Job Strain

What it is

Karasek's model, published in Administrative Science Quarterly in 1979, crosses two dimensions. Demands are what the job asks. Decision latitude is how much discretion the worker has over how to meet it. The combination that predicts harm is high demand with low latitude, and it is the combination rather than either dimension alone that carries the association.

The companion construct is effort-reward imbalance, Johannes Siegrist's, published in 1996, where workers see little return in pay, recognition or security for the effort they put in. The two models are usually studied together and they measure different things.

Stephen Stansfeld and Bridget Candy's 2006 meta-analytic review of the psychosocial work environment and mental health is the synthesis most often cited. Its finding is that job strain, low decision latitude, low social support, high demands, effort-reward imbalance and high job insecurity all predicted common mental disorders, with the strongest associations for job strain and effort-reward imbalance.

That last detail matters, because the popular rendering of this paper inverts part of it. A widely read account reports Stansfeld as finding that teamwork quality, social support and job security showed little correlation with depression. The review lists low social support and high job insecurity among the prospective risk factors. This publication does not repeat that reading.

In effect

The reason the construct travels beyond occupational health is that it relocates a problem people usually experience as personal. A worker who cannot change the pace of the work, and who is nonetheless held to its output, is not failing at coping. They are in a job with a known structural feature.

The connection to distraction that some writers draw is worth stating carefully as an argument rather than a finding: low control is an unpleasant state, an unpleasant state is a cue, and a phone is the nearest available relief. On that reading, addressing a distracted workforce by restricting devices treats the proximate cause and leaves the structural one in place. It is a reasonable argument and it is not what the occupational health literature measured.

The evidence base is observational throughout. People are not randomly assigned to jobs, and the association between job strain and mental health outcomes is consistent across later meta-analyses without that resolving the direction of causation. Causal wording needs care and this page avoids it.

What it does not say

It does not say that job strain causes depression. The literature is observational, the associations are robust and repeated, and the inference to causation is a further step the designs cannot take.

It does not say social support is irrelevant. That claim is a misreading of Stansfeld and Candy, whose review lists low social support among the risk factors.

It does not supply an economic figure. The estimate of annual cost to an economy that circulates with this material comes from an advocacy organisation, not from research, and should not be cited as a finding.

It does not diagnose anyone. Common mental disorders in this literature are measured by screening instruments in populations, and nothing here describes an individual's condition or a course of action for one.

And it does not belong to whoever most recently wrote about it. This is Karasek's construct, from 1979.

---