Effects of unemployment and insecure jobs on youth wellbeing in Europe: economic development and business cycle fluctuations

In terms of its societal impact, the global financial crisis of 2008 was considered to be the most significant crisis since the Great Depression in 1929 (Rollero and Tartaglia, 2009), and it is often referred to as the Great Recession. Ever since, Europeans, and particularly young Europeans, have been facing a threatening work situation because unemployment rates have increased substantially in most European countries (Chung et al, 2012; Eurostat, 2014). At the same time, temporary employment has followed suit, albeit at a lower speed, but affecting mostly young people. , Empirical evidence suggests that unemployment and job insecurity have detrimental effects on individuals’ well-being, and not only in the general population (for reviews see Sverke et al, 2002; De Witte, 2005; Cheng and Chan, 2008; see McKee-Ryan et al, 2005, for a meta-analysis) but also among young people (see Voßemer and Eunicke, 2015). Hence, it is important to investigate the outcomes of unfavourable labour market conditions for young people in Europe, because employment opportunities mark young individuals’ transition to adulthood (Bynner and Parsons, 2002). Any attempt to fully understand the conditions under which young people facing unemployment and job insecurity are particularly vulnerable must also account for the role of macrolevel moderators, given the large variations in social policies and economic growth observed across the different European countries (Voßemer et al, 2018).

Although it is plausible to think that the strength of the individual-level effects of unemployment and job insecurity on well-being may vary between poorer and richer countries, few studies have looked at potential cross-country differences and the role of macrolevel moderators in understanding this relationship (for example Eichhorn, 2013; Wulfgramm, 2014). Most of these studies focused on the moderating role of labour market policies or the countries’ economic conditions, often measuring either GDP per capita or the country-level unemployment rate (UR). However, these two measures may have a completely different meaning when examined as moderators, and they should not be viewed as being interchangeable.

First, GDP and UR do not substitute each other, because some countries often experience jobless growth, whereas others may experience a decrease in UR due to a concerted effort by the government without any increase in GDP per capita. Second, conceptually, GDP and UR underline different phenomena. Thus, when they are investigated simultaneously, different mechanisms are being discerned and tested. GDP per capita reflects the level of resources per citizen available in the country such as employment-related (for example, unemployment benefits) or health-related (including health expenditure) resources. The present analysis focuses on GDP as an overall measure that captures all kinds of societal support and the capacity of the government to provide such support. In this way, it captures the broader context in contrast to previous empirical attempts that looked at specific policy measures (Voßemer et al, 2018). Based on conservation of resources (COR) theory (Hobfoll, 1989), it is argued that higher levels of GDP may buffer the detrimental effects on well-being of unemployment and job insecurity, because in countries with a higher GDP, people have access to more resources when they face difficult times – irrespective of that country’s UR.

When the level of resources (as measured by the GDP) is held constant, there are two potential mechanisms to consider regarding the moderating role of UR. On the one hand, based on the principle of social comparison or the norm (see Clark, 2003), unemployment is likely to hurt less when there are more unemployed people around, because it is attributed externally and not to one’s own, personal failure. On the other hand, from the standard economics perspective, higher UR means fewer prospects of finding a job for those currently unemployed – hence, strengthening the unfavourable effect of own unemployment. In reality though, both mechanisms are likely to be in place, and one can determine only empirically which of the two effects prevails.

The current analysis investigates the moderating role of both GDP per capita and UR simultaneously as moderators of the relationship between unemployment/job insecurity and well-being. This attempts to distinguish the economic (that is, resource) effect from the social norm effect. The GDP per capita is expected to represent mostly an economic resource. Even though the UR is likely to combine both economic and social norm considerations, this study expects that the effect of UR will mainly capture the social norm effect when simultaneously controlling for GDP.

Furthermore, the study also distinguishes between the moderating role of long-term trajectories and business cycle fluctuations, given their different nature with respect to the formation of individual expectations. Macroeconomics has long been preoccupied with the separation of long-term trends in key indicators from short-term business cycle fluctuations. There are two reasons to think that such a separation is relevant when analysing the moderating effect of macroeconomic conditions on the relationship between employment status and well-being. If one considers the resource availability argument (Hobfoll, 1989), the underlying reason for the separation of the long-term trend from the business cycle fluctuation lies within the state budgeting process. The budget is set annually, based on historic information and future spending forecasts. Therefore, it is more likely that short-term fluctuations will send a weaker signal to individuals with regards to resource availability compared to that from the long-term GDP trajectory. A similar mechanism is likely to be in place with regards to UR: a long-term unemployment trajectory should send a stronger signal to individuals with regards to social comparison than the business cycle fluctuation.