Richard Wilkinson Reducing Inequality

Richard Wilkinson Reducing Inequality

Seminar Friday 31st October 2008 Sheffield Quaker Meeting House

Verbatim transcript of Richard Wilkinson’s contribution and discussion.

There are many references here to Richard’s graphs. Unfortunately these are not currently available for copyright reasons, but will be published in his new book, the Spirit Level in 2009.

MARTIN: Good morning everybody. Welcome to the Sheffield meeting house and to the Socialist Health Association for those of you who have not been before. We are expecting another dozen or so people but my experience is that some people wander in at 11.30 so I am not proposing to wait for them.

Richard Wilkinson: I think they are interested in hearing David Blunkett not me … little do they know …

MARTIN: The Socialist Health Association, for those who don’t know, is an august body – founded in 1930 when it was called the Socialist Medical Association and we like to claim credit for the starting of the National Health Service because that was the the aim of the people who set the thing up in 1930 – to campaign for the establishment of a free National Health Service. They thought, and quite a lot of people thought, in the 1930s and 40s that once we had a proper Health Service, the demand for healthcare and the health of the population and health inequalities would all be sorted out. It wasn’t really, was it? If you read the debates from about 1940 … 1948 to 55, there was this general amazement about how much healthcare people wanted to consume, particularly teeth. Teeth and glasses and spectacles which is why we have the charges we have.

Once we had a Health Service, our predecessors started to get interested in public health because they started thinking well, yes, this is all very well but people are still unhealthy and there’s still a big difference between the health of the rich and the health of the poor. A seminal moment in that debate was in 1980 the publication of the Black Report. Richard Wilkinson is blamed by some people for the Black Report because he wrote a letter to the then Secretary of State for Health when … were you still a student at that time?

RICHARD WILKINSON: I’d just finished. It was a New Society article in the form of an open letter. It wasn’t just a letter.

MARTIN: Pointing out that inequality, measured in almost any way you could measure it, was worse then in 1980 than it had been in any earlier recorded period. The timing of the Black Report was terrible really because it arrived on the desk of the Secretary of State just after the Thatcher Government had been got into power and they didn’t really want to know about this at all. In fact they refused to talk about it. They refused to use the word “inequality.” They did pay for some interesting research which was all called health variations, an interesting term which… you know variation is okay. We can have variation that’s okay, isn’t it? Nobody worries about variation, but inequality implies that there’s something wrong about this variation. It’s a different way of looking at the same thing. Richard has been involved in the debate about inequality and I think it’s fair to say that you have moved away from health specifically into inequality in a more general sense.

RICHARD WILKINSON: I haven’t dropped health, I’ve just expanded.

MARTIN: Public health people of course always say public health encompasses everything which it does in some ways. Richard has been involved in that debate for years and years and years and we are very proud that he is a member of the Socialist Health Association and I don’t think I will stand in his way any longer, except to read out the instructions which essentially is that the fire escape is behind you, should there be a fire. Nobody told me to expect a fire drill. The toilets are just round there by the coffee point.

A little bit about how we are going to do the day. Richard will talk until he collapses from exhaustion because this is what he likes to talk about.

RICHARD WILKINSON: That’s not fair …

MARTIN: We are expecting David Blunkett to arrive about 11.30 and to stay for about an hour. So he is not going to hear what Richard is going to say but we have already sent him the summary; so I am hoping that he will talk a bit about the things Richard wants to talk about. After Mr Blunkett departs, we will have lunch and then we have two speakers in the afternoon, Neil Goulbourne and a man from Sheffield whose name escapes me.

A delegate: Kieron Williams.

MARTIN: Hello, have a seat. Grab a badge before you sit down. They are just here.

KIERON WILLIAMS: I think you just said my name.

MARTIN: Are you Kieron?


MARTIN: Neil and Kieron are going to talk, I suppose a bit more practically about what can and has been done about inequality in health more specifically now and at a more local level. Richard and David Blunkett, I think, are both going to talk more on a macro level — is that the right word? — about society at large.

RICHARD WILKINSON: Martin told you a little bit about the background to health inequalities. I remember them being not called just class differences in health or inequalities in health, but even differential ageing and social subgroups as a way of sort of hiding it under the early Thatcher years.

I notice that actually Martin had titled my talk something about poverty or inequality and actually that’s not a bad title for what I’m going to start off by talking about. Basically I am going to explain why it is inequality, not poverty in absolute terms that really matters in our society now, why the objectives should be reducing inequality. Maybe stealing the thunder from people later today. Members of the jury my view is that there are hundreds of ways of reducing inequality and the problem is political will. The only way to get reductions in inequality is if people widely recognise how important it is. So I am going to take you through some of the evidence that we have dug up very recently showing how different more and less unequal societies are from each other, how many health and social problems are much, much worse in more unequal societies not simply because the poor suffer more from all these problems but because the vast majority of the population in more unequal societies seems to suffer more.

Let me start with a very sort of simple graph. Are most people here happy reading graphs? You know roughly how they work. On this one we’ve got all these black dots which are countries and there’s life expectancy up here going from 30 right at the bottom up to 90. Every country in the world is somewhere on that. I don’t mean we have dots for them all. On here is how much the countries per capita gross domestic product is. So here are the rich countries with high life expectancy and the poor countries with low life expectancy. But what you see, why I am showing you this slide, is also life expectancy improves very rapidly in the early stages of economic development, as soon as you get out amongst the developed countries it makes no difference to get richer. So countries here have life expectancy as good as countries out here, although this is $15,000 per capita and this is $30,000 or $35,000 per capita. So you can get huge increases in levels of wealth of income and so on and life expectancy does not go on up. Although life expectancy goes on improving it’s nothing to do any more with whether our economic growth is fast or slow or non existent. It is not just that life expectancy doesn’t improve any more. This is a very similar graph, exactly the same along the bottom but this is the per cent of people who say they are happy or satisfied. Again, it increases very fast in the early stages of development, more and more people as societies get richer say they are happy and then it levels off again.

Almost any level of welfare, well-being shows the same sort of pattern; that now in the rich world, as we get richer, we no longer get the real social benefits that we used to get. In a way, modern society is at a turning point. For thousands of years if you wanted to improve the real living standards for human beings, the thing to do would be to raise material standards of living. We are the first generation to have got to the end of that. Economic growth may matter to people interested in profits and so on but for most of us in the developed world, it does not improve health or happiness or other measures of well-being. That is really important. It’s particularly important because we know actually that now economic growth, it’s not just that it doesn’t produce the benefits for us, it actually creates more and more environmental problems.

You can see that on this graph, this is infant mortality up there. Again, the dots are countries. It is a bit like the graph that I had before turned upside down. So instead of going up and flat levelling off, infant mortality as countries get richer, comes down. Instead of levels of income, GDP per capita, we have energy use per capita along the bottom. So this simply shows that as energy use increases, initially infant mortality falls but then you can consume hugely much more energy, more carbon emissions and so on, and it doesn’t help.

So we’ve run into the environmental buffers and we are the first generation to have to think how now do we improve the real quality of life? What does matter? Most of what I want to say is about the social environment. You know, I think that people often feel our society is materially very successful but they have a lot of social failings. Perhaps social failures more generally and it is that that I want to talk about. But just to emphasise the problem, the sort of contrast, this is just the rich countries. Each dot is one of the rich countries. So that is the richest one — I don’t know whether that’s the USA. I think we don’t have Luxembourg in this data so that is the USA there. These are countries on the level part of the curve out here. So one of then is Britain, there’s France, there’s Japan in there. You see there is no relationship at all between GNP per capita and life expectancy. It is not that it is not a statistically relevant relationship, a weak relationship: There is no relationship amongst the richest countries.

But within each of them there is an extraordinary social gradient in death rates right across society. This is American data, these are – each column is a zip code area, the postal code areas in the United States, classified by how rich or poor they are. So here you have the rich zip code areas with low death rates and poor zip code areas with high death rates. You can see this extraordinary gradient running right across society. It’s not just the difference between the poor and the rest of society, it’s even the difference between the rich and the very rich.

So income means something quite different within societies from what it means between societies. That is a really important distinction. Actually what it tells us is really what I was saying before: Average living standards, whether the whole society, the whole of our rich society is a little bit richer or little bit poorer doesn’t matter. What matters is the differences between us. It’s relative income that matters, social position, status — something like that. How we are in relation to each other.

Just to sort of rub that point in, because I think you know when anyone talks about health inequality everyone’s mind moves automatically to poverty and thinking of the homeless and people who run out of benefits before the end of the week and don’t have much money for food. That does happen but it’s a tiny proportion of the population, a small fraction of 1% who are homeless, much less than 1%. It has nothing to do with this gradient that, you know, this gradient is amongst the rich, amongst the middle class and amongst those who are still adequately clothed, fed and housed. So although there are problems of homelessness, it’s not what most of this problem is about. It’s wrong to start thinking from health inequalities to problems of homelessness.

This slide really rubs that point in. The Americans use not a relative poverty definition, they use an absolute poverty definition and according to their poverty line the federal poverty line, nearly 13% of the population are poor, classified in the States as poor. You know America is the most unequal society, it does have real problems of poverty and so on, but if you take the people below the poverty line, 80% of them have some kind of air conditioning, nearly 80% of them have a DVD or VCR, nearly three-quarters own a car or truck, a third of them own more than one car, more than half of them have two or more TVs, over a 30 have a dish washer, over a third have a computer. Certainly some of these people might run out of food at the end of the week, but actually that’s because of the pressures on consumption to maintain standards, participation in society. You know, you might prefer to have a good-looking mobile phone and the right kind of clothes because that’s how you are seen and save on food. So even the absolute disadvantages, things like lack of food, are often driven by relative considerations.

Sorry, Martin can you see there? I will try and stand further to the side.

Our poor are not as – their living standards are not as high as that, but, you know, the American poor if you are going to find problems of violence or teenage births or obesity or poor health, you will find it amongst the American poor more than any of the other developed countries. They are people who have these levels of material wealth.

This is an old classic slide. This is the Whitehall study 17,000 civil servants working in London offices. It is just to emphasise these are the most senior civil servants, these determine status, professional executive level, then clerical level and then the most junior office workers. This is not the poor or unemployed or people on the streets, this is not even mainly manual workers. They are all office workers. Yet there are 4-fold differences in heart disease death rates between them – people working in the same offices together. We have to understand these problems in a way that can explain the differences even amongst the middle class. If you take all causes of death there would be a 3-fold difference and these coloured bars is just what is explained by the most common known risk factors. The blue is unexplained, the larger part.

As Martin said, I’ve rather expanded from health. We have been looking at – for a long time I worked on a tendency for more equal societies to be healthier overall. When I say that, I mean all that I am going to say is about the developed world. So the Nordic countries and Japan are the most equal amongst the rich developed democracies. The USA, Britain, Portugal are amongst the most unequal. So most of the comparisons I am going to show you are those.

So it’s not only that the countries have less good health than the more equal ones. We were very slow to recognise that ill health is just one of many problems which has a social gradient, which is worse in the poor areas of our societies, which are worse in more unequal societies. But actually that pattern seems to apply to almost all the social problems that have social gradients. If a social problem has a social gradient then it is probably worse in more unequal societies.

We collected data on violence, on imprisonment, on bullying, on mental illness, on physical health, obesity, drug abuse, teen births, all the sort of health and social problems we could get internationally comparable data for. They are all problems that are in the news all the time.

I think the next slide is out of place but a little aside. Of course, when we hear or read stuff about more crime or health problems or social problems, the response, the public response, is mainly to think in terms of more services. If people are frightened of crime they want more police; if they think there are health problems, they want more doctors; if there are social problems, more social workers or drug rehabilitation units. Mainly our reaction is to think of these problems just as, you know, society produces them somehow, they are there and we’ve got to try and cure and treat and prevent these – well, not prevent … that’s my point. All these services are very expensive and have surprisingly little impact on the problems they are aimed to deal with.

Anyway, we collected data from all these, as many of these social problems as we could get data for and we combined them into one index. So we gave each country a score depending on its sort of average level of all these things. I am going to block out this bit and take you…so we got data on life expectancy, on the international maths and literacy scores that show how well kids are doing in countries, infant mortality rates, homicide rates, imprisonment rates, what proportion of the population are in prison in each country, teenage birth rates, levels of trust, whether people feel they can trust others or not, obesity, mental illness (including drug and alcohol addiction which is in the World Health Organisation’s definition of mental illness) and social mobility.

We gave each country a sort of average score and there’s a statistical technique for combining things so each one is weighted equally. So where our country in its overall score is influenced as much by homicide rates as social mobility or life expectancy. They are all equally weighted.

Here we have related that index to income inequality. The measure of income inequality we have used is very simple. We chose it because it’s available on these international – well, it’s UN data actually. It’s simply how much richer are the richest 20% in each country than the poorest 20%. So amongst the most unequal countries here, the high inequality countries, they are 8 or 9 times as rich. The top 20% are 8 or 9 times as rich as the bottom 20%. Down here they are three or four times as rich. So that is the sort of scale of inequality we are talking about, you know. Most of them it’s 5, 6, 7 times as rich – something like that. But look what a striking relationship there is. The more inequality, the further out countries are in this line, the higher their health and social problems, the worse their health and social problems are. It’s an extraordinary clear lineup.

If you put that same index, if you relate that same index to GNP per capita there is no statistically significant relationship. You can have a country as – well, Japan and the UK have almost exactly the same level of gross domestic product, Japan has very low levels of health and social problems, as we know because it’s fairly egalitarian compared to us, and we have high levels. Or look at USA and Norway: Both very rich, Norway is fairly equal so has low levels of health and social problems and USA very high.

So average income explains nothing, as I told you right at the beginning. But inequality tells you a great deal.

This is the UNICEF – some people I fear sometimes think I’ve chosen the data to suit the argument, so I also use this Index of Child Well-Being. UNICEF made up an Index of Child Well-Being that contains 40 different components. How well they get on at school, what immunisation rates are like, what about relationships with parents and so on, what about conflicts between kids – lots and lots of things like that that are just taken as measures of indicators of different aspects of children’s well-being.

Here we have it related to a measure of inequality, although people always call it poverty it is relative poverty, it is the proportion of kids below 60% of the average income. So, again, it is showing whether most children are at the bottom end of inequality or not. So measures of child well-being are much lower where more kids are below 60% of the national average income. It does not make any difference to child well-being whether the country is rich or poor. Same measure of child well-being this time simply in relation to average income levels in each country — no relationship.

So we’ve actually done this for the 50 States of the US as well. All that I am going to show you we’ve done, looked at it amongst the rich developed countries and amongst the 50 states of the USA just to make sure that we’re not coming up with fluke associations. We find just the same pattern in the USA. If I had two hours I would show you all the American data as well. The same pattern tendency for more equal states to have fewer health and social problems.

I’m going to show you now some of the individual components of this index of health and social problems just to show you the sort of scale of differences. These are significant relationships in every case, often they are not as tight as the ones I have shown you. This is infant mortality and a tendency again for the more unequal countries to have higher infant mortality rates. Actually Singapore – nobody understands this and I just don’t believe the Singapore data. I remember once having to stop in Singapore on a flight to the States. It’s the only place where walking along the street in the evening I’ve been accosted by male and female prostitutes and drug pushers and so on. I do not believe their infant mortality rates are the lowest in the world. If you look at any of their other social problems they are up here where you might expect them to be.

But anyway, amongst the more equal countries, low levels of infant mortality and around maybe close to twice as high, not quite twice as high, amongst the less equal countries. This is one graph showing the American States and Canadian provinces. This is death rates of men of working age. This is somebody else’s work. It’s a measure of income inequality and in this case the more equal States and Canadian provinces are this end (right-hand side) and the more equal ones are that end (left-hand side) and the death rates are higher there. You see a surprisingly tight fit between the amount of inequality and death rates, rather like – well, it’s closer than the infant mortality stuff.

We have reviewed nearly 200 of those kinds of studies and found that pattern pretty consistent.

It’s not just physical health. It’s hard to compare mental health in different countries but WHO has put together some data on mental health using exactly the same measures of mental illness in different countries and this is that data. Here is the proportion of the population with any mental illness. So down here we’ve got 10% and actually the best countries are below that, a bit below 10%. Up here we’ve got 25 or 30% of the population with mental illness. So maybe around 3 fold differences in the frequency of mental illness amongst the whole population related, highly significantly, to the amount of inequality. The WHO also has not just prevalence of any mental illness but serious mental illness and, again, strong significance with outcome.

Obesity: This is male and female obesity combined in relation – all these graphs are now going to have inequality along the bottom and almost all of them the exact same measure. That question how much richer the top 20% than the bottom 20% and the more unequal ones will be this end. But obesity we have 10% of the population obese there. So below 10% and it goes up to about 30%. So sort of three-fold differences in the proportion of the population obese again is statistically significantly related to inequality and in the US 50 states as well.

These are homicide rates American States with Canadian Provinces there. This is the more unequal end, the more unequal States higher homicide rates. This is homicides per million and down here, I suppose, it’s about 15 homicides per million and there it’s 150. So when that data was put together by Dally and Wilson there was a 10 fold difference in homicide rates that strongly related to inequality. Since then homicide rates have come down a bit and it’s only a six-fold difference now.

If you read the newspapers, listen to the Today Programme, they are endlessly talking about these problems and yet never is inequality mentioned.

This is data on drug abuse internationally. The UN Office on Drugs & Crime has separate figures for what proportions of the population are using opiates, cocaine, cannabis, ecstasy, amphetamines. We put the data together in one index of drug use. So use of cannabis is weighted just the same as the use of cocaine or ecstasy and more unequal countries, more drug problems.

Child conflict: Questionnaires asking kids three different questions: Whether they have been involved in fighting in the last month, been bullied in the last month. I think that’s not only being bullied but bullying and the proportion of kids saying their peers are not kind and helpful. Again, a strong relationship with inequality. The UK – looks as if kids get on with each other worse in Britain than any of the other developed countries. That’s kids of 11, 13 and 15 years old included in these questionnaires.

Levels of trust, the world values survey has questions about trust. Do you feel most people can be trusted? Or the American governance general social survey asks do you think people would take advantage of you if they got the chance? Would people rip you off if they could? In this graph we’ve got the proportion of the population who say they feel they can trust others and it’s very low, below 20% only maybe 10% of the population feeling they can trust others going up to 60 or 70% in the more equal countries.

Similar scale of differences in the American States and it must feel quite different to be on the streets in countries where people feel they can’t trust each other or countries where they feel they can.

Teenage birth rates: Just the same problem. You are getting rather used to and bored with this. You know before I have shown you a graph what it is going to show. I do think it is important to emphasise the scale of differences. These are births per 1,000 women between 15 and 19. So teenage births and it goes from about 5% of women in that age group having babies up to well 40 or USA is over 50%. Again, when I say things are statistically significant, it is simply a statistical measure of how likely a relationship – supposing you just did … what is it? A Jackson Pollock painting – and threw these countries on the board, how likely are they to line up apparently along one line as this? When I say it’s statistically significant, I mean that it could only happen by chance once in a 100 throws or 1,000 throws or whatever. There are different levels of significance.

I am getting near the end of these graphs on just showing you how many problems have this pattern. This is the proportion of the population in prison. It goes from – this is a log scale rather complicated but if it wasn’t on a log scale this would zoom up off the top of the screen. But anyway they go from less than 50 because it’s a log scale only a little bit below would be 40 up to 400 or so. So maybe a 10-fold difference in the proportion of the population locked up.

There’s the same data for the American States. There’s a significant drift this way, higher levels of imprisonment and, again, big differences. There is 200 and there’s states with probably more like 150 prisoners per 100,000 population and it goes up to probably around 700. So I don’t know what’s that? A five-fold difference in rates of imprisonment. We’ve coloured them red, these red circles if they retain the death penalty just a bit of additional information and these dark triangles if they have abolished the death penalty. You see there’s a tendency for the dark triangles to be on the more equal end.

MARTIN: Even though the executed prisoners are not in the prison so it would have a small effect of reducing the number of people.

RICHARD WILKINSON: We, yes, I suppose that’s true. I hadn’t thought of that. I suppose also it means that despite the awfulness of the death penalty, there are not such vast numbers going through this. I don’t know how many have been executed in this particular time period.

Why I show that is because I think people always assume that more people in prison means more crime but actually there are a couple of papers, research papers, looking at the rise in prison populations in the United States and they show that only 20% of the rise is accounted for by more crime. 80% of it is more punitive sentencing and of course this tendency for the more unequal countries in all the ones more unequal than that retain the death penalty and fewer of the ones as equal of that who retain the death penalty.

So it’s more punitive response to crime that’s mainly driving these relationships and of course in Britain, here’s the UK, our prison populations have reached record levels while crime has been going down.

We have looked at basically – I have shown you most of this data now – this is basically all we have looked at internationally, all the stuff we have been able to get internationally comparable data for and American data for the 50 states. So all these things we have looked at in both settings. Internationally we have been able to look at these things as well. We haven’t been able to get American data for social mobility, hours worked, child conflict or that UNICEF index of child well-being. Those were only available internationally.

Amongst the US states, we have looked at the high school drop-out rate related to inequality which again is not available internationally.

The only things we found not related to inequality are smoking and suicides. You might be surprised by suicides because I think what happens is there’s some truth to the sort of cliché that violence either goes in or out. In a very famous paper on death rates in Harlem in New York, a poor black area, at most ages it showed that they were as high as in many parts of Bangladesh, one of the poorest countries in the world but the only cause of death which was not more common in Harlem than the rest of the United States was suicide. I think if you blame other people for your problems, you blame white society or whatever it is, then maybe you are violent but maybe in a more equal cohesive society you commit suicide because you’ve let down your family, you have shamed them or let down … whatever your company. So suicide is not related to inequality.

I want to show you now – sorry, I want to come back briefly to this graph which, remember, is putting all these problems together, all these health and social problems together and showing how strongly related they are to inequality because two points: one is you have probably noticed in all the graphs I have shown you it is always the same countries – do you want to?

NEW SPEAKER: No, it’s okay.

RICHARD WILKINSON: It’s almost always the same countries that do badly. Whether we are talking about health or prisons or obesity or drug problems, it’s always the USA, Portugal, Britain that are doing badly and it’s always the same countries that do well.

We’re talking about a sort of general social dysfunction that is related to inequality, almost everything seems to go wrong in more unequal societies. It’s stupid that we have different people doing research, different policies, different Government departments dealing with each of these problems because they are all rooted in things to do with inequality and relative deprivation and they all have the social gradients, they are all more on unequal societies. You know, we must get at the roots of them and these are the roots that I’m talking about.

The other thing is, you know, we should think whether there are other ways of explaining this. Am I wrong to think that’s all being driven by inequality? Is it some other obvious explanation in this line-up? People used to say it’s the Anglo-Saxon, the English-speaking countries that do badly. But here’s Portugal about as different from the UK and USA as it could be and on many outcomes and a little bit on this actually the line-up between how well or badly even the Anglo-Saxon countries do seems to be related to inequality which explains why America does even worse than us.

At the other end of the thing, here [indicated], although there are obvious similarities between the Nordic countries, cultural similarities as well as their equality, Japan is totally different and actually how Japan – just think of the contrast in the sort of position of women or the family structure in Japan and Sweden. Sweden you have equal political representation and the biggest departure from the traditional family structure. Japan quite the opposite. It’s the country in the OECD that is still most closely based on the nuclear family and where women have the most traditional role. And yet the fact that they are both more equal puts them at the good end of all these outcomes.

How they get their equality is totally different. Sweden has very big income differences, differences in earnings and so on and they redistribute with taxes and benefits and they have a big Welfare State, as you know. Japan has smaller differences in earnings to start with. They do less redistribution and they have a much smaller Welfare State. They spend less – there’s less public social expenditure in Japan than other OECD countries.

So I don’t think it’s easy to explain this away as all due to something else that I’ve sort of left out.

Now I want to take you quickly through data that shows, you know, I think the first response to looking at this stuff people say oh, yes more unequal countries have more poor people and that is why their outcomes are worse. It’s more poor dragging the average down.

It’s not that. Most of this is the result of the vast majority of the population doing less well in more unequal countries.

Let me start off by showing you literacy scores internationally. You know the OECD does surveys of literacy levels so it can compare how well kids are getting on in different countries. Here they are for Sweden, Canada and the US. I say kids’ scores but actually these ones are 16 to 25 so they are young adults. Literacy scores are scored by how many years of education their parents have had. So this end are the kids of well-educated parents, they are kids at the top of society and this end are the kids of badly educated parents nearer the bottom of society.

You see that the differences are much bigger – the scores are much bigger at the bottom end of society than the top, but you see even at the top you do a little bit better if you are in Sweden than in Canada or USA. Of course income inequality is greatest in the USA, then Canada and then Sweden of these three in this graph.

Think of, you know, ourselves in the middle somewhere. It clearly makes a difference to our kids which country we are in. So even a given position in society or as I shall show you later given income level, given a social class, people in the middle do better in more equal societies.

This one remember the fanning out downwards because the bottom end those are the poor scores with the next one. These are death rates and of course high scores are bad, high death rates are bad.

These two lines show the relationship between – it’s American data. It shows the relationship between average income in each county in the USA (there are something like 3,100 counties in the USA) and we simply looking at the relationship between the average death rate in the county and the average income in the county – actually the median. We have compared them in the 25 more unequal American States and the 25 more equal states. So about half the county is under that line in the more unequal states and the other the rest of the counties are in the more equal states. You see the relationship between county median income and the average death rate in each county is different. If you are on whatever level of income it is, your death rates are lower – sorry, I should have changed these labels they are rather technical labels but basically think of this axis as simply as death rates. Death rates are lower at all levels of income in the more equal states. US again, the differences fan out, they are bigger at the poorer end amongst the poorer counties but they are still visible amongst the richer counties. Incidentally although these may look like rather low levels of income, about 95% of the American population comes within this graph. So the vast majority of Americans do better if they are more equal states.

This is a comparison between infant mortality in Sweden which are these dark red columns and England and Wales. The infant mortality rates are arranged by social class. So here’s class 1 the professional, directors of larger companies and so on going down through to class 3 non-manual, that is the junior office workers, the skilled manual workers, the semi-skilled workers, the unskilled manual and then they are classified by the occupation of the father in the official statistics that we use – sorry, this is not a paper of ours this is other people – single parents and the unclassified.

But you see again there’s a suggestion of the fanning out. There’s our steep social gradient and if there is a social gradient in Sweden it is much flatter. But the differences are biggest at the bottom of society once more. So we had two lines it would show that same fanning out that you saw in the other graphs.

This does not show it so well but the authors of a paper going with this slide make the same point. This is a much more recent comparison between death rates of Americans and death rates in England. To avoid people saying that it was differences were due to race or anything like that, they just took the white population. They had lots of measures of health. They had cholesterol levels and blood pressure and a number of other biological measures. They had death rates, they had diabetes, hypertension, cancer, lung disease, heart disease and this looks rather complicated but let me just take – let’s take these big blocks here. This is hypertension. The ones in front are the death rates for England and the ones behind are the death rates for the USA. So the first thing is to see that in the USA – the taller blocks – higher death rates.

Now, they are arranged by thirds of education. So these blocks on the right are the highest educated third of the population in England and the US and this is the lowest educated third of the population in England and the US. So again if the reason why death rates are lower in England than the USA is because of the fact that we are not quite as unequal as the Americans, then it looks as if it’s the whole of the population that benefits.

You can’t see so clearly the pattern that I was describing of fanning out of the differences, but the authors looking across all their measures say in a later paper that there’s a strong tendency for the US to have steeper social gradients in their health outcomes – even steeper than we do.

I am now going to try and answer this question why we are so sensitive to inequality. You might be rather surprised what I have shown you because maybe most of us can’t tell whether we are more or less equal than any of the countries where you might go for holidays.

I come at all this through health and the big change in our understanding of the determinants of public health over the last 15/20 years has been a recognition of the importance of psychosocial issues. So it’s not just what poor material conditions do to you directly – whether damp housing maybe there are mould spores in the air they affect your respiratory function or air pollution outside that has affects on you whether you know it or not. That’s a little bit of the picture.

But much more the picture is what you feel about your circumstances and the most important psychosocial risk factors that could be grouped under social status. Things by which, as I say, I don’t mean simple material standards, I mean something like superiority and inferiority. Where you are in relation to others.

Any measure of friendship, social involvement, close confiding relationships, they all seem to be highly protective of health. That is the second category. So good social contact is important to our health.

Lastly stress in early life, maternal stress in pregnancy or in early childhood, poor attachment, domestic conflict, all those things seem to cast a long shadow forwards over health in later life. Indeed, you know, levels of a stress hormone like cortisol in men my age are related to our birth weight which is affected by maternal stress and so on in pregnancy.

I had looked at these groups of risk factors and thought maybe they are telling us about some underlying risk factor or psychosocial problem. I think it is interesting when we talk about stress in modern societies, we know that the psychosocial risk factors work through chronic stress. That is how they have their biological affect on us. Stress changes physiological responses. While you’ve got to deal with an emergency all sorts of functions that are not essential in a brief emergency are down regulated. So if the stress goes on for more than about an hour your immune system is down regulated. Even if the stress is very short, tissue maintenance and repair, wound healing, reproductive functions, growth, processes like that are put on hold. Don’t give them resources when you have got to save your skin. You need everything you’ve got to deal with the emergency so in the flight or fight response you become very alert, you are less aware of pain, your reaction times are speeded up and you mobilise energy resources from fat supplies and so on all tipped into the bloodstream.

So because we know that these are the main triggers to psychosocial triggers to poor health we also know that they are the most important causes of stress in the population as a whole. So when you think of stress think of things to do with low social status, lack of social affiliations and early childhood.

But it seemed to me they were indicators of an underlying source of stress. I suggest it like this: that the issues to do with early childhood you might have had a very difficult early childhood and that makes you more vulnerable to the stresses of low social status. They are rather like each other. We talk about insecurity in both contexts and we know in animals that – when I say animals, I mean other primates, non-human primates – they are both related to higher cortisol levels, the central stress hormone.

Friendship comes into that picture because of course if you have friends you feel appreciated and valued. If you are not included in things, people seem to leave you out, you start worrying whether you are unattractive, you’re boring, socially gauche, whatever it is. We all have those sorts of fears about how others see us. I suspect that is actually what these risk factors are about, that they are to do with the sort of stress of negotiating new relationships all the time, our worries about how we are seen, about whether we are valued and you can see how social status fits into that. You know, are you admired and respected or are you looked down on by everyone and you know people treat you as if you were dirt or you feel they do. Of course you can see how friendship fits into that and you can see how early childhood fits into that.

We will stop very soon.

I developed that kind of explanation and then I saw a study which went through – it was looking at hundreds of different studies sort of trying to glean the general patterns. It looked at studies in which people had been invited, volunteers invited to come, into a psychological laboratory and subjected to experimental stresses. So I say, “I want you all to do these arithmetic problems, write down the answers” and in some experiments you might have to read out your marks afterwards and I say actually you people did pretty pathetically and these are the smart ones, you’re a stupid lot. You know, stresses maybe like that or maybe writing about an unpleasant experience you have had to try and get your mind into a more stressed state or being videoed while you are doing things or a bit of public speaking or something like that.

What they were doing was trying to see what the cortisol response was, if you subject people to these stresses, what pushes people’s cortisol levels up?

Anyway, a study found 208 of those studies where cortisol levels had been measured after subjecting people to an experimental stressor and when they looked to see what kinds of stressor most reliably pushed up cortisol levels, they found out that included social evaluative threat such as threats to self-esteem or social status in which others could negatively judge performance, particularly when the performance was uncontrollable, the outcome was uncontrollable.

So that fits it almost exactly, I feel, into how I would explain those risk factors coming out of the social entomology. Social evaluative threat. Just one more point to sort of rub that home: there are now a lot of experiments in which you look and see how people’s performance on different tasks is affected by sort of social stereotypes. So, for instance, you might think that women are not good at spatial things or maths or something like that and you’d prime that in an experiment so that women were sort of thinking about whether they are less good at that kind of thing and you find actually that that performance is very much affected by whether or not that idea is primed beforehand.

Similarly with racial stereotyping. If blacks in America are made to think a test they are given is a test of ability, a test of intelligence, they do less well at it than if they are told it is not a test of ability, it’s an interesting problem.

You know the experiments with schoolchildren, one called blue eyes in the States, where one group of kids were told that the ones with blue eyes are brighter than the others and sure enough it very soon shows its affects on ability. A teacher a week later says, “I actually got it wrong. It’s the brown-eyed ones who are better” and suddenly the performance switch round. We are very sensitive to this social stereotyping.

One example is this one of Indian children in different castes. They were asked to do mazes. So they are given a bit of paper and they have to find their way into the middle of the maze. There are high and low caste kids and at first they do it without knowing each others castes. The results are equally good. These are high caste, these are low caste. Then they do them under a second condition where they have had to say which village they come from, what their father’s occupation is and so everyone knows what caste they are and suddenly the lower caste kids’ ability falls off.

This kind of way of understanding the effects of inequality fits very well also with the evidence on violence, that violence is triggered by people feeling looked down on, humiliation, loss of face. I must not take up more time but Andrew Gilligan, a prison psychologist said, “I have yet to see a serious act of violence not provoked by the experience of feeling shamed and humiliated, disrespected and ridiculed.” People in more unequal societies, because they are deprived of access to the markers of status in terms of jobs and housing and good incomes, and also because status is even more important in those societies, are more sensitive to feeling looked down on, humiliated and so these are the triggers for violence.

It’s the fact that having second class goods seems to show you are a second class person that matters. There are two responses in a way to being put down: one is to try and make people respect you by thumping them; the other is to accept your inferiority, and I use a quote of Alan Bennett talking about his parents to show that response. It’s worth reading. The son makes it difficult but his parents put down most of their imagined shortcoming to not having been educated. Remember they worked in a butcher’s shop in Barnsley, was it?


Martin: Even worse!

NEW SPEAKER: Now then!!

RICHARD WILKINSON: “They put down their shortcomings to not having been educated. Education to them was a passport to everything they lacked, self-confidence, social ease and above all the ability to be like other people. Put simply, and as they themselves would have put it, both my parents were shy, a shortcoming they thought of as an affliction at the same time enshrining it as a virtue. I assured them that everybody felt much as they did, but that social ease was something that could and should be faked. ‘Well, you can do that’ Dad would say, ‘you’ve been educated'” – remember Alan Bennett went to Cambridge – “adding how often he felt he had nothing to contribute. ‘I’m boring, I can’t understand why anybody likes us. I wonder sometimes whether they do, really.”

So a very different response from the violence one.

Can I just show two more?

MARTIN: Is anybody bored?


Again and again in this stuff friendship and social status come together but in opposite ways. If we look at levels of trust or violence or involvement in community life, it all gets worse in more unequal societies, status differences reduce that sort of the involvement in community life. It’s socially divisive and of course we all know that, you know, we socialise with our equals. You don’t have very much richer people to dinner and you don’t have very much poorer people. You know, this is entirely intuitive. But not only do social status and friendship interact in those two ways, they also interact as risk factors for your individual health, as I have been saying. Friendship is good for my health. Social status issues start wearing me down.

Why those two come together is because they are two sides of the same coin. In any species, there is always a serious threat or potential for conflict with other members of the same species. Because you all have the same needs and you compete not only for food and shelter, sexual partners, but everything that you need and there are two ways of dealing with that: either we sort it out on the basis that you are stronger than me and so I just have to wait and hope there’s some food left when you have eaten, sort it out on the basis of power or in animal studies obviously simple physical strength is usually the most important thing and you get to know which animals are higher than you so it avoids conflict because you just recognise position in the dominance hierarchy.

The other way though is sharing and co-operation and equality. So friendship and social status interact because they are the two sides of the same coin. Social status or dominance hierarchies, pecking orders, they are orderings based on power conversion, privileged access to resources regardless of other people’s needs. I have it because I can and you have to look after yourselves afterwards. Or friendship which in contrast is based on reciprocity, mutuality, social obligations, sharing and a recognition of each other’s needs. The gift is the symbol of friendship because it says you and I don’t compete for access to necessities and your feeling in indebtedness that makes you reciprocate the gift some anthropologists have suggested that is basic social compact. It gets over this problem of conflict for access to resources which is the basis of Hobbes politics saying life is nasty, brutish and short, the competition against all gets over that by sharing and equality.

Of course for 90% of human existence in hunting and gathering societies we lived almost always in highly egalitarian hunting and gathering bands. There are no cases like animal studies where the dominant males monopolise access to the females and others starve when there’s not enough food. Food sharing seems absolutely standard in these societies.

Human beings have lived in every kind of society and the kind of space we find ourselves living in has dramatic effects on our behaviour to each other.

This is the very last slide. Simply what has been happening to inequality. This is from 1979 Thatcher, Major, Blair the human raise in inequality under Thatcher and then little ups and downs. You may have heard recently been the biggest decline of inequality in Britain. That was just this decline [indicated]. I think that this figure has been revised so that it’s only the most recent one that is up and as Robert Chotts of the Institute for Fiscal Studies said Britain was rather lucky in the years over which OECD compared countries and said, you know, we had the biggest decline because they just happened to take that period where there was a decline.

But in the scale of these things, you know, none of this huge rise in equality has been reversed and what we are seeing in our societies is the long-term results of that.

Thank you. I am sorry to have gone on too long.


So it is political will that we need.

MARTIN: Questions?

Nick Gradwell: Professor, my name is Nick Gradwell, let’s get the equalities term. I work in the Equality & Human Rights Commission and we’ve very interested in health inequality. I’m sort of probably your biggest fan and biggest critic. I think at the macro level obviously what you demonstrate to us is that the cohesion of societies is fundamental to health as everything else. At the time where we’ve just had a credit crunch and, depending on the results of the election next Tuesday, your work might be quite seminal in that regard. It would be nice to get away from this ridiculous rat race of a world order that we’ve come into and the damage it’s causing. I also think that at a national level it gives hope we might actually one day address things like long term redistribution and fiscal policies that have the guts to do that.

However, I do find – and this is somebody working on equality like yourself – that where the historic association between equality and inequality is so predominant around money, around the socio-economic model which is very important and can be both an input to equality and an outcome of it.

I find whether it’s world stat figures or in terms of how ideas like yourselves, and I think they translate into influence on national policies, the process then can get quite muddled by the time it’s down within some of the spear head programmes that the Government try to do, to address actually health inequalities because it very, very much takes only sparse note of the impact of gender, of race, sexuality, disability, age, belief – six factors alone which any one of us in this room has and which we all have a differential hierarchical status upon.

I find that where you have opened up an issue, I hope others and yourselves other researchers will take forward kind of borrowing down far more from just what is socio-economic inequalities to how that has an extra impact when one is a man or a woman in society, as I say the effect of race and indeed of disability. Because in health inequalities one problem in this spearhead is the only things that work well are where there is specificity; in other words, an additional health risk such as early onset undiagnosed cancer by working age men, particularly in the BME communities, and also risk of cardiovascular disease. When there is that specificity and it is an additional health risk where issues like gender and age are taken into account, then the success of tackling health inequalities is very great. Where it is only that certain communities have been identified as poor, to be honest policy sometimes even make it worse by starting to see their behaviours as feckless. That is not so much a question as my comment to your work and obviously I am sure you can comment likewise.

RICHARD WILKINSON: I think the other dimensions gender and race and so on, I think it is really important to see that those dimensions cause problems insofar as these different groups are seen as inferior and, you know, whether —

Nick Gradwell: And experienced psychosocial inferiority of course. I take that point on board.

RICHARD WILKINSON: I do think that the monetary issues, the financial things and the power which goes with that, are key in an evolutionary sense, that dominance is about access to scarce resources and so it is the underpinnings through which we then start worrying about whether one group of the population, whether women or blacks or gays or whatever, get a worse lot.

But actually I think that those equal opportunities issues – I stick my neck out in saying it – but they are a side issue. They are arguments about who suffers a given level of inequality and is it much better if it is one shape or form of a person than another? What we’ve got to do is to get rid of the problem of low social status whoever is suffering, not just make sure that equal numbers of blacks and whites and gays and women and so on suffer. We are dealing with outcomes, equality of outcome not equality of opportunity. However, I could show a graph which shows that social mobility is lower in more unequal societies simply because I think the material differences are the back bone on which all the cultural forms of expression of social status hinge. That people, for instance, use their money to express social status, social differentiation and you can’t maintain it for very long without the money. That’s why it’s important.

DAVID BLUNKETT: Sorry, Richard. You will have to shout when I’m doing my bit.

RICHARD WILKINSON: So I think that it is about the economic things because it’s they that exacerbate the other dimensions that you are right to mention, but also this is not primarily an argument about who suffers inequality, it’s about how much inequality.

MARTIN: Shall we have a couple more questions and then ask David to speak?

NEW SPEAKER: On that chart there, if you charted your index of health and social problems would it follow pretty much or would there be a time-lag?

RICHARD WILKINSON: That’s an important question, I think. People have looked at time series relationships between changes in inequality and changes in outcomes. It’s difficult because we haven’t long had comparable data either on income inequality so you can really compare over long periods of time different countries and we have only just started looking at all these other outcomes, you know. Research on inequality in health which health inequalities took off about the Black Report but looking at the tendency of more equal societies to be healthier, that really took off from a paper of mine in 1992.

But even so, people have looked a little bit at time trends to see whether health inequalities and overall standards of health are affected by inequality. The results are fairly mixed. There are two or three papers that show quite clear relationships. There are one or two which don’t show such clear relationships but I think the issue is time lags. What one should do actually is not look at overall causes of death but you should look at different causes separately so that one time-lag for homicide, another time-lag for heart disease and when you get into all the social problems, a time-lag for I don’t know teenage pregnancies and a time-lag for drug abuse. They will all be different and that work just hasn’t been done yet.


Catherine Gleeson: My name is Catherine, I’m a practice nurse. I just wanted to ask – you mentioned obesity as one of the markers of inequality.


Catherine Gleeson: I wonder if you could say anything about the current targets for measuring and monitoring childhood obesity which is so time consuming that it’s not leaving school nurses, for instance, any time to do other work. So I’m asking really about targets affecting clinical improvement, as it were.

RICHARD WILKINSON: I don’t know anything really about Government targets in obesity except to assume they are downward and in most cases targets seem to me just projections of current trends, except in cases like obesity where they are upwards and you’ve got to hold out the possibility of reversing the trends.

But obesity is part of the picture. I showed you the international relationship with obesity showing enormously high rates in US and Britain and low rates in the more equal countries. We looked at the figures in the States as well and we find – well, at first we didn’t find a relationship and we were rather perplexed and then somebody pointed out that the figures we were using were self-reported obesity from telephone surveys in the US and someone showed us where we could get figures on obesity which were people were weighed and measured and of course those are related to inequality.

It is also childhood obesity that seems to be worse in more unequal societies.

MARTIN: This is an issue I think we come back to because one of the problems that Richard’s work seems to generate, if inequality is the fundamental driver of so many of these problems, is there much you can do by tinkering about? As a school nurse you are in no position to alter the inequality, if there is, in the schools where you work. All you can do is sort of …

RICHARD WILKINSON: I think one of the problems of talking about this inequality of the thing is people feel there’s nothing they can do about it, it’s for governments but actually at the most basic level it’s about people feeling valued and you can do something about that in every institution, the social environment in schools whether children feel valued and so on or whether there are pecking orders and bullying and so on as some of the data I have shown you suggests there is. And at work whether people feel ignored and taken advantage of or whether they feel valued.

You have a point to add?

Catherine Gleeson: I would agree entirely with what you are saying but my point was because of Government targets the primary care trusts are obliged really to do measurements for childhood obesity at 4 and 11 and because that’s such a time-consuming process instead of school nurse being able to do all the other interventions that could help improve school atmosphere and that kind of thing, they are having to spend the time measuring — measuring per se appears to be a waste of time but because it’s a Government target it has to be done. Do you see what I mean? It’s like a political thing to do whether it’s any use or not. It leaves no time for anything else.

RICHARD WILKINSON: Yes, I understand the problems.

Amanda Normand: Amanda. I’m Amanda Normand from North Tyneside council. I’m a councillor. I don’t know if you have seen the John Prescott TV programme last week.


Amanda Normand: It was absolutely brilliant.

RICHARD WILKINSON: My son sent me a text message to say I should watch it.

Amanda Normand: It was really good talking about how he didn’t feel as if he fitted in Government because of his class and I really related to him because I come from a really socially deprived estate in North Tyneside and I was the only one in four generations to go on to university from that estate. As a result of that, I lost my friends and my family and nobody speaks to us any more. It’s sad …

But as you are saying on there, you know, about friendships and stuff and how important it is, I think we should go into that not fitting in and what happens to you when you do move out of the areas and try to move on with your life and you still get discriminated against because I don’t fit in either camp now. I don’t fit in the middle class but I also can’t go back to my own family.

RICHARD WILKINSON: People have always known that inequality is divisive and all this is show that even between different American States make really large differences in outcomes.

It’s not just – many socialists have thought that equality is important if we could get to perfect equality but we all have to live in the real world with huge inequalities. So those sort of ideas or notions don’t mean very much. But the data shows that the differences in inequality between different developing countries make a huge differences to all these different health and social problems. Much, much more problems whenever there is more inequality.


Sam Pryke: I was going to make an observation in relation to the point that you were making about measurement.

I work in an equalities environment and whilst I am sympathetic of the different pressures that people in the health service are under, we very much come from the perspective that without having some form of statistics and qualitative evidence you will never be able to meet the needs of doubly disadvantaged groups of the population.

This also relates to a point you made earlier in response to my colleague Nick’s argument, that I personally believe that as long as a generalised approach exists for health inequalities focusing on one category of social status, for example, looking at social class we will never delve deeper into other ways of social stratification that exists and perhaps increasingly so with the demographics of our society, then we’re never going to be able to adequately treat those people who are in greatest need.

The Disability Rights Commission, we did an investigation into health inequalities which was a couple of years back and some of the findings were so stark. One of them was people with learning disabilities are 58% more likely to die before the age of 50. That really is sort of like a life sentence, being born with a learning disability.

But that’s only one of a very small raft of inequalities. Nick could quote far more. But I’m just making the argument that you need better measurement in the health sector in relation to our population. We’re not monolithic and we need to reflect that otherwise need will never be addressed.

MARTIN: I think we will come back to this because I think this is an interesting area.