Work won’t set us free.

Today, the Department for Work and Pensions [DWP] announced the release of a Green Paper and consultation on disability, social security and work. The Green Paper wants a debate on “recognising the value of work as a health outcome”, while Damien Green (and many others) actively pontificate the ‘work cure’ — the idea that a sick of disabled person’s health or disability will ‘improve’ if they become employed. This goes so far that the Chief Exec of Public Health England claimed health and happiness were “inextricably” linked with employment.

A generous interpretation of this state of affairs would be that those claiming work improves health have engaged with the relevant literature [there are decades of studies] in only the most basic and facile way. My personal interpretation is somewhat more cynical:

The claim that work improves health is a lie.

Given that there are decades of studies, I’m obviously not going to go through them all, or even a substantial amount. I’m going to go through a few that illustrate the flaws in the ‘work cure’ argument but, to get it out the way early: there, to my knowledge, has been no study that tested the hypothesis that employment improves health. What the evidence does say is that the consequences of unemployment — societal isolation, poverty and so on — are associated with poorer health.

This is an example of what is commonly called the “third variable problem”, where an observed relationship is actually caused by other variables. There is a correlation between the number of churches in an area and the number of crimes committed in that area. This is not because churches cause crime, but because a higher population causes both, more churches and more crime.

It’s not unemployment that is causing ill health, but the societally-inflicted consequences of unemployment that do.

Study 1: Two factories closed down. 100 employees were recruited into this study which measured informal social support. The study continued for two years. The ‘low support’ and ‘high support’ groups did not differ in how long they were unemployed, or in incomes. Despite that, during their period of unemployment the low support group exhibited significantly more physical and psychological symptoms but objective differences too, including elevated serum cholesterol. If it’s unemployment that is bad for health and not the social isolation that, for numerous reasons, accompanies it then we would expect no difference between the two groups on the various health measures.

Study 2 and Study 3 make similar points. Study 2 was A cross-sectional survey study of ~40,000 people living in the EU, measuring the association between health and employment in different social security systems. While unemployed people consistently reported lower health (because unwell people tend to not work), the association between health and unemployment varied significantly between the different social security systems. The odds of an unemployed person reporting ill health in the UK’s system (the highest association) were 2.7x that of an employed person, but only 1.8x higher in ‘Southern’ systems (Portugal, Spain, Italy, Greece).

Study 3 was a 20 year longitudinal study following a cohort in Germany and one in the US, about 20,000 people in total. There was a significant association between unemployment and mortality in America, with people unemployed at the time of the surveys being 2.4 times more likely to die before the study completed. A small effect (1.4x more likely) was found for Germany, but this was not statistically significant. When education was included in the model, low- and medium-skilled (i.e. working class) Americans were found to be the most at risk.

Assuming work improves health and thus unemployment is a direct cause of poor health, these two studies beg a question: why do the health effects of unemployment vary between countries? Smoking tobacco is a direct cause of lung cancer, but it’s not safer to smoke tobacco in Germany than America. Tobacco doesn’t become less carcinogenic if you smoke it in Germany. You shouldn’t be less likely to die while unemployed in Germany than America if it were unemployment itself causing ill health. Study 2 takes this further and directly links it to social security systems.

Like the social support in Study 1, Study 2 shows how preventing one of the inflicted punishments of unemployment is associated with a diminished health harm while being unemployed. First it was social support — with supported people showing objective blood chemistry improvements and now it’s poverty. In the UK (and basically everywhere), life expectancy and health are related to local wealth and income through numerous mechanisms. That the relationship between unemployment and health varies between different social security systems — which vary in generosity, conditionality, universality — again suggests unemployment isn’t the true cause here. If you could stop tobacco causing cancer by changing something else, that would be evidence against tobacco being a true cause of cancer. The same applies here.

Study 4 takes an old study and repeats it to see what’s happened over time. It repeated a 1996 study using the Office for National Statistic’s nationwide surveys. The original study found that working class people were the most likely to have negative health effects following unemployment. The updated study concluded that “between 1973 and 2009, the relation between good health and securing and sustaining employment has strengthened”. The study doesn’t directly discuss the causal relationship between health and work, but its findings are relevant here. Disabled people are being removed from the labour market; “slight disadvantages in terms of health can lead to large disadvantages in terms of employability”.

It’s not that unemployment is causing higher ill health, but that ill health (and discrimination) are causing higher unemployment. If it were unemployment causing ill health, why is the association getting stronger? Why is unemployment ‘less healthy’ now than it was in 1973? Cigarettes aren’t less healthy now than they were in 1973. The simple answer to that question is that the social and economic contexts are different — we’re awash with scrounger rhetoric and rising hate crime with a social security system with stigma at its core.

Disability directly (inability to work) and indirectly (discrimination) causes unemployment, and the treatment unemployed people are subjected to causes ill health. Poverty, isolation, lack of purpose and so on are not essential facets of unemployment. Rich people frequently don’t work, but they are the healthiest group. The relationship between unemployment and health has changed a lot since 1973, before the rise of Thatcherism and the widespread destruction of working class communities based around mining and manufacturing — exactly the group most at risk of the ‘negative effects of unemployment’. Sunderland, previously a shipbuilding hub, has one of the highest unemployment rates with 2.9% claiming Jobseeker’s Allowance (vs 1.6% nationally). It also has the highest national rate of antidepressant prescriptions, with 1 for every 4 adults.

The work cure is a lie. There has never been a study showing that unemployment is a direct cause of ill health or that employment directly improves health, and the studies support the the idea that the causality between unemployment and health is in the opposite direction. The punishments for unemployment like poverty and isolation can be avoided or at least ameliorated in the current economic system (e.g. via social security), but they are just tweaks to a harmful system.

In 2010, the then Secretary of State for Work and Pensions Iain Duncan Smith said, on BBC News, that “work actually helps free people”. By getting jobs, we’ll ‘work our way out of poverty’ despite more people in poverty being in “working families”. We’ll ‘improve our health’ in spite of our genetics, our absent limb, organs, chemicals or senses. Insidiously, they want to redefine employment as a ‘health outcome’. They want an open goal — if they redefine health as ‘being employed’, then of course people in employment will be healthier and becoming unemployed will be ‘losing’ health.

Unfortunately, work doesn’t actually help free people. But a society that values its members as more than economic production units, that doesn’t punish unemployment with destitution and isolation, that allows unemployed people to contribute — artistically, spiritually, interpersonally, scientifically, politically — just might.

Addendum: Since posting this, I’ve actually seen the Green Paper. It included this lovely.. diagram [Nov.1: now only accessible through Google’s cache and the PDFs].

“Good work leads to good health leads to good work leads to good health leads to…”

Three brief points to make. Firstly, “worklessness” is not the opposite of “good work”. What about ‘bad’ work (low pay, bad conditions, poor accommodations), which is what disabled people are most likely to get?

Second. Penicillin is an antibiotic. There isn’t ‘Penicillin’ and ‘Good Penicillin’, the latter of which is more effective. An intervention (‘employment’, penicillin) works, or it doesn’t. The intervention has specific rules to follow — min and max doses and timings for drugs, length and frequency of therapy sessions, and so on. Their proposal of ‘good’ work as a health intervention obviously fails — either all work is good for health (in the same all penicillin works the same on particular bacteria) or it’s something else that’s good for health (e.g. money.) And since effective interventions have an upper limit on dose or duration (e.g. for patient safety), I wonder when the DWP are going to publish the correct number of hours, salary, terms and conditions that need to be applied universally across all jobs, to ensure the Health of the Nation. Why isn’t work essential for the health of older people? *innocent-face*.

Lastly, the word ‘work’ itself. The DWP cares not one jot about ‘work’. They care about employment. Work is the expenditure of mental or physical effort towards a goal. Employment is (generally) doing it toward’s someone else’s goal in return for less money than your work is worth. Writing this was work — the research, the actual writing and so on. Unpaid work (i.e. not employment, like volunteering, or caring for family) can sometimes be good for health because they alleviate some of the enforced consequences of unemployment, like isolation, a general sense of purpose or just keeping your brain ticking over. The DWP don’t want you to get ‘work’, they want you to get ‘employment’ so you can be a Net Contributor instead of a Net Drain because, remember, all we are is what we contribute to the economy. The moment your personally and socially beneficial volunteering interferes in the slightest way with the Decision Maker’s random whims, or when you are compelled to give free labour to charities, is when the distinction becomes clear.