Validation of the Employment Precariousness Scale and its associations with mental health outcomes: results from a prospective community-based study of pregnant women and their partners in Dresden, Germany

Key findings and summary

According to our findings, the EPRES-Ge is a useful multidimensional instrument for assessing dimensions of PE in Germany for this sample of expectant mothers and their partners, with most participants having a permanent employment contract. The measure yielded a very good response for the respective items with very few missing values and good reliability in almost all subscales. The overall factorial structure was confirmed by exploratory factor analysis with high item-subscale correlations and factor loadings. Construct validity was examined by CFA and the hypothesised five-factor model showed a very good fit to the data. Finally, the total score of the EPRES-Ge predicted several mental health outcomes at 8 weeks postpartum, highlighting the negative impact that PE may have on employees’ mental health.

Concerning temporariness, 80% of participants stated having a permanent work contract, yielding a low subscale score. Nevertheless, PE also affects workers with a permanent work contract,59 which is reflected in other subscale scores within our sample. As the authors of the EPRES state, the scale temporariness might underestimate the EPRES’ association with worker’s health, meaning that employees might score low on this subscale, but the effects of PE in other dimensions could still have an impact on employees’ health.59 This was also supported by our findings, making the application of PE measures such as the EPRES in samples of standard employment just as important as in samples of non-standard forms of employment.

A more theoretical explanation may be needed for the low reliability of the rights subscale. Cronbach’s alpha can be described as a measure of the reliability and consistency of the sampling instrument. It examines whether all items measure the same underlying construct. German government legislation protecting certain employee rights could explain a low Cronbach’s alpha in this sample. Most employees in Germany are entitled to unemployment benefits as well as pension benefits from the social security system. Therefore, employers are not necessarily expected to provide these benefits themselves. In contrast, most employers are required by law to provide paid parental leave, so the rights scale measures different aspects of rights provided by employers and/or the welfare system in the context of the German work environment. However, the association with mental health outcomes, particularly for men, demonstrates the importance of this scale, even though it measures different aspects of the work environment.

Only the subscale disempowerment could not be used to distinguish between PE and non-PE in this highly skilled German sample, as participants negotiated working hours and salaries that were in some cases more advantageous for the individual than the conditions imposed by the unions (eg, fewer working hours). Future research should focus on adapting the scale of disempowerment to different economic backgrounds. The authors of the EPRES-Se suggested reformulating the items of disempowerment according to the laws in force in each country,38 and the authors of the EPRES-E have adapted the scale to assess whether trade unions or work councils are generally represented in the company, whether the employee can exercise control over his or her own working hours and whether regular meetings are held to express the employee’s opinions.60

In conclusion, participants in our well-educated sample in Germany scored relatively low on the EPRES as it has been found in other populations of permanent workers in Spain.36 47 This was expected due to the unique sample of expectant mothers and fathers who are likely to be in a position to financially support a newborn child.

However, these scores are relatively low compared with other study populations. In the Swedish version of the EPRES, participants scored higher on all dimensions of the EPRES and on the total score (EPRES-Se=1.9). Similarly, participants in Chile also scored higher (EPRES-Ch=1.32), although Chile has a different economic background than Germany, making a comparison more difficult. Recently, an adapted version, the EPRES-E, was validated in Spain during an interview with only 13 proxy indicators60 adding the new dimension of Uncertain working times to the theoretical structure. A comparison of these EPRES-E scores across 22 European countries, including Germany, has recently been published.61 Germany was found to be the third worst-performing country in terms of overall EPRES-E score within 22 European countries. A particularly poor performance was found in the dimension of wages,61 which was also seen within this sample, where the scale wages yielded higher scores in comparison with the other subscales within our population.

There are a few possible limitations that could provide valuable insights into our findings and considerably low subscale scores. First, although the selective sample consisting of expectant parents provides valuable insights into the peripartum period, it is a limitation in terms of the generalisability of the results to other populations. For example, Vives, Benmarhnia10 suggested that especially elder employees are more vulnerable to PE and negative general and mental health consequences. As a second limitation, the sample was highly educated (see study protocol; Kress et. al. 2019), making the application of findings to the general population more difficult. Since higher education can lead to more stable and less PE,10 39 the transferability from our sample with over 55% of participants holding at least a university degree and 80.0% of participants reporting a permanent contract to the general population might be impaired. Third, the study included only participants with sufficient German language skills and formal employment contracts (full-time, part-time, marginal employment or apprenticeship, employment ban or parental leave), making the sample less representative for working conditions of the foreign-born population and non-standard employment or ‘occasional’ or irregular work. Overall, this could explain the lower EPRES-Ge scores and considerable floor effects found in this investigation of a sample with a high educational background and permanent contracts. To ensure the applicability of the EPRES-Ge in the German work context, we recommend it use in a wider context and in future standardised employee surveys in Germany.

First prospective associations with mental health outcomes

EPRES-Ge during pregnancy predicted mental health outcomes even after childbirth, when women are typically on maternity leave. This association was evident even after controlling for the major life event of childbirth at T2, 8 weeks after childbirth, and was also true for men who were not protected by the employment ban for women. It should be noted that we did not control for possible changes in employment conditions between T1 and T2, which must be taken into account when interpreting the results. However, in Germany, a pregnant employee is protected against dismissal by the employer under national law. It can therefore be assumed that women’s employment conditions were unlikely to change significantly during pregnancy. In addition, with regard to the period after childbirth, women who have given birth in Germany are not allowed to work for the first 8 weeks after childbirth, while many men continue to work with possible changes in employment conditions. To our knowledge, the present investigation shows the most differentiated results using the EPRES and multiple measures as indicators of mental health, so far. Especially, the scales vulnerability, wages and exercise rights showed the strongest associations with mental health outcomes. There appears to be a clear gradient pattern between increasing PE and poor mental health.37 40 41 Previous research in this area has already suggested that particularly low wages may be a risk factor for pregnant women developing symptoms of PPD after giving birth.33 In addition, it has been shown that women in continental welfare states, such as Germany, also tend to score higher on the wage dimension.21

This finding is supported by our analysis, where wages were more strongly associated with mental health outcomes in women and men than the other EPRES subscales (except vulnerability). Low income appears to have a strong impact on family health, even in a high-income country and welfare state such as Germany. The research team using the adapted version of the EPRES-E in countries across Europe confirmed this finding.61 Besides Germany being a high-income country, it had the third worst scoring within the subscale wages. To add to that, Germany has experienced the strongest growth of in-work poverty in the European Union with close to 10% of workers to be at risk of in-work poverty in 2017.62 This suggests that people in Germany might experience a worse ratio of income and living expenses than their European neighbours. Padrosa, Bolíbar61 add to the discussion by stating that the rise in low-paid and part-time jobs due to the Hartz reforms might explain the considerable high scores within the German population compared with other European countries.

However, in our analysis, there was no association between temporariness and mental health. Wagenaar, Kompier17 suggested a possible explanation for this finding in terms of the bi-directional causality of the ‘healthy hiring effect’, whereby healthy individuals are more likely to be selected for permanent employment.

As mentioned above, research has found that the association between PE and mental health varies by sex. Women were more susceptible to negative consequences of PE, both in general and in relation to mental health.10 20 63 However, this difference was not evident in the correlation and regression analyses of the present investigation. Still, further research on possible sex differences is needed, especially since women are over-represented in precarious working conditions such as part-time and marginal employment64 and have experienced an increase in low wages during the past years due to economic changes leading to an increase in PE.63

Nonetheless, a recent review of atypical employment and mental health in Germany concludes that PE is associated with mental health outcomes; however, the fit with the individual needs of the employer needs to be assessed before this association can be assumed.65 But even in a sample with low levels of PE such as ours, the associations found with mental health outcomes make it remarkably important to examine PE in the context of mental health problems. Yet, this study must be interpreted within a wider context. Women’s employment contributes significantly to a country’s economic growth and development. When women participate in the labour force, it increases the total labour force, leading to higher productivity and economic output.66 To better understand the context of women’s participation in the labour force, it is crucial to use tools such as the EPRES to assess employment during this period to systematically evaluate the employment conditions that might affect parents during this time. As our results suggest, PE can have a negative impact on mental health even in the presence of relatively favourable working conditions with a low EPRES score in a comparably young population, as was the case in this sample. Given the demographic situation and the importance of mental health, it is important to assess and evaluate PE that may have a negative impact on individual and family health, with consequences for further employment and the upbringing of the next generation of employees, the children.

This post was originally published on https://bmjopen.bmj.com