ABSTRACT
The paper studies Slovak female elder care workers in Austria. It evaluates the relevance of the care drain metaphor using a representative survey of the carers. It asks two sets of questions: (1) How widespread is the care drain phenomenon? Does migration of elder care workers create care shortages in Slovakia? (2) How do care workers evaluate their employment setting? Is it influenced by experiencing care drain and/or transnational partnering? The higher average age of the care workers means that only a minority has small children in Slovakia. The potential for care drain is therefore limited. Moreover, very few have elderly family members in need of care. On the other hand, most live with partners in Slovakia, so transnational partnering problems could be expected. The average job evaluation of the care workers is highly positive and comparably more enthusiastic than job assessment of women employed in Slovakia. Care workers with small children or partners in Slovakia do not provide a less positive evaluation of the job or its impact on relationships within their families. However, higher earnings and relatively good health of the patient influence the evaluation positively. This focus on earnings and work conditions indicates that care work is probably more similar to other forms of labour migration than anticipated by the current approaches to female migration flows. Indices suggest that these findings could be valid beyond the studied case. A cautious application of the care drain metaphor in intra-EU migration where living standards between countries are not as divergent as between post industrial societies of the West and the Third world is advised.
The topic of migrant elder care workers is typically approached within two distinct research areas. One option is to study migration of care workers as a by-product of welfare policies in the ageing Western societies. Second alternative is to focus on the migrants themselves using insides from studies of female migration. Our approach will use the latter perspective on women from Slovakia working as live-in around the clock care providers for elderly in Austrian households.
Most of the existing academic research on migrant elder care workers in Austria can be subsumed under the ‘policies approach’. The research mostly focused on explaining the demand for migrant care workers as an outcome of the Austrian cash for care scheme introduced in 1993 and the post-2006 legalisation policies (e.g. Gendera 2011; Kaelin 2011; Österle and Bauer 2012). Insights from female migration research, like the lack of care in the families of the migrants (care drain) and difficulties in maintaining a partnership (transnational partnering), have been used less frequently. They have, however, found their way into research of care workers from other Central and Eastern European (CEE) countries in Germany and Italy (e.g. Lutz and Palenga-Möllenbeck 2012; Piperno 2012).
Our question within this study will be if the above mentioned problems expected to be experienced by female migrants and their families also apply to Slovak live-in elder care workers in Austria. Before introducing our approach in detail, we will first provide a short summary of the theoretical background that has shaped research on migration of elder care workers so far.
Female migration, care regimes and care drain
Initially, the analysis of gender in migration research was connected with a critique that traditional (i.e. economic) approaches to migration took no account of female migration, which was considered at best as an accompanying part of male labour migration (see Morokvasic 1984). This late discovery (according to Lutz [2010] rather a re-discovery) of female migration, however, has led to the emergence of a separate ‘female migration school’ – an approach that was built rather around the typical employment opportunities female migrants worked in than on traditional topics of labour migration study. Authors studying female migration often point away from the emphasis on economic rationales (e.g. Lutz 2010) or even explicitly separate it from labour migration studies. For example, Rostgaard et al. (2011) argue that:
The traditional push – pull factors associated with labour migration cannot automatically be applied to the study of care migration, as other important factors may come into play, meaning that care migration might not be reduced simply to a question of unfavourable conditions in one place that ‘push’ people out and favourable conditions in an external locations that ‘pull’ them in. (142)
A similar statement is made by Helma Lutz:
domestic work is not just another labour market … the logic of care work which is clearly different from that of other employment areas … domestic work cannot just be analyzed using the terminology of migration theories following the rationale of a global push-pull model. (2008: 2)
Indeed, while growing feminisation of international migration has been acknowledged by mainstream migration research (e.g. Castles and Miller 2003: 9), the study of female migration remains a specific domain within migration studies.
The above can be demonstrated by the fact that since the occurrence of gender perspective in international migration in the 1990s, the approach to the topic of female migration has been shaped by discussions connected with new concepts. With terms like the global care chains, care drain or transnational motherhood – the focus was on provision of care either at the starting or at the ending point of the migratory movement. Arlie Russel Hochschild, the author of the global care chains metaphor (Hochschild 2000) and of the juxtaposition of ‘brain drain’ and ‘care drain’ (defined in Hochschild 2004) draws a parallel between love (that comes with care provision) and gold that was once sought in the countries of the global south (Hochschild 2004).
Globally, care drain has become a topic of study in the early 2000s with studies conducted in the Philippines (e.g. Parreñas 2000, 2001). Despite being initially developed on studies of female migration from the third world counties, the care drain framework has been later used also for migration of female workers from Central and Eastern Europe. In the European context, the term ‘Euro-orphans’ for children of migrating mothers has been mentioned by several studies (e.g. Lutz 2010; Lutz and Palenga-Möllenbeck 2011, 2012). Care drain has also been studied in the cases of circular elder care migration from Central and Eastern Europe to Germany and Austria (Lutz 2011a) or from Ukraine to Poland and from Poland to Germany (Lutz and Palenga-Möllenbeck 2012). Recently, Flavia Piperno focused on care drain created by the migration of female care workers from Romania to Italy (Piperno 2007, 2012).
Another problem that families of female migrants have to deal with is the occurrence of ‘transnational partnering’ (Zontini 2010). While theoretically affecting both – male and female – partner migration, it has been observed that the effect – similarly as in the case of care drain – is not gender neutral. According to Zontini, couples usually experience more strain when wives are absent (Zontini 2010: 55).
As already mentioned, migration of care workers is seldom discussed without references to welfare policies in the receiving country. The policies mix in the area of care provision for the young and old in Europe has been described as ‘care regimes’ (Bettio and Plantenga 2004). Bettio and Plantenga identified five care regimes (or care strategies) in European countries identifying a special ‘familiaristic’ regime in Southern Europe. They argued that contrary to the case when emphasis is on labour market – i.e. in the Esping-Andersen (1990) typology – when focus is on care regimes, Southern European countries form a distinctive type that relies on care provided within the family. Later, Bettio et al. (2006) made the connection between the familiaristic model and the ‘migrant in the family model’ (Bettio et al.2006: 272). They suggest that the familiaristic care regime of the Mediterranean countries is more prone to migrant care labour: The state acts only as a ‘carer of last resort’ while the relatively generous pensions as well as cash payments for dependent persons make the care services of migrants affordable even for relatively low-income families. Indeed, migrant elder care workers in Italy – typically from Romania – have become an intensively studied topic in recent years (Da Roit 2007; Piperno 2007, 2012; Rostgaard et al.2011; Näre 2013). Italy, with its 840 thousand domestic workers of which 90% are foreign born (Rostgaard et al.2011: 140) has been used for comparison with other care regimes. However, when we look at the broader picture as the monograph edited by Pfau-Effinger and Rostgaard (2011) does, we can conclude that the importance of the immigrant care providers rises in every of the identified care regimes. Williams (2012) speaks of convergence across Europe. This is relevant also for our case. Austria is not part of the Southern European care model – the share of public home as well as residential care is triple compared to Italy (Saraceno and Keck 2010: 686) – yet it has a considerable migrant elder care workforce.1 Österle and Bauer (2012) write about co-existence of immigrant care work with formal home care services in Austria.
The emphasis on policies is strong also in the research of migrant elder care workers in Austria (e.g. Gendera 2007, 2011; Bachinger 2009; Kaelin 2011; Kretschmann and Pilgram 2011; Österle and Bauer 2012). Besides the focus on cash for care policies, the starting point of post-2006 research on migrant care workers in Austria is the around-the-clock home care legalisation policies in 2006–2008. Both policy areas are used to explain the influx of care workers from Central and Eastern Europe into Austria.
We will now, as mentioned in the introduction, focus on our topic from the perspective of studies of female migration.
The research question: is the care-drain metaphor relevant?
A potential problem when trying to measure care drain is its vague definition. When care drain is discussed the usual focus is on ‘small children’ that are left without sufficient care or with a surrogate care provider. We did not find any explicit definition of ‘small children’ in the literature. Based on the examples from the research material it seems that this could be somewhere between 2 to 15 years.
One of the more formal definitions of care drain is provided by Piperno, who defines care drain as:
a reduction in the level of practical care and emotional and educational guidance available to the most vulnerable members of a family (particularly minors and the elderly) due to the emigration of the family members most engaged in the provision of care. (Piperno 2012: 4)
As already noted, literature exploring care drain does not use any explicit age limit to define the ‘minors’ mentioned here. For example, when Hondagneu-Sotelo and Avila (1997) present statistics about how many women in their sample had children, they do not report their age.
Given this ambiguity, for the purposes of our study we decided to use the above stated definition of care drain while understanding ‘minors’ as children younger than 15 years. This boundary seems to be in line with the empirical material in the literature on care drain and also works well in the Slovak context because it is the age when transition from elementary to secondary school occurs and identity cards are issued.
As was noted in the literature review, the concept of care drain occurring in transnational families has been initially developed on female migration from the third world countries. In recent years, it has, however, been applied also to the CEE context. We ask therefore if circular elder care migration of women from CEE countries can be compared to the situation of women from the Philippines of which several have not seen their children for 10 or more years (e.g. Parreñas 2000). Helma Lutz (2011b) for example observed that:
eastern Europeans can operate a self-organized rotation system which enables them to commute rather than having to emigrate, whereas women from Latin America rely on the (usually vain) hope of being able to return home before too long. (11)
So far all reviewed studies focusing on care drain (not only in the European context) relied on small scale qualitative research. Using theoretical, convenience and snowball sampling methods, none aspired to quantify the care-drain phenomenon by providing basic demographic info about the care workers and their children. The very nature of this approach leaves the connections between the basic demographic variables and care drain rather unresearched. In a different context, Madziva and Zontini (2012: 429) state that the literature on global care chains formed by transnational families ‘tend to assume, rather than document, the obstacles to family life caused by separation’. This opens the question if care-drain problems occur in all migration flows dominated by women.
The purpose of our study is to evaluate the relevance of the care-drain metaphor using the case of migration of elder care workers from Slovakia to Austria. We do so by asking two sets of questions: (1) How widespread is the care drain phenomenon in migration to Austria? Does migration of elder care workers create care shortages in Slovakia? What are the consequences for the family? (2) How do migrant care workers evaluate their employment setting? Is this evaluation influenced by experiencing care drain and/or transnational partnering? Compared to previous research, the main advantage of our analysis is the quantitative nature of the data. Contrary to reviewed studies, we were able to obtain a representative sample of Slovak live-in elder care workers in Austria.
We will proceed in our argument in the following way: First, we provide a short overview of the history of migrant elder care workers in Austria. Then we examine the reliability of the available data, introduce our survey of Slovak care workers in Austria and discuss its representativeness. Next, we present descriptive statistics on the demographic composition of the care workers and their care obligations in Slovakia. We then look at job evaluation of care workers. We establish if care workers with care obligations and partners in Slovakia provide a less positive evaluation of their work than their counterparts without caring responsibilities and partners. A conclusion follows.
Migrant elder care workers in Austria and available data
Historically, there are three important milestones with regard to migrant elder care in Austria. First is the fall of the Iron curtain in 1989 that enabled visa free travel between Austria and its neighbours from the former Soviet Block. Second is the introduction of the Pflegegeld cash for care scheme in 1993, which strengthened the ability of the frail elderly to pay for care in their households (in 2011 the monthly amount of Pflegegeld was – depending on the level of dependency – between 154 and 1656 Euro). Third is the 2006–2008 period of gradual legalisation of around the clock care provided by foreign carers.
We have only sketchy knowledge about the beginnings of this cross border employment in the early 1990s. According to Austrian sources they are connected with Czech carers and care work agencies acting as charities (Bachinger 2009). It seems, however, that the source of care providers has later experienced a gradual shift from the Czech to the Slovak part of former Czechoslovakia. This could have been caused by higher unemployment and lower wages in Slovakia compared to the Czech Republic. Austrian data on legally employed carers from November 2011 show a clear dominance of Slovak carers among the elder care workers in Austria (69%) and a marginal role of carers from the Czech republic (1% among the carers in Vienna) (Richter 2011).
The legalisation of care work in Austria is important also with regard to available data about the – until then – semi or illegal work arrangement. Since 2008 the level of subsidies for the employment of around-the-clock care workers makes their legal employment even more convenient than the illegal variant (Bachinger 2009). We therefore consider the official figure of 40,000 care workers employed in Austria in November 2011 a relatively exact estimate (Richter 2011).
The legalisation in Austria has helped also in providing a clearer picture about the elder care workers in the Slovak Labour Force Survey (LFS). The usual commuting regime of the care workers consisting of two to three weeks of work in Austria followed by stays of the same length at home provides good potential for being captured by the LFS. We believe that even when the household is contacted by the survey while the care worker is in Austria, the respondents will inform the survey about their mothers/wives.2 However, until the legalisation of care work in Austria, doubts existed about the reliability of reporting (il)legal employment in the LFS. The legalisation could be therefore one of the reasons of the steep increase of care workers count in the LFS between 2007 and 2010. Fortunately, after 2010 the Austrian and Slovak data sources provide rather compatible figures on the numbers of care workers. This is important because the existence of reliable data from the LFS is crucial with regard to the assessment of the representativeness of the cAreworkers 2011 survey on which our analysis will be based.
The cAreworkers 2011 survey was carried out within VEGA grant no. 2/0115/11 and included almost 60 questions focusing on the work situation, tasks carried out and overall satisfaction in Austria as well as questions on prior employment, care obligations and family background in Slovakia. One of the explicit aims of the survey was to measure the impacts of the absence of the care workers on their families. The survey fieldwork was conducted in Slovakia in November 2011 by a professional research agency. Interviewers from the fieldwork agency's network were used to find and interview female care workers currently caring for seniors in Austria. The final sample includes 151 care workers.3 Although initially not planned as a representative survey, a comparison of the care workers in the survey with the profile of female care workers working in Austria in the Slovak LFS shows that there are no statistically significant differences between the two samples in age, education and regional background. We should note that with regard to the Slovak LFS, the care workers working in Austria are not a ‘hidden population’ as they might be in a survey fielded in Austria. Most of them spent half of the month in Slovakia and work in Austria legally (for a detailed discussion of the issue see Bahna 2014). We therefore consider the cAreworkers 2011 survey as representative for the population of Slovak female care workers in Austria.4 However, the size of the cAreworkers 2011 survey is significantly smaller than the usual size of a representative survey, which means that the presented descriptive results will have wider confidence intervals.
In the next part of our study we will use the cAreworkers 2011 survey (later referred to as ‘survey’) to answer the two sets of our research questions.
Does care drain occur when Slovak women care for elderly Austrians?
The average age of care workers in the survey is 47 years with a median age of 48 years. While 78% of them have children, among mothers only 17% have a child below 15 years of age and only 6% have a child below 6 years of age. So while 78% of the care workers have a child, only 13% have small children according to our definition. The most usual commuting regime – two weeks in Austria followed by two weeks at home – is employed by 74% of them. Only about 2% of the care workers stay in Austria for more than three weeks at a time.5
The relatively high average age of Slovak care workers in Austria creates a predisposition for care drain regarding their elderly family members. What if care for the elderly in Austria results in less care for the frail elderly in the families of the care workers in Slovakia? This does not seem to be the case. A vast majority (96%) of the care workers does not have a family member who needs special assistance or care similar to the care provided in Austria.
These results seem to indicate that the majority of the Slovak carers in Austria represent a particular group of women who already have less care responsibilities towards their grown up children yet still little care obligations towards their ageing parents. However, there still seems to be a relevant minority of care workers with young children in Slovakia. It is this group that could, despite the employed regular commuting regime, experience problems with care provision for their children.
Another group of care workers who could face problems caused by their migration are care workers that are married or are living with a partner in Slovakia. According to Zontini (2010), the situation when women are absent in the family creates more strain in the partnerships than the migration of men. According to the survey, 59% of the care workers were living in a common household with their partner at the time they decided to start working in Austria, 7% did not share a household with their partner and 33% did not have a partner. At the time of interview, 7% of the care workers reported that their family situation has changed since they started working in Austria. Among them five have been divorced/separated from their partner, three found a new partner, two have moved in with their partners and one partner died. While there seem to be some changes in the family situation, it is not clear in which direction the results point. Moreover, the sample size of the survey does not allow us to go here into further analysis.
We will now shift our focus to the evaluation of the employment as a care worker and to its impact on the care worker's family in Slovakia. The aim is to analyse if the overall evaluation is less positive in the groups of care workers that are at ‘risk’ of experiencing care drain or strains in their partnership.
Table 1 shows the overall evaluation. As we see, it is extremely positive – only 6% of the carers opted for one of the two ‘unsatisfied’ answers. While this may seem surprising in the context of our approach it is in line with findings of a high job satisfaction among migrant care workers (Iecovich 2011).
| . | Very satisfied . | Satisfied . | Neither satisfied nor unsatisfied . | Unsatisfied . | Very unsatisfied . | Total . |
|---|---|---|---|---|---|---|
| Has children aged 0–14 years | 37 | 53 | 11 | 0 | 0 | 100 |
| Lived with a partner when started work in Austria | 24 | 53 | 18 | 2 | 3 | 100 |
| All care workers | 20 | 58 | 16 | 4 | 2 | 100 |
| . | Very satisfied . | Satisfied . | Neither satisfied nor unsatisfied . | Unsatisfied . | Very unsatisfied . | Total . |
|---|---|---|---|---|---|---|
| Has children aged 0–14 years | 37 | 53 | 11 | 0 | 0 | 100 |
| Lived with a partner when started work in Austria | 24 | 53 | 18 | 2 | 3 | 100 |
| All care workers | 20 | 58 | 16 | 4 | 2 | 100 |
Note: Exact question wording: ‘All in all, would you say that you are with your work:’.
Source: cAreworkers 2011 survey.
How does this highly positive job evaluation of the care workers compare to job satisfaction of women of the same age profile employed in Slovakia? We used results from a similar question about work satisfaction asked in the ISSP Slovakia 2009 survey as a benchmark. To be able to compare, we weighted the answers of female respondents to match the age distribution of the carers. The interesting result is that the overall satisfaction with their job is significantly higher among the carers than among women of comparable age employed in Slovakia. Despite the slightly different question wording used in both surveys, this suggests that the job evaluation of Slovak carers is indeed very positive.
With regard to our argument, two groups of care workers are of particular interest: those with small children and care workers living with a partner in Slovakia. If we look at the former group we see that they provide a statistically significant more positive evaluation than the whole sample.6 On the other hand, there is no significant difference between the evaluation among all care workers and those living with a partner.
Table 2 offers another set of unexpected results. When asked about the influence of care work on the relationships in their family in Slovakia, the views are again overwhelmingly positive. Only 9% assess the influence as ‘rather’ or ‘definitely’ negative. And again – although this time the difference is not statistically significant – carers with small children are even more positive about the influence. Similarly, as in the previous case, the answers of the carers who live with a partner do not differ from the whole sample.
| . | The influence is definitely positive . | The influence is rather positive . | The influence is neither positive nor negative . | The influence is rather negative . | The influence is definitely negative . | Total . |
|---|---|---|---|---|---|---|
| Has children aged 0–14 years | 33 | 39 | 28 | 0 | 0 | 100 |
| Lived with a partner when started work in Austria | 18 | 35 | 38 | 5 | 5 | 100 |
| All care workers | 21 | 36 | 34 | 6 | 3 | 100 |
| . | The influence is definitely positive . | The influence is rather positive . | The influence is neither positive nor negative . | The influence is rather negative . | The influence is definitely negative . | Total . |
|---|---|---|---|---|---|---|
| Has children aged 0–14 years | 33 | 39 | 28 | 0 | 0 | 100 |
| Lived with a partner when started work in Austria | 18 | 35 | 38 | 5 | 5 | 100 |
| All care workers | 21 | 36 | 34 | 6 | 3 | 100 |
Note: Exact question wording: ‘All in all, would you say that your work influences the relations in your family’?Source: cAreworkers 2011 survey.
Those unexpectedly positive results call for a deeper analysis. We therefore look into which explanatory variables are behind the positive evaluation of care work and the positive assessment of the influence of care work on family relations. We present multivariate analysis for our two dependent variables – the overall evaluation of care work (Table 3) and its influence on the relations within the family (Table 4).
| . | OLS regression . | Logistic regression . | ||
|---|---|---|---|---|
| . | Beta . | Sig. . | Exp(B) . | Sig. . |
| Age | 0.04 | 0.65 | 0.98 | 0.49 |
| Education | −0.03 | 0.77 | 0.79 | 0.67 |
| Unemployment in region of origin | −0.04 | 0.63 | 1.02 | 0.63 |
| Was previously unemployed | −0.16 | 0.11 | 0.49 | 0.22 |
| Has children 0–14 years | 0.12 | 0.19 | 1.47 | 0.75 |
| Has a partner | −0.02 | 0.78 | 0.40 | 0.14 |
| Lives in east Slovakia, works in west Austria | 0.01 | 0.90 | ||
| Stays for more than two weeks | 0.03 | 0.77 | 3.07 | 0.18 |
| Works in Vienna | 0.22 | 0.01 | 7.57 | 0.00 |
| Income | 0.22 | 0.02 | 3.23 | 0.00 |
| Health of patient | 0.18 | 0.04 | 2.02 | 0.03 |
| Working time | −0.08 | 0.35 | 0.98 | 0.77 |
| R2 | 0.226 | |||
| R2 adj. | 0.143 | |||
| Nagelkerke R2 | 0.393 | |||
| N | 125 | 125 | ||
| . | OLS regression . | Logistic regression . | ||
|---|---|---|---|---|
| . | Beta . | Sig. . | Exp(B) . | Sig. . |
| Age | 0.04 | 0.65 | 0.98 | 0.49 |
| Education | −0.03 | 0.77 | 0.79 | 0.67 |
| Unemployment in region of origin | −0.04 | 0.63 | 1.02 | 0.63 |
| Was previously unemployed | −0.16 | 0.11 | 0.49 | 0.22 |
| Has children 0–14 years | 0.12 | 0.19 | 1.47 | 0.75 |
| Has a partner | −0.02 | 0.78 | 0.40 | 0.14 |
| Lives in east Slovakia, works in west Austria | 0.01 | 0.90 | ||
| Stays for more than two weeks | 0.03 | 0.77 | 3.07 | 0.18 |
| Works in Vienna | 0.22 | 0.01 | 7.57 | 0.00 |
| Income | 0.22 | 0.02 | 3.23 | 0.00 |
| Health of patient | 0.18 | 0.04 | 2.02 | 0.03 |
| Working time | −0.08 | 0.35 | 0.98 | 0.77 |
| R2 | 0.226 | |||
| R2 adj. | 0.143 | |||
| Nagelkerke R2 | 0.393 | |||
| N | 125 | 125 | ||
Note: Dependent variable in logistic regression is coded 1 for ‘very satisfied’ and ‘satisfied’ answers and 0 for all other answers. Reference category for the dependent variable is 0.
Source: cAreworkers 2011 survey.
| . | OLS regression . | Logistic regression . | ||
|---|---|---|---|---|
| . | Beta . | Sig. . | Exp (B) . | Sig. . |
| Age | 0.06 | 0.54 | 1.00 | 0.86 |
| Education | −0.06 | 0.51 | 0.74 | 0.44 |
| Unemployment in region of origin | −0.02 | 0.80 | 1.00 | 0.96 |
| Was previously unemployed | 0.07 | 0.49 | 1.06 | 0.90 |
| Has children 0–14 years | 0.14 | 0.16 | 1.64 | 0.49 |
| Has a partner | −0.19 | 0.04 | 0.55 | 0.17 |
| Lives in east Slovakia, works in west Austria | −0.05 | 0.60 | ||
| Stays for more than two weeks | −0.01 | 0.93 | 0.82 | 0.71 |
| Works in Vienna | 0.01 | 0.91 | 0.99 | 0.97 |
| Income | 0.15 | 0.13 | 1.42 | 0.16 |
| Health of patient | 0.30 | 0.00 | 1.89 | 0.02 |
| Working time | 0.02 | 0.86 | 1.03 | 0.54 |
| R2 | 0.190 | |||
| R2 adj. | 0.101 | |||
| Nagelkerke R2 | 0.136 | |||
| N | 123 | 125 | ||
| . | OLS regression . | Logistic regression . | ||
|---|---|---|---|---|
| . | Beta . | Sig. . | Exp (B) . | Sig. . |
| Age | 0.06 | 0.54 | 1.00 | 0.86 |
| Education | −0.06 | 0.51 | 0.74 | 0.44 |
| Unemployment in region of origin | −0.02 | 0.80 | 1.00 | 0.96 |
| Was previously unemployed | 0.07 | 0.49 | 1.06 | 0.90 |
| Has children 0–14 years | 0.14 | 0.16 | 1.64 | 0.49 |
| Has a partner | −0.19 | 0.04 | 0.55 | 0.17 |
| Lives in east Slovakia, works in west Austria | −0.05 | 0.60 | ||
| Stays for more than two weeks | −0.01 | 0.93 | 0.82 | 0.71 |
| Works in Vienna | 0.01 | 0.91 | 0.99 | 0.97 |
| Income | 0.15 | 0.13 | 1.42 | 0.16 |
| Health of patient | 0.30 | 0.00 | 1.89 | 0.02 |
| Working time | 0.02 | 0.86 | 1.03 | 0.54 |
| R2 | 0.190 | |||
| R2 adj. | 0.101 | |||
| Nagelkerke R2 | 0.136 | |||
| N | 123 | 125 | ||
Note: Dependent variable in logistic regression is coded 1 for ‘the influence is definitely positive’ and ‘the influence is rather positive’ answers and 0 for all other answers. Reference category for the dependent variable is 0.
Source: cAreworkers 2011 survey.
Both multivariate models presented in Tables 3 and 4 include four theoretical variables (having children below 15 years of age, having a partner in Slovakia, travelling long distance, staying for more than two weeks) and a set of control variables. The control variables include age, education, regional unemployment level and personal experience with unemployment. The last two variables serve as a proxy of the economic situation of the carer. Control variables further include income from care work, health status of the patient, overall working time and working in Vienna.
The dependent variables in both models have only five values which violates the strict data assumptions of the ordinary least squares (OLS) regression. We therefore decided to present our models using both OLS and logistic regression. The second method uses a dichotomised version of the dependent variables and is included as a ‘reliability check’ of the OLS regression results. With one exception both methods agree on the significance and direction of the connection between the independent and dependent variables.
If we look at the overall evaluation of care work in Table 3 we see that none of our theoretical variables has an influence. Contrary to the descriptive results, in the multivariate setting carers with small children do not evaluate their work more positively. However, even in this case, there is no sign of the expected negative impact of a care-drain experience. Similarly, carers living with a partner in Slovakia do not evaluate their work differently than the rest. Travelling almost thousand kilometres from eastern Slovakia to west Austria or staying for more than two weeks does not have an impact either.
There is also no connection with age, education, having been unemployed or the unemployment situation in the region of origin. There is, however, a clear connection between a high income level and a positive job evaluation. This is an obvious connection if we think of elder care as labour migration. It has, however, been made less explicit by the prevalent discourses in female migration. Another important explanation variable is the health situation of the patient. Carers caring for severely disabled or ill patients seem to be less satisfied with their jobs. The overall working time is not a significant predictor of care work evaluation. This makes sense if we think of the live-in arrangement of the care workers. Here perhaps less emphasis is on the amount of time spent with the patient himself – it is rather the amount of effort necessary in the daily caring routine that influences the overall evaluation. Working in Vienna is the last independent variable with a significant positive relation to the overall evaluation of care work. The inclusion of this variable is the result of our effort to include a measure of the rural – urban workplace continuum. For example, Bauer and Österle (2013: 469) found that feelings of loneliness were more pronounced among carers working in remote rural areas. However, in our case the models that included a rural–urban scale variable did not show a connection between settlement size and work evaluation of the carers. It seems therefore that it is working in Vienna that specifically results in a more positive evaluation. The exact nature of this connection remains unclear. Possible explanations include more leisure time activities and social contacts in the big city or better travel possibilities from and to Slovakia making the carers less dependent on ‘taxi’ services organised by the intermediaries.
The second model with influence of care work on relations within the family as the dependent variable (Table 4) offers less explanatory power (both in the OLS and in the logistic regression variant) than the first model. Again, none of the basic demographic variables is related to the evaluation of the influence of care work on family relationships. A similar statement can be made about our theoretical variables – with one notable exception. In the OLS results, having a partner seems to lead to a more critical assessment of the influence. This connection is, however, on the edge of statistical significance and is not confirmed in the logistic regression results.
Of the three variables that were significant in the previous model, only one remains significant in Table 4. The connection with income has vanished and so has working in Vienna. Care workers who care for a patient in relatively good health see the influence of their work on their family relationships as more positive. This probably indicates that care workers who care for a more independent patient do not return home to Slovakia exhausted and perhaps even have more time to communicate with their family while in Austria. Again, having a ‘healthy’ patient was more important than the actual working hours.
Both models offer little support for the importance of care drain and transnational partnering problems in the lives of Slovak care workers in Austria. Not only they are, on the average, extremely satisfied with their job, but when controlling for basic demographics, the group potentially experiencing care drain – care workers with children below 15 years – do not evaluate their job less enthusiastically than the rest. The results from the second model offer room for assumption that care workers living with a partner in Slovakia perceive the potential strain generated by their absence. However, looking at the descriptive results (Table 2), this means that care worker living with a partner prefer to say that the job does not have influence on their relationships while the rest holds the opinion that their work improves relationships in the family. The variables that influence the job evaluation (income, health of patient) bring us to the very beginnings of the female migration conceptualisation. It seems that at least in the studied case, the ‘paid work’ dimension of care work is far more important than the factors discussed within female migration and domestic work debates.
We could argue that Slovak care workers commuting between neighbouring Austria and Slovakia are a special case and therefore are not affected by problems experienced by transnational families. This may be true; however, even care workers commuting in longer than two week cycles and those who travel more than 800 km do not answer differently. Our findings may indeed indicate that – despite the attention care drain has received in academic writing – it is less important for the migrants. For example, Iecovich's (2011) analysis of job satisfaction of Philippine care workers in Israel (which was not aimed at exploring care drain) found also no connection between job satisfaction and marital status of the carers. On the other hand, one of the more speculative interpretations of our findings could be that they speak about successes of transnational mothering.
One might object that our approach cannot provide a satisfactory answer to all of our research questions as our research provides only the perspective of the carers. Research on care drain and transnational families has found that the views of the mothers and their children may differ. When mothers perceived a satisfactory mothering experience some children found such mothering from afar intrusive or unsatisfactory (Parreñas 2005; Madianou and Miller 2011, 2012).
Our answer to such objection is two-fold. First, it does not question our findings regarding the low share of women with small children and elderly dependent relatives among the care workers. We can therefore state that only a fraction of the care workers can be experiencing care drain problems in our understanding of care drain. Second, the survey included also a question regarding receiving support for the decision to become a care worker. Sixty-four percent of the carers reported receiving support from their family members, no support was reported by 19% and 17% reported that their decision was supported by acquaintances who were already working in Austria. Among care workers living with a partner, 47% answered that their partner supported them in the decision to take up care work in Austria. Support from their partners, children or from other family members was indicated by 69%. While this does not prove that the views of the carers and their families always align, it suggests a basic compatibility of the evaluation within the family.
Conclusion
Our study investigated the case of Slovak live-in around-the-clock elder care workers in Austria. Contrary to similar studies that used qualitative methods, our results are based on a representative sample of care workers.
The main research question of our study was if and to what extent they experience the problems transnational families and female migrants are expected to encounter. We first checked if there was a potential for experiencing care drain or transnational partnering. We found that there are relatively few women among Slovak care workers who are ‘at risk’ according to the care drain definition. Only 13% have children younger than 15 years and 96% do not have care responsibilities towards their frail family members. On the other hand, most of them live with a partner, so problems in partnerships could be expected. However, when asked about the overall evaluation of their employment as well as its impacts on relations within their families, care workers with small children or partners in Slovakia did not provide a less enthusiastic evaluation than the rest.
While the higher average age of the care workers explains the low share of mothers with young children, their overwhelmingly positive evaluation of care work is rather surprising in the context of studies of female migration. On the other hand, we were able to find indices that carers living with a partner in Slovakia evaluate the impact of their work on the relationships within their families less positively. This, however, means that – contrary to the generally perceived positive influence – they prefer to say that their work does not impact relationships in their family.
These findings open the question if the circular migration regime employed by Slovak care workers in Austria is not a special case creating a rare win-win situation for the care workers and their employers. This could be well the case. However, at least in the Slovak-Austrian setting even the carers staying for longer periods and travelling long distances do not offer a less positive evaluation. So, perhaps, our conclusions have some validity also beyond the studied case. Previous research has found that care workers from Eastern Europe manage to maintain family ties more easily than those from Latin America (Lutz 2011b). Sekulová (2012, 2013), for example, explored the mechanisms Slovak care workers employ to provide care for their family even during their absence. However, even research on job satisfaction of elder care workers from the Philippines in Israel did not find a link between family situation and job satisfaction (Iecovich 2011).
The results seem none the less counter intuitive without mentioning the variables that had an influence on the generally very positive evaluation of care work by the Slovak carers. Care workers with higher earnings and those with healthier patients were among the most content. The importance of the economic dimension is demonstrated also in the answers to the open-ended question if the family situation of the care workers has changed while doing care work. Several of the care workers who did not experience a change in their family situation answered with a ‘yes’ and stated that their family is better off now – that they finally live decently.7 Such answers support our findings that it is not having small children or a partner in Slovakia that determines job evaluation of the care workers, but rather the income that this job generates. While perhaps not being the primary expectation in female migration research, the finding is very intuitive if we treat care workers as labour migrants. The discussed results suggest that this ‘classical’ perspective should receive more consideration also in studies of female migration.
Our conclusions, however, should not be interpreted as contesting the existence of care or relationship problems that are caused by longer absences of the care workers in their own families. We are aware that our quantitative approach can not grasp the full complexity of what satisfaction with their work situation may mean for them. A combination of qualitative and quantitative techniques would be necessary to fully cover the issue in the researched context. Still, our results advise for a more cautious application of the care drain metaphor as it seems not to be the formative experience in all female migration flows. Special caution should be applied in cases when the living standards in the source and destination countries are not as widely divergent as between post-industrial societies of the West and the Third world. Such is the case of most intra-EU migration.
Acknowledgements
The research was supported by APVV grant no. 0309-11 and by the Housing and Home-care for the Elderly and Local Partnerships Strategies in Central European Cities (HELPS) project financed by the European Regional Development Fund within the Central Europe programme. The author would like to thank Martina Sekulová for her valuable comments and help with the cAreworkers 2011 survey questionnaire. An earlier version of the paper was presented at the 11th ESA conference in Turin, Italy.
Disclosure statement
No potential conflict of interest was reported by the authors.
Footnotes
If we adjust for Italy's higher population (factor 7.5) and share of people aged 65 years and more (20.8% and 18.9% in 2013), the 40,000 legal care workers in Austria equal 330,000 care workers in Italy.
For a more detailed discussion about measuring migration with LFS see Bahna (2013).
The data file with documentation and questionnaire from the survey is available in the Slovak Archive of Social Data (http://sasd.sav.sk).
The (statistically insignificant) higher average age of care workers in our survey can be explained by the hypothesis that Slovak care workers that draw a pension classify themselves as pensioners and not as care workers in the LFS. This is supported by the fact that respondents who stated their status prior to care work as 'retired' are over-represented among carers with an 'informal' employment status.
Further descriptive results from the survey can be found in Bahna (2014).
Later, in our multivariate model, we will test if this difference holds when controlling for their lower age.
Forty-two percent of the care workers were unemployed in Slovakia before they became care workers in Austria. The average income of the carers in our survey was around 770 euros while the average wages in the eight Slovak NUTS 3 regions ranged from 680 to 1157 euros in 2011.
References
Miloslav Bahna is a researcher at the Institute for Sociology, Slovak Academy of Sciences. He wrote a monograph about post EU enlargement migration from Slovakia and contributed to The Encyclopedia of Global Human Migration. He published in International Migration Review, European Societies, Journal of Contemporary European Studies and other journals.