Background and mechanisms

Studies on the relationship between economic downturns and mental health trace their intellectual origins to Durkheim's seminal work ‘Le suicide’ (1897) (Durkheim, Reference Durkheim1951), in which he argued that changes in the economy could increase psychiatric pathology potentially leading to suicide (Horwitz, Reference Horwitz1984) and noted that macroeconomic phenomena could have significant psychological effects on individuals. Durkheim's research led sociologists to conduct research examining suicide rates within and across societies based on sociodemographic characteristics (Hamermesh & Soss, Reference Hamermesh and Soss1974). Decades later, Brenner conducted empirical studies linking societal economic change to increased psychiatric hospital admissions, arguing that economic instability was the major factor influencing rates of mental hospitalization (Brenner, Reference Brenner1973; Trainor et al. Reference Trainor, Boydell and Tibshirani1987). He found that individuals who had lost employment were at heightened risk and individuals who had not lost employment but who worked in organizations that lost many employees (causing fear and stress to those who remained) could also experience negative mental health outcomes (Brenner, Reference Brenner1990).

Durkheim and Brenner's research highlighted the impact of macro-level social conditions on population mental health with important public policy implications (Dooley & Catalano, Reference Dooley and Catalano1984). This work suggests that if the economic environment influences the incidence of symptoms of mental disorders in populations, then policymakers should consider the psychological implications of their economic decisions (Dooley & Catalano, Reference Dooley and Catalano1984).

Other theoretical models could explain how macroeconomic conditions might influence mental disorders. Brenner's ‘Economic Change Model of Pathology’ examined the complex relationship between the economic environment and changes in physical, social and mental well-being (Brenner, Reference Brenner1987). Catalano & Dooley proposed that macroeconomic indicators influence the use of mental health facilities based on the ‘provocation’ and ‘uncovering’ hypotheses (Catalano & Dooley, Reference Catalano and Dooley1977). During economic downturns, family members and society have decreased tolerance for disordered behavior, hence ‘uncovering’ behavior resulting in more manifest psychiatric disorder seen by health providers. Alternately, economic change can cause new disorder by eliciting or ‘provoking’ maladaptive behavior. Their work suggests more evidence for the uncovering hypothesis (Catalano et al. Reference Catalano, Dooley and Jackson1981), but they could not reject the provocation hypothesis (Catalano et al. Reference Catalano, Dooley and Jackson1985).

Other relevant theories include: the ‘shift’ hypothesis which suggested that economic contractions may deprive certain populations of private treatment, thereby increasing use of public mental health facilities (Catalano et al. Reference Catalano, Dooley and Jackson1985), and the ‘stress’ hypothesis that community economic change is related to mental disorder via individual life changes and elevated symptoms of depression and stress (Dooley et al. Reference Dooley, Catalano, Jackson and Brownell1981). It has been suggested that the relationship between the economy and mental health can be evaluated on the continuum between the ‘unfettered enterprise’ hypothesis and the ‘social protection’ hypothesis (Dooley & Catalano, Reference Dooley and Catalano1984). Unfettered enterprise suggests that the human costs of economic change are negligible and that ‘normal’ people can adapt to such change, whereas social protection identifies high human costs of economic change that may be inequitably distributed over the population (Dooley & Catalano, Reference Dooley and Catalano1984). To the extent that research identifies results more consistent with the social protection hypothesis, policymakers should seek to prevent harm and provide support for those who are most vulnerable during economic downturns (Marshall & Dowdall, Reference Marshall and Dowdall1982).

Other factors could explain the relationship between economic downturns and poor mental health. During times of stress, including economic recessions, people are likely to increase their drug and alcohol use, which can result in poorer mental health (Vlahov et al. Reference Vlahov, Galea, Resnick, Ahern, Boscarino, Bucuvalas, Gold and Kilpatrick2002; Roberts et al. Reference Roberts, Friedman, Brady, Pouget, Tempalski and Galea2010). The financial strain that results from retiring or losing a job can affect marital satisfaction and further reduce mental well-being (Higginbottom et al. Reference Higginbottom, Barling and Kelloway1993). Financial strain can result in the loss of personal control which further inhibits emotional functioning and physical health (Price et al. Reference Price, Choi and Vinokur2002). Financial strain can lead to reduced social support and negative relationships with family (Lincoln et al. Reference Lincoln, Chatters and Taylor2005). Finally, workplaces are often a source of social support; therefore, job loss can negatively affect psychological well-being (Atkinson et al. Reference Atkinson, Liem and Liem1986; Peirce et al. Reference Peirce, Frone, Russell, Cooper and Mudar2000).

While these theoretical perspectives and existing research have offered reasons why there may be a link between economic processes and population mental health, the empirical literature must evaluate whether such a link exists.

Overview of empirical research

Articles examining population macroeconomic decline and mental health have frequently used the unemployment rate because it is accessible and may provide a general reflection of the overall well-being of the economy (Dooley & Catalano, Reference Dooley and Catalano1984). Empirical research studies typically examine three primary mental health indicators: (1) mental disorders; (2) admissions to mental health facilities; and (3) suicide.

Studies of the relationship between macroeconomic decline and mental disorders were primarily conducted in the 1970s and observed that people in the workforce were more likely to be negatively affected by the economic downturn than people not in the workforce (Dooley & Catalano, Reference Dooley and Catalano1984). These studies reported that people of lower socio-economic status were more likely to be negatively affected than middle- or upper-status groups (Dooley & Catalano, Reference Dooley and Catalano1979, Reference Dooley and Catalano1984; Dooley et al. Reference Dooley, Catalano, Jackson and Brownell1981).

Studies on admissions to mental health facilities primarily used older data but examined the use of these facilities for most of the 20th century. These studies examined differences in rates of admissions overall, and were based on a variety of factors, including sex, age, socio-economic status, educational attainment and prior mental health status. There were no consistent relationships between any of these particular subgroups and increases in admissions. First in-patient admissions increased for those with higher levels of education but decreased for those with lower levels of education (Brenner, Reference Brenner1969; Barling & Handal, Reference Barling and Handal1980). Most studies either found that economic downturns were associated with increased first admissions (Brenner, Reference Brenner1967, Reference Brenner1969; Brenner et al. Reference Brenner, Mandell, Blackman and Silberstein1967; Marshall & Dowdall, Reference Marshall and Dowdall1982; Kiernan et al. Reference Kiernan, Toro, Rappaport and Seidman1989) or there was no association (Barling & Handal, Reference Barling and Handal1980; Ahr et al. Reference Ahr, Gorodezky and Cho1981; Catalano et al. Reference Catalano, Dooley and Jackson1985).

Most research on the impact of unemployment on mental health focuses on studies of suicide. Findings differ by demographic characteristics (such as sex and age) as well as based on occupational status. One study showed that suicide and unemployment rates were positively correlated in employed males and negatively correlated in employed females, but no significant correlations were found between suicide and the unemployment rate among unemployed people of either sex (Platt et al. Reference Platt, Micciolo and Tansella1992). Studies have shown mixed results regarding whether men (Vigderhous & Fishman, Reference Vigderhous and Fishman1978; Lester et al. Reference Lester, Motohashi and Yang1992, Platt et al. Reference Platt, Micciolo and Tansella1992; Gerdtham & Johannesson, Reference Gerdtham and Johannesson2005, Inoue et al. Reference Inoue, Tanii, Abe, Nishimura, Kaiya, Nata, Okazaki and Fukunaga2008), or women (Dooley et al. Reference Dooley, Catalano, Rook and Serxner1989; Weyerer & Wiedenmann, Reference Weyerer and Wiedenmann1995; Gunnell et al. Reference Gunnell, Middleton, Whitley, Dorling and Frankel2003), have increased rates of suicide during economic downturns.

Where findings were unexpected (e.g. there was no relationship between suicide and economic downturns, or suicide rates appeared to decline during downturns), authors tried to explain their findings (Neumayer, Reference Neumayer2004). For example, if a government provided social support during economic crises, such as labor market programs, this social protection could have mitigated the impact of unemployment and suicide mortality (Waters et al. Reference Waters, Saadah and Pradhan2003; Stuckler et al. Reference Stuckler, Basu, Suhrcke, Coutts and McKee2009). Findings regarding suicide rates were often more nuanced than simply a negative or non-significant relationship between economic downturn and suicide.

Gaps, challenges, and priorities

The literature suggests that economic downturns are associated with increased psychological distress, use of mental health facilities and suicide. Economic declines may have a disproportionately negative impact on vulnerable population subgroups, such as those with pre-existing mental disorders, those of lower socio-economic status and the unemployed. While the literature is relatively consistent, there are several limitations associated with the existing work on this topic.

There are substantial differences in study quality, yet this fails to fully explain the heterogeneity in findings. Heterogeneity based on factors such as populations, methods for measuring employment, and reporting of suicide likely accounts for some of the observed differences in findings across studies. There are also several potential measurement problems associated with examining the relationship between economic downturns and population mental health. These include potential unmeasured variables, level of analysis, and use of aggregate economic measures which may not sufficiently capture an individual community's experiences.

The appropriate use of statistical methodology is a concern for analysis of the influence of macroeconomic forces on population health over time. Most research on population mental health trends in response to economic downturns uses time-series methods. When evaluating work that uses time-series analysis techniques it is important to consider the potential use of an inappropriate or post hoc lag structure (e.g. how much lag time should be included if any, and should time be measured in months or years) (Charlton et al. Reference Charlton, Bauer, Thakhore, Silver and Aristidou1987), whether data should be de-trended, the immediacy of change and the method used to measure the impact of economic change on community mental health, choice of method to account for correlated data, and suitable framing of these analytic methods within a theoretical base (Catalano & Dooley, Reference Catalano and Dooley1983). There is no universally agreed-upon ideal lag time, making comparisons across studies difficult (Kasl, Reference Kasl1979).

Additionally, the ‘ecological fallacy’ (Robinson, Reference Robinson1950), i.e. errors that may arise from attributing population-level phenomena to individual experiences, may prove to be a concern in this type of research (Dooley & Catalano, Reference Dooley and Catalano1979). When considering studies over a long period, concerns about study quality and comparability of research standards arise. Studies concerned with mental disorders or mental health utilization frequently used archival measures of mental health phenomena such as hospitalization (Dooley et al. Reference Dooley, Catalano, Jackson and Brownell1981), and these studies are rather dated. Future research should investigate specific symptoms or diagnoses rather than just hospitalizations and suicides, events that are much rarer than other mental health issues. However, while it may be more difficult to obtain community-wide mental health data now than it was in prior decades, due to increasing patient data protections based on the Health Insurance Portability and Accountability Act (HIPAA) (Nosowsky & Giordano, Reference Nosowsky and Giordano2006) or other mechanisms, this research is much needed given today's economic state.

A number of these gaps in the existing literature lay the framework for future research opportunities that could prove highly beneficial. The best research likely combines both aggregate-level and individual-level research (known also as cross-level analysis) (Dooley & Catalano, Reference Dooley and Catalano1984). This would avoid the ecological fallacy while also accounting for pre-existing mental disorders and individual life events, and could improve understanding of how economic conditions are associated with disorders. Limited examples of cross-level analyses found similar relationships – as did the ecological studies – that economic downturns were associated with increased psychological distress, but continued research in this area is still needed.

While the literature covers several indicators over a long period of time and from several different countries around the world, it is still overall very sparse, prohibiting careful comparisons of similar phenomena at the same time and place. The unavailability of data across time (either a point in time or time periods) and place (size and scope of regions studied) as well as across either economic or mental health measures indicates that there is a clear need for research that could fill this gap.

While we do not focus on economic expansions, they have not necessarily always led to improved mental health, just as economic downturns do not always lead to mental health declines. Sometimes ‘counter-cyclical’ findings have been documented – economic expansions were associated with worse mental health (Dear et al. Reference Dear, Clark and Clark1979; Dooley et al. Reference Dooley, Catalano, Jackson and Brownell1981).