Stressed in Sweden 2000-2011
Maria Stanfors, Lund University
Jeffrey Neilson, Lund University
Since the mid-1990s, stress-related illnesses have become increasingly prevalent in Sweden, particularly among women with detrimental consequences for both income and health. Although stress in Sweden may be particularly acute due to competing work and family demands, as many work full-time (including mothers), the topic has not been investigated using time use data. The Swedish Time Use Survey (2000/01 and 2010/11; N=8,564) combines a subjective measure of stress with objective measures of work- and family-characteristics and diary information on time spent on various activities, which enables us to capture multiple dimensions of the relationship between work (paid and unpaid), parenting and stress. We use econometric methods to study the role of both economic pressure and time pressure for experiences of stress from a gender perspective. Results illustrate who is stressed in Sweden, the primary work- and family-determinants, how this differs by gender and has changed over time.
Background:Since themid-1990s, stress-related illnesses have become increasingly prevalent inSweden, particularly among working women (and mothers), with detrimentalconsequences for both income and health since such illnesses often lead to longspells of sick leave (Danielsson et al. 2012). This warrants a betterunderstanding of stress, its determinants and association with time use, notleast because parenting stress can exert harmful effects on the well-being ofparents, as well as relationships with their children (Crnic, Gaze &Hoffman 2005). In this paper we investigate self-reported stress using nationally-representativedata from the Swedish Time Use Survey (2000/01 and 2010/11; N=8,564),taking a gender perspective thorughout. Research questions include: Who isstressed in Sweden? What are the primary work- and family-related determinantsof stress for men and women and how did it change between 2000 and 2010? Towhat extent are experiences of stress associated with mens and womens dailytime use? We also address the role of economic pressure and explore therelationship between income and stress. We believe our approach, which uses twowaves of data, combining a subjective measure of stress with objective measuresof work- and family-characteristics, plus time diary information, capturemultiple dimensions of the relationship between work, parenting and stress incontemporary Sweden. Theoreticalconsiderations and previous research: From an economic perspective, time and stress are interconnected,and can arise from tensions generated by feelings that the time available isinsufficient to complete the desired activities (Bonke & Gerstoft 2007; Hamermesh& Lee 2007). Since nobody is exempt from the time constraint of a 24-hourday, it stands to reason that in Sweden, where most men and women not only workfull-time, but combine this with active parenting, a significant proportion ofthe population may feel stressed given potential conflicts and time constraints.Economic constraints can also induce stress, not least because lower incomereduces the ability to offset time pressures (e.g. outsourcing domesticservices). Time pressures are experienced differentlyfor men and women (Mattingly & Sayer 2006), and working mothers tendto report more stress than working fathers (Hill 2005). Parenthood shouldeffect womens stress more than mens given that they more likely performemotionally demanding tasks with children (Hochschild 1997; Mattingly &Sayer 2006), family to work spillovers are more common for women and thus theyare more likely to feel rushed, and children inhibit womens free time morethan mens (Sayer 2005; Neilson & Stanfors 2017). Hence, it is notsurprising that women report more stress in the presence of children than men(Musick et al. 2016). Single mothers likely face higher parenting stress and greatertime and financial pressures (Cooper et al. 2009; Kendig & Bianchi 2008). Data: Our sample consists of prime working-aged (25-54) men and women(N=8,564), using data from the 2000/01 and 2010/11 Swedish Time Use Survey,conducted by Statistics Sweden. The stress indicator variable is derived froma diary-day specific question, which was worded identically in 2000/01 and2010/11: Have you felt stressed during the diaryday? The question had a 98.5 percentresponse rate. Results: 15.5 percent of men and 19.5 percent of women reportedfeeling stressed during the diary day. For those with children, the proportionsfeeling stressed were higher than the overall sample, with 16.2 of men, 21.2percent for women, and 24.9 percent of single mothers feeling stressed. The proportionof men and women (by age category) who reported having a stressful day areshown in Figure 1, where stress declines with age, and womens stress prevalenceis higher in 2010/11 compared to 2000/01 across several age groups. Resultsfrom logistic regression (not shown) find that, net of controls, women are morelikely than men to feel stressed (OR 1.41, 95% CI 1.25-1.58), and we also findsignificant gender differences when we examine the employed, parents, and thosein dual-worker couples independently.
Regarding time use activities and their association with stress (Table 1), OLSresults show that men who had a stressful day worked more minutes than men whodidnt, and they also did more housework, slept less and watched less tv, onaverage. For women, we find stress is associated with more paid work, less sleepand tv, but we also find it associated with more childcare. Our results showgender differences in experiencing stress consistent with mens and womensdifferent work and family roles, indicating that the Swedish model with a highdegree of work family compatibility has many upsides, such a high female laborforce participation and state support for families, but that such progress requiresbetter strategies for healthier working lives. Figure 1. Proportion ofmen and women reporting stress, by age category and wave
Table1. Timeallocation in seven activities and the effect of stress, men and women
Session 1102: Families and Households