Migrant women ’ s employment in paid reproductive work through the crisis : the case of Italy ( 2007-2012 )

Paid reproductive work, especially in the case of cleaning and home-care for elderly people, is an important sector for foreign women in Italy. For this reason, since the beginning of the current economic crisis, scholars have wondered about the impact of the recession on migrant domestic workers. They have looked particularly at possible competition with Italian women entering the sector for lack of better alternatives. Our paper takes this discussion a step further by assessing the overall changes affecting migrant women in the Italian labour market, 20072012. We will look at how their position has been transformed, by taking both an ethnic perspective, in relation to Italian women, and a gender perspective, in relation to migrant men. By way of a conclusion, the argument will be made that there is a substantial lack of competition between Italian and foreign women in the care and domestic sector due to differences in their earnings, hours of work and activities.


INTRODUCTION
Italy is known to be a privileged destination for women migrating to seek work to support their families through remittances.These are women who have often found employment in the labour sector based on the delegation of reproductive work from Italian women to foreign women.This shift, from native to non-native, has expanded enormously, in the country, in the last forty years: and paid reproductive work, especially in the case of cleaning and home-care for elderly people, is a crucial sector for foreign women's employment in Italy (Qualificare, 2013;Castagnone et al., 2013;Ambrosini, 2013).
For this reason, since the beginning of the current economic crisis, scholars have been concerned about the possible impact of the crisis and austerity measures.Despite the fear that, due to the crisis, Italian women might have started to compete with migrants for these jobs, data has shown the relative stability of migrant women's employment in this sector; they have certainly had fewer problems keeping their jobs than their male counterparts (Bonifazi and Marini, 2013;Pastore et al., 2013;Reyneri, 2010).The general conclusion was that these foreign women have kept their positions thanks to being placed in a relatively "safe" sector, paid domestic and care work (Picchi, 2012;ISTAT, 2013;Farris, 2015;Semenza, 2012).
Our paper takes this debate a step further by identifying the elements of change and continuity for migrant women in the Italian labour market; along with their participation in paid domestic and care work between 2007 and 2012.We will look into how features of complementarity and competition have changed, both in ethnic terms (foreign women vs Italian women) and in a gender terms (foreign women vs foreign men).In particular, we want to verify how the increased participation rates of foreign women have affected the job security of two specific groups: Italian women and foreign men.Do foreign women compete or rather complement Italian women and foreign men in the Italian labour market?How have these patterns changed during the recession?
Our findings support the idea that, while no significant association is observed between foreign women and foreign men, there is "ethnic complementarity" between Italian and foreign women.This largely depends on the fact that foreign women are mostly employed in specific labour market niches -reproductive work especially.Here foreign and Italian women complement each other as they are employed in different occupations within the same sector or they are performing different tasks within the same occupation.The presence of foreign women has less impact on the employment status of foreign men seen that foreign women and men generally speaking tend to work in different sectors.Although some competition effects can be observed at exceptional moments -e.g. at the very beginning of the crisis in 2008 and 2009 -these Investigaciones Feministas 59 Vol.7 Núm 1 (2016) 57-74 effects disappeared relatively quickly.Overall the economic crisis encouraged migrants to adapt to rapidly evolving situations with emergency strategies.These solutions are, however, only temporary and not structural.
In the following pages, we will first set out the scholarly debates on the labour participation of migrants and/or women during economic recessions and parallel debates on the employment of migrant women in the Italian domestic and home-care sector.We will then discuss most recent data on the presence of migrant women in Italy by highlighting the tendency for "double feminisation".This "double feminisation" takes place in relation to the contemporaneous increase in the numbers of foreign women, and the resulting transformation of gender models and household strategies.We will, then, enter into the heart of our argument by elaborating on statistical data, noting the relation between migrant women's work on the participation of both Italian women (ethnicity dimension) and migrant men (gender dimension).The last section elaborates on the complementarity of migrant and Italian women in the domestic and care sector by using sectorial data to show the differences between their occupational profiles.

DEBATES ON GENDER, ECONOMIC RECESSIONS, AND MIGRANT WOMEN IN PAID DOMESTIC AND CARE WORK
There is a long-standing debate on whether women and/or migrants are a "reserve army": i.e. whether women do paid work in times of expansion, but retreat into unemployment and home-life in times of economic contraction (Milkman, 1976).Scholars have shown how already during the Great Depression and as recently as the most recent Asian financial crisis in 1997, women labour participation did not follow economists' expectation that women would be a flexible labour reserve (see e.g.Lim, 2000).On the contrary, women's labour participation in times of recession and shrinking opportunities is quite unpredictable.It depends, ultimately, on the specific configuration of the labour markets in each context, and women's possible complementarity/competition with other workers (males, migrants, etc.).
Jill Rubery and Anthony Rafferty (2013) took up this question while looking at women's labour participation (vs that of men) in the 2008/2009 recession in the United Kingdom.They found, in measuring female employment, that patterns of sectorial gender segregation were crucial; the same patterns that, in the past, had channelled women into administrative and clerical work.They concluded that the condition of these women workers was extremely volatile, and depended on austerity reforms in the public administration sector.In other words, no sector is inevitably "protected".External forces may determine job-loss and employment downgrading in fields that are traditionally seen as "safe".Sara Farris (2015) has discussed the factors that determine the general stability in the occupation of migrant women as domestic and care workers between 2007 and 2011 in Italy.She found that the main determinant was the non-cyclical character of this sector since, regardless of general economic trends, the demand for personal care services is always high due to an ageing European population.This sector has, also, remained almost exclusively occupied by migrants due to its "very severe, unregulated, stigmatized and poor working conditions" (Farris, 2015: 12).Thus she defines domestic and care workers as a "regular" rather than a "reserve" army to emphasise the chronic necessity of these occupations in European societies.This phenomenon is particularly evident in the case of Italy.There the employment of migrant women in the domestic and care sector continues to be the most common solution given the care needs of a growing population of elderly and sick people.True, there have been changes.There are new actors in the provision of care such as charity organisations, local administrations, cooperatives and other market actors.However, family-based provisions remain the most important approach (Rusmini and Pasquinelli, 2013;Del Favero, 2011;Network Non Autosufficienza, 2011;Sarti, 2008).Families are seen as the natural setting for the support of sick and ageing members: this has persisted in Italian policies despite the increase in non-traditional (single, no children, etc.) households (ISTAT, 2013).Children and other relatives are thus ultimately responsible for providing full-time assistance to the elderly and the ill, directly and/or managing its delegation to paid workers (Ambrosini, 2013;Marchetti, 2015).This mechanism is what determines a permanent demand for a low paid, flexible and highly-gendered workforce, ready to be employed by Italian households when necessity arises.Such workers being typically most easily found amongst (undocumented) migrants, Lena Näre (2013) rightly points to a "transnational familism".
One can say that Italian state policies in the fields of welfare, migration and labour generally emphasise, rather than contrast, the continual demand for a foreign workforce for paid domestic and care work.For example, the government regularly establishes quotas for the employment of migrants in this sector as it would normally do for industrial and agricultural work.Moreover, in the regularisation procedures for undocumented migrants that Italy has launched in recent years, domestic and care workers have often received preferential treatment (Sarti, 2016).These migration policies go hand-in-hand with welfare and labour policies which favour the hiring of external help.Franca Van Hooren (2012) argues that the "migrant in the family model" is enhanced by state support mainly consisting of monetary allowances to households with dependents, with usually no conditions on how these funds are spent.It is also true, as Barbara Da Roit and Bernard Weicht (2013) have pointed out, that segregation in the Italian labour market and the presence of undocumented migrants willing to work as domestic workers would be alone, without care benefits, sufficient factors in creating this model.
Finally, in understanding the use of foreign migrant labour in the family, it is important to note that the increasing participation of Italian women in the labour market has not changed traditional gender roles within families.Migrant women are needed to substitute Italian women in their traditional role in unpaid reproductive work.In other terms, the traditional gendered division of labour has been maintained thanks to a new gendered and ethnicised workforce that, from the 1970s, has bolstered this mechanism (Andall, 2000;Marchetti, 2015).Italy is an example of the international division of care that has been so widely discussed by feminist scholars.It shows the inextricable connection between economic globalisation, transnational Investigaciones Feministas  61  Vol. 7 Núm 1 (2016) 57-74 migration and the reorganisation of reproductive work along gendered and ethnicised lines (see the recent book by Kofman and Raghuram 2015 for this debate).Our paper offers an example of how the global economic crisis has affected this situation.

THE "DOUBLE FEMINISATION" OF MIGRATION TO ITALY. SOME DATA FOR GENDER AND ETHNICITY
Women's migration to Italy is paradigmatic of the double meaning of the "feminization of migration": feminization refers to an increase in the quantity of women who migrate; but it also points to transformations in the quality of labour market participation, migrant household strategies and gendered models, behind the exceptional participation of foreign women in the Italian labour market.
Since the 1970s, the entrance of migrants to Italy has been predominantly related to work purposes.Especially since the 1990s their number has significantly increased (table 1).On the 1st of January 2015, women accounted for 52.7% of 5,014,437 foreign nationals residing in Italy (femininity ratio=1.11).If we consider that 20 years ago (1995), this value stayed at 43.9% (femininity ratio=0.78), it can be said that the feminization of migration has been a rapid growing phenomenon.In addition, in 2015, women represent 46.4% of the total foreign workforce -amounting to 3,874,726 -with slight differences according to area of residence (45.1% in the North, 48.5% in the Centre and 47.9% in the South) (Labour Force Survey data).This increase resulted from two parallel dynamics: the rise of admissions for family reasons -which has typically involved more women than men; and the growing number of women migrants reaching Italy for work reasons.The increased propensity of women to reach Italy for work, meanwhile, makes Italy atypical for Europe.
However, when we come to the question of what the feminisation of migration might mean for a transformation of gender roles, the issue becomes more complex.The migratory patterns, and the related role women have in their households, differs according to the background of foreign women in Italy (figure 1).By looking at the first ten nationalities living in Italy, a "women-led migratory model" is observed among foreign nationals from Eastern Europe (Ukraine and Moldova), Latin America (Peru and Ecuador) and the Philippines.Here, the proportion Women-led migratory model

Gender neutral model
Investigaciones Feministas 63 Vol.7 Núm 1 (2016) 57-74 of women ranges from 58% (Philippines) to 80% (Ukraine).This group is also characterized by a high proportion of migrant working women: at least 56% of women hold a residence permit for work reasons, a number which reaches, respectively, 77% and 68% for Ukrainian and Moldovan women.On the contrary, a "men-led migratory model" is observed among Maghrebi (Morocco, Tunisia) and Indian nationals with women representing, respectively, 44%, 37% and 39%.In such cases, the proportion of workers among women is negligible: 18%, 9% and 11%.Finally, a "gender neutral model" is observed among foreign nationals from China and Albania with women representing almost 50% of all permits.But while Chinese women are more likely to be in Italy for work reasons (50%), only 14% of Albanian workers are women.Therefore, among these nationalities, seven out of ten belong to either a "women-led migratory model" or a "gender-neutral model".
The Italian labour market is characterized by high segmentation dynamics in terms of both gender and ethnicity: which means different job opportunities, sectors of employment, and occupational levels.Therefore, when trying to understand women labour performances, it is important to look at the way migrant women begin working.It is important to do this comparing foreign women both to their male counterparts (gender), but also to Italian women (ethnicity).It is in this light that Table 2 should be considered.
The labour market indicators by sex highlight the disadvantaged position of foreign women with respect to their male counterparts.Foreign women are, indeed, less active, less employed and more often unemployed than foreign men.Specifically, the gender differential (foreign women -foreign men) equals -22 percentage points (p.p.) for the activity rate, -21 p.p. for the employment rate and +3 p.p. for the unemployment rate.It is worth mentioning that similar differentials also apply to Italians which means that women perform worse than men in the Italian labour market, in all geographical areas (North, Centre, South and Islands), regardless of their nationality.
However, if we look at labour market indicators of women by nationality (ethnicity), we see that foreign women are more likely to be working than their Italian counterparts: the only area where this is not true is the North of Italy.Specifically, ethnic differentials (foreign women vs. Italian women) equal 8 p.p. for the activity rate and 4 p.p. for the employment rate.It is along these lines that Italy can be grouped together with Greece, Portugal and Spain into the "Mediterranean model of migration" (Strozza et al., 2009;ISTAT, 2009).All of these countries share a high proportion of working women amongst first-generation migrants, with higher rates of activity and employment than their native counterparts.The unemployment rate is, meanwhile, higher for foreign women (15.7%) than for Italian ones (11.4%).The only exception is southern Italy where foreign women and men are less likely to be unemployed than Italians because of their higher propensity to move within regions to take new job opportunities (Reyneri, 2007;Strozza et al., 2009).We will see in the next section how this model has been affected by the economic crisis.

THE IMPACT OF THE CRISIS: COMPLEMENTARITY AND COMPETITION AMONG WOMEN, MEN AND MIGRANT WORKERS
The economic crisis has certainly affected labour market dynamics and, in particular, the choice about which family members seek work.Yet, when looking at participation rate trends (figure 2), no relevant changes are observed among Italian workers, with the exception of a slight increase in women's activity rates.On the contrary, by observing labour market participation dynamics among foreign nationals, two opposite trends are worth noting.The significant decrease in foreign men's activity rate is counterbalanced by a slight increase in that of foreign women.
Investigaciones Feministas 65 Vol.7 Núm 1 (2016) 57-74 How can we explain this?The decline of labour market participation observed among foreign men is certainly linked to the continuing reduction of migration flows for work reasons.What happens to foreign women deserves, instead, much more attention in as much as it marks a transformation in household labour supply strategies.Following the recession and consequent difficulties for low-income families, the labour supply of foreign women seems to have increased significantly.What, then, has been the effect of foreign women on the employment conditions of those who were already in the labour market?What role does the crisis play in this?Accordingly, we must empirically verify whether the increase in labour market participation for foreign women has had any negative effect -in terms of competition with either Italian women (ethnic approach) or foreign men (gender approach).The analysis was conducted, 2007-2012, using Italian Labour Force Survey data.
A first rudimentary overview of these dynamics is given by the correlation between the share of employed foreign women and the unemployment rates of Italian women (table 3) and of foreign men (table 4).As to the ethnic approach, the correlation between the share of employed foreign women and the unemployment rates of Italian women is negative with coefficients ranging from -0.7 and -0.8.These preliminary findings seem to suggest that the two populations do not compete, even in times of crisis.Figure 3 shows, indeed, that foreign women tend to be concentrated in regions characterized by lower unemployment rates.

Figure 3. Unemployment rates and shares of foreign women in employment by region, Italy, 2012
Source: Authors' elaboration on Labour Force Survey data.
This is partially explained by the scarce propensity of southern Italian women to move towards other Italian regions: something decided, in part, by low accommodation prices, and by extended-family support in their home region.Foreign women, of course, do not have any territorial attachment and more easily move towards regions with higher labour demand and low unemployment (Venturini and Villosio, 2006).The general lack of social or welfare support that characterises Italy makes it difficult for migrants to remain unemployed or inactive: in this sense "working or leaving" are the only options for them, whether this means going abroad or shifting their base within Italy (Di Bartolomeo et al., 2016).The same analysis conducted in gender terms gives, however, different results (see table 4).The correlation between the share of employed foreign women and the unemployment rates of foreign men is positive, though coefficients are low.Yet, being based on aggregates, these findings can hide differences by sex, age, education and civil status.In addition, local labour area characteristics significantly influence these trends.
Thus, employing micro-data from the Italian Labour Force Survey (LFS), we isolated the effect of foreign women on the Italian labour market using cross-sectional datasets1 for 2007-2012, i.e. before and during the crisis.For each year, we implemented two logistic regressions (with clustering by Italian regions) on two subpopulations: 1) Italian women (ethnic approach) and 2) foreign men (gender approach).The dependant variable is the displacement risk (transition from employment to unemployment status): it equals 0 if employed at time t-1 and employed at time t and 1 if employed at time t-1 and unemployed at time t.The key independent variable is, instead, the share of employed female foreign migrants on the two populations of interest, which proxies the presence of foreign women on the Italian labour market.More precisely, to look at ethnic dynamics, we used the share of employed foreign female migrants on the total female employed population while, to look at gender dynamics, we used the share of employed foreign female migrants on the total foreign employed population.The effect of this key variable (net of individual and regional macro-economic characteristics) provides evidence for the impact of female migration on the employment dynamics of both Italian women and foreign men.As control variables, age, educational level and marital status are included at the individual level.In order to control for labour market local conditions, at a regional level, we included the growth in added value between time t-1 and time t and the unemployment rate for controlling effects on the demand side (Venturini and Villosio, 2006).The gross firms' creation rate (i.e. the sum of firm mortality and birth rate) was included as well.
Table 5 summarizes the core results of this analysis, reporting the effect of female migration on the risk of Italian women and foreign men losing their jobs.

_____________
It is interesting to note how the risk of job-loss for Italian women is not negatively associated with the presence of foreign women on the Italian labour market.On the contrary, one can generally observe a positive association between the higher presence of foreign women and the lower risk of Italian women losing their jobs.This seems a quite stable trend which was interrupted only in 2008, the first acute year of the crisis, and again in 2012.On the other hand, in some periods, the job security of foreign men is significantly influenced by the presence of foreign women: in the first two years of the crisis (2008 and 2009), the higher the presence of foreign women the higher the displacement risk for foreign men.
In conclusion, when looking at "normal" times, our results seem to suggest ethnic complementarity between Italian and foreign women, while no significant association is observed between foreign women and foreign men.The former can be explained by the fact that foreign women are mostly employed in specific labour market nichesespecially reproductive work -where foreign and Italian women complement each other, "being employed in different occupations within the same sector" or "performing different tasks within the same occupation".The former occurs, for instance, in the long-term care sector, where different occupations (doctors, skilled nurses, caregivers, etc.) are clearly characterized, not only by ethnic segmentation but also by a high degree of complementarity.Even within the same occupation, workers' profiles may differ according to their tasks, in terms of availability and/or expertise.This is also the case in paid domestic work, as we will see in the next section.
The lack of significant association in gender dynamics can be explained, instead, by the fact that foreign women and foreign men act in separate labour markets.Some competition/substitution effects can be observed at exceptional points in time: e.g. at the very beginning of the crisis in 2008 and 2009.However, these were balanced out within a very short time.This may signal the fact that migrants have adopted emergency strategies to respond to transformations provoked by the crisis, but that these were temporary not structural changes.

THE COMPLEMENTARITY BETWEEN MIGRANT AND ITALIAN WOMEN IN PAID DOMESTIC AND CARE WORK
Figure 4 gives us an idea of the complementarity by ethnicity which exists in the Italian labour market and compares the occupational profile of foreign women with that of Italian women.Their occupational profiles rarely overlap as they differ according to required labour skill.Seven out of 10 foreign women are employed in low skilled jobs, whilst only 1 out of 10 Italian women are employed in such occupations: i.e. elementary occupations, like craft and related trade workers and as plant and machine operators and assemblers.Italian women are mostly concentrated in mediumhigh skilled jobs, with 7 out of 10 employed as technicians and associated professionals (28.9%), as clerks (22.8%), and professionals (13.6%).These same sectors employ only 11.4% of foreign women.Moreover, as suggested in the previous section, these two profiles seem to be largely complementary within this sector and are characterized by very different profiles.Compare, for instance, working conditions as proxied by number of hours worked per week and by annual income.In terms of working hours, Italian women typically work fewer hours per week than foreign ones: 9 out of 10 Italians work fewer than 29 hours per week, while 8 out of 10 foreign women work for more than 20 hours (figure 5).Earnings are consequently very different between the two populations (figure 6): 1 out of 2 Italian women earn fewer than 5,000 euros per year (vs. 3 out of 10 among foreign women) while 5 out of 10 foreign women earn more than 6,000 euros (vs. 3 out of 10 among Italian women).This shows that, for Italian women, jobs in the domestic sector are still seen as "extra" monetary entries used to compensate the salaries of other members of their households.On the contrary, for foreign women these same jobs often represent their main source of income, which leads them to work longer hours.This usually entails cohabitation with their employer, and thus they sacrifice their autonomy and independence, in line with the growing need for around the clock home-care service for elderly people.Italian women are reluctant to work under such conditions.
The employment of migrant women workers is characterized by higher intensity and less remunerated jobs than those taken on by Italian women.In fact, the profile of Italian and foreign women workers rarely overlap, even within the same sector, i.e. domestic work.Even comparing Italian and foreign women in the care and domestic sector, our earlier conclusions remain valid and there need be no fear of "stealing jobs" as long as labour profiles remain distinct.

CONCLUSIONS
In this article we described the evolution and patterns in women's migration to Italy and of migrant women in the labour market there.We focused on their performances in terms of employment, unemployment and activity rate.In so doing, we used a double approach by showing how both gender and ethnicity can explain the generally disadvantaged position of migrant women in labour market outcomes, with respect to migrant men and to Italian women.
Second, we asked whether the role of migrant women in the Italian labour market has changed with the economic crisis.We have shown that foreign women were relatively less affected by the recession than foreign men and Italian women.Indeed, in terms of unemployment and activity rates, both the gender and the ethnicity differentials had dropped significantly by 2007 and 2012.In this sense, when it comes to the crisis, one can actually talk of an "advantage" for migrant women.
A third issue dealt with in the article was the possible "competition" between groups in search of occupation: competition between migrant men and women, but also between migrant women and Italian women.With respect to the former, we have shown that it was the involvement of foreign men in such sectors as construction, manufacturing, financial services and trade that has gone against them: quite simply these have been hit more harshly by the crisis in comparison to others.The employment of foreign women in sectors such as care and domestic work has, instead, "protected" them, given the "rigidity" of demand in these sectors.In this sense, this paper supports the idea that even when there is a shortage of labour opportunities the labour participation of women is likely to increase.
Finally, we expanded our focus from the risk of male-female competition to the risk of ethnic competition: that is competition between non-Italian and Italian women.We asked to what extent and in which direction the role of foreign women has changed within the crisis in comparison to Italian women.By way of an answer, we have shown how foreign women do not actually compete with Italian women working in the care and domestic work sector.This is because the performance of the two groups in terms of working times, remuneration and assigned tasks are so radically different.

Figure 1 .
Figure 1.Foreign population holding residence permits by sex (a) and foreign women holding residence permits by type of permit, first 10 nationalities, Italy, 2011 (Jan 1st): migratory models

Figure 5 .
Figure 5. Foreign and Italian women domestic workers who paid at least one pension contribution by annual average working hours per week, 2005-2014 (% values)

Figure 6 .
Figure 6.Foreign and Italian women domestic workers who paid at least one pension contribution by annual average income, 2005-2014 (% values)

Table 2 .
Labour market indicators by geographical area of residence, nationality and sex,Italy, 2012 (*) Notes: (*) ITA stands for Italian nationals while FOR stands for foreign nationals.Source: authors' elaboration on Labour Force Survey data.

Table 3 .
Ethnic approach: Pearson's correlation coefficients between the share of employed foreign women (on total employed women) and unemployment rates of Italian

Table 5 .
The effect of the presence of female migrant workers on the risk of Italian women

Table 6 .
Foreign and Italian women domestic workers who paid at least one pension contribution bynationality, 2005-2014 AMBROSINI, MAURIZIO (2013): Irregular Migration and Invisible Welfare.New York: Palgrave Macmillan.ANDALL, JAQUELINE (2000): Gender, Migration and Domestic Service: The Politics of Black Women in Italy.Burlington: Ashgate.BONIFAZI, CORRADO & MARINI, CRISTIANO (2013): The Impact of the Economic Crisis on Foreigners in the Italian Labour Market.Journal of Ethnic and Migration Studies, 40(3): 1-19.