The impact of Active Labour Market Policies on health outcomes: a Scoping review

Abstract

Background

Since the last Western great recession of 2008, an increasing interest on the effects of Active Labour Market Policies (ALMPs) on improving health has emerged. The aim of our review is to synthesize evidence on the effectiveness of ALMPs and whether some types of ALMP are more effective.

Methods

Using the Scoping review methodology, we conducted a literature review in PubMed/MEDLINE, Scopus and Web of Science and selected articles published between 1990 and 2017 in high income countries. We applied four sequential phases of document screening to the list of retrieved articles.

Results

Of the 416 documents detected in the search, 36 documents were finally selected. Most of them focused on mental health and related components (72.2%) and found positive results at least in one outcome (80.6%). The ALMPs reported mainly attempt to build capacity through job search assistance (31.6%) with a component on mental health, showing positive results on health; some offer job training (16.7%) and few subsidized employment (8.3%), showing more controversial results, although mostly positive. The rest include a combination of different types of ALMPs.

Conclusions

This review shows that ALMPs have a positive impact on health and quality of life. There are relatively few studies of the impact of ALMPs on general health, and most of the studies found are focused on Anglo-Saxon and Nordic countries. The most significant knowledge gaps are the mechanisms involved in achieving this improvement, and above all the differential health impacts according to axes of inequality and welfare state.

Introduction

Employment and working conditions have long been recognized as key social determinants of health and health inequalities.^1–3 The lack of employment has been associated to poor health.⁴ The mechanisms underlying the relationship between unemployment and health can be grouped into psychosocial and material factors. Regarding psychosocial mechanisms, Jahoda⁵ argues that individual well-being relies on five fundamental needs: time structure, social contacts, collective effort or purpose, social identity or status and regular activity; which can be met by any activity but that most workers satisfy them by participating in the labour market. Moreover, job loss is associated with low self-esteem and stigmatization.⁶ Fryer⁷ also suggests that an individual’s subjective well-being depends on feeling that they are in control of their life course, as well as economic matters. Unemployed individuals are prevented from being economically self-sufficient and lose some control over their life course.⁸ A poorer economic situation has been linked to increased mental ill health for various reasons, including accumulating debts,⁹ or as a consequence of increased social isolation due to the need to save money, which in turn prevents unemployed individuals from satisfying psychosocial and self-reliance needs.¹⁰

The effects of unemployment on health can be reduced through passive and active labour market policies (ALMPs). Passive policies are based on income transfer during times of unemployment. Some studies have shown that economic unemployment benefits, especially those of a contributory nature,¹¹ can reverse or alleviate the impact of unemployment on mental health.^2,12 ALMPs are oriented towards creating employment and promoting labour reintegration. Various types of ALMPs can be implemented, mainly: (i) job search assistance aimed at increasing search efficiency and the probability of job-matching through training and counselling; (ii) job training to raise human capital and attenuate skills mismatch through classroom vocational and technical training, work practice, basic skills training such as language and life skills training excluding cognitive skills; (iii) subsidized public employment, which implies direct job creation, generally with a short-term effect; and (iv) subsidized private employment aimed at improving the job-matching process and increasing labour demand through financial incentives and other subsidies for newly hired or existing workers, self-employment assistance and start-up grants.^13,14 Some of these programmes (such as subsidized work, training and education) demand full-time participation over a long period of time (e.g. several months), while others (such as job search assistance and education) are part-time and have a short duration (e.g. few days/weeks).¹⁴

ALMPs can help unemployed individuals escape from the downward spiral of increasing problems with health and social and personal development.¹⁵ Extending Jahoda⁵ and Fryer’s⁷ theories on unemployment and health, Strandh¹⁶ has argued that some types of ALMPs, such as education or some forms of labour market activity can fulfil some of the needs met by employment, both the psychosocial needs and the sense of control. That is, beyond re-employment, participation in training and other forms of ALMPs in itself raises unemployed individuals’ well-being compared with just being unemployed. Thus, some authors consider that such programmes act as a transitional form of labour market status in which participants are neither properly employed or unemployed in the typical sense, what they call ‘open unemployment’.¹⁷

Since the 1980s most Western countries have shifted from passive to ALMPs. Compared to passive policies, the impact of active policies on health status has been less well evaluated. However, there has been increasing interest in this topic among public health researchers since the last Western great recession of 2008–09, resulting in various reviews of this topic in recent years.^17–20 However, most of these reviews have a somewhat narrow scope, in that either they are narrative reviews without systematized results, or they do not focus specifically on ALMPs, but rather on any intervention to improve the mental well-being of the unemployed. As far as we know, only one systematic review has been carried out, which included six studies, mostly referred to the most cited programmes: JOBS in the USA and Työdön in Finland conducted before 2006.¹⁸ The narrative reviews,^17,19,20 while mostly recent, did not conduct a systematic literature search or results selection.

The aim of our review is to synthesize evidence on the effectiveness of ALMPs, and whether some type of ALMPs are more effective.

Methods

We conducted a literature review using the Scoping review methodology to synthesize scientific knowledge and identify gaps in research.²¹

We consulted the PubMed/MEDLINE, Scopus and Web of Science databases, and focused on articles published between January 1990 and December 2017. Full details of the search strings used for the various databases are shown in Supplementary appendix figure S1. Note that we used three different search strings for PubMed/MEDLINE, in order to include all the articles that could potentially provide data of interest. For the Scopus and Web of Science databases, one search string was sufficient.

We included studies published in English, Spanish and Catalan, and applied geographical criterion to include studies from high income countries. We excluded books, communications to congresses and other studies for which a summary was not available.

We also selected only studies that evaluated or analyzed ALMPs that included at least one result on health or quality of life.²²

Our review focused on adults of working age, in their respective country, who were unemployed and participating in an ALMPs. We excluded studies conducted among people who suffered from some disability condition or pre-existing disease.

We applied four sequential phases of document screening to the list of articles retrieved by the searches. In Phases 1, 2 and 3, we screened the articles’ titles, abstracts and full text, respectively. In the last step (Phase 4), we manually retrieved additional articles that also met the inclusion criteria from the reference lists of the articles selected, as well as the reference lists of five previous literature reviews,^{17,18,23–25} and repeated the same procedure of screening. To ensure internal validity, we triangulated the results as follows: Phases 1 and 2 were carried out separately by two researchers. When these disagreed on the results, the most inclusive option was chosen, such that no study was discarded in which the criteria of the two researchers did not agree. Phase 3 was carried out in parallel between two researchers, discussing the inclusion/exclusion of documents that generated doubts. A third researcher was included if there were still doubts. For the articles retrieved in Phase 4, we replicate the same selection and triangulation process.

For each paper, the following information was collected by all the researchers: first author, year of publication, name and place of the programme, design of the study, number of cases, mean age and range, percentage of men, socioeconomic level of the study population, time of unemployment, health outcomes, programme description, type of ALMP, main results and mechanisms that explain the results.

Results

Figure 1 summarizes the search and selection process, and the documents included. Of the 416 documents detected in the search, 270 were discarded after reviewing the title, and a further 98 after reviewing the abstract because they did not meet the inclusion criteria. After reading the full text of the remaining 48 documents, 15 were included. Reviewing the references from these articles and from the literature reviews,^{17,18,23–25} we identified 12 documents from the search results and nine from the literature reviews, giving a final total of 36 documents.

Table 1 shows the main characteristics of these documents. Half had a quasi-experimental design. The rate of publication tended to increase over time, with almost 40% published after 2009. Most studies were conducted in Finland, the USA, Sweden and Germany. There were two types of studies, those which addressed the effect of ALMPs at individual level (N = 29), and those which addressed the general effect of ALMPs at country level (N = 7), such as the effect of national spending on ALMPs on suicide. Different types of ALMPs were reported at individual level. The programmes reported mainly attempt to build capacity through job search assistance (N = 13) most of them with a component on mental health, some offer job training (N = 6) and few subsidized employment (N = 3). A significant number of programmes include a combination of different types of ALMPs: (i) job search assistance (with or without a component on mental health) and job training (N = 2); (ii) job training and subsidized employment (N = 3); (iii) job search assistance, job training and subsidized employment (N = 2). With regard to health outcomes, 26 out of 36 focused on mental health and/or related components such as self-esteem, self-efficacy, self-confidence and inoculation against setbacks; six studies addressed general health and mental health; and four studies analyzed other outcomes (life satisfaction, social integration, well-being and quality of life). Most of the studies (80.6%) found positive results, at least for one of the health outcomes studied.

Supplementary appendixtables S1A–S3A show the main characteristics of the manuscripts included. The nine studies that used an experimental design (Supplementary table S1A) were based on the ‘Jobs project’ and ‘Jobs II’ in Michigan (USA) and the Työhön Job Search Program in Finland. These programmes are based on job search assistance and counselling and have an important psychological component based on gaining self-efficacy, inoculation against setbacks, receiving social support and positive feed-back from the trainers. All of these studies were published before 2006. All studies analyzed mental health and related outcomes and all found positive results.

Supplementary table S2A presents the 18 studies with a quasi-experimental design. With the exception of two studies from Australia, the rest were from European countries. They evaluated several ALMP types and combinations thereof. Most of subsidized employment is linked to economic benefits, some are salaries and some welfare payments. Other programmes are also linked to unemployment benefits. Participation in some of these programmes, such as the Madrid Regional Government’s Welfare Programme,²⁶ was a mandatory condition for economic benefits. Health outcomes were also very diverse. Two of three studies^26–28 analysing general health showed a reduction of physical problems.^26,27 Regarding mental health, several studies showed that ALMPs participants performed better coefficients of mental health and life satisfaction than those in open unemployment^16,29–32; Sage found similar coefficients of GHQ-12 and well-being than those employed.^20,33 Other studies found a decrease on levels of depression and psychological distress and an increase of life satisfaction, self-esteem through participation.^34–36 Three quasi-experimental studies could not found effects on mental health and related components, as well as, social integration.^37–39

The non-experimental studies are shown in Supplementary table S3A. All of studies are from European countries, five are from Nordic countries (Sweden and Finland), one is from Germany, and three compare more than 15 European countries. One study was published in 1999, while the others were published after 2004. Again, these studies evaluated various types of ALMP, and combinations thereof. Three ecological studies comparing several countries examined the relationship between health status of unemployed and spending on ALMP.^40–42 Two studies analyzed the effects on suicide^41,42 and one on depressive symptoms,⁴⁰ showing that an increase in government spending on ALMPs reduced the effect of unemployment on suicide^41,42 and was related to a narrowing of educational inequality in depressive symptoms.⁴⁰ Except Björklund et al.⁴³ who underwent a study with qualitative methodology to describe quality of life among unemployed young men participating in a job search assistance programme, the other eight studies analyzed mental health.^{40–42,44–48} Only one study failed to find a positive result in the health outcome examined.⁴⁴

Effectiveness according the type of ALMPs

Job search assistance

Programmes that offer job search assistance with a psychological component, e.g. inoculation against setbacks, improving self-confidence or self-image or increasing self-efficacy, have been shown to have positive effects on mental health and related components in all cases^{34,,43,49–56} in different countries (USA, Finland and Australia). ‘Jobs project’ (USA) produced higher quality re-employment in terms of earnings and job satisfaction and higher motivation,⁵⁴ reduced depression and anxiety at 1-month, 4-month and 28-month follow-up.^49,52–54 ‘Jobs II’ (USA) showed similar results, reducing financial stress and increasing sense of mastery which in turn had beneficial effects on depressive symptoms and role functioning.^50–55 ‘Työhön Job’ (Finland) showed an increase in self-efficacy, inoculation against setbacks and self-esteem; a decrease in financial stress and an increase in re-employment; and a reduce in psychological distress.^47,56,57 In Australia,³⁴ also participation in job search assistance reduced depression and negative mood, although no differences were found between participants and non-participants on GHQ-12 and self-esteem. Details on coefficients and statistical significance are found at Supplementary tables S1A–S3A.

The only study showing no effectiveness of job search assistance with a psychological component was Reynolds⁵⁸ in Ireland. However, it showed a statistical significant improvement in inoculation against setbacks.

This type of ALMPs were effective in different groups of people including sole-parenting females.³⁴ Most of the studies showed that those who benefitted more from the interventions were participants in high-risk groups, i.e. those who were at greatest risk of poor mental health at the baseline^{34,,49,52,54,55,57} Harry and Tiggemann³⁴ also found a greater effect among men than women.

Job training

For programmes that focus on job training,^{29,37,38,44,45,48} the results were mostly positive although more controversial than job assistance. The studies of Andersen²⁹ and Röjdalen et al.⁴⁵ showed positive results in mental health. In UK, Andersen²⁹ showed that those people participating in training reported statistically significant lower levels of poor mental health compared to those people who were in openly unemployment. In Sweden, Röjdalen et al.⁴⁵ showed that those who most judged mental health was affected positively as a result of training were women, the less educated and individuals with a positive attitude to competence development.

In Finland, Saloniemi et al.⁴⁸ did not find a significant improvement in mental health for the whole group; however, they did observe some changes when they analyzed the differences according to socioeconomic position. Training led to improve health and well-being among participants with a higher socioeconomic status, whereas for blue-collar workers the effects were neutral or even detrimental.⁴⁸ Breidahl’s study³⁸ of Danish unemployed individuals did not find any systematic correlation between participation in an ALMP and any of the social marginalization indicators examined. However, the participants’ own assessment of the impact of the ALMP on their self-esteem was quite positive.³⁸ In Germany, Behle⁴⁴ showed a positive impact on mental health for West German female participants; however, the associations got weaker when more variables were controlled for in the models. In Sweden, Reine et al.³⁷ did not find any statistically significant effect on psychological symptoms among young people (16–20 years old).

Subsidized employment

Eight of the studies combined different forms of subsidized employment with other types of ALMP such as job search assistance, job training or both.^{16,26,27,30–32,39,,46} Results were also mostly positive, but more controversial than job search assistance. The study of Vuori and Vesalainen,⁴⁶ which combined vocational training, job search assistance and subsidized employment, showed that male participants in subsidized employment reported fewer symptoms of psychological distress, and that after the period of subsidized employment, the symptoms returned to the level of others. Korpi³⁰ showed that unemployed individuals in temporary relief work (subsidized employment) seem to have an intermediate position in terms of well-being. They were clearly better than unemployed individuals, but worse than those in regular employment. In Sweden, Strandh¹⁶ showed that those unemployed individuals involved in subsidized employment had a statistically significant lower score in GHQ-12, proving a better mental health compared to those in open unemployment. In Spain, Ayala and Rodríguez²⁶ showed that participation in work-related activities produced only modest positive effects on mental health. Similarly, a programme in Serbia that combined vocational training and subsidized temporary employment²⁷ showed a particularly high health improvement among individuals participating in both activities.

The evaluation of ABM Germany subsidized employment programme suggested an offset of approximately 60% of the negative effect of unemployment on life satisfaction.³² The two studies evaluating Euro-One-Job, the subsidized employment programme that succeeded ABM in Germany, showed higher levels of life satisfaction in comparison with openly unemployed welfare recipients, although the programme could not fully substitute regular employment in terms of satisfaction³¹; and similar self-perceived integration between participants and non-participants.³⁹ The difference with ABM is that the former implied a regular paid wage and Euro-One-Job only an extra payment in the welfare benefit.

Discussion

Under the hypothesis that there is a positive association between ALMPs and health, the aim of this article was to describe the impact of active employment policies on health and quality of life and which types of ALMP were more effective in improving health. Most of the studies analyzed mental health and related components, and few general health and other outcomes. This review shows that in general ALMPs have a positive impact on health and quality of life. Regarding the type of ALMP, all studies except one that analyse job search assistance showed positive results on mental health and related components. Vocational training and subsidized employment showed more controversial results, although mostly positive. Aggregated studies aimed to analyse the effect of the total amount of GDP each state allocates to ALMPs and show that total spending in ALMPs decreases suicide risk and narrows social health inequalities.

There is a marked heterogeneity between the interventions, the design of the evaluation study, and the outcomes examined. Moreover, most of them lack a theoretical framework that systematically assesses the dimensions they address.

What works and why? Mechanisms and type of ALMPs

It is difficult to draw conclusions about the effectiveness of each type of policy and the related mechanisms as most programmes, and hence their evaluations, combine types. However, programmes that offer job search assistance with a psychological component, showed positive results in mental health and related components,^{34,43,49–56} with a partial exception in Ireland.⁵⁸ Most of these evaluations report positive results in terms of Fryer’s theory of gaining mastery and control over one’s life course.⁷ According to Coutts,¹⁷ ALMPs can have more positive health impacts if they focus on enhancing personal development (the self-efficacy model) rather than entirely on increasing occupational skills, ‘supply-side’ factors and getting individuals into any job as quickly as possible. In addition, a qualitative study that evaluated a resource centre in Finland,⁵⁹ also referred to psychosocial mechanisms tied to Jahoda’s latent function theoretical model,⁶⁰ such as giving structure to life. That is, programmes are also useful for transforming unemployment into a meaningful pattern of time use, especially through work-like activities or education. Moreover, in line with the idea of sense of mastery and economic control, participants reported seeing new possibilities in the labour market, and having economic support from the programme.

Some of these studies point out that unemployed individuals who benefit most from job search focused interventions were those with the poorest well-being at baseline (i.e. high-risk participants).^49,52 According to Vinokur,⁵³ low-risk participants by having higher self-esteem confidence and social skills may play a positive role model for the high-risk participants, creating socially supportive interactions.

Programmes that offer vocational training schemes enable participants to engage in new and more advanced forms of labour market activity, according to Strandh.¹⁶ These activities raise unemployed individuals’ competitiveness, and hence their sense of control, in line with Fryer’s theory on mastery and control over one’s life course.⁷ The effect of unemployment on mental health and well-being could be affected by being in line for a job or having good expectations of finding a job.¹⁶ In addition, vocational training courses provide the perception of being involved in a legitimate alternative status to employment, namely education status.¹⁶ Social identity or status is one of the five fundamental needs that underlie individual well-being according to Jahoda.⁵

However, many questions remain unanswered, such as whether acquiring technical skills increase the sense of control for all people equally and in any context. Improved skills may not improve one’s sense of control (at least to the same degree) for workers that are considered ‘secondary labour force’ due to axes of oppression, such as women, the migrant population and manual workers. Different forms of oppression contribute to the fact that, especially in times of economic crisis, women and migrants, even those with good qualifications, receive jobs later or receive more precarious jobs in the labour market. This could lead to a feeling of time loss, rather than an increase in control. That is, vocational training could lead to a situation where participants with better resources are provided with an edge over others in the labour market, i.e. courses provide more effective help to workers with better resources.⁴⁸

Subsidized employment could be the most similar to an employment-like life situation. According to Strandh,¹⁶ the more similar the better it will replace individual’s psychosocial needs for employment. Studies have shown mostly positive results at short term.^{16,26,27,30–32,39,,46} According to the authors, these improvements were explained by the experience of social interactions, daily routine, while participating in the programme and physical activity carried out during the programme. However, participants did not experience improved re-employment after the programme. Since the effects of the programme in replacing psychosocial needs for employment do not extend beyond the end of the programme, the positive impacts on mental health and well-being are likely to be only temporary.²⁷ Moreover, people may only receive psychological benefits from jobs with certain desirable characteristics, such as being perceived as having meaning or conferring social status.¹⁶ Since workfare jobs are often poorly paid and confer little social status, they may be poor substitutes for regular jobs when it comes to increasing life satisfaction.^31,32

For whom do ALMPs work? Some explanations on health inequalities

Both, the design of ALMPs as well as their evaluation in relation to health impacts, depart from a neoliberal approach where employment is the norm, central to everybody’s lives and sometimes the most important way to achieve a socially desirable identity. While this is mainly certain, especially since most private income comes from being employed in the labour market, there are differences within the population regarding one’s psychological needs for employment, i.e. time structure, social contacts, collective effort or purpose, social identity or status and regular activity.

From a feminist perspective, labour is understood as all processes that encompass life reproduction.⁶¹ This includes not only paid work but also care work in the private sphere, as well as community work in the public sphere. Thus, participation in other forms of work could also replace the psychosocial needs for paid employment. Even without going that far, gender differences in reactions to ALMPs have rarely been addressed, despite the clear gender division of labour, where women are still the main person responsible for care work in the private sphere, which sometimes changes their psychological needs for employment, most notably time structure and social identity or status. In addition, it is well known that women are typically over-represented among the unemployed, and in precarious jobs. There are important gender differences in the impact of unemployment on mental health,^2,12 which also interacts with other axes of oppression, such as social class and migration status, and other social characteristics such family composition.

In this regard, a study carried out in the UK³³ showed that while the feeling of life worth and happiness among male ALMPs participants was significantly higher than among their unemployed male-counterparts, there were no difference between the two groups amongst female respondents. Further, ALMPs had a negative effect on anxiety among women, but a positive effect among men. However, ALMPs appear to be more effective in improving life satisfaction of women than that of men. Consequently, while men seem to benefit more from ALMPs than women, the results equally suggest some advantages for women participants too.³³

Moreover, regarding social class, there have been few studies addressing the differences in the effectiveness of ALMPs among the unemployed. In a Finnish study, training led to improved health and well-being among participants with a higher socioeconomic status, whereas for blue-collar workers the effects were neutral or even detrimental.⁴⁸ This result raises questions about the role of ALMP measures, since they could increase socioeconomic inequalities in some contexts. On the contrary, an ecological study suggested that reduced social protection may have damaging effects on the mental health of particular groups, such as those with lower level of education and single parents.⁴⁰ We did not find any studies that address different cultural backgrounds and migrant status.

Geographical context is another key issue to take into account. Researchers commonly distinguish broadly between two national models of activation: a liberal model (work-first) and a universal model of activation (social investment), with the UK often given as an example of a liberal model, and Denmark as a universal model.⁶² In fact, most evaluations of ALMPs and their health effects have been conducted in Anglo-Saxon^{20,29,33–36,49–54,58} and Scandinavian^{16,30,37,38,,43,45–48,55–57,63} countries, with few exceptions in Germany, Serbia and Spain.^26,27,31,44 The rest of studies comprised all European countries.^40–42 Activation models from Southern and Eastern European countries probably have different characteristics from the liberal and universal models. Moreover, labour market characteristics also differ, as do other characteristics of these countries, such as welfare states, culture of work and family composition. For example, in terms of culture, it has been discussed that the protestant ethic and the moral virtues of work are central to the Nordic work ethic. In this culture, work provides good role models and stimulates good behaviour. Work ethics has an important influence on the significance of employment and probably on the unemployment experience. Paugam and Russel⁶⁴ also argued that family has a fundamental influence on the unemployment experience, suggesting that non-work will be less damaging in societies where there is a stronger emphasis on family life. Thus, employment programmes may have a different meaning depending on the geographical context, and this could lead to a differential impact on health. However, we only found one study from a European Southern country,²⁶ and one from a European Eastern country.²⁷ More evaluations on the effect of ALMPs on health in these countries are needed.

In conclusion, the significance of employment for different population groups, such as women or groups with different cultural backgrounds must be taken into account when evaluating the health effects of ALMPs in order to understand their limitations and the added value they can bring. They are also key issues for the transferability of ALMPs to different settings. It would be desirable to undertake new evaluations of the effects of ALMPs with a gender, class, intercultural and international perspective.

From ALMPs to re-employment: what is the role of precariousness?

Literature on the effect of ALMPs on health and well-being fills the gap between the extensive literature on the effectiveness of ALMPs on re-employment,⁶⁵ and that analysing the health effects of transitions from unemployment to employment,^66,67 according to the quality of the work. The ALMP literature has identified positive effects on employment in the medium term (2 or 3 years from training) but not in the short term, and the effects are also larger among females, among the long-term unemployed, and in times of deep recession.⁶⁵ Specifically, the effects of ALMPs in improving health tend to be immediate, such that, besides increasing employability, they can bridge well-being before the effectiveness of employment is reached. However, a recent branch of literature has stressed that, in the context of high precariousness, getting an insecure job does not necessarily lead to health improvement.⁶⁸ On the contrary, in a longitudinal study in UK, re-employed workers engaged in low-quality jobs were found to have higher levels of biomarkers for stress load compared to their unemployed peers.⁶⁹ In this sense, eventual improvements in health and well-being due to the ALMP may dissipate quickly after the transition to precarious employment. This possibility must be taken into account when implementing welfare-to-work polices in an ALMP.

Conclusions and knowledge gaps

This review shows that ALMPs have a positive impact on health and quality of life. There are relatively few studies of the impact of ALMPs on health status, and most of them are focused on Anglo-Saxon and Nordic countries. The most significant knowledge gaps are the lack of understanding of the mechanisms involved in achieving this improvement, and the differential health impacts according to axes of inequality and geographical, economical and political context.

Thus, we suggest that future studies that evaluate ALMPs in relation to health and well-being should provide: (i) an in-depth definition of programme, taking into account a consensual typology of ALMP to properly classify it; (ii) a theoretical framework that systematically assesses the dimensions they address and related mechanisms, and the programme theory underlying the changes; (iii) the target population, including its socioeconomic characteristics, and whether any differential impact of the ALMP is expected in this specific population in relation to the mechanisms; (iv) whether the target population is receiving unemployment benefits and whether participation is voluntary or linked to the receipt of unemployment benefits; and (v) both the geographical context and economic context when ALMP is implemented. It is desirable to have a feminist and intercultural approach in these evaluations. A feminist approach goes further gender-sensitive analysis, recognizing gender as a non-binary construction, and a complex and dynamic psycho-social phenomenon, which forces to problematize and interrogate the modes of historical, political and social constitution of sex differences.⁷⁰ In addition, the intersectional perspective (that comes from the feminism) complicates the notion of gender by assuming the impossibility of analysing gender as an isolated category and aims to articulate different categories or axes of domination, such as social class, ethnicity, sexual identity, migratory status and functional diversity.⁷¹ These characteristics are necessary to understand why and for whom the programme works, and to ensure that the results will be comparable to other studies.

Acknowledgements

This study has been possible thanks to the initial version of Pol Giró, as a product of his work on the Postgrade in Public Health, by IDEC-UPF, Barcelona, Spain.

Funding

The research leading to these results has received funding from RecerCaixa.

Conflicts of interest: None declared.

References