Which Potential Duration of Unemployment Benefits is Perceived as Fair? Results From a Factorial Survey Experiment

Welfare states redistribute between different groups of citizens – therefore, it is crucial for political actors to know to what extent citizens accept or reject those measures by the state. Existing empirical research in this field often focuses on quite general attitudes towards institutions of the welfare state. However, it is relatively unknown to what extent certain legislative actions or the content of laws is approved. Only little empirical evidence is available on the question how citizens evaluate specific of social or labor market policy regulations. To expand knowledge in this area, our study focuses on design alternatives for unemployment insurance. Our main research question is: “Which potential duration of unemployment benefit receipt do citizens perceive as fair for different groups of recipients?” In a broader sense, the general question is “Who deserves what?” from deservingness research. We draw a random sample of individuals employed (due to social security contributions) or registered as job seekers from the administrative data of Germany’s Federal Employment Agency (FEA). Among these, we conducted an online survey with a net response of around 1,300 people. To analyze our question of interest, we used a factorial survey design: Respondents were asked to evaluate different hypothetical situations (“vignettes”). We describe fictitious jobseekers who have different labor-market related characteristics that vary randomly: sex, age, reason for unemployment, labor market history, additional financial needs and support from spouse. Respondents should evaluate the length of unemployment insurance benefit they consider fair. Moreover, we randomly provided information on the current legal situation to half of the respondents to control for anchoring effects. We also are able to use additional information on political party preference and some attitudes towards social policy issues from the online survey and detailed information on socio-demographic characteristics and labor market history from administrative data, i.e. information on individual (un-)employment spells, wages, regional variables etc.. Our results show that different deservingness criteria like need and control matter for fairness judgments that go beyond the principles of risk insurance. While one’s gender makes no difference, respondents consider longer unemployment insurance benefits as fair if recipients had no control over the reason for their unemployment or if there are additional hardships like older relatives in need of care. In addition to that, reciprocity considerations matter: Older workers and workers continuously employed in the past deserve longer benefit payments in the respondents’ view. Shorter payments are considered fair if one’s spouse substantially contributes to the household’s income. If respondents receive information about the current legal situation, this also has an effect: better informed respondents grant less generous unemployment benefits. Moreover, results are quite stable if we control for some personal characteristics of the respondents. Our study contributes to the growing research on fairness considerations in different disciplines and provides information for possible design alternatives in the context of unemployment insurance systems.