Immigration, work and health in Spain: the influence of legal status and employment contract on reported health indicators

To the best of our knowledge, this study was the first to analyze the relationship between legal status, employment conditions, and immigrants’ health, providing an unprecedented look at the experiences of undocumented workers. In the findings presented, we see trends that worker health is more associated with employment conditions than legal status. This is reflected in the frequencies of both poor self-rated health and mental health across the seven legal/contract situation categories, in the odds ratio of mental health problems (male undocumented foreign-born workers, time in Spain ≤3 years; and male documented foreign-born workers with temporary contracts, time in Spain >3 years), and in the odds ratios of poor self-rated health (male Spanish-born workers with temporary contracts; male undocumented foreign-born workers, time in Spain <3 years; and female documented foreign-born workers, time in Spain >3 years, with temporary contracts or without contracts).

The observed association between employment conditions and health is supported by a recent review of relevant evidence (Benach et al. 2007), which highlighted the prevalence and health consequences of unfair and informal employment. Another recent study in Spain (Borrell et al. 2008) examined the influence of migration on health, and the role of employment conditions, along with social class and other factors. The authors found that migration affected self-rated health in both males and females, even after controlling for social class. Our findings regarding the effects of employment conditions on health agree with this existing literature. The less expected outcome—the relative importance of employment conditions over legal status—reemphasizes the effects of flexible employment in an already vulnerable population. Additional logistic regression models were constructed to compare Spanish-born and foreign-born workers with the same contract types and no statistically significant differences emerged between the two populations (data not shown).

Interestingly, we also observed variation in legal/contract situation’s effects by health measure between genders. In males, the 12-item mental health indicator was more associated with legal/contract situation than was the 1-item self-rated health measure. In females, the opposite was true, with self-rated health more sensitive than mental health to legal/contract situation. Similar differences have been observed in health outcome sensitivity to various factors among men and women in previous studies in Spain (Borrell et al. 2008; Rodriguez Alvarez et al. 2008). However, it is also important to study the influence of other potentially related variables, such as time since migration, working conditions (as they vary across social class and economic activity sectors), and perceived discrimination. Future research would be well guided to explore the health effects of these factors, as well as the potential effects of changes in individual immigrants’ social class between the origin and host societies.

The analyses yield surprising results regarding the relative health status of undocumented migrants, especially in females and foreign-born workers who have lived in Spain longer. These quantitative findings counter those of the earlier associated qualitative study, as well as those of other qualitative studies (Ahonen et al. 2009b; Simich et al. 2007). Although we cannot conclusively explain these differences, one potential explanation involves a possible relationship between legal status and time since migration. Immigrants who enter a host country undocumented and transition to a documented status have a longer history of cumulative precarious experiences than do more recent still-undocumented migrants. The change in documentation status may improve their situation, but they are still in vulnerable positions, and may retroactively be feeling the effects of their previous experiences as undocumented migrants. It is also conceivable that the transition to a more stable documented status acts as a catalyst for the expression of the effects of their previous more precarious situations.

All findings presented here should be interpreted with consideration of the study design constraints. As previously noted, migration research presents a variety of methodological issues, and this study is not without limitations. First, legal status is a sensitive issue for foreign-born subjects, with the potential for a perceived risk of legal consequences and/or deportation. To reduce the potential non-response bias and/or false responses, we chose to define legal status as permission to work rather than permission to reside. As this introduces the issue of measures of legal status, it is noted that findings presented here relate to undocumented workers; the degree to which undocumented worker and undocumented resident populations overlap, as well as a comparison and contrast of health effects between the two groups, requires future research. The time in Spain stratified analysis adds a dimension to the documented and undocumented overall situation. Time in Spain is considered to control for the healthy migrant, establishing differences between recent immigrant workers and the rest of the immigrants. However, sample size limitations prevent us from reaching reasonable statistical power, and it still insufficiently corrects for the “healthy immigrant effect” (which takes more than 3 years to disappear). Further analyses highlighting this situation and carried out in different immigrant groups would help in clarifying this situation.

Second, our sampling methods may compromise the external validity of the findings. The data for this study were collected under quota sampling methodology. The hidden nature of immigrant populations, especially the undocumented subgroup, has prevented a complete accurate description of the population of interest, prohibiting representative random sampling. Quota sampling offers a means to analyze across key demographic variables to try to get a meaningful look at heath of undocumented foreign-born populations (Bloch 2007). However, the potential effects on the ability to extrapolate findings to a larger population should be noted. In addition, it is important to take sample size limitations into account. A larger sample size would improve our statistical power. Nevertheless, characteristics of the immigrant population may be sufficiently represented by the included participants, given the nature of the research and the selected explanatory variables.

Finally, the cross-sectional nature of the data reduces the ability to assess causality in the observed relationships. We cannot be sure if working without contracts negatively affects workers’ health or if poor health status somehow acts as a limiting factor in attaining a job contract. The literature surrounding the healthy immigrant effect, however, supports health changes with time in the host country (Newbold 2005), boosting our confidence of the relationship’s direction and the conclusions drawn from our findings.

Accepting the above limitations, this study adds an examination of a large sample of undocumented immigrants to the existing literature on migration and health. We introduce a new combined analysis of employment conditions together with legal status to assess factors relevant to immigrants’ health. While the concrete findings should be interpreted considering the afore-mentioned limitations, we do get a glimpse into the growing global phenomenon of undocumented migration, about which much remains undiscovered. Although our population of interest’s hidden nature necessitated our deviation from ‘gold standard’ sampling methodology, the extent to which our findings agree with those of the preceding qualitative study (Ahonen et al. 2009a) increases our confidence in our current results. In both ITSAL phases, the importance of employment conditions, specifically contracts, emerged from the analyses. In the qualitative study, legal status also stood out as a factor increasing worker vulnerability and exploitation, while in this quantitative phase, legal status is observed as a less influential factor to health. An exploration of the assessment of legal status, and its effects on employment conditions and health outcomes has begun here, and future investigation would be well guided to continue to expand upon it towards more absolute understanding.

It should also be noted that this study was conducted at the end of an economic cycle with high rates of employment. It will be interesting to see how the situation changes as the upcoming economic dynamics impact a job market that already excludes many foreign-born workers, with 28.4% unemployment, as compared to 17.4% in Spanish-born workers (Spanish National Institute of Statistics, INE 2008).

This study presents an attempt to handle the methodological limitations inherent to migration research, and serves to aid the development of migration research methodology. Methodology developed from these experiences will potentially produce more generalizable findings via representative samples and inclusion of other representative collectives in Spain. Future research would be well guided to continue to explore the relationship between legal status, employment conditions, and health in immigrant workers. Other possible factors to consider include specific contract-related factors, such as duration, oral versus written nature, and the qualitative meaning of contract types in various economic sectors. This research can also be more rigorously compared with the existing knowledge of the effects of employment conditions on health (in general populations) to distill immigration-related health changes, and to help guide societal development, accommodating immigrants’ needs and protecting the rights and health of all people.