SOEP-IS 2013 - Factorial survey module on job preferences and job offer acceptance

Auspurg, Katrin; Hinz, Thomas Research Report SOEP-IS 2013 - Factorial survey module on job preferences and job offer acceptance SOEP Survey Papers, No. 1091 Provided in Cooperation with: German Institute for Economic Research (DIW Berlin) Suggested Citation: Auspurg, Katrin; Hinz, Thomas (2022) : SOEP-IS 2013 - Factorial survey module on job preferences and job offer acceptance, SOEP Survey Papers, No. 1091, Deutsches Institut für Wirtschaftsforschung (DIW), Berlin This Version is available at: https://hdl.handle.net/10419/253408 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-sa/4.0/ Katrin Auspurg and Thomas Hinz Series H - SOEP-IS Modules SOEP-IS 2013 – Factorial Survey Module on Job Preferences and Job Offer Acceptance 1091 2022 Running since 1984, the German Socio-Economic Panel (SOEP) is a wide-ranging representative longitudinal study of private households, located at the German Institute for Economic Research, DIW Berlin. The aim of the SOEP Survey Papers Series is to thoroughly document the survey’s data collection and data processing. The SOEP Survey Papers is comprised of the following series: Series A – Survey Instruments (Erhebungsinstrumente) Series B – Survey Reports (Methodenberichte) Series C – Data Documentation (Datendokumentationen) Series D – Variable Descriptions and Coding Series E – SOEPmonitors Series F – SOEP Newsletters Series G – General Issues and Teaching Materials Series H – SOEP-IS Modules The SOEP Survey Papers are available at http://www.diw.de/soepsurveypapers Editors: Dr. Jan Goebel, DIW Berlin Prof. Dr. Stefan Liebig, DIW Berlin and Freie Universität Berlin Prof. Dr. David Richter, DIW Berlin and Freie Universität Berlin Prof. Dr. Carsten Schröder, DIW Berlin and Freie Universität Berlin Prof. Dr. Jürgen Schupp, DIW Berlin and Freie Universität Berlin Prof. Dr. Sabine Zinn, DIW Berlin and Humboldt Universität zu Berlin Please cite this paper as follows: Katrin Auspurg and Thomas Hinz. 2022. SOEP-IS 2013 – Factorial Survey Module on Job Preferences and Job Offer Acceptance. SOEP Survey Papers 1091: Series H. Berlin: DIW/SOEP This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License. © 2022 by SOEP ISSN: 2193-5580 (online) DIW Berlin German Socio-Economic Panel (SOEP) Mohrenstr. 58 10117 Berlin Germany [email protected] SOEP-IS 2013 – Factorial Survey Module on Job Preferences and Job Offer Acceptance Module Title in SOEP Documentation: Job Preferences and Willingness to Accept Job Offers Katrin Auspurg and Thomas Hinz Factorial Survey Module on Job Preferences and Job Offer Acceptance (Proposal to add Questions to the SOEP Innovation Sample 2013) Katrin Auspurg and Thomas Hinz (Department for Sociology, University of Konstanz, Box 40, D-78457 Konstanz, Germany; [email protected]; [email protected]) (http://www.soziologie.uni-konstanz.de/en/professuren/prof-dr-thomas-hinz/) December 2012 Documents: ‐ Proposal & Appendix (Survey Questions) ‐ CV Katrin Auspurg ‐ CV Thomas Hinz SOEP Survey Papers 1091 SOEP-IS Factorial Survey Module on Job Preferences and Job Offer Acceptance 1 1. Background and Motivation We propose to include a Factorial Survey module on job preferences and conditions of labor market participation. This will be the first multidimensional, experimental measurement of job preferences and will permit to evaluate the impact of gender, household structures, working conditions and workfamily policies on inequalities in the labor market participation as well as on work-family conflicts in a large representative, longitudinal survey. The main motivation for the module is given with persisting gender inequalities in the labor market. Despite a rising participation of women in higher education in the last decades women and especially mothers are still more likely to be fixed-term or not employed, they have lower career prospects, and they earn about 20 to 30 percent lower wages (for the US: Blau/Kahn 2007; Cha 2010; for Germany: Gartner/Hinz 2009; Gundert 2007). Extensive controls of the human capital endowment, labor market positions and working hours explain only part of these inequalities (Blau/Kahn 2007; Gangl/Ziefle 2009; Gartner/Hinz 2009). The remaining gender gaps might be caused by omitted differences in performance or a (statistical) discrimination by employers (England 1992). Both of these explanations point to the demand side of the labor market. There are, however, also strong reasons to believe that the remaining gaps are caused by differences in the labor supply of men and women (Kirchmeyer 2005). According to this line of reasoning, labor market inequalities can only be explained by additionally considering personal job preferences (Hakim 2000, 2003).1 So far the research progress on preferences for working conditions is hindered by a misfit between well-elaborated theoretical models and empirical data. Labor market theories (like human capital and search approaches) and preference theories assume that people accept a job only in case the offered wages and working conditions meet their preferences. However, with existing labor market data, it is very difficult to isolate these preferences. One reason is the problem of sample selection: Especially for women and mothers, labor market data provide only biased estimates of preferences and reservation wages. This is because people participating in the labor market do not represent a random selection, but only a selected group where the job offers meet their preferences (Heckman 1979).2 Furthermore, to determine to what extent social groups like men and women differ in their preferences, and not only in their access to job offers, one would need information on all available job options, never minding if they were accepted or not. Yet survey and register data provide in general only information on accepted, and not on declined job offers, showing only half of the picture. For similar reasons men’s and women’s responsiveness to flexible working conditions and family-policies such as schedule flexibility and childcare is unclear. Those options might not exist or persons might not choose them when offered because of their fear of signaling lower work commitment and being punished with lower income and career prospects (Glass 2004). All in all, little is known on preferences for working conditions and the extent to which working conditions and work-family policies impact gender inequalities in the labor market. 1To name only one example: Wage penalties for mothers might stem from their special preference (and need) for working fewer hours, flexible schedules or jobs with lower commuting distances. All of these features restrict the range of available job options and are typically connected to jobs with lower earnings (Auspurg/Schönholzer, forthcoming; Gangl/Ziefle 2009). 2Corrections for the resulting selection bias rely on strong assumptions and sometimes lead to even more biased results (Stolzenberg/Relles 1997). SOEP Survey Papers 1091 SOEP-IS Factorial Survey Module on Job Preferences and Job Offer Acceptance 2 Similarly, the research on the decision making within partnerships or on choices between market and family work is hindered by a lack of adequate data. Quantitative studies on family decision making mostly research changing employment patterns over time or when moving to other locations (see, for example, Boyle et al. 2009; Geist and McManus 2012; Mincer 1978; Shauman 2010; Shauman and Noonan 2007; for studies with the SOEP: Jürges 1998, 2006; Lersch 2012; Nisic 2010). Most of these studies suggest that male careers are given priority. According to bargaining theories (Ott 1992), these gendered decisions are caused by less attractive earning prospects associated with female occupations (Davanzo and Hosek 1981; Mincer 1978; Shauman 2010). Following sociological gender display and role theories (Brines 1994; Fenstermaker 2002; West and Zimmerman 1987), the priority of male careers is rather triggered by gender-specific norms. Preference theories in contrast emphasize the impact of work orientations and values. Again, disentangling different theories would need a standardisation of job options available to men and women (Abraham et al. 2010). Observing what men and women would have decided in case they would have had similar attractive job options would most clearly reveal their pure willingness to accept different working conditions and show the existing heterogeneity in preferences.3 A promising research strategy to overcome these empirical problems in measuring labor preferences is to directly ask people on their reservation wages, preferences for job characteristics and willingness to move. This research strategy is meanwhile acknowledged in several surveys and was also implemented into the SOEP by means of single item questions.4 Analyses based on these data are very instructive (see e.g. the highly cited paper by Bielby/Bielby 1992; for some recent examples with the SOEP data: Caliendo 2010; Humpert/Pfeiffer 2012; Pannenberg 2007), but they hardly keep up with the complexity of real labor market decisions. The theory of compensating wage differentials (Rosen 1986) as well as theories on family decision making (like bargaining models) and preference models all posit fundamental trade-offs between monetary and non-monetary job characteristics that are definitely not captured with one-dimensional item-questions.5 In addition, methodological research suggests respondents to have problems to directly indicate monetary terms like (reservation) wages, or these measures being biased by social desirability (Jasso 2012; Breidert et al 2006). We suggest overcoming these limitations by extending survey questions on reservation wages and job preferences by means of a multidimensional, experimental Factorial Survey (FS) design. Within the FS module, all groups of respondents will be confronted with the same job offers. The multidimensional design will standardize the job offers and at the same time enable an estimation of trade-offs and reservation wages in a more subtle, indirect way. Research questions addressed by our design include: (1) To what extent will labor market inequalities between men and women still exist 3The analyses are additionally limited by the low case numbers of couples with long distance moves, even within large-scale population samples like the SOEP (Jürges 1998; Lersch 2012). 4 Examples for questions in the SOEP are: “Are you interested in full-time or part-time employment, or would both suit you?” “How high would your net income or salary have to be for you to take a position offered to you?” (TNS Infratest 2011). 5 To name only one example: Lower reservation wages of women might be caused by their higher preferences for fewer working hours or work-family policies like schedule flexibility and child care arrangements. With onedimensional questions the exact trade-offs between wages and working conditions cannot be estimated, which is one reason why simple measurements of reservation wages can be very misleading (Glass 2004; Humpert/Pfeiffer 2012). SOEP Survey Papers 1091 SOEP-IS Factorial Survey Module on Job Preferences and Job Offer Acceptance 3 when they are offered similar job options? (2) Are gender differences in the willingness to accept job offers in line with labor market approaches and theories on the family ? (3) What is the impact of practical measures like child care arrangements or flexible working hours on labor market participations? Revealed gender differences in the evaluation of job offers would be a very strong indicator for gendered differences in labor supply. In contrast to other methods, the FS module provides the exceptional possibility to confront men and women with exactly the same job options. Analyses on differences between social groups like men and women will therefore be more straightforward than with non-experimental data. 2. Planned Factorial Survey Module Factorial Surveys (FS) have come to be an enriching tool in many subfields of social sciences as well as in up-to-date survey research (Mutz 2012; Wallander 2009). Within FSs, respondents are asked to form a judgment in regard to hypothetical objects or situations (vignettes), consisting of several characteristics (dimensions). The values (levels) of dimensions are experimentally varied, so that their impact on respondents’ judgments can be estimated (Jasso 2006; Rossi and Anderson 1982). With all respondents getting similar vignettes, there are no sample selections or selectivity issues. By embedding the experiment in surveys like the SOEP-IS, a heterogeneous respondent population is reached. Thus, FSs borrow the advantages of both experimental and survey research. But surprisingly, FSs so far were rarely used for labor market research (for a review: Wallander 2009). In our FS module respondents will be confronted with hypothetical job offers. Each respondent is proposed to get five different vignettes describing job offers in Germany. Overall, half of the offers will be in a daily commuting distance (about one hour traveling distance from the actual place of living) while the other half will require moving household (four hours distance). Therewith it will be possible to study the willingness to move, which is—as indicated—especially instructive on the tuning of careers within couples. Besides this, several characteristics of the job offers like the monetary gain related to them, the working hours, the long-term career prospects, and child care opportunities will be varied. For people living in partnerships, additionally the employment prospects for their partners will be indicated. All dimensions are derived from core theories of social science (Abraham et al. 2010; Auspurg et al. 2010, 2011). Respondents will be asked to indicate for each offer (a) the attractiveness, (b) the likelihood that they will accept it, and (c) that the likelihood that they will move to the new location. Figure 1 in the appendix shows an example; Table 1 in the appendix displays the dimensions and levels that will be varied across the vignettes. The FS module is proposed to be presented to all respondents who are employed, on parental leave, or looking for employment at the time of the survey. In order to present job offers as realistic as possible and to better standardize the situations across different respondents, the monetary gains should be implemented as percentage increase of the actual net household income of respondents.6 For respondents living in a partnership, we suggest to work with a mirror-inverted design: One partner is put in the position of the “mover”, meaning that she or he is presented an own job offer. The partner is 6 This will enable us to indicate income gains also for respondents who are not employed at the time of the survey. In case of missing values on household income, the monetary gain will be indicated solely as percentage gain (“your net household income will rise by X percent”). SOEP Survey Papers 1091 SOEP-IS Factorial Survey Module on Job Preferences and Job Offer Acceptance 4 put in the position of a “tied mover”, meaning that he or she is asked to imagine that his or her partner gets the offer. All tied movers will then be asked to rate a) the attractiveness of the offers for their partner, b) their own level to which they would like their partner to accept the offer, and c) the likelihood that they will agree on a joint job move. This mirror inverted design will even more closely narrow the real decision processes in partnerships and enhance the possibilities to test bargaining theories. A similar design was already well proved to study the decision processes and conflict potential related to moving decisions within dual-earner households (Abraham et al. 2010). We propose to implement the vignettes as a self-completing module since the information seems too complex to be processed when only read by an interviewer. Different respondents should get different subsets of vignettes. Those experimental split-ups can be easily implemented in the CAPImode.7 The expected sample sizes of about 2,300 households (Richter/Schupp 2012) will provide enough power to study subgroups related to theoretical propositions (like respondents with or without children). With similar FS modules (see Section 3) the mean response time measured for five vignettes was 3.7 minutes (sd: 2.3 minutes). The gained evaluations will offer outstanding possibilities to isolate the responsiveness of labor supply to changes in working conditions. Questions that can be addressed include: How large have monetary gains (reservation wages) to be that job offers are seriously considered? Are there genderspecific differences in job preferences, and to what extent are these dampened by family-policies like schedule flexibility or child care opportunities? Additional insights could be gained by a repeated implementation in further waves of the SOEP-IS. For example, there is still a debate on the reasons for the highly gendered decisions in couples with young children (Blossfeld and Drobnič 2001). Longitudinal data could help to find out to what extent preferences for job characteristics and reservation wages change over time. Due to the large scale sample, there will be enough respondents who start a family, new partnership and/or change their employment status. These and other changes in living conditions would add even more possibilities to test theories of labor market and family research.8 Hence, the FS module would be ideally implemented in several waves of the SOEP-IS. An implementation in the SOEP main sample would enable to study even more fine-grained social subgroups like working mothers at the time of the survey. But even when implemented solely in the 2013 SOEP-IS, a considerable amount of information would be gained. 7 For doing the programming of FS modules somewhat more effort is needed than for standard survey modules. The research team has sound experience with the implementation of such modules and is willing to provide material that makes programming as easy as possible (for instance, a data base containing the vignette texts for all single respondents could be delivered and be used like other pre-load variables when programming the CAPIquestionnaires).  8For instance, it is well known that the probability of re-employment decreases with length of unemployment. Possible mechanisms underlying this association are a depreciation of human capital, a ‘scarring effect’ of unemployment that dampens the probability of getting re-employed, or a discouragement of the unemployed persons, meaning their preferences change (Dahl and Lorentzen 2003; Gangl 2004). By confronting people over time with exactly the same job offers one could disentangle these different propositions.  SOEP Survey Papers 1091 SOEP-IS Factorial Survey Module on Job Preferences and Job Offer Acceptance 11 Table 1: Overview on the Dimensions and Levels Dimension Levels 1 2 3 1 Gain of income a 6 levels : 0% | 20% | 30% | 40% | 50% | 60%a 2 Minimum working hours 20 hours 30 hours 40 hours 3 Career prospects none few many 4 Duration of employment contract unlimited limited for 1 year limited for 3 years 5 Demand to work over-time no sometimes often 6 Schedule flexibility no some high 7 Child care facilities no part-time full-time 8 Commuting distance (single way) 45 min 1 hour 4 hours 9 Partner’s chances of finding a job at new location c worse similar better a Oversampling of higher income gains intended; gain of income will be adapted to actual working hours of respondents (e.g., with 20 instead of 40 working hours, the percentage increase for the full-time employment will be twice as large). c Dimension will be presented only to respondents actually living in a partnership. SOEP Survey Papers 1091 SOEP-IS