austria is 20.3 while it is 41.1 in Hungary in 2024 according to our survey with potential migrants working slightly more, 42 hours/ week on average. In Austria, indeed, Hungarians work less hours than their non-migrant counterparts, 35.2 hours, which, nevertheless, remain above the Austrian average, including other immigrants from new EU member states (20.1 hours) (Figure 49). Beyond the differences in terms of working hours there is another, still persisting difference between the two countries in terms of the share of respondents regularly participating in a leisure activity (Figure 50). While 71,7% (in 2022) of respondents in Austria engage in such activities, only 37.6% of Hungarian respondents reported engaging in leisure activities (in 2024). As these activities, such as sport, cinema, or concerts, could entail certain costs (e.g., entrance fees and/or travel costs), this can be an obstacle for some: indeed, a higher share of Hungarian respondents does not participate in such activities because they cannot afford it (18% as opposed to 8% in Austria). Nevertheless, the share of those who do not have regular leisure %0 %10 %20 %30 %40 %50 %60 %70 %80 %90 %100 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Physical restriction No es Strongly Unknown FWF–NKFIH Joint Project 68 activities for other reasons remains significantly higher in Hungary (43.8%) than in Austria (20%). These results seem to be stable over time. However, Hungarians with migration intention are also people who have more leisure activities: 50.6% of them regularly participate in leisure activities, while the share of those who are deprived because of cost issues are similar to the Hungarian population (17.6%). An even higher share of Hungarians living in Austria engage in such activities (65.5%) although their share remains below the native Austrians. This is especially due to the fact that they can now afford it: only 6.1% said that costs would prevent them to do so. Figure 49. Total number of working hours per week. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. Figure 50. Percentage of people regularly participating in a leisure activity. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. 0 25 50 75 100 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 %0 %10 %20 %30 %40 %50 %60 %70 %80 %90 %100 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Leisure, detailed No other reason No cannot afford it es Unknown FWF–NKFIH Joint Project 69 4.9. External factors The quality of the living, working and economic environment directly contributes to quality of life, and their subjective perception is related to subjective measures of well-being. In the following the perception of Hungary’s overall situation and economic context, expectations for the future and trust in institutions will be explored. The perception of the economic situation in Hungary is by far the worst among those planning to migrate (Figure 51). Whereas 15.6% of the general population think the situation being very bad, 23.6% of potential migrants share the opinion, meaning that negative opinions are not just characterizing half of the population in question but two third of them. Future expectations regarding the country’s economic performance show similar tendencies, potential migrants are the least optimistic: only 5.2% of the latter think that the situation will slightly improve as opposed to 13% among the general population (Figure 52). Figure 51. Perception of the current economic situation of Hungary (%). Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. %0 %10 %20 %30 %40 %50 %60 %70 %80 %90 %100 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Current economical situation of the country very bad bad neither good nor bad good very good Unknown FWF–NKFIH Joint Project 70 Figure 52. Expectations for the future economic situation of Hungary (%). Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. The perception of the current situation in Hungary follows similar tendencies: when rated on a 0-10 scale, potential migrants are the least positive with an average of 2,6 as opposed to 3.9 among the general population (Figure 53). Figure 53. Perception of the current overall situation of Hungary: values on a 0-10 scale. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. 0.0 5.2 0.5 5.7 0.10 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Migration intention Current situation of the country HUNGARY AUSTRIA 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Economical situation of the country in the future much worse slightly worse the same slightly better much better Unknown FWF–NKFIH Joint Project 71 Trust in institutions is another indicator of the perception of a country’s performance and the overall context. Overall, the most trusted institution in Hungary is the police followed by the legal system in general, the Government, the political system and the Parliament come next while the trust in the media is the lowest. Potential migrants follow similar tendencies, however, at a lower level of trust. The order of trust in Hungarian institutions among Hungarians living in Austria is somewhat different, although the level of distrust is similar to potential migrants. They mostly trust the legal system, while least trust the Government. Trust in Austrian institution, however, is significantly higher in all aspect with the legal system and the media being the mostly trusted (Table 9). Trust in the... General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 HU in AT 2024 (Hungarian institutions) HU in AT 2024 (Austrian institutions) … police? 5.5 5.7 4.5 3.6 6.1 … legal system? 4.9 5.1 4.2 5.2 7.6 … Government? 3.9 4.0 3.0 3.1 5.5 … political system? 3.8 3.9 3.1 3.5 6.4 … Parliament? 3.8 3.9 3.1 … media? 3.6 3.7 3.1 4.6 7.6 Table 9. Trust in institutions: mean values on a 0-10 scale. Source of data: EU-SILC 2022, MIGWELL surveys. Own calculation. Compared to the results of the 2016 Microcensus, where 69% of the respondents declared to feel fairly or very safe the situation has improved, according to our survey, 83.6% of Hungarians felt fairly or very (35.6%) safe when walking alone in their area after dark (Figure 54). The share of those feeling very safe (35.6%) is even higher among potential migrants (41.2%) and Hungarians living in Austria (42.3%). However, perception of personal security is more polarized among the latters: feeling a bit unsafe in Austria is at a similar level (10.3%) than among the general Hungarian population (13.6%), both higher than among those planning to migrate (3.9%). FWF–NKFIH Joint Project 72 Figure 54. Perceived personal safety. Source of data: EU-SILC 2022, MIGWELL surveys. Own figure. 4.10. Transnational activity According to Kim and colleagues (2021), transnationalism can be broken down into several dimensions: social ties, transnational family status, cultural ties, economic ties, political ties, transnational attitude and identity, and finally transnational healthcare use. In the MIGWELL project, we mainly focused on social ties, transnational family status and economic ties (Figure 55). Although the EU regulations require the establishment of a residence in only one country at a time, 62.58% of the Hungarians in the sample a Hungarian address. They do so with only 20.65% of the respondents having a household in Hungary. 45.16% of the total sample can therefore be said to have retained their Hungarian address even without having a household in Hungary. It can also be seen that Hungarians in Austria have preserved their social contacts with those who have remained in the home country. Only 15.16% of the sample have no friends in Hungary (and 6.45% did not answered). For those who have no friends in Hungary, the median emigration year is 2010, and this increases with the number of friends (with at least 10 friends the median is 2019). The descriptive data also show some relationship between the number of friends and geographical distance, as only 38.29% of those with no friends live in the three provinces closest to the Hungarian border (Upper Austria, Vienna, Burgenland), while 80% of those with at least 10 friends live in these provinces. Most Hungarians in Austria have 2-4 friends in Hungary (35.48%). 66.77% of respondents can count on material support from Hungary, while 49.67% can rely on non-material support. In terms of visits to Hungary, 1-2 trips home per year are the most common (32.58%), but returning home two or three times a year and every month is also common (20.96% for both). The two extreme response options, returning home several times a month (11.93%) and not visiting home at all (12.9%), have almost the same proportion of answers. %0 %10 %20 %30 %40 %50 %60 %70 %80 %90 %100 General HU 2024 Pot. stayers Representative 2024 Pot. migrants Boost 2024 General AT 2022 Native in AT 2022 EU12 in AT 2022 HU in AT 2024 Subgroup Feel safe Very unsafe A bit unsafe Fairly safe Very safe Unknown FWF–NKFIH Joint Project 79 To summarise, the main conclusions of Table 10 and Figures 61-70 is that the satisfaction of Hungarians in Austria with material goods is quite high, approaching and even exceeding the Austrian average in terms of satisfaction with personal income. (Here the only exception is the satisfaction with housing conditions). The relatively high SWB values are particularly striking when compared to the subsamples in Hungary. For the non-material factors, however, the relationship is reversed: both in terms of satisfaction with social contacts and available leisure time, the lowest average values were measured among Hungarians in Austria. These indicators are not only worse than in Hungary, but also compared to immigrants from the new EU countries. Table 10. Average domain satisfaction scores by analytical subgroups. Avg. HU Pot. stayers HU Pot. migrants HU Avg. AT (2022) Native in AT (2022) EU12 in AT (2022) Hun in AT Diff.: Hun in AT vs. Avg. HU Diff.: Hun in AT vs. pot. migr. Diff.: Hun in AT vs. Avg. AT Material factors HH financial situation 6.2 6.4 6.0 7.3 7.6 6.7 7.1 +0.9 +1.1 –0.2 Personal income 5.9 6.1 5.3 6.9 7.1 5.9 7.1 +1.2 +1.8 +0.2 Job 6.8 7.0 6.4 7.5 7.6 6.9 7.1 +0.3 +0.7 –0.4 Accommodation 7.5 7.6 7.2 8.4 8.5 7.7 7.4 –0.1 +0.2 –1.0 Non-material factors Health 7.3 7.3 8.2 n.a. n.a. n.a. 7.7 +0.4 –0.5 n.a. Social relationships 7.7 7.8 7.8 8.6 8.7 8.3 7.5 –0.2 –0.3 –1.1 Time use 7.0 7.0 6.1 7.4 7.6 7.1 6.8 –0.2 +0.7 –0.6 External factors Living environment 7.4 7.5 7.2 n.a. n.a. n.a. 7.4 0 +0.2 n.a. Green areas 6.8 6.8 6.9 n.a. n.a. n.a. 7.4 +0.6 +0.5 n.a. Public services 5.4 5.5 5.0 n.a. n.a. n.a. 7.5 +2.1 +2.5 n.a. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own table. Figure 61. Satisfaction with the financial situation of the household by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 80 Figure 62. Satisfaction with personal income by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. Figure 63. Satisfaction with job by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 81 Figure 64. Satisfaction with accommodation by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. Figure 65. Satisfaction with health by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 82 Figure 66. Satisfaction with social relationships by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own figure. Figure 67. Satisfaction with time use by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 83 Figure 68. Satisfaction with living environment by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. Figure 69. Satisfaction with green areas by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 84 Figure 70. Satisfaction with public services by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. 5.3. Affective well-being Incorporating affective well-being metrics into surveys enables researchers to capture emotiondriven effects, offering a deeper understanding of how individuals truly feel in their daily lives and how various factors influence subjective well-being (SWB). This concept focuses on the frequency of positive and negative emotions, such as happiness, anxiety, or stress, shedding light on aspects of human behaviour that life satisfaction measures alone may overlook The affective dimension of the SWB gap between the two countries is a long-standing phenomenon. In the previous Research Report, we summarised the changes in the nationallevel gap after 2013 (Németh et al., 2023: 59-61). This time, we focus on the analytical subgroups, as shown in the table and figures below. Table 11 shows the percentage of respondents who have chosen the options "all of the time" or "most of the time" in the questionnaires of the EU-SILC or MIGWELL surveys. Perhaps the most striking conclusion is that the affective well-being scores of potential migrants are more favourable than that of those planning to stay, in all the aspects examined. This is an important part of the descriptive data analysis because it has become clear that affective well-being is not the main driver of emigration intentions. In Austria, national averages are available for only two variables, as the 2022 SILC questionnaire focused solely on happiness and loneliness. Based on our non-representative sample, Hungarians in Austria scored lower in both categories. However, when compared to data from Hungary, a more nuanced pattern emerges. Hungarians in Austria experience both positive and negative emotional states more frequently. They are more likely to feel happy and FWF–NKFIH Joint Project 85 calm in Austria than in Hungary, yet also more prone to feelings of nervousness and depression. Notably, their average scores for loneliness and stress are worse, at 10.7% and 16.3% respectively, compared to the Hungarian averages of 8.7% and 9.9% (Table 11, Figure 71-77). Table 11. Affective SWB scores by analytical subgroups. (Frequency: all of the time + most of the time answers, %, non-respondents are excluded) Avg. HU Pot. stayers HU Pot. migrants HU Avg. AT (2022) Native in AT (2022) EU12 in AT (2022) Hun in AT Diff.: Hun in AT vs. Avg. HU Diff.: Hun in AT vs. pot. migr. Diff.: Hun in AT vs. Avg. AT Positive emotions Happy 56.2 56.7 70.4 74.0 76.4 62.8 70.4 +14.2 0 –3.6 Calm, peaceful 52.3 52.2 61.0 n.a. n.a. n.a. 69.3 +17.0 +8.3 n.a. Negative emotions Nervous 7.7 6.9 5.2 n.a. n.a. n.a. 8.1 +0.4 +2.9 n.a. Downhearted,depressed 7.0 5.8 3.9 n.a. n.a. n.a. 7.5 +0.5 +3.6 n.a. Feeling down 6.4 5.5 2.1 n.a. n.a. n.a. 4.6 –1.8 +2.5 n.a. Lonely 8.7 7.4 2.6 4.3 3.4 6.9 10.7 +2.0 +8.1 +6.4 Stressed 9.9 9.0 8.6 n.a. n.a. n.a. 16.3 +6.4 +7.7 n.a. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own table. Figure 71. Being happy: the share of respondents by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 86 Figure 72. Feeling calm and peaceful: the share of respondents by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. Figure 73. Being nervous: the share of respondents by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 87 Figure 74. Feeling downhearted or depressed: the share of respondents by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. Figure 75. Feeling down in the dumps: the share of respondents by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 88 Figure 76. Feeling lonely: the share of respondents by analytical subgroups. Source of data: EU-SILC 2022, MIGWELL surveys 2024. Own figure. Figure 77. Feeling stressed: the share of respondents by analytical subgroups. Source of data: MIGWELL surveys 2024. Own figure. 5.4. Eudaimonic well-being Eudaimonic variables receive far less attention in international surveys compared to the cognitive and affective dimensions. However, incorporating measures of eudaimonia – which reflects a meaningful life and includes aspects such as autonomy, self-esteem, and selfactualization – could offer a more comprehensive understanding of SWB. This approach would FWF–NKFIH Joint Project 95 living in Austria, having a partner is more important, with a marginal effect of 1.25 (CI: 0.541.96), while for potential migrants the value is 0.47 (-0.05-1). The former is significant at the p<0.01 level, the latter at the p<0.1 level. The relationship between partner and job satisfaction is also moderated by the Hungarian group subcategory. The marginal effect of having a partner is significantly different between those who live in Austria and those who want to stay (Figure 84). Figure 84. Average marginal effects of having a partner by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Job satisfaction is related to the variable of how well an individual can cover his or her household expenses, as a proxy for salary. Compared to the reference „with major difficulties” category (5.22; CI: 4.53-5.92), all other variables show a significant positive marginal effect, except for "with difficulties" (mean marginal effect 0.87; CI: -0.13-1.87), at the p<0.001 level. It can be observed that, apart from the category „with difficulties”, the average marginal effect increases in all other categories as the ease of payment rises. The average marginal effect in the „very easily-with major difficulties comparison” is a remarkable 3.59 (CI: 2.46-4.71). After breaking down the groups, it becomes clear that the most important issue among Hungarians in Austria is the secure coverage of expenses. This is already apparent in the „with some difficulties-with major difficulties comparison”, with the marginal effect being significantly larger for Hungarians in Austria than for the other two groups in Hungary (5.2; CI: 1.43-9 in comparison with those intending to stay, 4.21; CI: 0.32-8.1 in comparison with potential migrants). The reason for this large difference is that having major difficulty paying for Partner: Yes Pot.stayers Pot.migrants in AT -0.5 0.0 0.5 1.0 1.5 2.0 estimate Reference: No No association with partner among pot.stayers FWF–NKFIH Joint Project 96 expenses is associated with markedly low job satisfaction among Hungarians in Austria. For Hungarians in Austria, the first step is the important one of achieving a level of „with some difficulties”. In comparison, even the average marginal effect of „easily-with major difficulties” is not significantly different from this „with some difficulties-with major difficulties step”. However, we can observe a significant jump at the „very easily-with major difficulties comparison”. This makes it clear that for Hungarians in Austria, „very easily” is the second important step up after the „with some difficulties” one (Figure 85). Figure 85. Average marginal effects of the ease of covering the expenses by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. The last variable with which we found a link is being looked down because of work or financial situation. Those who feel that society looks down on them have lower mean estimates of job satisfaction (5.77; CI: 5.39-6.15) than those who do not experience this (7.47; CI: 7.35-7.58). Being looked down upon versus not being looked down upon is associated with lower job satisfaction by an average of 0.93 (CI: 0.47-1.39) (no significant difference in comparison with the "so-so" category). Being looked down upon is the most negative in Austria, with an average marginal effect of -1.62 (CI: -2.72-(-0.52)) among Hungarians living there, while being looked down upon is the least strongly associated with job satisfaction among those planning to stay in Hungary (-0.7; CI: -1.33-(-0.06)). Furthermore, for those living in Austria, even the „so-so being looked down-not being looked down comparison” is significant, indicating a disadvantage of even some degree of disdain (Figure 86). Rather easily Very easily I don't know With difficulties With some difficulties Easily Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT 0 4 8 0 4 8 estimate Reference: With major difficulties Covering expenses is important in AT FWF–NKFIH Joint Project 97 Figure 86. Average marginal effects of being looked down upon by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. 6.1.2. Extended model for Austria Those who are overqualified for their current job have lower mean estimates (6.67; CI: 6.167.18) than those who have a suitable „education-job pairing” or were able to find a job requiring higher skills than they have based on their education (7.46; CI: 7.22-7.69). The mean marginal effect of overqualification is negative (-0.78; CI: -1.41-(-0.16)), with a significant association with job satisfaction. The negative association with overeducation is strongest for those who were maximally satisfied with their job in Hungary (-2.95; CI: -5.36-(-0.53)). This effect is not observed for those who were previously dissatisfied. The effect of overeducation is significantly different at the p<0.1 level for those who were previously absolutely dissatisfied and those who were maximally satisfied, with a marginal effect difference of 4.65 (CI: -0.42-9.73). Of the new variables included in the model, only the frequency of visits to Hungary showed a significant relationship with job satisfaction. Contrary to expectations, the mean estimates for those who visited home frequently were lower (6.96; CI: 6.63-7.3) than those who visited home infrequently (7.52; CI: 7.22-7.82). The mean marginal effect was -0.55 (CI: -1.08-(-0.03)) for frequent visits home. One could speculate that those who have no relatives in Hungary indicate a negative relationship with frequent home visits, while those with family members left behind are more satisfied with their job if they can travel home frequently. However, this is not the So-so No Pot.stayers Pot.migrants in AT Pot.stayers Pot.migrants in AT -1 0 1 2 estimate Reference: Yes Being looked down hurts in AT FWF–NKFIH Joint Project 98 case. In fact, the negative relationship of frequent home visits is even higher for those who have children or a partner in Hungary. However, the difference is not significant in this regard. Differences between subjective and objective integration among Hungarians in Austria This analysis is based on Karin Amit and Svetlana Chachasvili-Bolotin's article published in 2018 "Satisfied With Less? Mismatch Between Subjective and Objective Position of Immigrants and Native-Born Men and Women in the Labour Market". The authors introduced the idea that, in addition to the possible mismatch between skills required for a job and qualification, another mismatch may be present in the labour market. This arises when a person's labour market position based on objective data does not match the position occupied by subjective job satisfaction within a society. Previous research has shown that for women, job satisfaction is often higher than would be indicated by the conditions and prestige of their jobs. This discrepancy is often explained by the presence of low job-related expectations. In Israel, Amit and Chachasvili-Bolotin (2018) also showed that women have a higher proportion of this positive mismatch, yet men are more likely to have a negative mismatch (not satisfied with their otherwise good job and salary). In their study, this positive mismatch was also observed for some socioeconomically disadvantaged ethnic groups who were satisfied with their low material wealth. We did the same analysis for Hungarians in Austria, placing them in Austrian society. Calculation procedure and limitations: • First, we needed the subjective measure. In the cited article, this was the average of job satisfaction and income satisfaction. But since EU-SILC did not measure satisfaction with income, we replaced it with satisfaction with the household's financial situation. The wave that included both satisfaction with household income and satisfaction with work is 2018. The average of the two satisfaction levels was used as the subjective indicator, and then z-score standardization was used to produce the final score. For Hungarians in Austria, these are the 2024 satisfaction scores, so even this six-year difference could be a bias if there is a trend change. • The easier, work-related part of the objective indicator is connected to the prestige of the job. If the person is working in ISCO08 1,2 or 3, then code 1, if not, then 0. • In the objective financial part, since we have replaced satisfaction with the household's financial situation with satisfaction with income, we should also look at household income. However, our household income category data is either corrupted or the respondents not indicate the true category due to the personal nature of the question. Thus, for Hungarians, only the personal income response is meaningful. Then the Austrian 2018 EU-SILC personal income values were adjusted for inflation to 2023 FWF–NKFIH Joint Project 99 levels. Then the Austrian values were classified according to the salary response categories of our own survey questionnaire. Later on, as in the study by Amit and Chachasvili-Bolotin (2018), the median value of the categories was converted from category to salary value (e.g., a category between €250 and €500 became €375). This figure was then converted to a z-value. • The 0 or 1 and the z-value were added together, then another z-value conversion was performed. If the difference between the objective indicator and the subjective indicator was greater than 0.5, we can talk about a negative mismatch, if less than -0.5, about a positive match, and if between the two, about a good match. On average, the Hungarians are the second worst after the Turks on the subjective indicator. On the objective indicator, Hungarians are the third worst (ahead of Turks and Yugoslavs). Of all the country of birth groups, Hungarians have the best subjective and objective match. 49% of men and 36% of women belong to this correctly categorising group (Figure 87-88). Negative pairing is less common for women in all groups by country of birth, while positive pairing is more common for them. Negative pairing is true for 28% of Hungarian women, which is the third lowest. For Hungarian men it is 30%, the lowest of all country of birth groups. However, Hungarian women are more likely to be positively matched (35%) than Hungarian men (20%). Positive incorrect pairing is more common than correct pairing: entrepreneurs, low educated, married, living in Vorarlberg. However, this is not the case for Hungarians, who are 11.6% less likely to belong to this group. Negative incorrect pairing is more common than correct pairing: highly educated, native Austrians, Turks. FWF–NKFIH Joint Project 100 Figure 87. Negative and positive mismatch by gender and country of birth group. Source of data: MIGWELL surveys 2024. Own figure. Figure 88. Negative, correct, and positive mismatch in the country of birth groups by gender. Source of data: MIGWELL surveys 2024. Own figure. 6.2. Satisfaction with income 6.2.1. Model for the whole sample Based on the model, Hungarians living in Austria have the highest average income satisfaction estimate (7.2; CI: 6.96-7.44), while potential migrants have the lowest (5.03; CI: 4.78-5.27) (Figure 89). Compared with the estimates for the job satisfaction variable, it is striking that the gap between Hungarians living in Austria and those intending to stay in Hungary has widened. The average estimate for those intending to stay is 5.94 (CI: 5.81-6.07). The model indicates that living in Austria is associated with an average 0.54 higher income satisfaction compared Female Male Native EU15, EFTA EU12 HU Turkey Yugosla. Other 0% 20% 40% 60% 0% 20% 40% 60% Country of birth Percentage Subjective integration match negative matched positive FWF–NKFIH Joint Project 101 to being a potential migrant and 0.49 higher income satisfaction compared to intending to stay in Hungary. However, these marginal effects are not significant. Figure 89. Mean financial satisfaction estimates by Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. In terms of economic activity, entrepreneurs (7.28; CI: 6.89-7.67) and employed persons (6.27; CI: 6.15-6.4) have the highest average estimates, while the disabled (3.81; CI: 2.97-4.66) and the unemployed (2.57; CI: 1.93-3.21) have the lowest estimated income satisfaction. The average marginal effect compared to the employed reference category is significant for the disabled (-1.58, CI: -2.58-(-0.59)), student (-1.64; CI: -2.31-(-0.97)) and unemployed (-2.15; CI: -2.94-(-1.41)) groups in increasing order of magnitude. There are some interesting findings regarding the moderating effect of the Hungarian group. Being an entrepreneur in comparison with being employed is associated with significantly higher income satisfaction (0.62; CI: 0.11.15) among those who intend to stay, while this effect is negative but not significant in Austria. Also, the negative student financial satisfaction already presented is the lowest in Austria (- 1.15; CI: -2.27-(-0.03)) (Figure 90). 5.0 5.5 6.0 6.5 7.0 7.5 Pot.stayers Pot.migrants in AT Income satisfaction Enjoying the benefits of Austria FWF–NKFIH Joint Project 102 Figure 90. Average marginal effects of economic status by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. There is a very slight U-shaped correlation between age and income satisfaction, but for Hungarians living in Austria, a stagnation is expected until the age of 40, and then a slight decrease is observed. The model estimates that men have a slightly higher average income satisfaction (6.11; CI: 5.97-6.26 compared to 5.93; CI: 5.79-6.07 for women). In the counterfactual analysis, women can expect to have a slightly lower income satisfaction, with the largest gender gap observed for those living in Austria. However, the average marginal effect in this case is only -0.27 and the 95% confidence interval covers both negative and positive values, so no gender difference is observed. As expected, those with tertiary education have higher average estimates of income satisfaction (6.97; CI: 6.73-7.21) than those with a certificate (6.08; CI: 5.91-6.24) or a vocational qualification (5.89; CI: 5.72-6.07). Those with primary education have the lowest estimates (4.54, CI: 4.22-4.85). Looking at the marginal effects, there is a significant relationship between education and income satisfaction. Tertiary education is associated with higher satisfaction by an average of 0.62 (CI: 0.1-1.14) compared to the primary education category. The average marginal effect is also significant in the „vocational-primary education comparison”, with a value of 0.44. The Hungarian group has a moderating effect. For potential migrants, the effect is not significant in any of the comparisons. The mean marginal effect of „tertiary educationprimary education” is largest for Hungarians living in Austria (1.53; CI: 0.09-2.97) (Figure 91). Retired Disabled Other inactive Self-employed Unemployed Student Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -5.0 -2.5 0.0 -5.0 -2.5 0.0 estimate Reference: Employee Self-employed benefits for pot. stayers FWF–NKFIH Joint Project 103 Figure 91. Average marginal effects of education by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. By type of settlement, those living in densely (6.15; CI: 5.96-6.34) and intermediate (6.19; CI: 6.04-6.35) populated settlements have higher estimated average satisfaction than those living in sparsely (5.57; CI: 5.37-5.77) populated settlements. The counterfactual analysis does not support the positive returns to satisfaction of living in densely populated settlements. However, a positive (0.48; CI: 0.21-0.75) significant marginal effect is observed for living in intermediate densely versus sparsely populated settlements. This difference shows a positive significant association for those intending to stay, with an average marginal effect of 0.57 (CI: 0.24-0.89) (Figure 92). Secondary: vocational Secondary: certificate Tertiary Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -2 -1 0 1 2 3 estimate Reference: Primary High differences by education in AT FWF–NKFIH Joint Project 104 Figure 92. Average marginal effects of settlement type by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. On the partnership variable, those with a partner have higher average estimates of satisfaction not only with work but also with income. In their case, the average estimate is 6.2 (CI: 6.086.33), while for singles it is 5.63 (CI: 5.45-5.81). Having a partner is found to be associated with an average increase in satisfaction of 0.16 (CI: -0.09-0.42), the relationship is not significant. However, the role of the partner is moderated by the Hungarian group variable, with a significant positive association only in the potential migrant group (0.74; CI: 0.08-1.4) (Figure 93). Figure 93. Average marginal effects of having a partner by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Intermediate Dense Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -0.5 0.0 0.5 1.0 estimate Reference: Thin Importance of urbanity for pot. stayer FWF–NKFIH Joint Project 111 Figure 100. Average marginal effects of education by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. In the settlement type variable, the surplus in income-related satisfaction disappears for densely populated settlements once housing satisfaction is considered. The model estimates that the highest estimated housing satisfaction is observed in intermediately dense settlements (7.63; CI: 7.51-7.75). They are followed by densely populated (7.44, CI: 7.29-7.59) and sparsely populated (7.3; CI: 7.15-7.45) settlements. Living in an intermediately dense municipality is associated with an increase in housing satisfaction compared to a sparsely populated municipality. The average marginal effect is 0.24 (CI: 0.04-0.45). When examining the moderating effect of the Hungarian groups, it can be stated that this effect is mainly due to those who has the intention to stay (0.31; CI: 0.06-0.56) (although the average marginal effect is also positive in the other two groups). For the intention to stay group, the „dense-thin settlement comparison” (0.45; CI: 0.15-0.76) is also significant in the expected dense settlement direction. However, for Hungarians living in Austria, the „dense-thin comparison” shows a significant negative mean margin (-1.06; CI: -1.66-(-0.45)), i.e., living in rural areas is associated with a satisfaction premium (Figure 101). Secondary: vocational Secondary: certificate Tertiary Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT 0 1 2 estimate Reference: Primary Secondary in HU, Tertiary in AT FWF–NKFIH Joint Project 112 Figure 101. Average marginal effects of settlement type by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. In line with the usual trends, those with a partner also have a higher average estimate of satisfaction with housing. For those without a partner, the value is 7.19 (CI: 7.15-7.33), while for those with a partner it is 7.63 (CI: 7.53-7.72). The average marginal effect is also more favourable for those with a partner, with a higher estimate of housing satisfaction of 0.44 (CI: 0.24-0.64). The positive effect is evident in all Hungarian groups, the weakest for those intending to stay (0.38; CI: 0.14-0.62) and the strongest among Hungarians living in Austria (0.6; CI: 0.1-1.1) (Figure 102). Intermediate Dense Pot.stayers Pot.migrants in AT Pot.stayers Pot.migrants in AT -1 0 estimate Reference: Thin Opposite trends for pot. stayers and HU in AT FWF–NKFIH Joint Project 113 Figure 102. Average marginal effects of having a partner by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Previous research has shown that people who live in their own property report higher levels of housing satisfaction. Based on the mean estimates, this is also true for the sample of Hungarians, as those who live in their own property (7.64; CI: 7.55-7.73) have higher scores than those who do not own their home (7.05; CI: 6.89-7.21). The counterfactual model also suggests a significant negative relationship between not owning the property and housing satisfaction. The average marginal effect is -0.37 (CI: -0.61-(-0.13)). The negative effect is observed in all Hungarian groups, but is significant only in the two groups currently in Hungary. It is slightly stronger for potential migrants (-0.53; CI: -0.06-(-1)) than for those intending to stay (-0.32; CI: -0.01-(-0.63)) (Figure 103). FWF–NKFIH Joint Project 114 Figure 103. Average marginal effects of house/flat ownership by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. There is a negative relationship between the number of household members per room and housing satisfaction. Thus, the more household members per room, the more dissatisfied the respondent is. An increase in the ratio by one is associated with a decrease in satisfaction of 0.49 (CI: 0.22-0.76). The decrease is steepest for those intending to stay, and smallest for those living in Austria. Similarly, lower housing satisfaction is associated with more problems with the dwelling and its surroundings (leaking, darkness, noise, environmental problems, crime). A decrease in satisfaction of 0.26 (CI: 0.11-0.41) is observed when one more problem is present. The relationship is not moderated by Hungarian group variable. Closely related to housing satisfaction is the financial situation of the respondent, which is still represented by the variable which measures the difficulty of covering expenses. The lowest estimated average satisfaction is for those who have major difficulties in financing their expenses (5.5; CI: 5.08-5.92), while the highest estimated satisfaction is for those who can cover their expenses very easily (8.91; CI: 8.49-9.33). Compared to the reference group, all other response categories (even the " I don't know") show a significant positive mean marginal effect. The strongest is of course in the „very easily-with major difficulties comparison” (3.14; CI: 2.42-3.86). While for the previous domains, covering expenses was a very important indicator in Austria, it was not a relevant variable for housing satisfaction for Hungarians in Austria. Compared to the major difficulties category, only the response „I do not know” shows a significant positive difference for this migrant group. The positive marginal effects are thus due to the two groups in Hungary, especially the potential migrants (Figure 104). FWF–NKFIH Joint Project 115 Figure 104. Average marginal effects of the ease of covering the expenses by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. 6.3.2. Extended model for Austria For Hungarians in Austria, pre-migration housing satisfaction moderates the already presented association in several ways. The non-significant male gender advantage is strongest when the level of satisfaction before migration was low. The model suggests that men in Austria who were dissatisfied with their housing before migration are more satisfied with their housing than women who were equally dissatisfied before migration (mean marginal effect is 1.8). Similarly, for the low level of dissatisfaction before migration, the most significant additional satisfaction is related to living in a sparsely populated settlement. As expected, the higher the pre-migration satisfaction, the lower the estimates for post-migration satisfaction (but the relationship is not significant). This is only not evident for those living intermediately dense settlements, where the slope is positive. Also, in terms of education, the satisfaction advantage associated with high education is highest when satisfaction before migration was low. For the low-educated and those with vocational education, low pre-migration satisfaction is associated with low postmigration satisfaction and high satisfaction with high satisfaction (positive slope). The slope for housing-related problems and the household to room ratio is described by the most negative slope when pre-migration satisfaction was high. Thus, if an individual was satisfied with their housing before migration, but afterwards experiences many problems with housing or lives in an overcrowded dwelling, they will have a low estimate of their housing satisfaction. Conversely, if they were dissatisfied before migration, these problems are less bothersome. Rather easily Very easily I don't know With difficulties With some difficulties Easily Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -2.5 0.0 2.5 5.0 -2.5 0.0 2.5 5.0 estimate Reference: With major difficulties Importance of finances in HU FWF–NKFIH Joint Project 116 Those who send money home have higher average satisfaction with housing (7.34; CI: 7.067.63) than those who do not (7.2; CI: 6.84-7.94), but lower than those who did not answer to this question (7.46; CI: 6.98-7.94). The mean marginal effects are not significant, so we found no evidence that sending home causes individuals to spend so little on their housing that it is associated with dissatisfaction. In terms of reasons for migration, indicating the reasons of financial and helping one's future in Hungary (7.44 and 7.37) resulted in higher estimates than not indicating these reasons (7.07 and 6.08). For the reason of helping one's future in Hungary, the average marginal effect is 0.528 (CI: -0.05-1.11), which is significant at the p<0.1 level. Thus, those who know they will return home are more satisfied with their current housing. There is a significant effect for year of emigration, the earlier the emigration, the higher the satisfaction with housing. 6.4. Satisfaction with social relationships 6.4.1. Model for the whole sample Satisfaction with social relationships is the first domain for which Hungarians living in Austria have the lowest mean estimate (7.61; CI: 7.41-7-81), followed by potential migrants (7.69; CI: 7.52-7.86), while those intending to stay have the highest mean estimate (7.87; CI: 7.78-7.97). Although the mean marginal effect is less favourable for those living in Austria compared to those intending to stay (-0.27; CI: -0.81-0.26), the counterfactual analysis did not reveal any significant difference between Hungarians intending to stay and Hungarians living in Austria. Being a potential migrant is also associated with lower satisfaction compared to those intending to stay, although this relationship is neither significant (-0.26; CI: -0.61-0.08) (Figure 105). Figure 105. Mean relationship satisfaction estimates by Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 117 In the economic activity categories, a high average estimate satisfaction appears among students (8.38; CI: 8.05-8.71). Also, high scores are found for the other inactive (8.16; CI: 7.8-8.51) and, again, estimates for the self-employed (8.01; CI: 7.72-8.29) are higher than those for the employed (7.86; CI: 7.77-7.96). The estimates for the retired (7.19; CI: 6.97-7.41), the unemployed (7.1; CI: 6.63-7.58) and the disabled (6.75; CI: 6.15-7.35) are well below these values. Average marginal effects compared to employed only for the disabled show a significant difference in satisfaction at the p<0.1 level (-0.59; CI: -1.29-0.1). For entrepreneurs (0.02; CI: -0.28-0.34) and students (0.25; CI: -0.21-0.72) the model also indicates a positive average marginal effect compared to employed, but the difference is not significant. The Hungarian group significantly moderates the relationship between economic activity and relationship satisfaction in only one case. Those intending to stay show lower satisfaction among retired compared to employed (-0.53; CI: -0.93-0.14) (Figure 106). Figure 106. Average marginal effects of economic status by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. The average estimated satisfaction of women (7.84; CI: 7.73-7.95) is slightly higher than that of men (7.75; CI: 7.64-7.86). The analysis does not show a general relationship between gender and satisfaction. After moderating effects are taken into account, the group intending to stay has the strongest female satisfaction surplus (0.16; CI: -0.03-0.35), while the average marginal effect among Hungarians living in Austria is negative (-0.27; CI: -0.71-0.17). Although the Retired Disabled Other inactive Self-employed Unemployed Student Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -5.0 -2.5 0.0 2.5 -5.0 -2.5 0.0 2.5 estimate Reference: Employee Retired disadvantage among pot. stayers FWF–NKFIH Joint Project 118 moderating effect is not significant, it points to a slightly less favourable social relations of Hungarian women (including transnational mothers) living in Austria (Figure 107). Figure 107. Average marginal effects of gender by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. For the education variable, the average estimates follow the usual ranking. The highest scores are for the highly educated (8.18; CI: 8.01-8.36), while the lowest are for the low educated (7; CI: 6.76-7.24). Belonging to any other category compared to the low-skilled was associated with significantly higher relationship satisfaction. The average marginal effect compared with the primary education category is 0.33 (CI: 0.00-0.66) in the case of vocational education; 0.38 (CI: 0.05-0.71) in the case of certificate; and 0.5 (CI: 0.12-0.87) in the case of tertiary education. The moderating effect appears in a way that those significant differences are only found among those who intend to stay and Hungarians living in Austria (in this group the average marginal effects are larger) (Figure 108). FWF–NKFIH Joint Project 119 Figure 108. Average marginal effects of education by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. People living in densely populated areas (8.01; CI: 7.86-8.16) and those living in intermediately dense areas (7.89; CI: 7.77-8) are more satisfied with their relationships based on the estimates, than those living in rural, thinly populated areas (7.44; CI: 7.3-7.59). The average marginal effects also show a clear positive relationship in favour of intermediate density (0.26; CI: 0.060.47) and high density (0.47; CI: 0.24-0.7) settlements. The advantage of these settlement types is most pronounced for those intending to stay (intermediate: 0.33; dense: 0.81), with a significant moderating effect. For Hungarians living in Austria, the effects are directed towards thinly populated settlements. In the „densely-thinly comparison”, the average marginal effect is -0.4 (CI: -1.08-0.27), but the relationship is not significant (Figure 109). Secondary: vocational Secondary: certificate Tertiary Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -1 0 1 2 estimate Reference: Primary Importance of education FWF–NKFIH Joint Project 120 Figure 109. Average marginal effects of settlement type by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Satisfaction with relationships is also higher for those with larger households. An extra household member is expected to be connected with a 0.14 higher satisfaction value. This positive slope is pronounced for the two groups in Hungary, while the slope is much steeper for Hungarians in Austria. This may be due to the fact that in their case the household also includes people who have stayed in Hungary, so they may not live together. As expected, those with a partner (7.94; CI: 7.85-8.03) have a higher estimated average satisfaction with relationships than singles (7.49; CI: 7.35-7.63). The mean marginal effect is significant and directed towards those with a partner (0.42; CI: 0.23-0.62). The relationship is also positive in all of the Hungarian groups (not significant for potential migrants), but the role of the partner is more prominent for Hungarians in Austria. The average marginal effect in this group is 1.18 (CI: 0.63-1.73) (Figure 110). FWF–NKFIH Joint Project 127 satisfaction. No significant relationship appears when generalised trust is replaced by localised trust (trust in the residents of your neighbourhood) in the model. 6.5. Satisfaction with health 6.5.1. Model for the whole sample The average estimates of health satisfaction are an excellent reflection of the healthy migrant phenomenon, i.e., that potential migrants enjoy excellent (subjective) health. The mean estimate for this group is 8.17 (CI: 7.99-8.34), which is well above that of those who want to stay (7.41; CI: 7.31-7.51). Hungarians living in Austria are in between the two groups, but closer to the potential migrants' score (7.9; CI: 7.7-8.11). After controlling for the other variables, there is no evidence of a surplus by being a potential migrant, but Hungarians living in Austria show a negative average marginal effect compared to the other two Hungarian groups, even if these differences are not significant at the p<0.05 level (-0.48 in comparison with those intending to stay; CI: -1.03-0.06; p=0.08) (Figure 116). Figure 116. Mean health satisfaction estimates by Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Estimated differences in economic activity are also affected by age. The average estimate for students is the highest (8.57; CI: 8.22-8.92), while the average estimate for pensioners (5.72; CI: 5.49-5.95) and disabled (4.08; CI: 3.47-4.7) is the lowest. The mean estimate for employed persons is 7.92 (CI: 7.83-8.02). In contrast to employed persons, pensioners (-1.34; CI: -2.2-(- FWF–NKFIH Joint Project 128 0.49)) and disabled persons (-2.27; CI: -2.99-(-1.53)) show a negative mean marginal effect. A few moderating effects are worth highlighting. Other inactive persons enjoy a satisfaction surplus among those intending to stay (0.65; CI: 0.2-1.1) compared to employed persons, while among Hungarians in Austria this other inactive group have a negative average marginal effect (-1.64; CI: -3.2-(-0.07)). Potential migrants show an unemployment satisfaction surplus (1.55; CI: 0.6-2.5) (Figure 117). Figure 117. Average marginal effects of economic status by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Most studies describe the relationship between age and health in terms of a U-shaped curve. This U-shape is most pronounced in the group of Hungarians in Austria. In contrast, for those who intend to stay, a slope is visible after the age of 40. This trend at the end of the life curve differs significantly between Hungarians intending to stay and Hungarians in Austria. The average estimate of health satisfaction for men is higher (7.75; CI: 7.64-7.86) than for women (7.53; CI: 7.42-7.64). The average marginal effect is also in favour of men (-0.131; CI: -0.30.03), but the difference is not significant. The average marginal effect between genders is closest to 0 in the group of Hungarians in Austria. The ranking of educational attainment observed for the previous domains is also maintained in the average estimates of health satisfaction. The low-educated (6.59; CI: 6.35-6.84) are outperformed by those with a vocational degree (7.38; CI: 7.24-7.52), those with a certificate (7.84; CI: 7.71-7.97) and those with a tertiary education (8.2; CI: 8.02-8.38). The average marginal effects, while pointing in the expected direction towards higher education, are not Retired Disabled Other inactive Self-employed Unemployed Student Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -9 -6 -3 0 3 -9 -6 -3 0 3 estimate Reference: Employee Opposite effects for other inactive FWF–NKFIH Joint Project 129 significant (0.31 in the „tertiary education-primary education comparison”; CI: -0.08-0.7). No significant relationship with the moderation of the Hungarian groups is found, with the only noteworthy difference being the larger „tertiary education-primary education” difference in the sample of Hungarians in Austria (Figure 118). Figure 118. Average marginal effects of education by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. By type of settlement, the highest average satisfaction estimates are for those living in densely populated areas (7.73; CI: 7.58-7.89), followed by those living in intermediately dense areas (7.65; CI: 7.53-7.77) and thinly populated areas (7.52, CI: 7.37-7.67). The average marginal effects show an intermediate density superiority compared to living in sparsely populated areas (0.2; CI: -0.0009-0.4). This positive effect is observed in the two groups in Hungary. The „densely populated-thinly populated comparison” shows the opposite direction for Hungarians intending to stay and for Hungarians in Austria. In Austria, there is a perceived health advantage associated with living in rural areas (-0.97; CI: -1.6-(-0.32)), while in Hungary the effect is in the direction of densely populated settlements (0.33; CI: 0.03-0.64) (Figure 119). Secondary: vocational Secondary: certificate Tertiary Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -1 0 1 2 estimate Reference: Primary Slightly higher importance of education in AT FWF–NKFIH Joint Project 130 Figure 119. Average marginal effects of settlement type by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. The average estimate for those living with a partner is higher (7.69; CI: 7.59-7.78) than for singles (7.53; CI: 7.39-7.67), but the average marginal effect (0.04; CI: -0.14-0.23) does not show a significant advantage of having a partner. The existence of peer relationships can have multiple effects on health. They may encourage joint exercise and health promotion activities, and may also protect mental health by reducing loneliness and depression. In the sample, those who can count on some help from friends and family (7.69; CI: 7.61-7.77) have higher mean satisfaction estimates than those who cannot (7.04; CI: 6.77-7.32). The average marginal effect is not significant in the total sample (0.17; CI: -0.16-0.52), but the moderating effect of the Hungarian groups is present and a significant positive effect is observed for Hungarians in Austria (1.4; CI: 0.12-2.69) (Figure 120). FWF–NKFIH Joint Project 131 Figure 120. Average marginal effects of the potential of getting help by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. The analysis also shows a positive relationship with the number of friends. Having just one friend is associated with a higher mean estimate (7.26; CI: 7.04-7.47) than having no friends (6.6; CI: 6.29-6.91). A further larger jump appears in the 2-4 friend category (7.75; CI: 7.647.85), and from there the differences level off (5-9 friends 7.96; 10 or more friends 7.82). The average marginal effect for the „2-4 friend-no friend comparison” (0.45; CI: 0.02-0.88) is significant at the p<0.05 level and indicates a positive return for having friends. The moderating effect of the Hungarian groups does not appear. Those who do leisure activities have significantly higher mean estimate (8.26; CI: 8.14-8.38) than those who do not do so due to financial or time constraints (6.86; CI: 6.66-7.06) or other reasons (7.24; CI: 7.11-7.36). Recreational activity has an average marginal effect of 0.34 (CI: 0.04-0.64) compared to passivity due to limitation. The mean marginal effect of leisure activity is also positive against the mean marginal effect of not participating in leisure activities due to other reasons, but it is no longer significant (0.1; CI: -0.09-0.31). The absence of leisure activities due to limitation is the strongest in the sample of Hungarians in Austria. Even when compared to other reasons, dropping out due to lack of money or time has a significant negative (-1.33; CI: -2.5 (-0.41)) relationship with health-related satisfaction (Figure 121). FWF–NKFIH Joint Project 132 Figure 121. Average marginal effects of spending leisure time by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Significant differences in health satisfaction are found in the perception of social exclusion. The average estimate for those who experience total exclusion is 6.52 (5.89-7.15), while for those who absolutely do not experience exclusion it is 8.17 (CI: 8.05-8.28). The average marginal effect between these two categories is significant at the p<0.1 level (0.63; CI: -0.1-1.38). After moderation, it is found that this relationship is strong for Hungarians in Austria. In their case, in contrast to total exclusion, even „rather not being excluded” is significantly associated with higher satisfaction at the p<0.1 level (1.33; CI: -0.16-2.82) (Figure 122). Figure 122. Average marginal effects of social exclusion by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 133 Membership in CSOs is also a variable that can reflect health-conscious behaviour. Those who are members of a CSO have higher mean estimated satisfaction (7.75; CI: 7.59-7.92) compared to non-members (7.6; CI: 7.51-7.69). The average marginal effect is also positive, but not significant (0.16; CI: -0.04-0.36) for the total sample. This is because a negative effect (-0.02; CI: -0.28-0.22) is observed among those intending to stay, while the effect is positive as expected at p<0.05 and p<0.1 levels among potential migrants (0.53; CI: 0.04-1.02) and Hungarians in Austria (0.47; CI: -0.02-0.97) (Figure 123). Figure 123. Average marginal effects of being a member in a civil society organisation by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. A one higher number of problems in the living environment is associated with lower health satisfaction for those intending to stay (-0.15; CI: -0.36-0.04, not significant) and potential migrants (-0.35; CI: -0.63-(-0.08)). Significant differences in satisfaction with health status are observed with respect to the financial situation of the respondent. Those who find it very difficult to meet expenses have a much lower mean estimated satisfaction (5.65; CI: 5.22-6.08) than those who find it very easy to cover expenses (9.1; CI: 8.68-9.52). Even the „with difficulties-with major difficulties comparison” is associated with higher satisfaction (0.59; CI: -0.01-1.2), which is significant at the p<0.1 level. The average marginal effect for the „very easily-with major difficulties comparison” is 2.29 (CI: 1.55-3.03). The moderating effect indicates that the link between finances is especially closely associated with health-related satisfaction, among those who intend to stay in Hungary (Figure 124). FWF–NKFIH Joint Project 134 Figure 124. Average marginal effects of the ease of covering the expenses by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. There is a large difference in mean estimated satisfaction between those who experienced a long-term negative health effect from COVID (6.38; CI: 6.12-6.63) and those who did not (7.77; CI: 7.68-7.85). The mean marginal effect is 0.53 (CI: 0.21-0.86), with no long-term COVID effect associated with significantly higher health satisfaction. The effect is positive in all Hungarian subsamples, but is significant only for those intending to stay (0.51; CI: 0.15-0.88). 6.5.2. Extended model for Austria The level of satisfaction with health before migration shows a significant positive relationship with current satisfaction. For the group of Hungarians in Austria, the moderating effect of premigration health status is reflected in the urbanity, gender and long COVID variables. For those who were dissatisfied with their health before migration, densely populated settlements are associated with higher health satisfaction (2.77 for pre-migration satisfaction score of 2; CI: 0.38-5.15). Conversely, those who were fully satisfied show a positive association with living in thinly populated settlements. Similarly, in the case of pre-migration satisfaction, women (premigration satisfaction score of 2: 1.9; CI: 0.24-3.56) show higher current satisfaction. For premigration satisfaction the relationship is not clear. The long-term health effects of COVID are most severe for those who were satisfied with their health before migration. The mean estimated satisfaction of frequent visitors to Hungary (7.82; CI: 7.6-8.04) is lower than that of infrequent visitors (7.99; CI: 7.78-8.21), and the mean marginal effect is also against Rather easily Very easily I don't know With difficulties With some difficulties Easily Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -4 -2 0 2 4 -4 -2 0 2 4 estimate Reference: With major difficulties Importance of finances in HU FWF–NKFIH Joint Project 135 frequent visitors (-0.29; CI: -0.76-0.18), but not significant. In addition to the possible tensions associated with frequent home visits, the possible purpose of home visits to seek medical care in Hungary may also point towards lower satisfaction. The positive impact of home visits is reflected in the emotional charge provided by the family. Those who send money home have a lower average satisfaction (7.82; CI: 7.61-8.03) than those who do not (8.03; CI: 7.75-8.31). However, the average marginal effect is not significant, pointing minimally against sending money home (-0.14; CI: -1.15-0.02). The analysis therefore does not suggest that sending money home takes significant resources away from spending on maintaining health or treating illness. There is evidence of lower satisfaction with health among those emigrating to establish a future in Hungary. The mean estimate for this group is 7.55 (CI: 7.15-7.95), while for those migrating for other reasons it is 7.97 (CI: 7.81-8.14). The relationship is significant at the p<0.1 level (-0.55; CI: -1.14-0.02). This may suggest the role of overwork related to the need to meet goals as soon as possible. 6.6. Satisfaction with work-life balance 6.6.1. Model for the whole sample Looking at the mean estimates of satisfaction with the use of time, the group of potential migrants shows a considerable lag (6.17; CI: 6.96-6.39) compared to Hungarians in Austria (6.8; CI: 6.56-7.04) and those intending to stay in Hungary (6.97; CI: 6.85-7.09). The difference between potential migrants and those intending to stay is significant, with the average marginal effect (-0.58; CI: -1.01-(-0.16)) showing lower satisfaction among potential migrants (Figure 125). Figure 125. Mean time use satisfaction estimates by Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. FWF–NKFIH Joint Project 136 In terms of economic activity, the average estimates for this domain do not show an employment advantage. The average estimated satisfaction with time use of employed persons is relatively low compared to the other groups (6.6; CI: 6.49-6.72). Only students have a lower value than this (6.53; CI: 6.12-6.94). Unsurprisingly, the average estimated satisfaction of retired people (7.58; CI: 7.31-7.86), other inactive people (7.29; CI: 6.85-7.72) and unemployed people (7.08; CI: 6.48-7.67) is also higher than that of employed or self-employed people (6.86; CI: 6.5-7.21). However, when comparing the mean effects with those of being employed, only the disadvantage of being a student (-0.79; CI: -1.4-(-0.18)) and the advantage of being unemployed (1.57; CI: 0.73-2.41) are significant after controlling for other variables. Student dissatisfaction is strongest among Hungarians in Austria (-1.074; CI: -2.01-(-0.13)). In contrast, the unemployment advantage is clearly due to the two groups in Hungary (1.46 for those intending to stay and 1.1 for potential migrants). The group of those intending to stay also shows a significant advantage of other inactive (0.56; CI: 0.006-1.15) and retired (0.56; CI: 0.06-1.05) in comparison with employed at the p<0.05 level. In contrast, Hungarians in Austria show a more favourable entrepreneurial situation (1.07; CI: -0.05-2.2) (Figure 126). Figure 126. Average marginal effects of economic activity by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. There is also a U-shaped relationship between satisfaction with the use of time and age. The upward trend for potential migrants in the 30-50 age range is lower than the upward trend for those who intend to stay. Retired Disabled Other inactive Self-employed Unemployed Student Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -4 0 4 -4 0 4 estimate Reference: Employee Major moderator effect in economic status FWF–NKFIH Joint Project 143 Figure 134. Average marginal effects of the ease of covering the expenses by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. 6.6.2. Extended model for Austria Satisfaction with the pre-migration period does not appear as a moderator. No significant marginal effects are found for reasons for migration. The largest difference is found in the category of helping to establish a future in Hungary, where the average estimate for those who selected this response (6.31; CI: 5.72-6.9) is lower than for those who migrated for other reasons (6.87; CI: 6.63-7.11). The average marginal effect is -0.42 (CI: -1.28-0.42). The mean estimated satisfaction of those who rarely move home is slightly higher (6.8; CI: 6.49-7.12) than that of those who move home frequently (6.77; CI: 6.45-7.1). The non-significant relationship (-0.46; CI: -1.17-0.24) provides no evidence of a potential time burden associated with moving home. There is also a negative association between family member staying at home and satisfaction, but this effect is not significant (-0.58; CI: -1.34-0.16). 6.7. Satisfaction with public services 6.7.1. Model for the whole sample The satisfaction advantage of Hungarians in Austria is greatly reflected in their satisfaction with public services. The average estimated satisfaction of Hungarians in Austria is 7.42 (CI: 7.167.68), of those who want to stay 5.51 (5.39-5.64), while that of those who want to leave Hungary is only 5.02 (4.79-5.24). The average marginal effect is significant for those living in Austria compared to both groups in Hungary. In comparison with those who want to stay, the average marginal effect is 1.6 (CI: 0.9-2.38), and in relation to potential migrants it is even higher at Rather easily Very easily I don't know With difficulties With some difficulties Easily Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -3 0 3 6 -3 0 3 6 estimate Reference: With major difficulties Importance of finances in HU FWF–NKFIH Joint Project 144 2.04 (CI: 1.2-2.8). Potential migrants also suffer a satisfaction disadvantage compared to those who want to stay, with an average marginal effect of -0.39 (CI: -0.84-0.05), significant at the p<0.1 level (Figure 135). Figure 135. Mean public service satisfaction estimates by Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. In terms of economic activity, disabled people have the lowest mean estimate (4.92; CI: 4.125.72), while entrepreneurs have the highest (6.11; CI: 5.72-6.49). The mean marginal effects show no significant relationship between economic activity and satisfaction with public services. However, with the introduction of the moderating effect, a satisfaction surplus associated with the other inactives appear compared to those who are employed, among those who intend to stay (0.77; CI: 0.19-1.35). The average estimated satisfaction is highest in young adulthood, then falls sharply and starts to rise again with a modest slope around age 70. For Hungarians living in Austria, however, there is no such collapse in their 30s and 40s. Around age 40, for example, the satisfaction surplus of Hungarians living in Austria is 2.15 (CI: 0.873.44) compared to those who want to stay (Figure 136). FWF–NKFIH Joint Project 145 Figure 136. Average marginal effects of economic status by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. The average estimated satisfaction is higher for men (5.74; CI: 5.6-5.89) than for women (5.67; CI: 5.53-5.81). The average marginal effect also confirms the association with the gender variable, with women being less satisfied (-0.22; CI: -0.48-(-0.005)). The effect is mainly due to the two groups in Hungary, with the average marginal effect for Hungarians in Austria being only -0.02. Turning to the education variable, although the satisfaction of highly educated people is higher (6.16; CI: 5.93-6.39) than that of those with a certificate (5.67; CI: 5.5-5.83), those with a vocational school degree (5.61; CI: 5.42-5.79) or those with low qualifications (5.27; CI: 4.95-5.59), no significant relationship between education and satisfaction with public services can be found. There is no moderating effect of Hungarian group variable. Satisfaction once again reflects the disadvantaged role of being at the bottom of the settlement hierarchy. Those living in sparsely populated settlements (5.49; CI: 5.3-5.69) have the lowest average estimate, followed by those living in densely populated settlements (5.74; CI: 5.545.93) and then those living in intermediately populated settlements (5.81; CI: 5.66-5.97). The average marginal effect in the „intermediate-thin comparison” is significant (0.24; CI: -0.020.5) at the p<0.1 level. This difference is due to the group of those intending to stay, for whom the effect is 0.35 (CI: 0.04-0.67). In the „dense-thin comparison”, the potential migrant group is the only one where the average marginal effect is in the direction of dense settlements (0.83; CI: 0.19-1.47), furthermore this relationship is significant (Figure 137). Retired Disabled Other inactive Self-employed Unemployed Student Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -5.0 -2.5 0.0 2.5 5.0 -5.0 -2.5 0.0 2.5 5.0 estimate Reference: Employee Other inactive surplus among pot. stayers FWF–NKFIH Joint Project 146 Figure 137. Average marginal effects of settlement type by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Again, the presence of a partner was associated with a higher mean satisfaction estimate (5.78; CI: 5.66-5.91) than its absence (5.35-5.71). However, there is no significant relationship between this variable and satisfaction with public services (0.09; CI: -0.16-0.34). In terms of social exclusion, the average estimates do not show a clear trend, with those who are totally excluded (5.48; CI: 4.66-6.29) being higher than those who are rather excluded (4.82; CI: 4.415.22). As well as those indicating a feeling of rather not being excluded have higher mean estimates (6.01; CI: 5.82-6.21) than those who experience absolutely no exclusion (5.59; CI: 5.44-5.74). The mean marginal effects do not show a significant relationship. However, the moderating effect of the Hungarian groups appears and for Hungarians in Austria, a relationship appears to exist between feelings of exclusion and satisfaction with public services. For this group, even uncertainty ("so-so") is associated with significantly higher satisfaction (2.62; CI: 0.64-4.6) compared to total exclusion. Compared to this effect, the „absolutely not excludedtotally excluded comparison” provides only a slightly higher average marginal effect (2.98; CI: 0.93-5.03) (Figure 138). FWF–NKFIH Joint Project 147 Figure 138. Average marginal effects of social exclusion by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Average estimates of satisfaction with public services increase as an individual's financial situation improves. The lowest score is for those who find it a major difficulty to meet their expenses (3.4; CI 2.92-4.02), while the highest score is for those who find it easy to meet their expenses (6.42; CI: 6.1-6.74). Even in the „with difficulties-with major difficulties comparison” (1.16; CI: 0.38-1.92), a significant relationship between better financial situation and higher satisfaction is already evident. The average marginal effect is even larger in the „very easilywith major difficulties comparison”, 2.12 (CI: 1.17-3.07). The relationship is the strongest for those intending to stay, and for potential migrants it is significant in only one comparison (with difficulties-with major difficulties) (Figure 139). FWF–NKFIH Joint Project 148 Figure 139. Average marginal effects of the ease of covering the expenses by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Based on health status, the mean estimates for those in bad health are lower (5.08; CI: 4.685.48) than those in very good health (5.84; CI: 5.69-5.99). However, the mean marginal effects do not show a significant relationship. In fact, for those who want to stay, a relationship appears where those in very good health are more dissatisfied with public services than those in poor health (-0.76; CI: -1.4-(-0.12)). There is also a strong positive association between the level of institutional trust and satisfaction with public services (0.18; CI: 0.13-0.23). The association is slightly stronger for those intending to stay (0.23; CI: 0.18-0.28) than for potential migrants (0.14; CI: 0.04-0.24). In contrast, the association is not significant for Hungarians in Austria (0.02; CI: -0.16-0.21). 6.7.2. Extended model for Austria In general, there is a slight downward slope in the relationship between satisfaction before migration and current satisfaction. Low satisfaction before migration is associated with high satisfaction now. And satisfaction before migration is associated with current dissatisfaction. This relationship is very strong for the low skilled, with a much steeper slope than for the other groups. Rather easily Very easily I don't know With difficulties With some difficulties Easily Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT 0.0 2.5 5.0 7.5 0.0 2.5 5.0 7.5 estimate Reference: With major difficulties Importance of finances in HU FWF–NKFIH Joint Project 149 6.8. Satisfaction with green areas 6.8.1. Model for the whole sample Hungarians in Austria were the most satisfied with green spaces (7.16; CI: 6.9-7.42), based on average estimates. Potential migrants (6.84; CI: 6.6-7.07) and those intending to stay (6.9; CI: 6.76-7.03) have almost the same values. The mean marginal effects do not indicate a significant relationship between the Hungarian groups and satisfaction (Figure 140). Figure 140. Mean green area satisfaction estimates by Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. Differences in economic activity and gender variables are not significantly associated with satisfaction. In terms of moderating effects, among those intending to stay, students are associated with significantly lower satisfaction than employed at the p<0.1 level (-0.87; CI: - 1.7-0.03). Among the categories of education, the highest mean estimate is for highly educated (7.4; CI: 7.17-7.64), while the lowest is for low-educated (5.99; CI: 5.65-6.32). Having a certificate (0.45; CI: 0.003-0.9) and being highly qualified (0.63; CI: 0.12-1.14) are associated with significantly higher satisfaction in comparison with primary education. The difference in the „vocational-primary qualified comparison” (0.38; CI: -0.06-0.83) is significant at the p<0.1 level. These effects are pointing towards the more highly educated categories in all Hungarian groups, but are significant only for the potential migrant group (Figure 141). FWF–NKFIH Joint Project 150 Figure 141. Average marginal effects of education by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. On average, those living in densely populated settlements (7.27; CI: 7.07-7.47) are the most satisfied with green spaces, followed by those living in intermediately dense settlements (7.16; CI: 7.01-7.32) and then by those living in sparsely populated settlements (6.19; CI: 5.99-6.39). Living in an intermediately (0.91; CI: 0.64-1.19), and in a densely (1.15; CI: 0.83-1.47) populated settlement is associated with a significant satisfaction surplus compared to living in a thinly populated settlement. These effects are strongest for those who want to stay in Hungary. For Hungarians in Austria, there is no significant relationship, and the mean marginal effect in the „densely-thinly comparison” is negative (-0.35; CI: -1.17-0.45). As the burden of financial expenditure decreases, higher average estimates of satisfaction with green spaces can be expected. The value for the „with major difficulties” category is 5.18 (CI: 4.62-5.73), while the value for the „very easily” group is 7.73 (CI: 7.18-8.28). The average estimated satisfaction is the highest for the „easily” option (7.74; CI: 7.42-8.07). Based on the average marginal effects, a significant relationship is found between easier coverage of financial burdens and higher satisfaction with green spaces. Already in the „with difficultieswith major difficulties comparison” a significant satisfaction surplus is found for the „with difficulties” group (1.12; CI: 0.33-1.91). In the „easily-with major difficulties comparison” this value rises to 2.16 (CI: 1.36-2.96). The effect is due to the two groups in Hungary. For Hungarians in Austria, there is no significant relationship between financial situation and satisfaction with green spaces. However, a significant mean marginal effect appears in the „I Secondary: vocational Secondary: certificate Tertiary Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -1 0 1 2 estimate Reference: Primary Importance of education for pot. migrants FWF–NKFIH Joint Project 151 do not know-with major difficulties comparison” for Hungarians who want to stay (3.03; CI: 1.34-4.73) and Hungarians in Austria (3.03; CI: 1.34-4.73). In relation to health, a relationship emerges among potential migrants in terms of self-reported better health being associated with less satisfaction with green spaces. The average marginal effect in „very good health-bad health comparison” is (-1.88; CI: -3.71-(-0.05)) (Figure 142). Figure 142. Average marginal effects of perceived health by the Hungarian groups. Source of data: MIGWELL surveys 2024. Own figure. The average estimate for those with a partner is slightly higher (6.96; CI: 6.83-7.09) than for those without a partner (6.85; CI: 6.67-7.04). The average marginal effect is not significant for the whole sample (0.09; CI: -0.16-0.35), driven by the opposite trends between those who want to stay (-0.23; CI: -0.54-0.08) on the one hand and potential migrants (0.74; CI: 0.12-1.35) and Hungarians in Austria (0.57; CI: -0.09-1.2) on the other. The positive effect is significant at the p<0.1 level for Hungarians in Austria, and at the p<0.05 level for potential migrants. The presence of a child is significantly (p<0.1) associated with higher green space satisfaction (0.61; CI: -0.02-1.26) only in the sample of Hungarians in Austria. A relationship between feelings of safety and green space satisfaction is also found. Those who think that the neighbourhood is very unsafe (6.09; CI: 5.4-6.78) have a lower mean estimated satisfaction value, while the highest estimate is observed in the group of those who gave the very safe answer (7.25; CI: 7.08-7.42). According to the average marginal effects, the „fairly safe-very unsafe comparison” (0.72; CI: -0.04-1.5) is significant at the p<0.1 level, while the „very safe-very unsafe comparison” (1.08; CI: 0.3-1.8) is significant at the p<0.05 level. These effects are observed Fair Good Very good Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT Pot. stayers Pot. migrants in AT -4 -2 0 estimate Reference: Bad Strange relationship among pot. stayers FWF–NKFIH Joint Project 152 for those intending to stay, with no association between safety and green space satisfaction for Hungarians in Austria. Also, in the two groups in Hungary, the model shows that noise pollution in the neighbourhood is associated with lower green space satisfaction (-0.892; CI: -1.26-(- 0.52)). The average marginal effect for Hungarians in Austria also points in the same direction (-0.51; CI: -1.23-0.19), but the relationship is not significant. 6.8.2. Extended model for Austria The impact of noise pollution is more closely related to satisfaction with green space in cases where satisfaction was higher before migration. The average estimated satisfaction of those who emigrated to Austria to help establishing a future back in Hungary is lower (6.31; CI: 6.786.85) than those who did not (7.35; CI: 7.13-7.57). The relationship is significant, with a mean marginal effect of -0.88 (CI: -1.61-(-0.15)). This effect is particularly strong for those who were satisfied with green spaces in Hungary. The opposite trend is observed for those who moved out because of a relationship. When this response is marked, the average marginal effect is positive, but not significant (0.57; CI: -0.13-1.28).