Thursday, 8 September 2011

Does inequality make sick?

Tony Blair was famously 'not bothered' about the super rich. However, more recently, the Labour Party has found its passion for a more equal income structure. There have also been some academics who argued for a while that there is a link between income inequality and health (The Spirit Level: why equality is better for everyone by Richard Wilkinso, Kate Pickett). In other words, they claim that there is evidence that the bigger the gap between top and bottom earners in terms of take home pay, the less healthy a society is. Health is often measured by morbidity, i.e. incidences of ill-health.

While for the Labour Party this new agenda may be prompted by its desperate search for its left-leaning voters, the case of academics is more puzzling. I believe there are some problems with their argument and here is why.

First, the evidence available is not in their favour. The Agency for Statistics in Germany has just released the figures of morbidity for the German states. The stats make interesting reading. Hamburg, Bremen and Bavaria, all states with the highest income gap, have the best population health. Thuringia and Saxony, states with the smallest income gap (half that of Bavaria), have the worst health outcomes for their populations.

So the evidence is at least contradictory. More importantly, however, there is, second, another problem with the supposed link between income inequality and health.

There has always been clear evidence that health is influenced by factors (amongst others) such as poverty and deprivation, access to prevention programmes, and health care quality. Now, the income gap says little about poverty. In fact, states with a large income gap between top and bottom earners might still have earners at the bottom of the scale that earn far more than those at the top of the scale in other states that have a small income gap.

In other words, income gap does not mean poverty. It's a relative measure of how far top and bottom are apart, which leaves the possibility (for example Hamburg), that the bottom earners are still better off than those in other areas.

Also, much has been made of the bad health outcomes figures of the US. However the figures there are skewed. Until recently, a small but sizeable minority of people had not access to health care at all, except for emergency care. It is difficult to see how this could NOT influence health outcomes. Early detection and treatment are a central pillar of good health care.

So, the argument will continue, but so far I am unconvinced that we have incontrovertible evidence that the income gap itself makes people sick.


  1. Why do you name only five states in your example?

    And all those you name with the smallest income gap and worst morbidity states are from the former Communist East Germany, whereas all the others you name which the largest income gap and the best morbidity stats are from West Germany!

    Prima facie doesn't seem like much of a "refutation" of Wilkinson.

  2. That's an interesting point you raised. Yes, the fact that they are from the former East Germany may have something to do with it. Yet this is exactly my point: health disparities are influenced by many factors, not just by income inequality. Some scientists emphasize path dependency (of health systems) to tease out some of the contributory factors for poor outcomes.

    What I argue against is any monocausal link between income inequality and health outcomes.

    On top of this we should look carefully at the unit of comparison (something that was raised when the authors of the 'Spirit Level' visited Cardiff. Why should we take national indicators, and where does that leave the fact that we have four home nations in the UK? Shall we measure Wales, Scotland, NI and England separately? And if so, should we do the same with Italian and German states/Laender?

  3. I don't see how that was exactly your point. You nowhere mentioned the fact that the two less well off states were until recently communist states. I find this misleading since most non Germans such as myself will not know this unless they are skeptical and check it out as you did.

    You also don't say why you chose only those five states. Wilkinson uses countries precisely to avoid picking only a few examples of something. He goes where there is a lot of data so he can get statistically significant results instead of relying on cherry-picking special cases.

    He does not restrict himself to countries when good data is available. For example, he uses the data for ALL 50 US states as well.

    And Wilkinson does not believe that inequality is something that explains everything. In fact he explicitly says he is only considering wealthy countries past a certain point of income where absolute poverty is no longer such a big concern. He lays out these criteria ahead of time and sticks to them.

  4. Many thanks for your comment! What I find important in the discussion about health inequalities is that we do not operate with singular chains of causality. Hence my remark that your point about the history of East Germany strengthens my argument rather than the authors's thesis. It is exactly that I am happy to include the fact that there is a different history between East and West German states which distinguishes my approach from that of Wilkinson et al. Obviously this is just a blog entry which summarises some methodological concerns.

    But I do think it is echoed in the analysis of the Rowntree Foundation (you can find the report here which says there needs to be more research to say WHY there appears to be a correlation between income gap and health inequality. It seems to me reasonable to differentiate between correlation and cause.