Unemployment, measured and perceived decline of economic resources: Contrasting three measures of recessionary hardships and their implications for adopting negative health behaviors

Highlights

• Different types of recessionary experiences may affect health behaviors differently.

• Unemployment experience was associated with becoming a marijuana user.

• Measured decline in resources was associated with smoking cigarettes.

• Measured decline in resources was associated with lower likelihood of marijuana use.

• Perceived decline in resources was associated with harmful and hazardous alcohol use.

Abstract

Economic downturns could have long-term impacts on population health if they promote changes in health behaviors, but the evidence for whether people are more or less likely to adopt negative health behaviors in economically challenging times has been mixed. This paper argues that researchers need to draw more careful distinctions amongst different types of recessionary hardships and the mechanisms that may underlie their associations with health behaviors. We focus on unemployment experience, measured decline in economic resources, and perceived decline in economic resources, all of which are likely to occur more often during recessions, and explore whether their associations with health behaviors are consistent or different. We use population-based longitudinal data collected by the Michigan Recession and Recovery Study in the wake of the Great Recession in the United States. We evaluate whether those who had experienced each of these three hardships were more likely to adopt new negative health behaviors, specifically cigarette smoking, harmful and hazardous alcohol consumption, or marijuana consumption. We find that, net of controls and the other two recessionary hardships, unemployment experience was associated with increased hazard of starting marijuana use. Measured decline in economic resources was associated with increased hazard of cigarette smoking and lower hazard of starting marijuana use. Perceived decline in economic resources was linked to taking up harmful and hazardous drinking. Our results suggest heterogeneity in the pathways that connect hardship experiences and different health behaviors. They also indicate that relying on only one measure of hardship, as many past studies have done, could lead to an incomplete understanding of the relationship between economic distress and health behaviors.

Introduction

Are people more or less likely to adopt negative health behaviors during economic downturns? In the aftermath of the late 2000s “Great Recession,” a range of research has been directed at this question. The findings have been heterogeneous (Burgard et al., 2013, Catalano et al., 2011). At the population level, lower use of harmful substances and healthier lifestyles have been documented by Ruhm (1995) and Ettner (1997), among others, and most recently by Xu (2013). A different set of studies has focused on the consequences for change in individuals' behaviors of directly experiencing events that typically happen more often during recessions, such as unemployment or decrease in economic resources. Some of these studies have shown that consumption of alcohol and drugs is higher among those who experience unemployment (Henkel, 2011). Others have argued that people are less likely to engage in health-conscious behaviors, such as eating a healthy diet, when they experience financial strain (Macy, Chassin, & Presson, 2013). What is still missing from this literature is a careful distinction among the different underlying mechanisms that could be connecting typical recessionary hardships to changes in different health behaviors. In this paper, we use prospective data from a sample of individuals in Southeast Michigan in the wake of the Great Recession to focus on the potentially varying associations between negative health behaviors, namely smoking, high alcohol consumption, and marijuana use, with three typical recessionary hardships: unemployment, decrease in economic resources, and perceived worsened economic situation.

Researchers studying changes in health behaviors that follow after an individual experiences a typical recessionary shock have often conceptualized job loss or unemployment as a catalyst of adopting new negative health behaviors or relapsing to old ones. For example, in a prospective study of young Swedes, Janlert and Hammarstrom (1992) showed that those who lost jobs and experienced unemployment had four times the rate of hazardous drinking compared to their peers, net of other characteristics that could be associated with both job loss and hazardous drinking. Their finding has also been supported by a prospective study of younger workers in the Norwegian population, who were more likely to consume alcohol as well as marijuana in the aftermath of a job loss (Hammer, 1992). In a more recent paper, Khlat, Sermet, and Le Pape (2004) showed that French men who experienced unemployment had increased risk of heavy drinking, smoking, and consumption of psychoactive drugs. Moreover, Black, Devereau, and Slavanes (2012) demonstrated that displaced workers had greater rates of smoking compared to those not displaced, and that smoking increases translated to their higher incidence of cardiovascular disease over time. These conclusions support the explanatory framework of the stress hypothesis, which suggests that adopting negative health behaviors after a job loss is a coping strategy to deal with the stress of being stripped of a major social role and social networks (Hayes, 1982).

Additionally, increased stress may follow unemployment because of concurrent decrease in economic resources that can be stressful for the job loser or their family. However, it is important to recognize that unemployment and decrease in economic resources are two theoretically related but distinct constructs, and each may have its own consequences and therefore implications for health behaviors (Macy et al., 2013). Unemployment may bring a significant amount of anguish to the individuals who experience related financial trouble, but even net of financial hardship, it entails the loss of other positive aspects of work, from social interaction to the benefits associated with the enhancement of social status through employment (Burgard & Lin, 2013). Furthermore, financial struggles can occur for reasons other than unemployment, and a decrease in economic resources not directly tied to a personal loss of employment could also generate change in health behaviors. Importantly, health behavioral consequences of a decrease in economic resources vs. the experience of unemployment could differ in some instances. While unemployment may be associated with negative coping behaviors in response to stress, a decrease in economic resources may compel individuals to become more conservative in their spending on non-essential items, such as alcohol (Sutton & Godfrey, 1995). At the same time, longitudinal evidence presented by Dee (2001) shows that the prevalence of binge drinking rises during tougher economic periods, both among those who lose jobs and those who remain employed. Unemployment should therefore be considered independently from, as well as in connection with, changes in economic resources.

In this study, we build upon recent work by Macy et al. (2013) that called for a more careful distinction between experiences of unemployment vs. decline in economic resources and examined how lowered working hours, unemployment and financial strain each were associated with changing health behaviors. While a commendable first step, their study did not consider another potentially important distinction. We argue that it is also important to differentiate between a perception of financial strain that was reported by survey respondents vs. one that was more objectively measured by assessing a change in their economic resources over time. The perception of worsened financial situation may not perfectly overlap with a decrease in economic resources, and the two could be associated with health behaviors differently. Objective decrease in economic resources need not be associated with negative health behaviors, especially if it is not perceived as stressful by the individual. By contrast, self-reported financial strain is a direct expression of the stressfulness and pessimistic appraisal of one's financial situation, and could increase negative behaviors according to the stress hypothesis. The theoretical and analytical distinction between a measured decrease in economic resources and a perceived decrease in economic resources could have important implications for how we construct hypotheses about the adoption of negative health behaviors.

In this paper, we distinguish between three types of individual recessionary hardships, examining the associations between unemployment experiences, measured decrease in economic resources, and perceived decrease in economic resources, and the risk of subsequent take up of any one of the three negative health behaviors we focus on in this study: smoking, alcohol abuse, and marijuana use. We use the Michigan Recession and Recovery Study (MRRS), a longitudinal dataset initiated in the Detroit metro area in 2009–2010, with a follow-up interview in 2011, to explore the following research questions: (1) Are people who have had an unemployment experience more likely to adopt harmful health behaviors? (2) Are people who have experienced a measured decrease in economic resources more likely to adopt harmful health behaviors? (3) Are people who perceive a decrease in economic resources more likely to adopt harmful health behaviors?