Navigation – Plan du site

AccueilNuméros17-1Labour market integration of disa...

Labour market integration of disability insurance benefit applicants in Switzerland

Evidence from linked survey and administrative data
Intégration sur le marché du travail des demandeurs de prestations de l’assurance-invalidité en Suisse: observations tirées d’enquêtes et de données administratives couplées
Szilvia Altwicker-Hámori
p. 69-86

Résumés

La participation accrue des personnes handicapées au marché du travail reste une priorité politique pour les gouvernements européens. Cette étude visait donc à examiner les facteurs favorisant l’intégration sur le marché du travail des demandeurs de prestations d’assurance-invalidité (PAI) et des personnes ayant droit à des mesures d’intégration et d’occupation en Suisse. Un ensemble de données inédit a été créé via la mise en relation de l’Enquête suisse sur la population active 2018 avec les données des registres administratifs (2000-2017). La régression logistique multiple a été utilisée pour analyser les associations entre l’activité économique et les caractéristiques démographiques, socioéconomiques, de santé et régionales chez les demandeurs de PAI en âge de travailler (N=9,309) et les personnes ayant droit à des mesures d’intégration et d’occupation (N=5,753). Il en ressort que le fait d’être un homme, jeune, de vivre avec un partenaire ayant un emploi, d’avoir fait des études secondaires supérieures ou tertiaires, d’avoir réalisé l’entretien dans la langue de la région linguistique, de ne pas déclarer de limitations d’activité liées à la santé à long terme ou de morbidité chronique, est associé à une probabilité plus élevée d’activité économique. Les demandeurs résidant dans la région lémanique et au Tessin étaient moins susceptibles d’être économiquement actifs que leurs homologues vivant dans l’Espace Mittelland. Plus important encore, les résultats impliquent que l’éducation au-delà du niveau secondaire inférieur ainsi que la maîtrise de la langue sont essentielles pour l’activité économique des demandeurs de RAI et de ceux qui ont droit à des mesures d’intégration et d’occupation.

Haut de page

Notes de l’auteur

Acknowledgements

The author is grateful to the Swiss Federal Statistical Office and the Swiss Federal Social Insurance Office for providing data; to Christian Cajacob, Jonas Deplazes, and Beat Schmid for their cooperation during the data linkage process; and to Prof. Dr. med. Julia Dratva for her valuable contribution to the funding acquisition phase and throughout the project.

Availability of data and materials

The data that support the findings of this study are available from the Swiss Federal Statistical Office and the Swiss Federal Social Insurance Office but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available.

Funding

Funding from the “Societal Integration at the ZHAW” is gratefully acknowledged. The funding body played no role in the design of the current study; in the collection, analysis, and interpretation of the data; and in the writing of the manuscript.

Declarations

Ethics approval and consent to participate.

No formal ethics approval was needed in accordance with national guideline/statutory law [Human Research Act, Art. 2 (2)(c)].

Texte intégral

1. Introduction

1The aim of the European Disability Strategy 2010-2020 was to promote a barrier-free Europe and to empower individuals with disabilities so that they can enjoy their rights and participate fully in society and the economy; one priority being the increased participation of individuals with disabilities in the labour market (European Commission, 2020). Switzerland shares this priority. It has been high on the political agenda since 2004, leading to a number of revisions to the Federal Act on Disability Insurance (Altwicker-Hámori & Dratva, 2019; Bundesrat, 2018; OECD, 2014), including the development of measures for early detection, early intervention, and integration in order to identify affected persons as early as possible and to support them in keeping their current jobs (Bundesrat, 2005). In essence, the guiding principle of disability insurance (DI) in Switzerland is “rehabilitation before pension” (AVS/AI Information Centre, Swiss Federal Social Insurance Office, & State Secretariat for Economic Affairs, 2020; Swiss Federal Social Insurance Office, 2022). Therefore, rehabilitation measures to restore, maintain, or improve earning capacity or the ability to perform day-to-day activities are always preferred to disability pension (DP) (Swiss Federal Social Insurance Office, 2022).

2In light of the “rehabilitation before pension principle,” access to disability insurance benefits (DB) is essential. Going a step further and identifying factors promoting the labour market integration of DB applicants also appears crucial: it may help in designing effective integrative measures for DB applicants, which in turn may mitigate the major public health issue arising from labour market withdrawal or non-entry.

3Despite its significance, the group of DB applicants is understudied. In light of this, the aim of this paper was to examine, to the best of my knowledge for the first time, the association between demographic and socioeconomic characteristics, long-term health-related activity limitation, chronic morbidity, and region of residence and economic activity in working-age individuals who applied for DB in Switzerland between 2000 and 2017. For the analysis, a novel dataset was created via linking microdata from the Swiss Labour Force Survey (SLFS) and administrative registers.

4Taking advantage of the richness of the administrative data, the subsample of DB applicants entitled to integration and occupational measures for the first time between 2000 and 2017 was analysed separately. The rationale for analysing this subsample was threefold. First, integration and occupational measures – aimed at restoring, maintaining, or improving earning capacity – are two vital components of DB (Bundesamt für Sozialversicherungen, 2019). Addressing factors that may facilitate the effectiveness of these rehabilitation measures in terms of labour market integration is thus crucial. The second rationale was a theoretical one, that is, to use different definitions of “disability.” Grönvik (2009) differentiates between three definitions of disability: functional (based on a medical understanding of disability), administrative (based on the eligibility of benefits), and subjective (based on whether the individual conceives of himself/herself as “disabled”). Whereas in the sample of DB applicants the subjective definition of disability dominates, the administrative definition dominates in the subsample of DB applicants entitled to integration and occupational measures – as the individuals in the latter group fulfilled the criteria necessary for DB entitlement according to the cantonal DI office (Swiss Federal Social Insurance Office, 2022). The third rationale follows from the second one: using different definitions may facilitate future cross-country comparisons. Ideal cross-country comparisons are based on the same sample definition. Therefore, using two definitions in the present study broadens the possibility for appropriate cross-country comparisons in the context at hand.

5On an empirical level, the subsample comprised all integration measures (e.g., occupational integration measures, auxiliary aids) and all occupational measures (e.g., career counselling, further vocational training) between 2000 and 2017. Further detail on integration and occupational measures is provided in the Federal Social Insurance Office codebooks (Bundesamt für Sozialversicherungen, 2020). Hence, the following groups of DB applicants were excluded: DB applicants whose DB application was rejected (e.g., because certain requirements were not fulfilled), for whom the procedure was terminated (e.g., due to the assignment to another institution or the applicant’s withdrawal), for whom a clarification measure for the assessment procedure was needed (e.g., polydisciplinary medical expert opinion), or who were entitled to integration and occupational measures for the first time in 2018.

6International studies in the context of labour market integration of individuals with disabilities with a similar research design were used to select the independent variables (Altwicker-Hámori & Dratva, 2019; Boman et al., 2015; OECD, 2009). Accordingly, the models included standard demographic variables (sex and age), household characteristics (presence of a child and a working partner in the household), socioeconomic variables (educational and occupational qualifications), health-related variables that are likely to affect employment opportunities (long-term health-related activity limitation and chronic morbidity), and region of residence. In addition, the individual’s language skills were integrated into the analysis – a factor which has not been analysed in the current context but proved to be vital for the labour market success/integration of the “vulnerable” group of immigrants. For example, causal effects of language skills on wages have been estimated for immigrants in Germany, the UK, and the US (Isphording, 2014). Language skills are particularly important in Switzerland, a multilingual country with four language regions (Swiss Federal Statistical Office, 2021a). In 2000, 75 % of the Swiss population spoke German, 20 % French, 4 % Italian, and 1 % Romansh (Eugster et al., 2017). The distribution of the national languages has changed only moderately over the past four decades (Swiss Federal Statistical Office, 2022b). Accordingly, the present study accounted for the ability to speak the language of the language region of residence.

7Most importantly, the results suggest that education beyond the lower secondary level, occupational qualifications, and regional language proficiency are essential for the labour market integration of DB applicants and those entitled to integration and occupational measures.

2. Methods

2.1. Data

8For the statistical analysis, different data sources were used. The first source was the 2018 cross-section of the SLFS, provided by the Swiss Federal Statistical Office (FSO). The SLFS, a household survey carried out since 1991, is based on sampling of individuals whose addresses are selected at random from the FSO’s sample register (Swiss Federal Statistical Office, 2021b). Only one person per household is selected for the interview (Swiss Federal Statistical Office, 2021c). The SLFS covers around 1 % of Switzerland’s permanent resident population aged 15 and over. The permanent resident population includes all individuals who officially reside in Switzerland for the entire year: Swiss citizens and foreign citizens holding a permanent residence permit or a residence permit valid for at least one year (Swiss Federal Statistical Office, 2021c). Apart from the relatively large sample size, a further advantage of the SLFS is that it is part of the European Union Labour Force Survey (EU LFS). Accordingly, the SLFS adheres to international concepts and definitions, follows the International Labour Organisation’s (ILO) guidelines, uses common classifications (e.g., Nomenclature of Territorial Units for Statistics, NUTS), and records the same set of characteristics as other participating countries, thereby facilitating international comparisons (Eurostat, 2022a, 2022b).

9The SLFS contains detailed information on the demographic and socioeconomic background of the household members as well as minimal information on health status. However, the SLFS does not include information on DB application or integration and occupational measures granted by the cantonal DI offices; the key variables for sample selection (see Section “Sample selection”). The latter information for the period 2000-2018 was therefore retrieved from administrative registers by the FSIO and linked to the SLFS.

10Data linkage for the purposes of this study was requested via the FSO’s data linkage request form (Swiss Federal Statistical Office, 2022a). The approval process followed the standard procedures outlined by the FSO (Swiss Federal Statistical Office, 2022a). Linkage of the individual administrative register data files and the SLFS was based on individual social insurance numbers. The data files were prepared for linkage by the FSO in collaboration with the FSIO and the Central Compensation Office (CCO). All data files were made available in an anonymised format for the purposes of this study.

2.2. Sample selection

11The sample of interest included working-age adults in the 2018 cross-section of the SLFS who applied for DB between 2000 and 2017. In Switzerland, women and men are entitled to old-age pension at age 64 and 65, respectively. Accordingly, 18-to-63-year-old women and 18-to-64-year-old men were selected for the analysis (N=9,309). Furthermore, a subsample was defined to include working-age adults in 2018 who were entitled to either an integration or an occupational measure between 2000 and 2017 (N=5,753). These individuals fulfilled the criteria necessary for DB entitlement according to the cantonal DI office. That is, they were “disabled owing to a congenital or other illness, or as the result of an accident” (Swiss Federal Social Insurance Office, 2022). Hence, the analysis of the latter sample allows us to shed light on the labour market integration of individuals with disabilities who were entitled to measures aimed at restoring, maintaining, or improving their earning capacity.

2.3. Dependent variable

12The dependent variable captured labour market integration, in particular, being part of the workforce. The dependent variable was thus a binary variable equal to one for economically active individuals in 2018 and zero for economically inactive individuals in 2018. The economically active included employed individuals (self-employed, family workers, employees, and apprentices) and the unemployed as defined by the ILO. The economically inactive included those in education, homemakers, pensioners, and other economically inactive individuals.

2.4. Independent variables

13The benchmark model included information on sex, children aged 14 or younger living in the same household, employed partner living in the same household, age, language proficiency, highest educational attainment, long-term health-related activity limitation, chronic morbidity, and region of residence.

14Dichotomous variables for the respondent’s sex (“Male” versus “Female”) and the presence or absence of own children or stepchildren, aged 14 years or younger, living in the same household were included. Similarly to a Swiss study on DP receipt (Altwicker-Hámori & Dratva, 2019), a dichotomous variable for the presence or absence of an employed partner in the household (cohabiting or married) was generated by combining (1) information on the relationship of household members to the reference person and (2) the respective household member’s employment status. The employed category included self-employed, family workers, employees, and apprentices.

15Age was categorised to differentiate between young adults (“18-39”), middle-aged adults (“40-54”), and individuals outside prime working age (“55-retirement age”). The use of these age categories was motivated by the evidence that among adults with disabilities, middle-aged persons have better opportunities in the labour market than their younger and older counterparts (Boman et al., 2015).

16A dichotomous variable indicating whether or not the respondent carried out the interview in the official language of the language region of residence was generated by using two variables in the SLFS: (1) the respondent’s interview language (“German,” “French,” “Italian,” or “English”) and (2) the respondent’s language region of residence (“German,” “French,” “Italian,” or “Romansh”). The facts that (1) the interview could not be carried out in Romansh and (2) the Romansh-speaking region is officially trilingual (German, Italian, and Romansh) (Swiss Federal Statistical Office, 2021a) were taken into account.

17A dichotomous variable for highest educational attainment (“Lower secondary” versus “Upper secondary and tertiary”) was included in the benchmark model. The dichotomisation was chosen for two reasons. First, lower secondary education corresponds to the end of compulsory education in Switzerland, thereby providing a suitable cut-off point. Second, this dichotomisation has been found to be relevant in previous studies in a similar context, namely, in the analysis of risk factors for DP receipt in young adults with disabilities living in Switzerland (Baer et al., 2015) and in young adults with long-term health-related activity limitations living in Switzerland (Altwicker-Hámori & Dratva, 2019). Both studies found higher disability retirement odds for those with only a lower secondary education relative to their counterparts with a higher secondary/tertiary degree.

18To capture the respondent’s health, two domains of the Minimum European Health Module (MEHM) were used. First, a dichotomous variable was generated to differentiate between those reporting and not reporting long-term health-related activity limitation. The former category included individuals who gave the answers “Severely limited” or “Limited but not severely” to the following question: “For at least the past six months, to what extent have you been limited because of a health problem in daily activities people usually do?” Furthermore, an indicator variable capturing self-reported chronic morbidity was generated based on the question “Do you have any long-standing illness or health problem?” Long-standing was defined as lasting for at least six months prior to the survey or likely to affect the respondent for at least six months after the survey (Bundesamt für Statistik, 2015).

19Finally, a set of indicator variables for region of residence was included, classified at the NUTS2 level (“Lake Geneva Region,” “Espace Mittelland,” “Northwestern Switzerland,” “Zurich,” “Eastern Switzerland,” “Central Switzerland,” and “Ticino”).

20Building on a Swiss study on DP receipt (Altwicker-Hámori & Dratva, 2019) and exploiting the richness of the survey data, an additional model was estimated replacing highest educational attainment by learnt occupation as classified by the Swiss Standard Classification of Occupations (SSCO) 2000 (Swiss Federal Statistical Office, 2019). Aggregation at the one-digit level – the “nine Divisions of professions” – was used for individuals with a non-missing value in the learnt occupation variable; while those with the codes “Not appliable” (21 % of the sample) and “No answer/did not know” (less than 1 % of the sample) were merged into a separate “Not applicable” category. Therefore, the following ten categories were generated: (1) “Not applicable”; (2) “Agriculture, forestry, and livestock production” [henceforth “Agriculture”]; (3) “Manufacturing”; (4) “Technical activities and ICT”; (5) “Construction and mining”; (6) “Trade and transport”; (7) “Hotels and catering, and other personal services” [henceforth “Personal services”]; (8) “Management, administration, finance, insurance, and law” [henceforth “Business and administration”]; (9) “Health, education, culture, and science”; and (10) “Not classifiable.” The difference between the “Not applicable” and “Not classifiable” categories merits comment. The educational background variable revealed that the majority (61 %) of individuals in the “Not applicable” category completed at most lower secondary education. On the other hand, the “Not classifiable” category included individuals who were educated beyond the lower secondary level but for whom a specific learnt occupation could not be assigned by the SSCO 2000.

2.5. Analysis

21Characteristics of economically active and inactive individuals were compared using Pearson’s chi-square tests. Multiple logistic regression models were applied to analyse the associations between demographic, socioeconomic, health, and regional characteristics and economic activity. Three models were estimated. Model 1 corresponds to the benchmark model adjusting for all independent variables. Model 2 represents an alternative specification of Model 1, replacing highest educational attainment by learnt occupation. Both Models 1 and 2 were estimated in the sample of working-age adults who applied for DB between 2000 and 2017. Model 3 corresponds to Model 1 estimated in the subsample of DB applicants who were entitled to an integration or occupational measure between 2000 and 2017.

22Categories for missing values (relevant only for long-term health-related activity limitation, chronic morbidity, and learnt occupation) were included in all analyses. A p-value of <5 % was regarded as statistically significant in all analyses. Results of the multiple logistic regression models are presented as odds ratios (ORs) with 95 % confidence intervals (95 % CIs) and p-values. All statistical analyses were carried out and are reported in line with FSO regulations. Accordingly, weights were used in all analyses and statistics based on more than four but less than 50 observations are reported in brackets. In percentage calculations, the latter FSO regulation applies to the numerator. All statistical analyses were performed using Stata 17.

3. Results

3.1. Descriptive statistics

23Characteristics of DB applicants by economic activity are presented in Table 1. Approximately half of the full sample was female (50 %) and about half of the full sample lived with an employed partner (51 %). The majority of the full sample did not live with a child aged 0-14 years (74 %), carried out the interview in the regional language (90 %), had an upper secondary or tertiary degree (87 %), and reported no long-term activity limitation (81 %) or chronic morbidity (76 %). The full sample consisted of around 45 % young adults, 35 % middle-aged adults, and 20 % individuals outside prime working age. For approximately 22 % of the full sample, learnt occupation was recorded as “Not applicable.” Around 23 % of the full sample had a learnt occupation in the field of “Health, education, culture, and science,” followed by “Business and administration” (13 %), “Trade and transport” (9 %), “Manufacturing” (9 %), “Technical activities and ICT” (9 %), “Construction and mining” (5 %), “Personal services” (5 %), and “Agriculture” (2 %). For around 4 % of the full sample, the learnt occupation could not be classified. Approximately 19 % of the full sample resided in the Lake Geneva Region, 22 % in Espace Mittelland, 14 % in Northwestern Switzerland, 18 % in Zurich, 13 % in Eastern Switzerland, 10 % in Central Switzerland, and 4 % in Ticino.

24There were statistically significant differences between economically active and inactive DB applicants with regard to all the characteristics. A higher proportion of the economically active DB applicants was male (53 %), was living with an employed partner (53 %), and was living with a child aged 0-14 years (27 %) than of the inactive DB applicants (35 %, 39 %, and 21 %, respectively). Economically active DB applicants were younger than economically inactive DB applicants: 17 % and 33 % of the respective groups belonged to the “55-retirement age” category. The vast majority of economically active DB applicants (91%) carried out the interview in the regional language; this figure was 84 % in economically inactive DB applicants. Economically active DB applicants were better educated than economically inactive DB applicants: 88 % and 77 % obtained an upper secondary qualification in the two groups, respectively. This is mirrored in the learnt occupational composition of the two subsamples: merely 19 % of the economically active DB applicants had no specific learnt occupation; the corresponding figure was around 41 % in economically inactive DB applicants. Economically active DB applicants reported long-term health-related activity limitation and chronic morbidity less often (15 % and 21 %, respectively) than economically inactive DB applicants (39 % and 42 %, respectively). A lower share of economically active DB applicants was residing in the Lake Geneva Region and Ticino (18 %, and 3 %, respectively) than of economically inactive DB applicants (25 % and 6 %, respectively).

Table 1. Descriptive statistics for disability insurance benefit applicants by economic activity (%)

Table 1. Descriptive statistics for disability insurance benefit applicants by economic activity (%)

p-values based on Pearson’s chi-square tests; comparing characteristics of the economically active and inactive subsamples. A p-value of <5 % was regarded as statistically significant in all analyses. Figures in brackets: Extrapolation based on less than 50 observations. The results should be interpreted with great caution.

25Table 2 provides more detail on the labour market status of DB applicants. The vast majority of the economically active subsample were employed (approximately 95 %) and 5 % were unemployed. The economically inactive subsample showed an even distribution according to labour market status: around 24 % were in education, 26 % were homemakers, 29 % were pensioners, and 22 % belonged to the “Other economically inactive” category.

Table 2. Labour market status of disability insurance benefit applicants (%)

Table 2. Labour market status of disability insurance benefit applicants (%)

3.2. Regression analysis

26Table 3 reports the results of the regression analysis for Model 1. The results indicate that men were more likely to be economically active than their female counterparts (OR 2.15; 95 % CI 1.83-2.52; p-value<0.001). Individuals with an employed partner in the same household were also more likely to be economically active (OR 1.60; 95 % CI 1.37-1.87; p-value<0.001) than their counterparts living without an employed partner. Younger individuals had higher odds of being economically active than their older counterparts (OR 1.84; 95 % CI 1.50-2.24; p-value<0.001 for those aged 18-39 and OR 2.49; 95 % CI 2.02-3.07; p-value<0.001 for those aged 40-54 relative to those aged 55 to retirement age). Individuals who carried out the interview in the language of the language region were more likely to be economically active than their counterparts who did not carry out the interview in the language of the language region (OR 1.90; 95 % CI 1.49-2.43; p-value<0.001). Those with upper secondary or tertiary degrees showed higher odds of economic activity than their counterparts in the “Lower secondary” education category (OR 1.44; 95 % CI 1.17-1.76; p-value<0.001). Individuals not reporting long-term health-related activity limitations were more likely to be economically active than their counterparts with such limitations (OR 2.45; 95 % CI 2.04-2.94; p-value<0.001). Individuals not reporting chronic morbidity were also more likely to be economically active compared to their counterparts reporting chronic morbidity (OR 1.62; 95 % CI 1.36-1.94; p-value<0.001). Finally, individuals residing in the Lake Geneva Region (OR 0.55; 95 % CI 0.43-0.71; p-value<0.001) and Ticino (OR 0.40; 95 % CI 0.29-0.55; p-value<0.001) were less likely to be economically active relative to their counterparts living in Espace Mittelland. The coefficient estimate on living with a child aged 14 years or younger was statistically not significant.

Table 3. Regression analysis results with economic activity as dependent variable, Model 1

Table 3. Regression analysis results with economic activity as dependent variable, Model 1

Dummy variables for missing vales were included. Model 1 was estimated in the sample individuals who applied for a disability insurance benefit between 2000 and 2017. OR: odds ratio. CI: confidence interval. A p-value of <5% was regarded as statistically significant in all analyses.

27Table 4 reports the results of the regression analysis for Models 2 and 3. Starting with Model 2, all occupation groups were more likely to be economically active than the “Not applicable” group and the remaining coefficient estimates stayed robust. The only exception was “Northwestern Switzerland”: the coefficient estimate indicated that those residing in Northwestern Switzerland were less likely to be economically active relative to their counterparts in Espace Mittelland (OR 0.73; 95 % CI 0.56-0.97; p-value 0.029). The coefficient estimates also stayed robust in the subsample of working-age adults who were entitled to an integration or occupational measure between 2000 and 2017 (Model 3).

Table 4. Regression analysis results with economic activity as dependent variable, Models 2 and 3

Table 4. Regression analysis results with economic activity as dependent variable, Models 2 and 3

Dummy variables for missing vales were included in both models. Model 2 was estimated in the sample of individuals who applied for a disability insurance benefit between 2000 and 2017. Model 3 was estimated in the sample of individuals entitled to integration and occupational measures between 2000 and 2017. OR: odds ratio. CI: confidence interval. A p-value of <5% was regarded as statistically significant in all analyses.

28Sensitivity analysis (regarding the benchmark model) was performed to test the robustness of the results to different dependent and independent variables (available upon request). First, the unemployed were excluded from the sample, rendering the dependent variable equal to one for the employed and zero for the economically inactive in 2018 (N=8,971). The estimation results were in line with those in Model 1. Second, in Model 1 the binary variable capturing the presence or absence of an employed partner living in the same household was replaced by more detailed categories in order to separate the effect of the employment of the partner from that of “simply” living with a partner. Higher odds of economic activity were found for both married/cohabiting DB applicants whose partner was not working (OR 1.31; 95% CI 1.03-1.67; p-value 0.026) and DB applicants whose partner was working (OR 1.75; 95% CI 1.47-2.08; p-value<0.001) relative to their counterparts living without a partner. The remaining estimation results were in line with those in Model 1. As a final robustness check, Models 1-3 were estimated without weights. The results remained consistent in magnitude and statistical significance with only one exception regarding statistical significance: in Model 1 the coefficient estimate on Northwestern Switzerland was significant (OR 0.73; 95 % CI 0.58-0.92; p-value 0.008).

4. Discussion

4.1. Main findings

29This study addressed the economic activity of working-age DB applicants and those entitled to integration and occupational measures – a high-profile political issue – using linked survey and administrative data. Higher odds of economic activity were found for DB applicants living with an employed partner, with upper secondary or tertiary education, who carried out the interview in the language of their language region of residence, and those who did not report long-term health-related activity limitations or chronic morbidity. Male DB applicants and younger DB applicants also showed higher odds of economic activity. DB applicants residing in the Lake Geneva Region and Ticino were less likely to be economically active relative to their counterparts living in Espace Mittelland. These results remained robust in the group entitled to integration and occupational measures and in the specification addressing differences in economic activity by learnt occupation. The latter specification revealed higher odds of economic activity for those with a specific learnt occupation relative to their counterparts without occupational qualifications.

30While a vast number of European studies analysed the factors for DP receipt (Borg, Hensing & Alexanderson, 2001; Bruusgaard, Smeby & Claussen, 2010; Canivet et al., 2013; Galaasen, Bruusgaard & Claussen, 2012; Helgersson et al., 2015; Ishtiak-Ahmed, Perski & Mittendorfer-Rutz, 2014; Karlsson et al., 2008; Krause et al., 1997; Lallukka et al., 2015; Perhoniemi, Blomgren & Laaksonen, 2020; Salonen et al., 2018; Samuelssonet al., 2012; Stapelfeldt et al., 2014), fewer studies focused on economic activity/employment as the central outcome – despite its importance for inclusion. Of these latter studies, none addressed DB applicants and those entitled to integration and occupational measures. Nevertheless, they represent a useful basis for comparison. Therefore, the present findings will be discussed mainly in the context of these earlier findings.

31First, the results indicating that female DB applicants and women entitled to integration and occupational measures were less likely to be economically active than their male counterparts are consistent with European evidence. For instance, according to a Swedish study, in 2004 women aged 20-64 with disabilities were less likely to be “working” than their male counterparts (Boman et al., 2015). The authors point out that a similar pattern in gender differences exists among individuals without disabilities and that society treats men and women differently (e.g., division of labour and fewer career opportunities for women). This also applies to Switzerland – despite the reduction gender inequality in the labour market over the past decades (KOF Swiss Economic Institute, 2022).

32Second, the finding that living with an employed partner is associated with higher odds of economic activity augments empirical evidence for Switzerland. Namely, a Swiss study analysing young adults (aged 18-39) with long-term health-related activity limitations living in Switzerland based on the 2010-2015 cross-sections of the Social Protection and Labour Market (SESAM) found that young adults with long-term health-related activity limitations living with an employed partner had higher employment odds than their counterparts not living with an employed partner (Altwicker-Hámori & Dratva, 2019). The present finding in combination with that of the sensitivity analysis – indicating higher odds of economic activity for those with partners independent of the partner’s working status – appears to support the general notion that marriage/living in a partnership protects against marginalization (Ishtiak-Ahmed, Perski & Mittendorfer-Rutz, 2014; Karlsson et al., 2008; Samuelsson et al., 2012).

33Third, the result that younger age groups were more likely to be economically active than their counterparts aged 55 to retirement age are in line with a UK study analysing individuals with disabilities (Meager & Hill, 2005). The aforementioned study found employment rates to be falling in individuals with disabilities after the age of 50 in 2005. According to the authors, this finding may be explained by the differences in the types and severity of impairments among older people with disabilities compared with their younger counterparts.

34Fourth, the present study augments research findings indicating that language skills are essential for the labour market integration of vulnerable groups. Language proficiency has been shown to be key for the successful labour market integration of immigrants in a number of European countries (Dustmann & Fabbri, 2003; Isphording, 2015; Pieroni, D’Agostino & Lanari, 2019). The present study revealed its importance in a different setting, thereby indicating the significance of measures promoting the regional language proficiency of individuals with disabilities in a multilingual country.

35Fifth, it was expected that educational attainment and occupational qualifications increase the odds of economic activity in light of the large body of international evidence for people with and without disabilities. For example, individuals with disabilities with only primary education were found to be less likely to be employed than their counterparts with secondary or higher education in Sweden (Boman et al., 2015). Furthermore, an OECD study revealed that low-educated individuals living in Switzerland (2002-2006), Germany (1994-2006), and the UK (1991-2006) were more likely to transition from employment into unemployment than their highly-educated counterparts (OECD, 2009).

36Sixth, the results concerning both health indicators are in line with existing research findings demonstrating that worse health status increases the likelihood of economic inactivity. For example, the aforementioned OECD study showed that experiencing impediments in daily activities increased the probability of transitioning into all types of inactivity in the UK and into “disability” in Switzerland (OECD, 2009). Furthermore, the transition from employment to inactivity has been documented to be quick for those with chronic and long-standing health problems in the EU; posing a serious problem as the transition from inactivity to employment seems to be particularly complicated for these groups (Corral, Durán & Isusi, 2014).

37Finally, the findings regarding regional differences in economic activity – that is, relatively low odds of economic activity in the Lake Geneva Region and Ticino – for both samples under analysis add to those of a recent Swiss report (Schmidlin et al., 2020). Using administrative register data, the report evaluated the success of specific integration measures for individuals entitled to integration measures since 2012 in Switzerland. According to the report, the odds of achieving vocational integration capacity are significantly lower in the “Latin-speaking” regions of Switzerland than in the German-speaking regions. These regional differences are also consistent with “general trends.” For example, a Swiss study indicated similar regional differences in labour market performance between 2000 and 2018; in particular, lower labour force participation in the Lake Geneva-Ticino cluster than in the rest of Switzerland (Lalive & Lehmann, 2020). Another recent Swiss study provided survey and voting evidence revealing substantial differences between language regions in beliefs, norms, and values related to work (Eugster et al., 2017). Thus, the relatively low odds of economic activity in the Lake Geneva Region and Ticino documented in the present paper may be grounded in an economic disadvantage and/or different work-related values of these regions.

4.2. Methodological considerations

38One strength of this study stems from the promising and innovative method of linking highly reliable administrative records to detailed survey information (Künn, 2015). It enables, to the best of my knowledge for the first time, the analysis of factors associated with the economic activity of DB applicants and those entitled to integration and occupational measures in Switzerland, thereby not only generating novel results but also demonstrating the potential of the dataset for research in disability studies. Nevertheless, the dataset has its limitations. The health measures are self-reported and only available for 2018; a number of potentially important explanatory variables, such as the type (Boman et al., 2015; Meager & Hill, 2005) and onset (Jones, 2016) of disability could not be modelled. Moreover, different indicators to capture language proficiency and social support would be valuable to solidify the present results. In addition, further subsample analyses would be valuable, for example, by learnt occupation and language proficiency. A requirement for the latter analyses however is a larger dataset. This could be achieved, for instance, by linking the SLFS with the most recent administrative register data (i.e., beyond 2017). Finally, the present study does not allow for causal interpretation. Accordingly, future research should focus on overcoming these limitations. In particular, a valuable extension to the present study would be to account for potential self-selection bias.

5. Conclusion

39Most importantly, the results suggest that education beyond the lower secondary level, occupational qualifications, and language proficiency are essential for the economic activity of DB applicants and those entitled to integration and occupational measures. This in turn implies that access to well-targeted training and educational measures is essential for these groups of individuals to facilitate their labour market integration.

Haut de page

Bibliographie

Altwicker-Hámori Szilvia & Julia Dratva. 2019. Disability pension receipt in young adults: An analysis of the Swiss social protection and labour market (SESAM) data. BMC Public Health, 19(831). Online: doi.org/10.1186/s12889-019-7098-1.

AVS/AI Information Centre, Swiss Federal Social Insurance Office & State Secretariat for Economic Affairs. 2020. Social security in Switzerland. Bern. Online: www.ahv-iv.ch/p/890.e.

Baer Niklas, Altwicker-Hámori Szilvia, Juvalta Sibylle, Frick Ulrich & Peter Rüesch. 2015. Profile von jungen IV-Neurentenbeziehenden mit psychischen Krankheiten. Bern: Bundesamt für Sozialversicherungen.

Boman Tomas, Kjellberg Anders, Danermark Berth & Eva Boman. 2015. Employment opportunities for persons with different types of disability. Alter. European Journal of Disability Research/Revue européenne de recherche sur le handicap, 9(2): 116-29. Online: doi.org/10.1016/j.alter.2014.11.003.

Borg Karine, Hensing Gunnel & Kristina Alexanderson. 2001. Predictive factors for disability pension. An 11-year follow up of young persons on sick leave due to neck, shoulder, or back diagnoses. Scandinavian Journal of Public Health, 29(2): 104-12.

Bruusgaard Dag, Smeby Lisbeth & Bjørgulf Claussen. 2010. Education and disability pension: A stronger association than previously found. Scandinavian Journal of Public Health, 38(7): 686-90.

Bundesamt für Sozialversicherungen. 2019. IV-Statistik 2018. Bern.

Bundesamt für Sozialversicherungen. 2020. Codes zur Gebrechens- und Leistungsstatistik (KSGLS-C). Bern.

Bundesamt für Statistik. 2015. SAKE – Schweizerische Arbeitskräfteerhebung Fragebogen 2015: Vollständiger Fragebogen und Modul “Soziale Sicherheit”. Bern: Bundesamt für Statistik.

Bundesrat. 2005. Botschaft zur Änderung des Bundesgesetzes über die Invalidenversicherung (5. Revision) vom 22. Juni 2005. Retrieved from: BBl 2005 4459.

Bundesrat. 2018. Bundesgesetz vom 19. Juni 1959 über die Invalidenversicherung (IVG) – Chronologie. Retrieved December 4, 2018, from: www.admin.ch/opc/de/classified-compilation/19590131/history.html.

Canivet Catarina, Choi BongKyoo, Karasek Robert, Moghaddassi Mahnaz, Staland-Nyman Carin & Per-Olof Östergren. 2013. Can high psychological job demands, low decision latitude, and high job strain predict disability pensions? A 12-year follow-up of middle-aged Swedish workers. International Archives of Occupational and Environmental Health, 86(3): 307-19. Online: doi.org/10.1007/s00420-012-0766-4.

Corral Alza Antonio, Durán López Jessica & Iñigo Isusi. 2014. Employment opportunities for people with chronic diseases. Dublin: Eurofound.

Dustmann Christian & Francesca Fabbri. 2003. Language proficiency and labour market performance of immigrants in the UK. The Economic Journal, 113 (489): 695-717. doi.org/http://www.jstor.org/stable/3590195.

Eugster Beatrix, Lalive Rafael, Steinhauer Andreas & Josef Zweimüller. 2017. Culture, work attitudes, and job search: Evidence from the Swiss language border. Journal of the European Economic Association, 15(5): 1056-100. Online: doi.org/https://doi.org/10.1093/jeea/jvw024.

European Commission. 2020. Employment, social affairs & inclusion. Retrieved December 15, 2020, from: ec.europa.eu/social/main.jsp?catId=1484&langId=en.

Eurostat. 2022a. European Union Labour Force Survey (EU LFS). Retrieved May 16, 2022, from: ec.europa.eu/eurostat/web/microdata/european-union-labour-force-survey.

Eurostat. 2022b. Labour Force Survey in the EU, EFTA and candidate countries: Main characteristics of national surveys, 2020. Luxembourg.

Galaasen Anders Mølster, Bruusgaard Dag & Bjørgulf Claussen. 2012. Excluded from social security: rejections of disability pension applications in Norway 1998-2004. Scandinavian Journal of Public Health, 40(2): 142-9. Online: doi.org/10.1177/1403494811435497.

Grönvik Lars. 2009. Defining disability: effects of disability concepts on research outcomes. International Journal of Social Research Methodology, 12(1): 1-18. Online: doi.org/10.1080/13645570701621977.

Helgersson Magnus, Johansson Bo, Nodqvist Tobias, Lundberg Ingvar & Eva Vingård. 2015. Sickness absence at a young age and later sickness absence, disability pension, death, unemployment and income in native Swedes and immigrants. European Journal of Public Health, 25(4): 688-92. Online: doi.org/10.1093/eurpub/cku250.

Ishtiak-Ahmed Kazi, Perski Aleksander & Ellenor Mittendorfer-Rutz. 2014. Risk markers of all-cause and diagnosis-specific disability pension – a prospective cohort study of individuals sickness absent due to stress-related mental disorders. BMC Public Health, 14(805). Online: doi.org/10.1186/1471-2458-14-805.

Isphording Ingo E. 2014. Language and labor market success (IZA Discussion Paper No. 8572). Bonn.

Isphording Ingo E. 2015. What drives the language proficiency of immigrants? IZA World of Labor: 1-10. Online: doi.org/doi: 10.15185/izawol.177.

Jones Melanie. 2016. Disability and labor market outcomes. IZA World of Labor: 1-10. Online: doi.org/10.15185/izawol.253.

Karlsson Nadine E., Carstensen John M., Gjesdal Sturla & Kristina A. E. Alexanderson. 2008. Risk factors for disability pension in a population-based cohort of men and women on long-term sick leave in Sweden. European Journal of Public Health, 18(3): 224-31. Online: doi.org/10.1093/eurpub/ckm128.

KOF Swiss Economic Institute. 2022. What progress has been made on gender equality in the labour market? Vol. no158.

Krause Niklas, Lynch John, Kaplan George A., Cohen Richard D., Goldberg Debbie E. & Jukka T. Salonen. 1997. Predictors of disability retirement. Scandinavian Journal of Work Environment and Health, 23(6): 403-13. Online: doi.org/10.5271/sjweh.262.

Künn Steffen. 2015. The challenges of linking survey and administrative data. IZA World of Labor: 1-10. Online: doi.org/10.15185/izawol.214.

Lalive Rafael & Tobias Lehmann. 2020. The labor market in Switzerland, 2000-2018. IZA World of Labor: 1-11. Online: doi.org/10.15185/izawol.402.v2.

Lallukka Tea, Mauramo Elina, Lahelma Eero & Ossi Rahkonen. 2015. Economic difficulties and subsequent disability retirement. Scandinavian Journal of Public Health, 43(2): 169-75. Online: doi.org/10.1177/1403494814567028.

Meager Nigel & Darcy Hill. 2005. The labour market participation and employment of disabled people in the UK. Brighton.

OECD. 2009. Pathways onto (and off) disability benefits: Assessing the role of policy and individual circumstances. In OECD Employment Outlook 2009: 211-48. Paris: OECD Publishing.

OECD. 2014. Mental health and work: Switzerland. Paris: OECD Publishing.

Perhoniemi Riku, Blomgren Jenni & Mikko Laaksonen. 2020. Determinants of disability pension applications and awarded disability pensions in Finland, 2009 and 2014. Scandinavian Journal of Public Health, 48(2): 172-80. Online: doi.org/10.1177/1403494819843778.

Pieroni Luca, D’Agostino Giorgio & Donatella Lanari. 2019. The effects of language skills on immigrant employment and wages in Italy (Munich Personal RePEc Archive No91725). Munich.

Salonen Laura, Blomgren Jenni, Laaksonen Mikko & Mikko Niemelä. 2018. Sickness absence as a predictor of disability retirement in different occupational classes: A register-based study of a working-age cohort in Finland in 2007-2014. BMJ Open, 8: 1-8. Online: doi.org/10.1136/bmjopen-2017-020491.

Samuelsson Åsa, Alexanderson Kristina, Ropponen Annina, Lichtenstein Paul & Pia Svedberg. 2012. Incidence of disability pension and associations with socio-demographic factors in a Swedish twin cohort. Social Psychiatry and Psychiatric Epidemiology, 47: 1999-2009. Online: doi.org/10.1007/s00127-012-0498-5.

Schmidlin Sabina, Bühlmann Eva, Muharremi Fitore, Kobelt Emilienne & Cyrielle Champion. 2020. Evaluation der Integrationsmassnahmen zur Vorbereitung auf die berufliche Eingliederung. Bern.

Stapelfeldt Christina Malmose, Nielsen Claus Vinther, Andersen Niels Trolle, Krane Line, Borg Vilhelm, Fleten Nils & Chris Jensen. 2014. Sick leave patterns as predictors of disability pension or long-term sick leave: A 6.75-year follow-up study in municipal eldercare workers. BMJ Open, 4: 1-9. Online: doi.org/10.1136/bmjopen-2013- 003941.

Swiss Federal Social Insurance Office. 2022. Invalidity insurance benefits. Retrieved June 21, 2022, from: www.bsv.admin.ch/bsv/en/home/social-insurance/iv/grundlagen-gesetze/leistungen-iv.html.

Swiss Federal Statistical Office. 2019. Swiss standard classification of occupations 2000. Neuchâtel: Swiss Federal Statistical Office.

Swiss Federal Statistical Office. 2021a. Language - facts and figures. Retrieved January 5, 2021, from: www.eda.admin.ch/aboutswitzerland/en/home/gesellschaft/sprachen/die-sprachen---fakten-und-zahlen.html.

Swiss Federal Statistical Office. 2021b. Swiss labour force survey. Neuchâtel.

Swiss Federal Statistical Office. 2021c. Swiss labour force survey quality report 2020. Neuchâtel.

Swiss Federal Statistical Office. 2022a. Data linkages for third parties: Data linkage requirements. Retrieved May 13, 2022, from www.bfs.admin.ch/bfs/en/home/services/data-linkages/for-third-parties.html.

Swiss Federal Statistical Office. 2022b. Languages. Retrieved June 21, 2022, from www.bfs.admin.ch/bfs/en/home/statistics/population/languages-religions/languages.html.

Haut de page

Table des illustrations

Titre Table 1. Descriptive statistics for disability insurance benefit applicants by economic activity (%)
Légende p-values based on Pearson’s chi-square tests; comparing characteristics of the economically active and inactive subsamples. A p-value of <5 % was regarded as statistically significant in all analyses. Figures in brackets: Extrapolation based on less than 50 observations. The results should be interpreted with great caution.
URL http://journals.openedition.org/alterjdr/docannexe/image/1829/img-1.jpg
Fichier image/jpeg, 597k
Titre Table 2. Labour market status of disability insurance benefit applicants (%)
URL http://journals.openedition.org/alterjdr/docannexe/image/1829/img-2.jpg
Fichier image/jpeg, 134k
Titre Table 3. Regression analysis results with economic activity as dependent variable, Model 1
Légende Dummy variables for missing vales were included. Model 1 was estimated in the sample individuals who applied for a disability insurance benefit between 2000 and 2017. OR: odds ratio. CI: confidence interval. A p-value of <5% was regarded as statistically significant in all analyses.
URL http://journals.openedition.org/alterjdr/docannexe/image/1829/img-3.jpg
Fichier image/jpeg, 451k
Titre Table 4. Regression analysis results with economic activity as dependent variable, Models 2 and 3
Légende Dummy variables for missing vales were included in both models. Model 2 was estimated in the sample of individuals who applied for a disability insurance benefit between 2000 and 2017. Model 3 was estimated in the sample of individuals entitled to integration and occupational measures between 2000 and 2017. OR: odds ratio. CI: confidence interval. A p-value of <5% was regarded as statistically significant in all analyses.
URL http://journals.openedition.org/alterjdr/docannexe/image/1829/img-4.jpg
Fichier image/jpeg, 643k
Haut de page

Pour citer cet article

Référence papier

Szilvia Altwicker-Hámori, « Labour market integration of disability insurance benefit applicants in Switzerland »Alter, 17-1 | 2023, 69-86.

Référence électronique

Szilvia Altwicker-Hámori, « Labour market integration of disability insurance benefit applicants in Switzerland »Alter [En ligne], 17-1 | 2023, mis en ligne le 17 mars 2023, consulté le 29 avril 2025. URL : http://journals.openedition.org/alterjdr/1829 ; DOI : https://doi.org/10.4000/9ifi

Haut de page

Auteur

Szilvia Altwicker-Hámori

Institute of Public Health, School of Health Sciences, ZHAW Zurich University of Applied Sciences, Winterthur, Switzerland
Altw[at]zhaw.ch

Haut de page

Droits d’auteur

CC-BY-NC-ND-4.0

Le texte seul est utilisable sous licence CC BY-NC-ND 4.0. Les autres éléments (illustrations, fichiers annexes importés) sont « Tous droits réservés », sauf mention contraire.

Haut de page
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search