The labor market integration of immigrant women in Europe: context, theory, and evidence

In this overview, we seek to provide a comprehensive resource for scholars of female immigrant labor market integration in Europe, to act both as a reference and a roadmap for future studies in this domain. We begin by presenting a contextual history of immigration to and within Europe since the Second World War, before outlining the major theoretical assumptions about immigrant women’s labor market disadvantage. We then synthesize the empirical findings from quantitative studies published between 2000 and 2020 and analyze how they line up with the theoretical predictions. We supplement the review with descriptive analyses using data from 2019, which expose any discrepancies between the current situation in European countries and the situation during the time periods considered in the reviewed studies. Our review has three main take-aways. First, the theoretically relevant determinants of immigrant women’s labor market integration are generally supported by empirical evidence, but the unexplained heterogeneity that remains in many cases between immigrant women and other groups on the labor market calls for more systematic and comprehensive investigations. Second, quantitative studies which take a holistic approach to studying the labor market disadvantages of immigrant women—and all the considerations related to their gender and nativity that this entails—are rare in this body of literature, and future studies should address this. Third, fruitful avenues for future contributions to this field include expanding on certain overlooked outcomes, like immigrant women’s self-employment, as well as geographic regions that until now have received little attention, especially by employing the most recent data.


Introduction
In labor market research, both in economics and sociology, the economic integration of male immigrants in their new country of residence has always held a prominent position. Immigrant women's labor market integration, on the other hand, has rather been seen as "subservient" to the activities of their male counterparts (Boyle et al. 2001, p. 201), a perception that has had repercussions on immigrant women's place in the field of research. This early perception of immigrant women exclusively as invisible dependents meant not only neglecting them in research, but also inhibited opportunities to adopt a gender-specific perspective on the labor market performance of immigrants (Kofman et al. 2005;Mushaben 2009). Even for researchers who included women in their analyses, gender was often just treated as one of many covariates in the estimated model. In the case where only women were investigated, researchers applied the models to them that initially had been derived for men, ignoring that the lives of immigrant women (including their labor market integration) were likely to differ substantially from those of their male counterparts (Kofman et al. 2005). Research has become much more intersectional in the past two decades, but the investigation of the particular labor market integration patterns of female immigrants remains a work in progress.
Considering both gender and nativity, immigrant women often occupy the most disadvantaged spot on European labor markets. The sparse research on the determinants of these disadvantages is hardly understandable, especially considering the European demography. In 2018 alone, 2.4 million immigrants arrived in the EU from a non-EU member country, with an additional 1.4 million EU citizens moving from one EU country to another (Eurostat 2020a). Females account for 46 percent of the total. Clearly, investigating the labor market performance of immigrant women is not only of macroeconomic importance, particularly in light of Europe's demographic change (d'Albis et al. 2016;Fuchs and Kubis 2016;Barslund et al. 2017), but also addresses the fundamental question of ethnic and gender equality. Emancipative instruments might work very differently for immigrants, taking into account specific gender dynamics arising from the direct experience of immigration and integration as well as cultural differences (Kofman et al. 2005;Mushaben 2009). Understanding the causes as well as mitigating the detrimental consequences of the labor market disadvantages of immigrant women could substantially improve their quality of life, for instance by helping their family establish greater economic independence.
Thus, the aim of this paper is to provide a review of the quantitative research on first-generation female immigrant labor market outcomes across Europe generated between 2000 and 2020. We have three main objectives: The first objective is to contextualize female immigrant labor market integration in Europe both geo-politically and theoretically. The second objective is to summarize the recent findings on immigrant women's labor market integration patterns, identifying broad trends in their outcomes across Europe and connecting the theoretical determinants of their labor market integration to actual empirical evidence. The final objective of our review is to evaluate the intersectionality of the disadvantages that immigrant women face-based on their nativity and gender-and to highlight how different groups of immigrant women in different contexts occupy unique positions on the labor market, meriting appropriate research approaches. 1 We enhance our review of the existing literature by including descriptive analyses using the most current data.

General framework and focus of this review
The comparability of migration-related research rests on a clear understanding of the terms used and their applicability to the presented results. In this review, we adopt the working definition of immigrants as individuals who are living in a country in which they were not born (irrespective of the citizenship they hold). Given their foreign country of birth, these first-generation immigrants differ in their nativity and immigration experience from both the native population as well as secondgeneration immigrants, who were born in the host country. 2 In sociology, the term integration is a descriptive one, typically referring to the cohesion of a larger entity (Esser 2001). However, in practice, integration is often equated with immigrants' assimilation into the static host society of a nation state. Scholars like Hartmut Esser emphasize that assimilation only refers to a convergence and is not necessarily one-sided, but the terms integration and assimilation nonetheless often carry a normative character (Canan 2015). We acknowledge this debate, but since immigrant outcomes cannot be evaluated without a reference group (especially when employing quantitative methods), we adopt a concept of socioeconomic integration understood as the assimilation of immigrant labor market outcomes to those of natives for the purposes of this review. Most of the quantitative studies in the social sciences either implicitly or explicitly consider the (structural) integration 3 into the labor market to be successful when immigrants do not suffer any (additional) disadvantages vis-a-vis natives that are comparable in terms of their individual characteristics. The same logic applies to the comparison between women and men or any other groups.
As previously stated, research on immigrant women's labor market integration necessitates a well thought-out approach, since disadvantages by nativity and gender both need to be considered. Moreover, the disadvantages related to these dimensions can arise at multiple stages along the immigration and integration trajectory. We therefore present the contextual and theoretical underpinnings of these disadvantages below. In the "Empirical findings" section, we structure our review of the empirical findings by the outcome studied. We start with studies of labor force participation and then consider a comprehensive set of indicators in terms of labor market success, including (un-)employment rates, working hours/part-time work, contract type, occupational status, and earnings/wages.
Based on the definitions of 'immigrant' and 'integration' and given the set of labor market indicators to be considered, the scope of this review extends to all quantitative studies of first-generation female immigrant labor market integration in a European country, published in English between 2000 and early 2020. Although we look at female labor market integration after migration, this does not mean that we only consider female labor immigrants. Instead, we focus on the labor market outcomes of immigrant women irrespective of their individual migration paths or motives (including refugees). The studies included in this review were found via extensive database and web searches, as well as bibliographic searches within these initial articles.

A brief history of European immigration since 1950
There has been a long tradition of migration from, to, and across the European continent, but compared to other countries such as the U.S., Canada, or Australia, large-scale immigration and its political recognition are relatively recent phenomena for European countries, which started after the end of the Second World War (Dustmann and Frattini 2011). Migration experiences have been quite heterogeneous across countries within this period owing to historical, political, and economic differences, resulting in immigrant populations that differ considerably in their origins and composition. These aspects are important to consider in a review of labor market outcomes, since the characteristics of a country's immigration policy and the coethnic community that an immigrant encounters after immigration play an important role in their labor market integration. Thus, we provide a brief history of migration to and within Europe, grouping countries into three main regions (north-western Europe (NWE), southern Europe (SE), and central and eastern Europe (CEE)) with similar experiences. 4 While refugee movements were an important component of immigration flows across Europe after the Second World War (Berlinghoff 2018), immigration in the ensuing decades was particularly characterized by bilateral foreign labor recruitment ("guest worker") schemes between NWE-the 'receiving' countries-and SE (and also countries on the periphery of Europe)-the 'sending' countries. These labor inflows to fill low-wage positions were predominantly male, coming from poor agricultural areas with high unemployment (Van Mol and de Valk 2016), though females destined for jobs in the garment and electronics sectors also made up a considerable share (Mattes 2005). Coupled with independence movements in overseas colonies, this resulted in substantial increases in the foreign-born population of NWE countries between 1950 and 1970. For countries in CEE, on the other hand, their isolation behind the Iron Curtain hampered any sizable in-or outflows.
Immigration flows changed course in the early 1970s. Due to a global economic downturn, the formal "guest worker" schemes were largely terminated by the receiving countries. Nonetheless, many of these foreign workers remained in their host country and decided to bring their families to join them (Hansen 2003). Family reunification became one of the defining features of immigrant inflows in NWE from the 1970s to the 1990s, and these countries started grappling with immigrant integration as a political priority (to different extents). SE countries, on the other hand, gradually transformed from 'sending' to 'receiving' countries themselves, absorbing some of the labor immigration deflected from NWE to fill the demands for low-wage labor in construction, agriculture, and care services. However, the absence of recruitment schemes in this case and the increasing policy harmonization as European political integration progressed meant that many of these labor migrants (mostly coming from North Africa and, later, CEE) had irregular legal status, prompting several collective regularizations over time (Doomernik and Bruquetas-Callejo 2016;Geddes and Scholten 2016). Again, there was little movement into or out of CEE for the majority of this period, besides the return migration of some ethnic German populations to West Germany (Doomernik and Bruquetas-Callejo 2016). Yet flows of asylum seekers skyrocketed out of this region with the fall of the Iron Curtain, and the conflict in the Balkans in the early 1990s also resulted in considerable flows of refugees mainly to NWE (Hansen 2003).
Since the 1990s, one of the biggest changes to the European immigration context has been the introduction of the universal freedom of movement for EU citizens, especially as membership has expanded. Fewer legal restrictions have facilitated the mobility of migrants with various skill profiles, especially from CEE to countries in NWE and SE, although these migration patterns are increasingly temporary or circular. For non-EU citizens, labor immigration channels like the Blue Card scheme and student migration highlight the EU's focus on the global competition for talent. These channels are nonetheless fairly restrictive, so that family migration remains one of the most popular pathways for immigration from outside of the EU (de Haas ern Europe) or North Macedonia and Albania (as southern Europe), as typically used, for example, by the UN (see, e.g., UNDESA 2020).
Footnote 4 (continued) et al. 2020). International conflicts during this period have also resulted in a regular stream of asylum applicants to countries across Europe, for example, following the Kosovo War in 1998/99, the war in Afghanistan beginning in 2001, the Gulf War II in Iraq in 2003, and, most notably, following the Syrian civil war, when over 1 million refugees entered the EU in 2015 (de Haas et al. 2020).
In sum, migration flows to Europe have become much more diverse since the Second World War, reflecting changes in immigration policies as well as general global geopolitical trends. The classic channels of labor migration, family reunification, and humanitarian migration remain the most important, but additional migration motives such as education are increasingly relevant (King 2002). In terms of immigration policy, the EU has had a large influence, but the nation states still hold considerable sovereignty over many issues. As of 2019, the share of the foreign-born population among NWE countries ranges between 7 percent in Finland and 48 percent in Luxembourg, most countries having shares of 10-20 percent (see Fig. 1). The largest ethnic communities originate from, for example, India, Poland, and Pakistan in the United Kingdom; Morocco, Algeria, and Portugal in France; Turkey, Suriname and Morocco in the Netherlands; Poland, Turkey, and Russia in Germany; and Syria, Finland, and Iraq in Sweden (see Fig. A2 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/). In all SE countries except Portugal, the share of the foreign-born population exceeds 10 percent, clearly reflecting the substantial inflows since the 1990s. The main origins are Albania, Germany, and Georgia in Greece; Romania, Albania, and Morocco in Italy; Angola, Brazil, and France in Portugal; and Morocco, Romania, and Ecuador in Spain. In CEE countries, immigrant numbers are typically lower in absolute and relative terms, foreign-born population shares range between 2 percent in Poland and 14 percent in Estonia, with most immigrants coming from neighboring countries.
Understanding the (recent) history of immigration in Europe is one essential component for contextualizing the research on female immigrant integration. Immigration policies have changed substantially over the years, and these changes have had an important effect on the composition of the immigrant groups that arrived, including their skill level, country of origin, and most importantly here, their gender. In terms of the gender composition of the immigrant populations in 2019, it looks fairly balanced across European countries (see Fig. 1). However, it is important to acknowledge that stock statistics hide the gendered nature of specific migration flows over the years. For example, whereas the labor recruitment of NWE countries favored men, a larger proportion of the following family migration was made up of women. Another example is the gendered demand for low-wage immigrant labor in SE countries during the recession after 2008. Whereas (male) construction work was virtually non-existent in the wake of the crisis, the demand for (female) workers in care services was unbroken (de Haas et al. 2020). The history of immigration to Europe provided in this section calls into question the claim of a general 'feminization' of migration over time. Although varying by year and country, averaged across the OECD women have always made up close to half of immigrant inflows since 1950 (de Haas et al. 2019). Given this conclusion, it is critical to understand the gender-specific processes involved in immigrant women's labor market outcomes. In the following subsection, we present the second important component for the contextualization of this broad topic-the theoretical underpinnings of female immigrant labor market disadvantage.

Explaining female immigrant disadvantage
Empirical research rests heavily upon its theoretical foundations, meaning that the methodological approaches to studying female immigrants and the interpretation of analytical results are sensitive to the initial theoretical assumptions made. Generally speaking, most studies of immigrant disadvantage on the labor market start out from human capital theory, which highlights the importance of human capital endowments like education and work experience for labor market success (Becker 1985). For immigrants this is particularly important, because human capital is not perfectly transferable from one context to another. Assimilation theory thus states that immigrants remain disadvantaged on the labor market compared to natives even when their level of education and prior work experience are accounted for, and that their labor market outcomes converge towards those of natives as they spend more time in the host country and gain capital relevant to this new labor market (Chiswick 1978;Friedberg 2000). This not only relates to host-country educational degrees and job experience, but also proficiency in the language of the host country (Dustmann 1994;Chiswick andMiller 2002, 2003), the establishment of new social networks (Putnam 2000, from Kanas et al. 2011, and other forms of softer socio-cultural capital (Reimers 1985).
Assimilation theory itself has not been able to explain the full extent of immigrant disadvantages, and thus other approaches have been developed to extend this theoretical framework. Segmented assimilation theory, for instance, spells out that immigrants' integration trajectories are not just determined by their labor market capital endowments, but that group-specific reception contexts in a destination country also play a decisive role (Portes and Zhou 1993). These reception contexts are formed by government immigration and integration policies, the perceptions of the majority society, and the existing co-ethnic community. Institutional barriers inherent in the migration channel contribute to the disadvantage of some 'unwanted' groups compared to others, as do the ties that some immigrant communities have to low-wage sectors and ethnic job markets. Discrimination by members of the majority population, both socially and on the labor market, also creates part of the reception context that influences immigrant labor market outcomes.
While these theories have proven useful for explaining immigrant labor market integration in general, they do not include considerations of gender in and of themselves. In most cases, researchers who approached this topic from a 'household specialization' perspective (Becker 1985) often presumed that men occupy the traditional breadwinner role, and that women, especially those who are married with children, occupy a place outside of the labor force. Over time, though, researchers have acknowledged that women in general have heterogeneous labor market profiles, and that factors such as marital status and the presence/number of children in the household have a gender-specific impact. Most house and care work is borne by women, which negatively impacts their labor market outcomes, and immigrant women pose no exception. Moreover, immigrants may arrive with different sets of norms and values, specifically relating to gender roles, which can influence their subsequent labor market integration. With this realization came increasing efforts to explain female immigrant labor market behavior in itself, not just as a caveat to males.
Several investigations of labor market disadvantages by nativity and gender beginning in the 1980s highlighted the intersectionality of both dimensions and gave foundation to the 'double disadvantage' hypothesis. Based on the observation that immigrant women in Canada were on average the most disadvantaged group on the labor market in terms of occupational status compared to native women, native men, and their male compatriots, Boyd (1984) first concluded that… ...in addition to the status of being a migrant, migrant women experience additional difficulties in the labor force as women [...]. Overall, the position of migrant women in the labor force can be understood as reflecting the combined impact of sex and birthplace or the 'double negative' effect. (p. 1092f.) The assertion of such a 'double disadvantage' (Raijman and Semyonov 1997) or 'double jeopardy' (Greenman and Xie 2008) rests on the assumption that, on average, both women and immigrants experience penalties on the labor market and that these penalties coincide, leaving immigrant women in the least favorable position. That being said, while the double disadvantage approach represents a suitable intersectional setting for investigating female immigrant labor market disadvantages, intersections do not just involve the dimensions of gender and nativity. Empirical evidence points to very heterogeneous experiences for particular groups of immigrant women, thus hearkening back to the group-specific outcomes predicted by segmented assimilation theorists.
These general theoretical approaches and the extensions for female immigrants specifically are helpful for identifying the most important drivers and mechanisms that propel disadvantages by gender and nativity. The way these disadvantages translate into inferior labor market outcomes, however, is not instantaneous, but arises as these women move through different stages of the immigration and integration trajectory. In the following, we discuss how a disadvantage along these two dimensions arises for sequential outcomes, like labor force participation and success on the country's labor market.

Labor force participation
Typically, the earliest indicator of immigrants' labor market integration is their labor force participation (LFP), that is, the process of actively looking for and/or engaging in wage-earning employment. Immigrants have different LFP intentions after migration, which link back to the selection processes they have undergone up to this point. On one hand, labor immigrants are positively selected with regards to their labor force participation rates, in the sense that the vast majority intend to work, and for the so-called third country nationals (TCNs), they often already have a job lined up prior to arrival as a precondition to their work visa. However, the LFP intentions of other types of immigrants can be more ambiguous from the beginning of their trajectory. For 'unwanted' groups of TCNs, who often do not qualify for labor migration channels, other migration channels that are not conditional on labor market potential are the only way to receive the authorization to immigrate. Thus, TCN immigrants arriving as family immigrants, students, etc., may have intentions to participate in the labor force after arrival, and use these other channels to gain the initial immigration authorization they would have otherwise been denied. Still, non-labor migration channels may entail additional labor market access restrictions.
Clearly, the selection effects of immigration policy have an important impact on the LFP of these groups after arrival, often resulting in a disadvantaged situation for certain groups of immigrants on the labor market based on their origin. Gender constitutes an additional dimension of disadvantage. When it comes to LFP, more often than not, males' participation is taken as a fact. In most cases, this makes sensedata show that male immigrants had high labor force participation rates before immigration, and that this continues after immigration (for evidence from Germany, see Höhne 2016;Bürmann et al. 2018) regardless of the country of origin. Very few do not try to find a job at some point in their early integration trajectory. Compared to male immigrants, female immigrants undergo a much more deliberative process in their LFP decision, involving the following considerations: economic necessity, household duties, social norms, and perceptions of labor market success (including discrimination).
Regarding the economic necessity of female labor force participation, this often relates to their living circumstances. For those who live in households with other sources of revenue (from a spouse or parents), it might not be a requirement for them to contribute monetarily. Even so, it is becoming increasingly difficult to support a household on a single salary, creating economic incentives for female immigrants to seek paid work and driving up the female LFP rates even in the majority populations: the average female LFP rate increased across all European countries in the last decades (OECD 2020). Thus, the culture of the male breadwinner seems to be becoming less prevalent in Europe and other high-income countries, though females often continue to undertake household tasks (Altintas and Sullivan 2016), even if they are working (see Sofer and Thibout 2019 for evidence from France).
Along with economic incentives for immigrant women's LFP comes the opportunity to delegate time away from household activities, which is especially difficult when there are children to take care of. Immigrants often delay family formation until after migration, which is evident from a high birth rate in the year following migration (Milewski 2007). Thus, recent immigrants can have more strenuous household burdens than natives, which impacts their labor force participation. Nonetheless, many European countries encourage female LFP, and particularly mothers, via different types of activation policies such as expanded funding for childcare, but immigrant women remain difficult to engage. Some sources claim that they are less likely than native women to take advantage of public daycare schemes (Gambaro et al. 2017), though other studies find that employed immigrant women use public daycare just as much as employed native women (Rendall et al. 2010). In any case, immigrants are less likely to have extended family present to help with household tasks, which plausibly reduces their likelihood of allocating time away from the household.
Additionally, culture-specific gender values can factor into the decision to search for work. For women (and/or their spouses) who hold conservative ideas about the appropriateness of females on the labor market, the wage which would be earned from being active in the labor force often does not compensate for the perceived utility loss in household production, and therefore discourages their labor force participation (Bertrand et al. 2018). Typically, these conservative gender values are associated with immigrants from Middle Eastern/Muslim-majority origins. This being said, the correlation between Muslim religiosity and conservative gender values has not received widespread empirical support in recent studies (Koenig et al. 2016;Schieckoff and Diehl 2021), so this consideration needs to be treated with caution when examining the dimensions of nativity and gender.
The final aspect that female immigrants may take into account when deliberating whether to enter the labor force or not is their perception of being successful on the labor market. This involves, mainly, perceptions of discrimination based on sex and ethnicity/nativity, which could impede their ability to find a job. While perceptions are not necessarily accurate indicators of actual discrimination (Diehl and Liebau 2017;Small and Pager 2020), they likely influence the decision of whether participation in the labor force participation is individually worthwhile or not.
Thus, on top of the usual considerations, the additional determinants to consider when studying female immigrants' labor force participation specifically include their migration motive, the economic standing of their household, their marital status and number of children, their gender values and/or level of religiosity, and their perceptions of the inclusiveness of the host country's society and labor market.

Labor market success
Assuming that female immigrants manage to access the host country's labor market and they make the decision to participate in the labor force, those who are active on the labor market continue to experience disadvantages stemming from their nativity and gender (among other factors). Success on the labor market relates to several measurable outcomes. The employment/unemployment rates of immigrants are an important topic of study, though self-employment is also an important option for many immigrants (OECD 2017). Additional outcomes describe the conditions of employment, such as the contract type and number of hours worked, as well as characteristics of the work found, such as the occupational status and the wage paid.
Besides the immigrant disadvantage mentioned earlier involving human capital and the differences between home and host-country labor market contexts, there are gender considerations that particularly impact the labor market outcomes of female immigrants. For one, in the case of family immigration, the decision to immigrate often favors the husband's job prospects in the destination country (Mincer 1978). Although migration patterns have evolved since the eras of the male 'guest worker' and female 'trailing spouse,' female immigrants continue to be overrepresented in family reunification channels (OECD 2017), meaning that their immigration is to a larger extent conditioned on their family connections and not on their future labor market success.
Another aspect that affects the success of female immigrants on the labor market is their occupational segregation. Women in general, not just female immigrants, are typically found in service, care, and teaching occupations (Das and Kotikula 2019). These occupations tend to require significant customer service and communication skills. When it comes to female immigrants, their proficiency in the host-country language and familiarity with social norms can present a barrier to them finding jobs in their trained occupation (Raijman and Semyonov 1997). Moreover, the additional domestic tasks they may encounter after immigration in setting up a home may direct them to the types of jobs that can give them the flexibility in working hours they need. Thus, these factors work together to dictate the quality and characteristics of the job found after immigration.
Finally, discrimination is an important consideration when examining the outcomes of female immigrants. Audit studies back up the existence of discrimination against ethnic and racial minorities on European labor markets (Quillian et al. 2019). Discrimination is one factor that relates to both a nativity and gender dimension. Not only do immigrants experience labor market discrimination, but females also-they may be less desirable to employers if they have a spouse or children at home (Correll et al. 2007), and they sometimes carry an ethno-religious marker of 'otherness' like the hijab for Muslim women (Weichselbaumer 2020).
Thus, for studies relating to female immigrants' labor market success, additional factors that are important considerations are the migration motivation and/or the order of migration, the trained occupation and the occupational structure of the host country, and hiring discrimination.
Throughout this section, we have presented the most important contextual information needed to approach the topic of female immigrant labor market integration in Europe; the historical development of immigration policies and the immigrant community across Europe, and the theoretical foundations explaining why this group differs from other groups on the labor market. We have shown that women have always been an important component of the immigrant flows since the Second World War and that immigration regimes hold a lot of sway over the characteristics of those immigrants who arrive. We have also shown that empirical studies of female immigrant labor market outcomes need to look at additional factors because of the interplay between the disadvantages relating to their nativity and those relating to their gender. In the following section, we dive into the findings from quantitative studies between 2000 and 2020 and examine how the recent body of literature lines up with these considerations.

Empirical findings
In this section, we organize the empirical findings by the outcome studied. We start by reviewing studies on female immigrants' labor force participation, and then move on to their labor market success conditional on participation. Since labor market success is such a broad topic with multiple indicators, the review of this portion of the literature will be broken into two subsections: The first subsection deals with immigrant women's employment status (including their (un)employment rates, working hours/parttime work and contract type) and the second subsection with the quality of and returns to their employment once it is found (including occupational status, earnings/wages).
When seeking to investigate the coinciding labor market disadvantages that immigrant women face because of their gender and nativity, both of these dimensions have to be explicitly considered. Yet, only a few of the reviewed studies do so; most compare either women and men among immigrants, or immigrants and natives among women. As each of these comparisons corresponds to a different type of disadvantage, we first structure the findings by the comparison group. We then provide an integrated perspective that explicitly focuses on the intersectional nature of gender and nativity and how their interplay affects the labor market outcomes of immigrant women.
Many of the reviewed studies rely on data that are, by now, decades old. The extent to which these findings reflect current immigrant experiences is unknown. Thus, we supplement our review by situating the studies within the current European context. We provide additional descriptive statistics on gender and nativity gaps for the most common labor market indicators by country of destination for 2019 based on Eurostat data. In an interactive online appendix, we also present figures on how these gaps have developed over time.

Nativity gap among women
According to 2019 Eurostat data, immigrant women show lower average LFP rates than native women in many of the European countries covered by the studies included in this review (see Fig. 2). This disadvantage is observed for all of NWE except Ireland and Luxembourg, and has also been confirmed in other studies for the UK, France, Belgium, the Netherlands, Germany, Austria, Denmark, and Sweden from the past two decades (Rubin et al. 2008;Fleischmann and Höhne 2013;Koopmans 2016;Gorodzeisky and Semyonov 2017;Knize Estrada 2018;Neuman 2018). By contrast, in SE countries the average LFP rates of immigrant women in the Eurostat data consistently exceed those of native women (see also Amuedo-Dorantes and de la Rica 2007; Rubin et al. 2008;Rendall et al. 2010;Bernardi et al. 2011;Fullin and Reyneri 2011). We observe a similar pattern in some CEE countries, e.g., the Czech Republic, Poland, and Hungary, but the situation appears much more heterogeneous. Advantages in LFP need not indicate a disparity in the type of immigrants moving to each region, but can also point to the fact that, in SE and CEE countries, the LFP rates of native women are usually lower compared to NWE countries. For example, nearly 80 percent of native women in the Netherlands participated in the labor force in 2019, but only about 60 percent did so in Greece or Hungary (see Fig A4 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/). Besides general differences in female LFP in the destination countries, nativity gaps also vary greatly by immigrants' country of origin. In NWE countries, as of 2019, the nativity gap is generally smaller for immigrant women from EU countries compared to those from outside the EU, but the picture is much less consistent across SE and CEE countries (see Fig. A41 in the online appendix). Looking at immigrants from specific origin countries, several patterns emerge. Women from Muslim-majority countries show the lowest LFP rates across destination countries, Immigrants include all immigrants for which there are data available, irrespective of origin. Markers represent the values obtained by subtracting the comparison group value from the value for immigrant women. The dashed vertical lines separate northern-western Europe (NWE) on the left, southern Europe (SE) in the center, and central and eastern Europe (CEE) on the right. Lighter shades of markers represent the Eurostat flag 'low reliability.' Source Eurostat (2020b) e.g., Turkish women in Germany, or Moroccan women in Italy (Fullin and Reyneri 2011;Fleischmann and Höhne 2013). By contrast, immigrant women from eastern European countries rank among the highest in LFP, e.g., women from Bulgaria and Romania in Italy and Germany (Fullin and Reyneri 2011;Fleischmann and Höhne 2013).
These patterns suggest that female immigrant populations across Europe, and the groups of native women they are compared to, are quite heterogeneous in terms of the characteristics determining their LFP. Several relevant factors have been repeatedly identified in the literature that explain parts of the observed (positive or negative) nativity gaps in LFP among women. A large part of the gap usually disappears after accounting for compositional differences in age, human capital, and household factors between immigrant and native women. The most important human capital factor is education, and its distribution among immigrants is highly conditional on the selection processes shaping migration flows. Regarding the family context, the marital status and the presence of small children matter most, as marriage and birth rates among immigrant women often exceed those among native women (Adserà and Ferrer 2015). In some cases, after accounting for compositional differences in these relevant factors between immigrant populations in destination countries, but also between origin groups within the immigrant population, nativity gaps vanish completely. For instance, Rendall et al. (2010) show that the LFP nativity advantage of immigrant women in SE countries disappears net of these factors for Greece, Spain, and (almost) for Portugal. Gorodzeisky and Semyonov (2017) find a LFP nativity disadvantage for non-European immigrant women, and a nativity advantage for European immigrant women in France-both gaps narrow and become insignificant after adding socio-demographic controls. 5 While these examples show that human capital and the household context are important factors, they are in many cases insufficient to fully explain the nativity gap in female LFP. As the varying LFP patterns across origin countries suggest, additional factors are related to the cultural and social norms of immigrants and the corresponding perceptions of appropriate roles of women and men in the household and the labor market. Koopmans (2016) shows for Muslim immigrant women (from Turkey, Morocco, former Yugoslavia, and Pakistan) in NWE that traditional gender values are associated with significantly lower LFP rates. Khoudja and Fleischmann (2017) confirm this result for immigrant women from Turkey, Morocco, and Surinam/Antilles to the Netherlands, highlighting that the gender role attitudes held by the women's male partners play a significant role too. They also show that traditional women tend not to participate in the labor market if the labor market resources of their partners permit, whereas this is much less the case for egalitarian women. In the absence of direct measures, religiosity is often used as a proxy 5 In some cases, gaps also increase (or even change direction). For example, Khoudja and Fleischmann (2017) show for the Netherlands that immigrant women of Surinamese/Antillean origin exhibit slightly higher LFP rates as native women (+ 3 percent) in their sample. After controlling for age, education, health, language skills, and children in the household, the predicted Surinamese/Antillean advantage is 19 percent (p. 525).
for traditional gender roles, based on the assumption that higher levels of religiosity are associated with more conservative gender roles, particularly for Muslim women (see, e.g., Knize Estrada 2018, for Germany). However, Khoudja and Platt (2018) challenge this interpretation by considering both religiosity and gender role attitudes of immigrant women in the UK. Net of gender role attitudes, labor market entry rates of immigrant women in their sample even increase with religiosity (and exit rates decrease vice versa).
Assimilation into the host society is relevant for all the considered (as well as unobserved) determinants of LFP. With longer durations of stay, nativity gaps in female LFP typically narrow across European countries, as immigrant women acquire host-country-specific resources such as language proficiency and adapt to the regional culture and social norms (Amuedo-Dorantes and de la Rica 2007; Rendall et al. 2010;Fleischmann and Höhne 2013;Koopmans 2016;Neuman 2018). Still, in most studies modeling assimilation trajectories, immigrant women starting out at a disadvantage in LFP do not manage to catch up to native women over the analysis period (see, e.g., Rendall et al. 2010;Neuman 2018).
Several studies aim to consider as many factors relevant to female LFP as possible (e.g., Khoudja and Platt 2018), but even then some 'unexplained' gaps remain. This net disadvantage might be due to the fact that immigrant women experience lower returns to their characteristics on the labor market than natives, and the expected disadvantage in returns discourages them from participating in the labor market (Rendall et al. 2010). In such a perspective, factors relevant for labor market success also play a larger role, such as individual language skills (Koopmans 2016) or employer discrimination (Quillian et al. 2019).
That being said, the possibly far-reaching effects that migration motivations and available migration channels have on immigrant women's LFP seem to be quite underexplored. In a rare study, Cangiano (2014) shows that the migration channels under which immigrant women entered EU15 countries are strong determinants of their LFP. Both female family migrants and refugees show substantially lower odds of being active on the labor market in comparison to female labor migrants and native women in general, conditional on age, human capital, household characteristics, language skills, and duration of stay. This result is a clear example of the selection effects involved in migration regimes and the corresponding individual migration trajectories.

Gender gap among immigrants
This section reviews the evidence on differences in LFP rates between women and men among immigrants, thereby highlighting the gender-specific component of immigrant women's labor market (dis)advantages. Across Europe, women participate less in the labor market than men and immigrants pose no exception. As of 2019, the average gender gaps in immigrant LFP range from about 5 pp in Lithuania to over 25 pp in Greece, but there is no clear geographical pattern; most gaps lie between 10 and 20 pp (see Fig. 2). In NWE, gender gaps in LFP are larger for immigrants from non-EU compared to EU countries, whereas such a difference is not consistently present in SE and CEE countries (see Figure A4 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/).
In most studies, immigrant women and men are considered separately (usually in comparison to natives), making it difficult to pin down the factors behind the observed gender inequalities. Nevertheless, judging from their differential impact on women's and men's LFP, family context and cultural norms seem to be among the most important drivers. For example, controlling for age, education, and origin, Fullin and Reyneri (2011) estimate that in comparison to immigrant women living alone, the LFP rates of women living with a partner are significantly lower, particularly in the presence of children. By contrast, immigrant men living with a partner and children are significantly more likely to participate in the labor force than men living alone. Koopmans (2016) reports a 'marriage bonus' for men and a 'children penalty' for women in terms of LFP for Muslim immigrants of Turkish, Moroccan, former Yugoslavian, and Pakistani origin in six NWE countries. He also shows that the gendered effects of the family context persist when including socio-cultural factors in the model. However, these socio-cultural factors themselves only affect immigrant women: their LFP is estimated to significantly increase in host-country language proficiency, host-country media use, and, particularly, in holding liberal gender role attitudes (Koopmans 2016).
To the best of our knowledge, in the timeframe we consider there is only one study for the European context that compares the LFP rates of immigrant women and men in one model, explicitly testing the significance of their difference. Fleischmann and Höhne (2013) estimate gender gaps in LFP (net of age, education, marital status, children in the household, and duration of stay) across origin groups of immigrants in Germany in 2009. They find participation rates of women to be lower than those of men for all immigrant groups except ethnic Germans and Greeks, significantly so for immigrants from Poland, North Africa, Turkey, the Middle East, and a group of Western countries (in the order of decreasing gap magnitude). Clearly, nativity-specific factors not considered in their model seem to play a role here, such as differences in gender role attitudes or language skills.

Coinciding nativity and gender gaps
When examining labor market outcomes by nativity (among women) or by gender (among immigrants), we do not know whether nativity gaps are also present among men and whether gender gaps are also present among natives. To test for coinciding gender and nativity gaps, native men are a necessary additional comparison group in an increasingly complex framework. Across Europe, the LFP disadvantages of women compared to men are ubiquitous in the 2019 Eurostat data, but they differ considerably in magnitude between immigrants and natives as well as by country (see Fig. A3 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/). Whereas gender gaps in LFP have a uniform direction, this is not true for the nativity gap. Whether immigrants show higher or lower LFP rates than natives varies by country and gender. For example, in France as of 2019 immigrant women are the least likely to participate in the labor market, which suggests both a nativity and a gender disadvantage (see Fig. A3 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/). Yet, immigrant men actually show an advantage in LFP over native men.
As we have seen, comparing immigrant and native women as well as immigrant women and men, the raw, descriptive gaps usually narrow after accounting for human capital, the family context, and socio-cultural factors. Thus, any model attempting to estimate a coinciding disadvantage needs to consider such factors as much as possible, while comparing immigrant women, native women, immigrant men, and native men. In terms of LFP, the only available study employing such a design is again the one by Fleischmann and Höhne (2013) for Germany. Their results have two main implications. First, the gender gap varies by nativity or, vice versa, the nativity gap by gender, so that a possible 'double disadvantage' cannot always be inferred from estimating a general nativity and a general gender gap and adding them together. For example, they estimate that native women and Turkish men have higher conditional LFP rates than native men, yet Turkish women exhibit the lowest rates in comparison to all other groups. Second, conditional nativity and gender gaps vary in their direction by immigrant origin, so that immigrant women do not generally face a double disadvantage in terms of LFP in Germany. A tentative interpretation of Fleischmann and Höhne's (2013) (not precisely interpretable) results suggests that ethnic German immigrant women enjoy a double advantage in LFP, whereas immigrant women from North Africa and the Middle East suffer from a double disadvantage. This result could be driven by a combination of the selectivity inherent in Germany's migration regime (ethnic Germans received preferential treatment) and the differences in gender role attitudes between Germans and immigrant origin groups (ethnic Germans being rather egalitarian, immigrants from North Africa and the Middle East being rather traditional).

Summary
The reviewed literature clearly shows that labor market integration in Europe in terms of LFP is stratified by gender and nativity, and the considerable heterogeneity across destination and origin countries suggests that immigrant women experience a selective disadvantage in their LFP rates vis-a-vis native women, immigrant men, and native men. Studies seeking to explain these disadvantages have generally done a good job of investigating the theoretical considerations mentioned earlier that are particularly relevant to female immigrant LFP. Some structural factors of the receiving context like the current and historical migration policies and the labor market demand, which are inherent in immigrants' channel of entry and/or the motive they state for migration, seem to be strong drivers of nativity gaps in LFP and explain why immigrant women often fare better than natives in SE and CEE but not NWE countries. Household factors like marital status and children join human capital factors as the main determinants of LFP gaps, all of which have gendered outcomes that are additionally stratified by nativity. Gender role attitudes also appear to be particularly relevant for female immigrant LFP and should ideally be included in any such estimation. However, household economic standing as a motivating factor for female immigrant LFP, as well as the potential barrier represented by perceptions of ethnic discrimination, are two mechanisms which remain to be explored in this body of literature. In many of the reviewed studies, unexplained gaps in LFP remain after accounting for a range of explanatory factors, suggesting either omitted variables like these or unequal returns to observed characteristics between women and men, immigrants and natives. Moreover, among these reviewed studies, few combine both the nativity perspective with the gender perspective for a holistic understanding of immigrant women's LFP. This type of tailored approach is necessary for future studies investigating the coinciding disadvantage for immigrant women in terms of LFP in the European context, especially one that allows to assess the magnitude of inequalities by nativity vis-a-vis inequalities by gender.

Labor market success I: employment status
The next studies look at the (un)employment rates of immigrant women and the conditions of their employment once it is found (i.e. the contract type and number of hours worked). As outlined in the theoretical section above, employment outcomes differ from labor force participation, in that successfully finding employment not only requires a woman to decide to look for employment, but also deals with the local labor market conditions and hiring tastes of employers.
A conceptual challenge here is the comparability of different measures, particularly unemployment rates and employment rates. Unemployment rates are usually defined as the share of the unemployed relative to the labor force, whereas employment rates denote the percentage of the employed relative to the full (working-age) population. Any differences in employment rates might, thus, represent differences in LFP as well as differences in unemployment rates. In cases where unemployment rates have not been investigated, we report results on employment rates charged with this caveat.

Nativity gap among women
Similar to LFP rates, female immigrant unemployment rates are not uniform across European countries and, as of 2019, range between about 5 percent in Ireland, Germany, Croatia, and the Czech Republic, to over 15 percent in Spain, Italy, and Sweden, and even 35 percent in Greece (see Fig. A7 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/). Although there is no clear geographical pattern in the unemployment rates of immigrant women, one striking commonality is the observable disadvantage of immigrant women compared to native women in almost all countries (see Fig. 3). This result has been reported in other studies over the past two decades for countries such as the UK, France, Belgium, and Sweden (Gorodzeisky and Semyonov 2017), Germany (Kogan 2011), and Spain (Amuedo-Dorantes and de la Rica 2007; Bernardi et al. 2011). Despite the general pattern, the magnitude of the nativity disadvantage in unemployment among women varies greatly across countries (see Fig. 3). The disadvantage is less than 3 pp in Germany, Estonia, and the Czech Republic; between 5 and 8 pp in Finland, France, and Italy; and in Sweden and Greece we observe a disadvantage of 12 and 14 pp, respectively.
Studies since 2000 that present data disaggregated by an immigrant's geographical origin reveal that the aggregate figures for immigrant women conceal considerable variation between the groups. In NWE, nativity disadvantages in unemployment are typically higher for women of non-EU origin compared to those of EU origin in the 2019 Eurostat data (see Fig. A7 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/) and in earlier studies (Kogan 2006;Rubin et al. 2008;Keyser et al. 2012;Gorodzeisky and Semyonov 2017). In Germany, immigrant women from EU or OECD countries even exhibit the same unemployment rates as native women (Kogan 2011). With a few exceptions, the situation is similar in SE and CEE in 2019 (see Fig. 3), though in countries such as Portugal or the Czech Republic, immigrant women from non-EU countries fared better than those of EU origin around 2005, a pattern that reversed over time (Amuedo-Dorantes and de la Rica 2007; Rubin et al. 2008;Bernardi et al. 2011;Fullin and Reyneri 2011, see also Fig. A8 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/). Moreover, when looking at particular origin countries, there seem to be many deviations from the aggregate picture that only distinguishes between EU and non-EU origin. For example, in Italy women from eastern Asia show the lowest unemployment rates of all women, in comparison to other non-EU immigrants, EU immigrants, and natives (Fullin and Reyneri 2011). When immigrant women do successfully find employment, the conditions of this employment often differ from those of native women. In terms of their working hours, immigrant women in SE (except Malta) have higher rates of part-time employment compared to native women according to Eurostat data (see Fig. 4). The same is true in some NWE countries such as Germany, Finland, or France, but the opposite is found in the Netherlands, Switzerland, and the UK (see also Bevelander and Groeneveld 2006;Fleischmann and Höhne 2013). In CEE, where part-time employment is less prevalent in general, immigrant women also seem more likely to work part-time than native women. As working part-time can be desirable for individuals in many situations, a better measure to reflect disadvantage would be involuntary part-time employment rates-information which is rarely available in large-scale datasets. However, for 2005, Rubin et al. (2008) observe a quite consistent pattern across Europe, involuntary part-time employment being more common among non-EU immigrant women compared to native women, particularly in Austria, Denmark, Portugal, Spain, and Greece. In terms of contract type, there is evidence that immigrant women have a lower share of permanent contracts compared to native women in all of SE and many NWE and CEE countries (see Fig. 5), though there is again considerable heterogeneity by country of origin. For example, in the Netherlands, women from Suriname actually more often hold permanent contracts than native women (Bevelander and Groeneveld 2006). Overall, the descriptive data of female immigrants' employment situation across Europe does not present a very cohesive picture.  Eurostat (2020d) Quantitative studies investigating the determinants of these heterogeneous outcomes typically start by considering differences in human capital factors. Among immigrant women in Germany, accounting for their level of education reduces some of the nativity disadvantage that immigrant groups face in terms of unemployment, but significant differences remain for all groups besides older arrival cohorts coming from Western countries (Kogan 2011). For female immigrants in Spain, controlling for education reduces some of the biggest unemployment gaps that exist between native women and women from the African continent, though these gaps remain sizable (Bernardi et al. 2011). Algan et al. (2010) show that net of human capital differences, female immigrants from certain countries of origin still experience persistent disadvantages in employment compared to native women: Turkish women in France and Germany have consistently lower employment rates (33 and 16 pp lower, respectively), as do women from Bangladesh and Pakistan in the UK (36 pp lower), yet large parts of the remaining gaps stem from differences in LFP.
Moving on, unemployment rates (or employment rates conditional on LFP), seem to be fairly unaffected by household characteristics, like the presence of children, so that general and also origin-specific nativity gaps still persist after accounting for these factors on top of differences in human capital. Ballarino and Panichella (2018) estimate that the remaining nativity disadvantages in employment rates among women active on the labor market range from 6-8 pp in the UK, Italy, and Spain; over 11 pp in the Netherlands and Germany; up to 13 pp in France. Gorodzeisky and Semyonov (2017) report that European female immigrants have similar conditional  Eurostat (2020e) odds of being unemployed compared to native women in the UK and France, but a nativity disadvantage remains in Belgium and Sweden. By contrast, female immigrants originating from non-European countries show significantly higher net odds of unemployment in all four countries. In Italy, women from China and the Philippines are observed to have higher probabilities of avoiding unemployment compared to native women after controlling for education and household characteristics, but all other origin groups remain disadvantaged (Fullin and Reyneri 2011). Household factors play a larger role for women's working hours. Immigrant women in Germany exhibit on average higher rates of part-time employment compared to native (West German) women-a pattern that reverses after accounting for household factors and education (Fleischmann and Höhne 2013).
Extending the investigation to include host-country-specific capital, one study from Sweden found that particularly female refugees from countries with a large perceived cultural and linguistic distance from Sweden had, relative to other immigrant groups and native women, a low probability of being employed (Bevelander 2005). Language proficiency and cultural norms were also found to be important for Muslim women in other NWE countries (including Germany, the UK, the Netherlands, Belgium, and Switzerland): Koopmans (2016) shows that first-generation immigrant women experience a large disadvantage compared to native women in their unemployment rates conditional on human capital and household factors. After additionally controlling for socio-cultural variables (like language proficiency, social contacts, and gender values), there is a much smaller and statistically insignificant difference between the groups. A similar result is obtained by a German study which compares native women with women from Turkey and the former Soviet Union (Salikutluk et al. 2020). Accounting for German language skills and contacts to natives in addition to human capital and household characteristics further reduces the predicted nativity gap in unemployment by about 2 pp. Only for FSU women does a small (and insignificant) disadvantage remain.
Unfortunately, many studies do not consider specific factors relating to an individual's host-country-specific capital, but they do consider an immigrant's length of residence in the host country as a proxy. In many cases, there is evidence of female immigrant assimilation in the labor market, which is depicted by shrinking nativity disadvantages (or growing advantages) over time. Studies typically look at employment probabilities (vs. un-and non-employment) rather than unemployment rates, so that effects can stem from changes in labor force participation as well as employment conditional on participation. For example, Ballarino and Panichella (2018) observe the assimilation of immigrant women in terms of employment rates in France, Germany, the Netherlands, Italy, and Spain (less so for the UK). Yet, whereas the average gap between their employment rates and those of native women (almost) closes after 10 years in Italy and Spain, it is still 22 pp in Germany, 24 pp in France, and 32 pp in the Netherlands. According to Kesler (2006), female immigrants in Sweden close the nativity gap in employment rates quicker than those in Germany and Great Britain, although in Norway, the gap never closes for women from the most disadvantaged group-female refugees (Bratsberg et al. 2017). In Spain as well, immigrant women initially experience disadvantage in employment likelihood compared to native women with similar characteristics (− 25.9 pp for women from EU15 countries, − 12.9 pp for women from Africa), but these disadvantages decrease over time, particularly during the first year (Amuedo-Dorantes and de la Rica 2007). One available study on unemployment rates among women in Spain shows that particularly immigrant women from eastern Europe and Africa manage to reduce their unemployment risks over the first few years of stay (Bernardi et al. 2011). In terms of working hours, Basilio et al. (2009) find that immigrant wives in Germany work 230 h less than native wives per year directly after their arrival, but after 10 years of residence, they succeed in closing the gap completely (or even in overtaking native wives).
To this long list of explanatory factors some studies also add naturalization. Naturalization can be seen as one of the outcomes of the assimilation process, but naturalization in itself can also be a facilitator of labor market integration, for example, when holding the host country's citizenship entails particular rights (such as not being subject to labor market testing) or provides a positive signal to employers. Using a causal design based on a change in eligibility criteria for German citizenship, Gathmann and Keller (2018) estimate that immigrant women who obtained the German citizenship have a 5.9 pp higher employment probability compared to immigrant women who did not (net of age and years of residence).
Since variation in female immigrants' outcomes is observed not just between immigrants from different origins, but also between different regions of Europe, the unique reception context that immigrants encounter in the host country likely plays an important role. Comparing three countries in NWE, Sweden, Germany, and Great Britain, immigrant women were found more likely to be employed in Sweden compared to the other countries once family factors were taken into account (Kesler 2006). This could be related to the extensive social support that is provided to wives and mothers in the Swedish labor market. For SE countries, on the other hand, the smaller nativity disadvantages could likely be related to the large low-wage sector in these countries, which benefits particular groups of immigrant women in their search for employment (Kogan 2006). Although not always clear-cut, Cangiano (2014) emphasizes the importance of national migration regimes by concluding that lower restrictions on labor migration go along with smaller nativity gaps in employment rates across EU15 countries.
Cross-country differences may be related not just to the specific institutional context, but also different levels of discrimination which immigrants may face. Kogan (2006) finds that immigrant women across Europe encounter, on average, 148 percent higher odds of unemployment compared to native women, even after controlling for human capital and host-country institutional characteristics (including the type of welfare regime, strictness of employment protection, and the size of the low-wage sector). Labor market discrimination has also been documented by audit studies in several NWE countries. In Germany, Turkish women wearing a headscarf in the photo on their CV are 15 pp less likely to receive a callback compared to 'native German' female applicants, while bare-headed Turkish applicants are about 6 pp less likely (Weichselbaumer 2020). Moreover, an audit study in the Dutch labor market shows that the callback rate for females from non-EU countries (Turkey, Morocco, Suriname, Dutch Antilles) is 11 percent lower than the callback rate for native women (Andriessen et al. 2012). None of the audit studies included in our review focus on immigrant discrimination in SE or CEE countries, which makes it difficult to estimate the impact of discrimination on immigrant employment in those contexts.

Gender gap among immigrants
When it comes to the gender gap between male and female immigrants, there is considerably less research, and the findings again produce an incohesive picture. According to Eurostat data from 2019, the average unemployment rates of immigrant women across Europe are typically greater than those of immigrant men, but there is considerable heterogeneity by country of residence (see Fig. 3). In countries such as Germany, Ireland, or Norway, immigrant women have a similar or lower risk of being unemployed as men, but they face a much greater risk in Greece and North Macedonia. Gender gaps in the relative success of avoiding unemployment seem largely comparable between immigrants of European vs. non-European origin across countries (see Fig. A7 in the online appendix from https:// pages. cms. hu-berlin. de/ spren max/ fem-lit-review/), though the figures presented in the literature vary by data source and time (see, e.g., Rubin et al. 2008;Keyser et al. 2012;Gorodzeisky and Semyonov 2017). In terms of working hours, immigrant women consistently show higher rates of part-time employment than immigrant men in European countries, with gaps ranging between 2.8 pp in Hungary to 42.3 pp in Switzerland as of 2019 (see Fig. 4). Temporary employment is also more common among immigrant women than men in most of NWE, but the picture is less uniform in SE and CEE (see Fig. 5).
As far as quantitative analyses go from the past two decades, Fleischmann and Dronkers (2010) find that, for immigrants in 13 European countries, there is generally no difference between the conditional unemployment rates of women and men, and groups of individual-level or macro-level factors do not have a differential impact. Similarly, investigating additional determinants, Höhne and Koopmans (2010) find that the hazard of finding a job improves in host-country-specific capital for immigrant women and men in Germany to a similar extent, whereas strong Muslim religiosity has the opposite effect. For the German case, there is even evidence of female advantage in employment probability for immigrants from Turkey (+ 7.5 pp) and the FSU (+ 3.9 pp) conditional on LFP and net of differences in human capital, host-country-specific capital, and household context (Salikutluk et al. 2020). The authors interpret these results as pointing towards the positive selectivity of women from Turkey and issues of lower returns to education for FSU women, possibly due to difficulties in the recognition of qualifications.
Household factors rarely have a gender-specific effect on unemployment rates, but they play an important role for the working hours of women and men. Ala-Mantila and Fleischmann (2018) show for the Netherlands not only that female immigrants generally work fewer hours than their male counterparts, but immigrant mothers in particular work fewer hours, while immigrant fathers actually work more hours. Fleischmann and Höhne (2013) confirm these results for Germany based on an empirical model in which the effects of household factors are allowed to vary by gender. In addition to marital status, having (small) children in the household increases the probability to work part-time much more for women than men. As a result, they estimate conditional gender gaps in part-time employment among almost all of the different immigrant origin groups they investigate, with particularly strong gaps among ethnic German, Polish, and Western immigrants.
Another important gender consideration for partnered immigrant women's employment is their migration order, though to our knowledge, few studies investigate this in the European context. Krieger (2019) shows that both male and female lead immigrants enter the German labor market earlier than their tied (or equal) counterparts, but only male lead immigrants benefit in a higher employment probability. Thus, the gender imbalance in the selection into tied migration likely contributes to the gender gap in outcomes on the new labor market (if they choose to enter at all) (Krieger 2019).
The gender gap in hiring discrimination has also been explored in several European countries. An audit study on the Dutch labor market shows that immigrant women from non-EU countries (Turkey, Morocco, Suriname, Dutch Antilles) experience a smaller nativity disadvantage compared to immigrant men (11 percent vs 20 percent) (Andriessen et al. 2012). In France, another audit study had similar findings. Edo et al. (2019) found that women with non-French sounding names were less discriminated against (i.e., received more call-backs) than men with such names. In this case, including a signal of French language proficiency in the CV reduced the discrimination that non-French women faced, but did not have the same effect for non-French men.

Coinciding nativity and gender gaps
When it comes to looking at the coinciding nativity and gender disadvantages in employment and employment conditions, little evidence is available, which presents a largely heterogeneous picture across indicators. Descriptive evidence from the 2019 Eurostat data shows that, for most countries across Europe, the nativity gap in women's unemployment rates and prevalence of temporary job contracts is generally greater than the gender gap within the immigrant group (see Figs. 3 and 5)-that is, female immigrants are most disadvantaged because of their status as immigrants, not as women (see also Rubin et al. 2008). The opposite seems true for differences in part-time employment rates, which are for the most part driven by gender gaps (see Fig. 4).
Moving beyond pure description, two studies for the German case estimate nativity and gender gaps accounting for the relevant determinants. Salikutluk et al. (2020) show that the unconditional unemployment rates of first-generation Turkish women in Germany indeed are higher than those of native women (13.5 pp), immigrant men (1.2 pp), and native men (12.4 pp). Women from the FSU also show a (much smaller) nativity disadvantage, but still fare better than their male compatriots. Accounting for differences in human capital explains much of the observed gaps, and in fact, once education, labor force experience, German proficiency, and social contacts are all accounted for in the analyses, both groups of immigrant women are estimated to have lower unemployment probabilities than native and compatriot men. Moreover, no nativity gap remains among women, so this difference stems entirely from a gender advantage. 6 Fleischmann and Höhne (2013) investigate part-time employment rates by nativity and gender in Germany. Without conditioning on other factors, they find immigrant women to be more likely to work part-time than native women, immigrant men, and native men. However, net of age, education, marital status, children in the household, and duration of stay, they estimate that the part-time employment probabilities of immigrant women of most origins are actually lower than that of native women. As native and immigrant men do not differ for the most part, this makes for smaller conditional gender gaps among immigrants in terms of part-time employment rates.

Summary
Overall, the picture of female immigrant employment in Europe remains highly complex. The disadvantage that immigrant women experience is selective, and indeed in some cases an advantage is observed, where the outcome often depends on the studied country and employment indicator, as well as groups being compared. When considering descriptive data, unemployment rates vary more by nativity (with particular penalties for non-EU immigrants) than gender, and working hours more by gender than nativity. However, accounting for explanatory variables from the theory does not only explain large parts of these gaps, but makes the rather uniform descriptive pattern much more heterogeneous and less conclusive. Unemployment is often conditional on education, language skills, and inter-ethnic contacts, whereas marital status, the presence of children in the household, and cultural understanding are particularly relevant in determining working hours. Macro-level factors that describe the institutional context immigrants find themselves in are also important to take into account, especially when comparing different countries. However, few studies consider the migration order as a contributing factor to labor market disadvantage, and of the studies we reviewed, none addressed the potential difficulties that women may experience in accessing their trained occupation in the destination country. Thus, there is some potential to extend the field in these directions, especially for the cases where nativity and/or gender disadvantages still remain for female immigrants even after controlling for all these observable variables. This could also be evidence of labor market discrimination, which is difficult to quantify, and has not widely been explored in countries across Europe. In general, the heterogeneity in unconditional descriptive statistics is met with a considerable lack of conditional estimates, and this highlights the problem of overlooking gender-specific considerations when choosing a methodological approach-after all, studying labor force outcomes conditional on LFP is not such an important consideration for men. This oversight in the existing literature makes it tough to draw conclusions about more than a handful of countries, especially at the intersection of both nativity and gender disadvantages. At the moment, a general coinciding disadvantage for immigrant women seems neither supported in terms of their success in finding work when they are motivated to do so, nor in terms of working hours. The quality of the employment may be one relevant difference between groups, to be explored in the following subsection.

Labor market success II-Employment quality
In this final part, we turn to the literature which examines the quality of female immigrants' employment for those who have found it. This includes studies that focus on immigrant women's wages and their occupational status. Often these measures go hand in hand, since jobs with a lower occupational status also tend to pay less. In fact, in some studies of immigrant women's wages, occupational status is included as an explanatory variable. Evidence from Germany (Stypińska andGordo 2018), Sweden (Grand andSzulkin 2002), and Spain (Nicodemo and Ramos 2012) all show that occupational status factors into female immigrants' wage disadvantage. Nonetheless, studies also examine occupational status in its own right, either with regard to occupational segregation, over-qualification, or occupational mobility over time-all issues that impact the position of immigrant women on the labor market.

Nativity gap among women
Looking first at the nativity disadvantage among women across Europe, descriptive data again show that the success of immigrant women on the labor market is largely heterogeneous by their country/region of origin. Some groups of immigrant women experience a disadvantage because of their immigrant status, and this varies between the regions of Europe where they reside. In many NWE countries, immigrant women from western Europe and other industrialized countries experience a lower nativity penalty in their wages (or none at all) compared to women from other regions. In Finland, for instance, immigrant women from OECD countries earn on average 81 percent of the annual earnings of native women, though immigrant women from non-OECD countries have average annual earnings less than 50 percent those of native women (Sarvimäki 2011). For immigrant women in Denmark, data show they earn on average lower hourly wages than native women, though immigrant women from other Nordic countries actually earn higher wages (Husted et al. 2000). In SE, variation by country of origin has also been observed. The earnings of female immigrants in Spain who originate from Africa are especially low compared to native women, and compared to immigrant women from other origins (Amuedo-Dorantes and de la Rica 2007).
Similar descriptive evidence has been found for occupational status. Pooled data for several NWE countries show that female European immigrants are generally less disadvantaged in their occupational attainment compared to native women than non-European women (Gorodzeisky and Semyonov 2017). Disaggregated analyses from the same study show that the UK is an exceptional case, as European immigrant women actually have higher chances than native women of holding a high status job.
More evidence from Germany shows that female immigrants from Western countries do not differ in their average occupational status from native women, but females from most other countries do (Fleischmann and Höhne 2013). Findings from Rubin et al. (2008) also show that across Europe, immigrant women who originate from outside of Europe work more often in low-wage occupations than native women. Their results emphasize considerable occupational segregation between immigrants and natives, with sales and services, elementary occupations, and personal and protective services accounting for 40 percent of all immigrant women's employment.
In terms of human capital, two main findings emerge from studies across Europe. First, controlling for the level of education does reduce the nativity gap among women in earnings/wages and occupational status, but substantial differences often remain (Husted et al. 2000;Hayfron 2002;Bernardi et al. 2011;Kogan 2011;Sarvimäki 2011;Pichler 2011). Second, the remaining nativity gaps appear to trace back to unequal returns to qualifications. Female immigrants' education is significantly less correlated with their occupational attainment than that of native women in European countries (Pichler 2011). Studies show that particularly the returns to foreign education credentials of immigrant women are lower compared to native women in terms of occupational status (for Germany, see Kogan 2011) and in terms of earnings (for Norway, see Hayfron 2002). It comes as no surprise that immigrant women experience higher levels of over-qualification than native women throughout Europe, as descriptive Eurostat data from 2014 show (see Fig. 6). In all countries but the Czech Republic, Latvia, and Slovakia, a higher share of immigrants than native Nativity and gender gaps in over-qualification rates by country, 2014. Over-qualification measure based on respondent's self-assessment that qualifications and skills would allow more demanding tasks than current job. Immigrants include all immigrants for which there are data available, irrespective of origin. Age 15-64. Markers represent the values obtained by subtracting the comparison group value from the value for immigrant women. The dashed vertical lines separate northern-western Europe (NWE) on the left, southern Europe (SE) in the center, and central and eastern Europe (CEE) on the right. Lighter shades of markers represent the Eurostat flag 'low reliability.' Source Eurostat (2020f) women states that their individual qualifications would allow for more demanding tasks than required in the current job. The largest nativity gaps in over-qualification rates are present in Austria, Sweden, Italy, and Estonia, and range between 15 and 20 pp.
On top of an individual's human capital resources, household characteristics are again important determinants. Evidence from Norway shows that (net of education and age) marital status impacts female immigrants' earnings-married immigrant women in Norway earn about 8 percent less than their unmarried counterparts (Hayfron 2002). Whereas it is possible that these married women feel less economic pressure if they have a partner who is also contributing to the household income, children in the household play a role as well. Salikutluk et al. (2020) present evidence for Germany that immigrant women of Turkish or FSU origin earn about 4 percent less per hour than native women. Yet, this gap narrows after controlling for human capital, marital status, and the number of small children; for women of Turkish origin the gap vanishes completely. Moreover, Ballarino and Panichella (2018) show that across Europe, immigrant women whose husbands have a high status job also tend to have high status positions themselves, compared to women whose husbands have an unskilled job or are unemployed. This could be evidence of a network effect, that with high status jobs come high status connections that can be used when the partner is also seeking work, though this is not something explored by the study.
Most studies which control for both human capital and household characteristics find that the nativity gap reduces, but to differing degrees for each origin group. For immigrant women in Sweden, those from western Europe and North America on average do not differ from Swedes in their wages, while the wages of women from all other origins are significantly lower (Grand and Szulkin 2002;Behtoui 2008). This disadvantage is robust to controls for labor market resources and household characteristics. A similar picture is observed for non-Western immigrant women in other NWE countries, like France (Algan et al. 2010), Germany (Adsera and Chiswick 2006;Algan et al. 2010), and the UK (Algan et al. 2010), whereas European women are sometimes predicted to earn more than comparable natives (Algan et al. 2010). In Denmark, the considerable nativity disadvantage in wages is completely explained by differences in qualifications for women from Turkey and India, which is not true for women from Pakistan and Africa (Nielsen et al. 2004).
Something similar is seen with occupational status. In France and Sweden, non-European and European immigrant women have significantly lower net odds of holding a high status job compared to native women, even after accounting for education and household characteristics (Gorodzeisky and Semyonov 2017). In Belgium, this is only true for women of non-European origin, and in the UK no significant nativity gap is observed at all. However, in all countries, non-European immigrant women seem to fare worse than European immigrant women, though this difference is rarely significant (Gorodzeisky and Semyonov 2017). Fleischmann and Höhne (2013) find that for Germany, the nativity disadvantage applies to all groups of immigrant women in their occupational attainment, regardless of origin. Such patterns are found for Western and non-Western female immigrants in SE countries as well, in Spain (Amuedo-Dorantes and de la Rica 2007) and Italy (Fullin and Reyneri 2011).
The effects of country-specific migration regimes on the labor market outcomes of immigrant women are particularly salient for the SE case, as the demand for lowwage labor manifests in non-Western immigrant women being advantaged in terms of employment rates while being disadvantaged in terms of the wages or status of their employment. That being said, Pichler (2011) shows that the larger nativity gaps in occupational status among women in SE compared to other parts of Europe mostly stem from differences in welfare regimes (and associated demands in immigrant welfare workers), whereas differences in migration integration policy-such as access to citizenship, anti-discrimination legislation, or immigrant labor market access-do not have much explanatory power.
So far the literature indicates that for some disadvantaged groups of immigrant women, the compositional differences in human capital, household characteristics, and reception contexts do have some impact on the nativity gap, but cannot explain away the disadvantage in many cases. The final aspect that many of these studies consider is the impact of time on these outcomes. Thus, several longitudinal studies have looked at both the assimilation of female immigrants' wages and their occupational mobility compared to native women. Adsera and Chiswick (2006) find that, for female immigrants in the EU15 countries, their average earnings catch up to those of native women 18 years after immigration. For Germany specifically, Basilio et al. (2009) find some slight assimilation over the years, but immigrant women never attain the same wage as natives. In Finland and Sweden, different trends are seen for European and Western immigrants compared to women from other regions. Generally, women from Europe and other developed regions start with a smaller disadvantage in earnings right after arrival and assimilate towards native women, while women from other regions have a much larger disadvantage from the beginning, and their pace of assimilation tends to be much slower (Grand and Szulkin 2002;Sarvimäki 2011). Evidence from Denmark emphasizes that the assimilation trajectories vary among non-EU women. Nielsen et al. (2004) predict that, with 10 years of working experience, immigrant women from Turkey would close the wage gap with native women, but for women from Africa and Pakistan, any gains made by assimilation cannot offset the low remuneration of their qualifications.
Regarding occupational mobility, assimilation seems to be conditional on the labor market context. Ballarino and Panichella (2018) find that immigrant women in Italy tend to enter positions with a low occupational status and remain there. This same pattern is observed for immigrant women in Spain, while immigrant women in the UK and the Netherlands show evidence of starting in similarly low positions, but experience positive occupational mobility over time. These mobility trends may expose the strong demand for low-wage labor that is especially characteristic of SE labor markets. Evidence from Spain also shows that mobility differs by region of origin. Over the first five years of residence, women from Africa do not experience a significant increase in their likelihood to hold a higher ranked occupation vis-a-vis native women, whereas women from EU15, non-EU15, and Latin America do. Even so, though EU15 immigrant women exhibit a nativity advantage after this period, immigrant women of other origins remain disadvantaged in their occupational attainment (Amuedo-Dorantes and de la Rica 2007). Thus, while assimilation may occur, it is also very selective.

Gender gap among immigrants
Turning to the gender disadvantage in labor market success among immigrants, we might expect to find some similar trends between males and females. Foreign education and work experience is a commonality for both immigrant women and men, so the disadvantage that comes with transferring these endowments to the new labor market context should impact those with similar qualifications equally. Nonetheless, compositional differences in the proportion of immigrant men versus women that possess this type of capital from the country of origin will still be important considerations-something that is also tied to social norms in their country of origin. Moreover, immigrant women may face gender discrimination on the labor market in the host country in their wage and occupational status, something that women generally face on European labor markets (see, e.g., Weichselbaumer and Winter-Ebmer 2005).
Looking at descriptive data from the past two decades, female immigrants are almost consistently observed to earn less than male immigrants. This is shown by Husted et al. (2000) in Denmark for the earnings of female immigrants from all regions. In Finland, France, Germany, and the UK as well, there is a gender gap in wages among immigrants regardless of origin (except for Turkish women in France and Bangladeshi women in the UK), which highlights that this is a general disadvantage that most immigrant women face, though to a varying extent (Algan et al. 2010;Sarvimäki 2011;Salikutluk et al. 2020). The direction of gender gaps in occupational status, on the other hand, seems to differ by region of origin. Data from Germany show that ethnic German women, along with women from Poland, the FSU, and Spain/Portugal, on average do not hold significantly lower status positions than men from the same origin (Fleischmann and Höhne 2013). Female immigrants from most other origins do, however. Moreover, women from the Middle East are actually reported to hold on average higher status positions compared to men from the same origin (Fleischmann and Höhne 2013). As women from the Middle East also show lower average LFP rates in comparison to other immigrant women, it appears that only a selective and rather high-skilled group of women from this origin is employed in Germany, possibly being an effect of a selectively operating migration and labor market access policy. Moreover, evidence of gender-specific occupational segregation again appears for labor markets in SE. Immigrant women in Spain are found to end up in lower service occupations after immigration, regardless of their occupational status before migrating, while immigrant men tend to end up in lower industrial occupations (Fernández-Macías et al. 2015).
Again, the first step in investigating these gender gaps is accounting for human capital and household characteristics. Hayfron (2002) estimates for Norway that just about 10 percent of the gender earnings gap among immigrants can be explained by differences in human capital and marital status (along with controls for the region of residence, industry, and origin). This study attributes one-quarter of the remaining unexplained part to an overvaluation of men's endowments and three-quarters to an undervaluation of women's endowments. Based on a model accounting for differences in human capital and occupational characteristics, as well as selection into employment, Stypinska and Gordo (2018) report for Germany that immigrant women earn about 24 percent lower hourly wages than immigrant men-a disadvantage that persists over the life course.
In terms of occupational status, studies conducted in different countries have produced different findings. In the Netherlands, accounting for human capital and household factors explains the gender differences in occupational status for immigrants from Turkey, Bulgaria, Poland, and Spain (Ala-Mantila and Fleischmann 2018). For immigrants in Germany, however, the gender disadvantage in occupational status persists for most origin groups (except for immigrants from the Balkans) once these factors are taken into account (Fleischmann and Höhne 2013). This could be indicative that immigrant women tend to enter low-skilled occupations after immigration, regardless of their pre-migration status and qualification. In fact, the descriptive Eurostat data confirm that in 2014, in all countries except Finland and Sweden, immigrant women were more likely to be overqualified than their male compatriots (see Fig. 6). Cangiano (2014) suggests two different explanations for this picture across EU15 countries. First, immigrants entering as family migrants or refugees are often more likely to be overqualified than labor migrants, and women make up a larger part of family migrants. Second, female labor migrants show higher conditional over-qualification rates than male labor migrants, particularly in SE. When it comes to SE countries, the demands for female domestic and care workers are important drivers of gender-specific occupational segregation. In Italy, the gender gap in occupational status varies by duration of stay. Female immigrants on average start out in a position with a higher occupational status than their male counterparts, but they experience little occupational mobility over time because of overarching gender discrimination in the Italian labor market (Avola and Piccitto 2020). This being said, the occupational mobility of most immigrant men is not predicted to be substantial.

Coinciding nativity and gender gaps
As with the other indicators, there is scant literature comparing the wage and occupational status of immigrant women to that of native men. Nonetheless, the few studies that do make this explicit comparison provide a glimpse of the nature of the overall disadvantage that immigrant women face in their labor market success. As could be expected, descriptive data show that immigrant women have much lower average earnings than native men-in Norway, for example, immigrant women were observed to earn 76.3 percent of what native men earned (Hayfron 2002). Again, this disadvantage is selective and impacts some origin groups more than others. In Finland, immigrant women from OECD countries earn on average 61 percent of the annual earnings of native men, while immigrant women from non-OECD countries earn on average just 37 percent (Sarvimäki 2011). Algan et al. (2010) compare the hourly wages between native men and immigrant women in France, Germany, and the UK. They report the highest relative wages for north European immigrant women in France (114 percent), white immigrant women in the UK (114 percent), and EU16 immigrant women in Germany (118 percent). The lowest relative wages, on the other hand, show immigrant women of African origin in France (71 percent), of Pakistani and Bangladeshi origin in the UK (70/71 percent), and of Greek or Turkish origin in Germany (69/72 percent). In Germany, it is also women from Turkey who seem particularly disadvantaged in their occupational status relative to native men, even more so than female immigrants from the FSU (Salikutluk et al. 2020).
When it comes to investigating the disadvantage, many studies consider both the nativity and the gender aspect and the nature of their interaction while controlling for important compositional differences between groups. For instance, the analysis by Hayfron (2002) of immigrant earnings in Norway indicates that immigrant women face a coinciding disadvantage net of human capital and household factors, but the gender gap is between four and eight times larger than the nativity gap (depending on the model used). Another study from Denmark suggests that only some groups of immigrant women do experience a coinciding disadvantage in their wage: Husted et al. (2000) find that female immigrants from Pakistan are clearly disadvantaged based on their immigrant status and gender, while immigrant women from Nordic countries or India/Sri Lanka are only disadvantaged because of their gender. This result takes into account education, household characteristics, and time since migration. The heterogeneity in estimated nativity and gender gaps across immigrant origin groups as well as the empirical models used is also illustrated by two German studies. Stypinska and Gordo (2018) find no intersectionality in negative wage outcomes between the factors of gender and migration status for younger immigrants controlling for human capital and household factors. Immigrant women face wage disadvantages only because of their gender, to a similar extent as native women. However, Salikutluk et al. (2020) come to a different conclusion for immigrants of Turkish and FSU origin, incorporating additional controls for language skills and social inter-ethnic contact (though not accounting for selection into employment). They find that Turkish immigrant women's predicted wages do not significantly differ from the wages of native and Turkish men, but significantly exceed the wages of native women-corresponding to a nativity advantage for immigrant women. By contrast, immigrant women from the FSU face a double disadvantage as their conditional wages are significantly lower compared to native women (− 15 percent), FSU men (− 25 percent), and native men (− 23 percent) (Salikutluk et al. 2020).
Regarding occupational status, Salikutluk et al. (2020) find no gender gap among natives as well as Turkish and FSU immigrants, but immigrant women and men experience a strong nativity disadvantage. This nativity disadvantage can partly be explained by different levels of human capital, but only for women from Turkey and not the FSU. For immigrant women of both origins, the remaining gaps persist even when controlling for job characteristics, household factors, and language proficiency. Fleischmann and Höhne (2013) confirm a strong nativity disadvantage in occupational status in Germany for both immigrant women and men regardless of their origin, though they also report a gender disadvantage for immigrant women of Polish, FSU, eastern EU, and Western origin.
Gender differences in wages or occupational status among immigrants that remain after accounting for relevant factors can point to labor market discrimination. It is a well-established finding that women face additional obstacles on the labor market compared to men, but less is known if gender discrimination varies by nativity or vice versa. With regard to hiring discrimination, evidence suggest that there is, if anything, less discrimination based on nativity against women, though this finding might not extend to wages or occupational status after employment is found.

Summary
The reviewed studies on employment quality again do not point to a general or uniform nativity disadvantage for immigrant women in their wages and occupational status, but to a selective disadvantage for specific groups. Women from European countries and other Western nations tend to experience a smaller disadvantage (or none at all) compared to native women in their host country, and this disadvantage often comes down to compositional differences in human capital and additional resources (such as language proficiency and social networks). The nativity disadvantages for other groups of women, however, remain even after accounting for a broad range of relevant factors, and require further research to investigate other explanations. Particularly salient are the disadvantages in the returns to qualifications that immigrant women experience, many of whom are overqualified for their jobs throughout Europe. Moreover, the mobility in wages and occupational status that immigrant women experience over time varies widely across destination countries and origin groups, but in many cases disadvantages persist in comparison to immigrant men, as well as native women and men. A gender disadvantage is particularly present in terms of wages (less so for occupational status) in many host countries, and is often not just a matter of differences in human capital between immigrant men and women. Findings from the literature show that the gender disadvantage does not differ as much between immigrant groups within a given country, as it does differ based on the country where the immigrant settles. Thus, institutional factors seem to be important, such as the difference in gender structures between the labor market in the country of origin and the host country, and this consideration should be further explored in future research. Moreover, the household pressures that specifically immigrant women may be subjected to after immigration may contribute to the disadvantage they experience compared to their male counterparts.
Finally, the current literature does not provide a cohesive picture of the overall coinciding disadvantage faced by immigrant women in their labor market success. Some findings point to a gender disadvantage that is far greater than the nativity disadvantage. Other studies produce contradicting evidence of the intersectionality of these two dimensions of disadvantage. Given the fact that only a small proportion of the literature looks at this comparison, it is no surprise that the picture remains muddled.

Discussion and outlook
We have presented the most important contextualization for female immigrants' labor market integration in Europe and have reviewed the existing quantitative studies published within the past two decades which focus on this dimension of their integration trajectory. Our review of the empirical findings has been broken down by the outcome studied-labor force participation, (un)employment and working hours, and measures of employment quality-and in each section we examine two dimensions of the disadvantage that female immigrants face-the nativity disadvantage and the gender disadvantage-and the interrelation between the two.
As could be expected, the findings from the studies included in this review point to a very heterogeneous experience of immigrant women on the European labor market. A multitude of factors likely contribute to this variation, including the regional labor market specificities and different integration policies among the countries studied, different data sources and analysis periods, and different sample populations. Nonetheless, we have attempted to gather the important findings on female labor market integration in one place to expose overarching trends and potential gaps in the existing research, or oversights in theoretical/methodological approaches. We have identified three main take-aways from the review which we discuss below.

While the theoretically relevant determinants of immigrant women's labor market integration are often supported by empirical evidence, the extent to which these factors succeed in explaining the entirety of immigrant women's labor market experiences varies greatly across studies, calling for more systematic and comparative investigations in terms of intertwined individual and contextual factors
One phenomenon which is quite evident for all of the labor market outcomes reviewed is the importance of human capital factors in explaining the disadvantages that immigrant women (and men) are observed to encounter. The large majority of studies support the theoretical assertion that the level of education and prior work experience represent important contributions to immigrant women's (lack of) disadvantages at all stages of their labor market integration, but especially in terms of unemployment, wages, and occupational status. Household factors are important determinants of immigrant women's labor market outcomes too, although they appear to matter most in the decision to become active on the labor market and the average hours spent on paid work. Being married and the presence of children in the household significantly contribute to a gender gap in LFP among immigrants and natives, but the magnitude of this effect is also conditional on social norms and the associated gender roles, which can differ between immigrants and natives. Another commonality of studies is the finding that the disadvantages immigrant women face are often largest shortly after immigration and narrow with longer duration of stay, which is consistent with the predictions of assimilation theory. These trends can usually be attributed to adaptation processes in terms of acquiring host-countryspecific capital such as language skills or inter-ethnic contacts. However, although immigrant women experience upward mobility over time in many cases, significant nativity and gender gaps often remain.
The complexity of the issue often makes it difficult to pin down the reasons for immigrant women's disadvantages that remain after controlling for compositional differences in relevant determinants between women and men, immigrants and natives. In some studies, unexplained differences seem to stem from omitted variable bias, something that is most apparent in the case of host-country-specific capital (or duration of stay as a proxy), but also in terms of the relationship between traditional gender role attitudes and female LFP. However, omitted variable bias is rarely the only reason for persistent nativity and gender gaps in estimated labor market outcomes, and many studies suggest that immigrant women experience lower returns to the same characteristics compared to native women as well as immigrant and native men. Although there might be various reasons for such unequal returns, they invariably mean that immigrant women benefit comparatively less from their skills. Ethnic discrimination has been shown to disadvantage both immigrant women and men in hiring decisions, and there is much reason to believe that this discrimination extends to labor market outcomes as well. Moreover, immigrant women may face additional discrimination because of their gender.
Despite the persisting nativity and gender gaps in conditional labor market outcomes in many cases, there is little evidence for a general double disadvantage faced by immigrant women. While the results of intersectional studies acknowledge that the labor market outcomes of immigrant women are stratified by nativity and gender, the specific effects of the single dimensions as well as their interplay are by no means clear-cut and universal, but very selective. One of the most important conclusions from this review is the heterogeneous experience of immigrant women from different origins. Studying immigrant women as one aggregate group, or even in broad sub-groups, glazes over substantial differences. In some cases, it appears that distinguishing between EU internal immigrants and non-EU immigrants makes sense, because of the lack of institutional barriers that EU citizens face within Europe. However, studies show that European women themselves have different experiences when moving from one country to another, their experiences shaped, among others, by specific labor market demands, differences in the human or cultural capital they bring with them, or the different communities they encounter after arrival. Even more variation appears between immigrant groups from outside of Europe. These groups tend to be much more selective and their characteristics are heavily influenced by the immigration policy of the country they immigrate to. Immigration opportunities for women differ widely across destinations and over time, and the particular circumstances are reflected in the labor market outcomes of these women. One example is immigrant women showing high employment rates but low occupational status compared to native women in SE countries-countries with a high demand for care workers. The opposite seems true for countries in which immigration policy is so restrictive that only high-skilled individuals are able to immigrate via labor migration channels (e.g. Germany). The few immigrant women who meet these requirements are naturally successful on the host-country labor market. Yet, women entering via alternative channels such as family migration may also have ambitions to work, but do not undergo such rigorous selection processes. As a result, their labor market integration does not progress as quickly as labor migrants, and they are more likely to remain inactive.
Thus, the theoretical assumptions about the selectivity processes (or lack thereof) involved in immigration and their effects in terms of segmented assimilation trajectories have generally been supported. Still, we arrive at this conclusion in light of the historical and theoretical considerations we made at the beginning of this review. Studies that estimate to what extent the labor market outcomes of immigrant women are actually affected by the policies in place governing their access to a country and a country's labor market remain rare (see, however, Pichler 2011; Cangiano 2014). There seems to be much room to adopt a more gender-specific lens to selection issues in quantitative approaches to studying the outcomes of immigrants, particularly when this comes to measures of labor market success (conditional on being active in the labor force). Not only is immigration itself a very selective process, but the decision to participate in the labor market is another possible stage of selectivity, especially for women. Of the studies reviewed on labor market success, only a subset account for selection into labor force participation in their estimations. 7 Selection effects do not stop there. The data on immigrant outcomes can be further biased because of remigration, or naturalization (if the data only distinguishes immigrants by nationality). While this kind of selection is an important consideration for female immigrant labor market outcomes acknowledged in the literature (see, e.g., Dustmann and Görlach 2015), its gendered effects are rarely accounted for in empirical studies.
Quantitative studies which take a holistic approach to studying the labor market disadvantages of immigrant women-and all the considerations related to their gender and nativity that this entails-are rare in this body of literature, and future studies should address this A general oversight in the literature is the lack of studies which try to comprehensively understand the labor market disadvantage of immigrant women. Almost all studies of immigrant integration make use of a reference group in order to contextualize and evaluate the experiences of immigrant women, but by selecting only one comparison group, this pigeon-holes the analyses into one dimension: when immigrant women are compared to native women, it is only the nativity disadvantage among women which is investigated, and when immigrant women are compared to immigrant men, it is only the gender disadvantage among immigrants. While each one of these approaches is valuable, it is only when both of these disadvantages are examined together that we learn about the overall disadvantage of immigrant women in the labor force. Our review has made clear that the intersectional nature of disadvantage renders any empirical investigation increasingly complex. However, as we have repeatedly noted, intersecting inequalities cannot be properly estimated by focusing on different dimensions of disadvantage in sequence rather than conjunction. The studies by Fleischmann and Höhne (2013), Hayfron (2002), Husted et al. (2000), Nielsen et al. (2004), Salikutluk et al. (2020), and Stypinska and Gordo (2018) make a strong case for using quantitative approaches that allow to disentangle coinciding disadvantages. All these studies show, in one way or another, that nativity and gender gaps are not additive but interactive in many cases, and suggest that their specific relationship varies across destination countries, immigrant groups, and labor market outcomes. The limited number of studies and their exclusive focus on NWE mean that our preliminary and tentative conclusions are certainly subject to change as more and more studies adopt strategies suited to such endeavors, something we explicitly urge researchers to do.
Fruitful avenues for future contributions to this field include expanding on certain overlooked outcomes, like immigrant self-employment, as well as geographic regions that until now have received little attention, especially by employing the most recent data As far as the scope of the existing literature is concerned, it is fairly comprehensive in addressing the multifaceted nature of female immigrants' labor market integration. While certain outcomes, like labor force participation, employment, and wages, may receive the most attention, studies exist for most other relevant (measurable) outcomes too. Nonetheless, in our search of the literature, we noticed that at least one important outcome is generally overlooked: self-employment. The OECD reports that, in 2016, 10 percent of individuals who were self-employed in the EU were immigrants (2017). In many EU countries, the percentage of immigrants who are self-employed is comparable to the percentage of the native population that is self-employed, although in countries like Poland, the UK, Slovakia, Czechia, Croatia, Malta, and Lithuania, immigrants have a higher probability of self-employment than natives (OECD 2017). Yet, in the majority of studies we have included, selfemployment is not studied in and of itself. In some cases, self-employed individuals are grouped together into a general 'employed' dependent variable alongside individuals who were hired at a firm (Kanas et al. 2011;Gathmann and Keller 2018), while in other cases they are explicitly excluded from the analysis (Nielsen et al. 2004). Our searches returned only one study of female immigrant entrepreneurs. Constant (2009) finds that, after controlling for relevant factors, the wages of selfemployed immigrant women in Germany are no less than those of self-employed native women. For immigrant women who do choose to follow this path, it appears to be a lucrative choice. However, the lack of additional research limits the conclusions that can be made about female immigrant disadvantage in this area.
Furthermore, our review shows that for some countries in Europe, very little is known about immigrant integration, and for others, only older studies relying on rather outdated data have examined this topic. This situation calls into question the comparability of results, and their relevance to today's immigrant populations in Europe. The only countries with studies using data from 2010 or later are Germany, France, Norway, the Netherlands, Belgium, the UK, Switzerland, Spain, and Italy. No quantitative studies exist for countries in CEE (although some are no longer emigration but immigration countries), while other countries like Portugal, Greece, Austria, and Switzerland do not feature prominently in the literature. Of course, these conclusions are only based on the studies we have included in this review, and it is quite possible that a robust discussion exists on this topic in the native language of a given country. Overall, we can say that the 'internationally-accessible' literature is in need of research using more recent and more comprehensive data to construct a better picture of immigrants' labor market integration as we enter a new decade.
As a jumping-off point for future studies, we have supplemented the review above with descriptive analyses based on the most recently available European data for many of the labor market outcomes considered. In comparison to these data, it is evident that much of the heterogeneity across destination countries and origin groups is empirically underexplored, and that the situation in many countries has changed since several of the reviewed studies were first published (see also the trend Figs. A5, A8, A11, A14 in the online appendix from 8 . Going forward, important contributions to the field would include in-depth studies of countries for which there is less available evidence that not only investigate a wide range of labor market outcomes and their determinants, but explicitly consider different immigrant origin groups and arrival cohorts. In addition, there is still a lot of knowledge to be gained in pursuing comparative research. Cross-country studies allow one to examine the effects of structural differences in migration regimes, labor markets, and welfare states on immigrant women's labor market outcomes. Moreover, as nativity and gender gaps are conditional on the characteristics of the native and immigrant populations in the respective destinations investigated, comparing these estimates across countries can help us to understand the effects that the specific migration histories of these countries have had on the current labor market outcomes of immigrant women residents. Trend studies within single countries can additionally highlight the effects of temporal conjunctures and policy changes on the labor market success of immigrant women. In many cases, the reviewed studies provided only a snapshot of the current situation for a particular time point, estimates being subject to cyclical or other kinds of variation (Dustmann and Görlach 2015). Finally, adopting a temporal perspective can also help in disentangling age, period, and cohort effects when comparing labor market outcomes across immigrant groups (Altman 2015). All of these approaches can enrich the picture of immigrants' experiences across Europe, point to similar trends/barriers in their integration processes, and can promote more cooperation between countries on policy improvements.
Finally, we cannot present the findings of this review without also acknowledging its limitations. Most importantly, our review has only considered the findings from quantitative research in this field. This certainly has its drawbacks, because quantitative approaches can only attempt to capture the multitude of processes at play in immigrant labor market integration, both measurable and unmeasurable. A particular shortcoming of quantitative approaches might also be that disadvantages are usually equated with unexplained gaps in outcomes, running risk to the conflation of equality of opportunity with equality of outcomes. This would be the case if nativity and gender gaps in labor market outcomes were not disadvantages but the result of individual motivations, preferences, and choices. Luthra et al. (2016) emphasize this point by showing that migration duration intentions and individual motivations are important determinants of both socio-economic as well as social and subjective integration of Polish migrants across western Europe. Similarly, the concept of transnationalism, understood as being in a constant in-between state with no fixed place to settle, precludes assimilation as goal as well as end (Lutz and Amelina 2017). Qualitative research seems in many such cases to be particularly suited to illuminate the type of intersectional disadvantages that our review has attempted to present. Qualitative studies can also emphasize the considerable heterogeneity in employment trajectories and labor market disadvantages of immigrant women that are possibly uncovered by quantitative approaches such as shuttle migration (Morokvasic-Müller 2014). Moreover, qualitative studies are better positioned to investigate relevant topics like undocumented employment, which is difficult to gather reliable data on. Thus, investigating the findings from this second body of literature could be one important step in broadening our understanding of the issues we have uncovered.
Funding Open Access funding enabled and organized by Projekt DEAL. The research leading to these results has received funding from the German Science Foundation (DFG) under the grant agreement number 403158126.
Data availability All materials to reproduce the descriptive results are available as a Git repository (https:// scm. cms. hu-berlin. de/ spren max/ fem-lit-review). The data from both Eurostat as well as UNDESA are Open Access and are automatically fetched within the written programs, but we provide the respective URLs also in the reference list.

Declarations
Competing interests The authors declare that they have no conflict of interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.