How Being Exploited at Work Undermines Your Mental Health

Seth J. Prins

A new study shows that exploitative working conditions have severe impacts on mental health – and makes the case for empowering workers through unions and workplace democracy.

Interview by
Meagan Day

It’s pretty intuitive that being underpaid and overworked leads to psychological distress.

A huge body of public health research backs up this intuition by observing a strong link between mental illness and socioeconomic status. Socioeconomic status in turn tracks pretty closely to who works hardest and gets paid the least, who has the fewest spare hours and resources to care for themselves, and who is most alienated from their work, often performing rote tasks under strict supervision.

A new paper published in the journal Epidemiology affirms the linkage, but argues that the research to date is limited by its emphasis on surface markers of socioeconomic status, like income, education, and occupational prestige. The study’s authors contend that to really understand the social causes of mental illness, researchers need to look under the hood at the process that gives rise to socioeconomic stratification to begin with: exploitation.

To that end, the researchers sought to measure the relationship between mental illness and degree of exploitation, defined as the value produced by a worker’s labour that is not returned to that worker. It found a strong relationship between the two.

The paper’s title, ‘‘The Serpent of their Agonies’: Exploitation As Structural Determinant of Mental Illness’, is a reference to a passage written by Karl Marx. For Marx, exploitation was the motor of capitalism, the process that generates profits and keeps the whole system going. To the extent that exploitation both exacerbates mental illness and is endemic to capitalism, solutions to public mental health crises must involve anti-capitalist politics.

Meagan Day spoke to Seth J. Prins, lead author of the paper and Assistant Professor of Epidemiology and Sociomedical Sciences at Columbia University’s Mailman School of Public Health, about how and why the study was conducted and its academic and political implications.


What is the main question you set out to answer? And what are the deficits in the research to date?


The impetus for this research, and the big question for me, is how does capitalism affect people’s health?

Mainstream public health research tends to focus on a more downstream question, looking at the so-called social determinants of health. And I think a lot of people who study the social determinants of health really don’t understand class. Often they view class as an individual attribute, either an inherited or achieved one. A person’s income, a person’s years of education, or the prestige of their job are all sort of imagined to be characteristics of the person.

But anyone who reads this will know that class is not an individual attribute. It’s a social relation, as Marx taught us. And one of the defining features of that social relation is exploitation. Exploitation doesn’t mean your annual income, but how much of the value you produce is being appropriated by others.

A mainstream social science approach is to look at individuals’ annual income and years of education, and put that in a model and see how it’s associated with physical or mental health. There’s a huge body of research that does that kind of work, and we do see very clear social gradients in health outcomes based on those sorts of things.

But an insight from Marxist theory is that focusing on this gradient really hides the mechanism. Exploitation is what’s structuring the types of occupations people have, their income, their opportunities for education, and so on. Since exploitation is actually what creates inequality, surely that’s what’s making people sick. That might sound obvious, but it’s not something that’s really demonstrated very often empirically in the literature. So that was the goal of the paper.


The study found a strong correlation between exploitation and mental illness. To measure mental illness you used the Kessler scale, which is something that a lot of people in this field use. But exploitation is a more unconventional thing to measure. How did you go about measuring exploitation?


One of the challenges of doing research with data at the level of populations rather than small samples is that a lot of those surveys are long-standing US National Institutes of Health-sponsored surveys, or long-standing university-sponsored surveys.

We’ve learned a lot from that data, but the questions about occupations and jobs and work tend to be, for lack of a better description, the bosses’ questions. The sorts of questions that workers, leftists, and union activists might be interested in don’t tend to make it into those surveys. And so we’re forced to get creative with approximating the things that we really want to measure.

In this case, actually measuring the classic definition of exploitation is really difficult, because you would need to know not only how much money a person has been paid, but also the profit they’re generating, and how much their employer is spending on wages, capital outlays, facilities, and so forth. Being able to do that is a long-term goal of mine, but right now it just doesn’t really exist in any form that we could then link to good health data.

We figured we could approximate exploitation by looking at what we’re calling unconcealed exploitation. Exploitation is inherent to all wage labour, but it’s concealed in someone’s salary or wage. But people often are exploited above and beyond their wage. You can look for example at how much someone works. Many people work well above the standard forty-hour work week.

We looked at people’s income based on a standard forty-hour work week compared to their income if you actually counted the total hours they worked. When you do that, assuming that they’re being as productive during those extra hours as they are normally, everything that they’re producing in those extra hours is going straight to their employers if they’re not being paid overtime. That’s how we tried to get a ratio of the hypothetical amount that someone’s not being paid.


I want to go back to this idea of exploitation as the mechanism that structures the conditions of people’s lives at and outside of work. I assume it’s not the case that the correlation you found means that people are sitting around thinking about how much of the value they create is being stolen from them and getting sad or angry, which then exacerbates mental illness.

Rather it’s more likely that the degree to which you’re being exploited tracks with the degree to which you’re alienated from your work, or have to work long or strange hours, or of course get less take-home pay to cover living costs much less anything nice. Am I right that these are the kinds of real links that exist between exploitation and mental illness?


Absolutely. So this measure is kind of like a distant upstream indicator that, as you point out, does not depend on whether people are aware of it or not. It’s not based on any sort of individual perceptions.

There is a huge body of work that shows how the organisation and structure of work influences mental health. In the mid-twentieth century, there was a lot of research that looked at feelings of powerlessness and alienation on the job. It was often operationalised more specifically as imbalances between the demands of the job and the amount of control that you have on the job. All of those things are shown to be associated with increased mental illness and increased substance use. And we actually have fairly good descriptions of how different types of jobs are better or worse at things like autonomy, control over your own schedule, and so forth.

That research is great, and those sorts of measures really do depend on individual perception. But we think that exploitation is something that happens upstream of those factors. The more exploitative a job is, the more the work environment is going to suck. It just became apparent to us that nobody had ever really actually established that the more distinct, upstream factor was itself associated with mental health. But of course, we think that it operates through the characteristics of people’s jobs of which they are much more consciously aware.


If this link between exploitation and mental illness is established, then many people will surely respond that if you want to solve this problem you should eliminate exploitation. But if we’re following Marx, exploitation is endemic to capitalism, not something that can simply be excised from our economic system. What are the implications of that for the mental health of people in our capitalist society?


This type of research is kind of a reaction to another variety of public health research that tends to individualise these issues. For example, you end up with a lot of recommendations for like workplace wellness and seminars on work-life balance, essentially making it the responsibility of individual workers to try to deal with the conditions of wage labour. Obviously those things are going to have limited success, and I think they’re emblematic of a neoliberal approach to thinking about public health and public health interventions.

Instead of work-life balance seminars or essentially therapy for being exploited, let’s talk more about workplace democracy. Let’s talk about worker ownership. Let’s talk about non-reformist reforms that chip away at the way we’ve structured our labour markets. The point of this research is to say that actually the thing that’s making people sick is inherent to the nature of wage labour, and you can’t really change it unless you’re willing to confront capitalism.

About the Author

Seth J. Prins is an assistant professor of epidemiology and sociomedical sciences at Columbia University. He studies the collateral consequences of mass incarceration for public health, and how the division and structure of labour influence depression, anxiety, and substance use.

About the Interviewer

Meagan Day is a staff writer at Jacobin.