UNCORRECTED TRANSCRIPT OF ORAL EVIDENCE To be published as HC 422-ii

House of COMMONS

MINUTES OF EVIDENCE

TAKEN BEFORE

HEALTH COMMITTEE

HEALTH INEQUALITIES

Thursday 27 March 2008

MS ANNA DIXON, PROFESSOR HILARY GRAHAM and PROFESSOR MARGARET WHITEHEAD

PROFESSOR KAY-TEE KHAW, PROFESSOR SIR MICHAEL MARMOT
and PROFESSOR RICHARD WILKINSON

Evidence heard in Public Questions 87 - 196

USE OF THE TRANSCRIPT

Oral Evidence

Taken before the Health Committee

on Thursday 27 March 2008

Members present

Mr Kevin Barron, in the Chair

Charlotte Atkins

Sandra Gidley

Stephen Hesford

Dr Doug Naysmith

Mr Lee Scott

Dr Howard Stoate

Mr Robert Syms

Dr Richard Taylor

________________

Examination of Witnesses

Witnesses: Ms Anna Dixon, Acting Director of Policy, The King's Fund, Professor Hilary Graham, Professor of Health Sciences, University of York, and Professor Margaret Whitehead, Professor of Public Health, University of Liverpool, gave evidence.

Q87 Chairman: Good morning. Welcome to the second day of our inquiry into health inequalities. Perhaps I could ask you to give your name and the position you hold, for the sake of the record, please.

Ms Dixon: Anna Dixon, Acting Director of Policy at the King's Fund.

Professor Graham: Hilary Graham. I am Professor of Health Sciences at the University of York.

Professor Whitehead: I am Margaret Whitehead. I am the Professor of Public Health at the University of Liverpool.

Q88 Chairman: Thank you. Welcome. I have a question for all of you really: health inequalities have widened since the 1950s and the widening seems to have been particularly marked since the 1970s. We have a figure in front of us that I think everybody has seen. What, in your view, are the main reasons underlying the trends in health inequalities over the past 50 years? Are there any particular policies or trends in health behaviour to which the trends in health inequalities can be linked?

Professor Graham: As you probably know, over the last 50 years, for the major causes of death - what we call the big killers: stroke, coronary heart disease, cancer, particularly lung cancer - the gradients have widened. In the 1950s they were not diseases that were particularly linked to poor circumstances - there was not much of what we would call a gradient in those conditions - but over the last 50 years those gradients have emerged and become stronger. Clearly, if we have greater inequalities in the underlying major causes of death then we are going to have widening health inequalities, so the question then comes: Why have inequalities in the major causes of death widened? That is the point at which public health researchers start to look at what has been happening to the social factors that determine those causes of death. As two examples, one could take health behaviours like smoking and a wider determinant like living standards. If we look at smoking, in the 1950s about 60% of men across the social classes smoked cigarettes and about 40% of women across the social classes smoked cigarettes and it was not a habit that differentiated between rich and poor, but over the last 50 years it has increasingly become one that has. As prevalence has fallen, the rates have fallen much more rapidly among better off groups than among poorer groups, so the inequalities in this major cause of illness and this major cause of premature mortality have widened. If we look at living standards, over the 1940, 1950s and 1960s living standards remained roughly equitable. There was a gap, but the gap did not widen between poorer groups and richer groups, and, if anything, the proceeds of economic growth were used to the benefit of the poor: their living standards were raised relative to better off groups. Since the late 1970s and particularly through the 1980s and 1990s, inequalities in living standards have widened appreciably. For the period from 1970 onwards, if you like, we have a double-whammy: widening gradients in a major risk factor like smoking and widening gradients in an underlying determinant like living standards. If we are looking for why inequalities have widened, we need to ask the question: What has happened to the determinants of the major causes of death? And they have widened. Then, to comment briefly on the question about policies, in about 1950 we began to get evidence that smoking was a major cause of lung cancer from major British studies but it took nearly two decades before we had the beginnings of any health promotion interventions in the 1970s. It has actually taken us 50 years to get to smoke-free legislation. I think we can raise questions as to whether a more effective series of tobacco control policies from day one, say from the 1950s, would have had an impact on reducing the extent to which inequalities in smoking have widened. There is some evidence to suggest that poorer groups are particularly sensitive to comprehensive policies around tobacco; namely, they have a bigger effect on them. In relation to living standards, we know there was a policy shift in the late 1970s and early 1980s which moved the tax and benefit system from a more rigid redistributive model to one that allowed inequalities in income to widen. I think there are policies that we can identify which may be contributing to the widening in determinants of health across the last few decades.

Q89 Chairman: Would it be unfair to say on that basis that the major determinant of living standards is income?

Professor Graham: The major index of living standards is real incomes for households. There are other obvious contributors that come from welfare services but, when we are looking at living standards, researchers take household income as their major measure.

Q90 Chairman: Do you have anything to add?

Ms Dixon: I do not think on that particular issue.

Professor Whitehead: No. I agree.

Chairman: Let us move on then.

Q91 Dr Stoate: Thank you, Chairman. That was a very interesting opening statement. It certainly set the scene very nicely. I would like to look at evaluating action to tackle this obvious difficulty. We have all seen the 2007 Status Report which was published in March which indicates that 75 out of 82 government commitments in the Programme for Action have now been implemented or substantially met. There is also evidence that QuOf payments have pretty much been across the board. It is not that GPs are very wide in their spread of QuOf payments - most GPs have done fairly well in QuOf. Why is it, given the fact that the Status Report shows that many targets have been met and given the fact that GPs are, broadly speaking, delivering on QuOf, why is it that the health inequalities have continued to increase?

Professor Whitehead: As far as the targets are concerned they are quite short-term: 2010 for the health inequalities targets. I think that to have an impact on a very deep-seated problem like inequalities in health, it is too short a time to expect a major impact on health itself. There is a lag period between changing the determinants of health and changing health. I think that some of those targets are short and there is also the issue of the way that the targets are set, comparing the most disadvantaged to the most affluent, and there may be changes happening in different social groups between the worst off and the best off that we are not seeing because we are not measuring those in the targets. There may be a different approach to take.

Q92 Dr Stoate: Is it possible that we are measuring the wrong things or is it simply too early to tell?

Professor Whitehead: I think it is too early to tell. The sorts of changes that are expected and needed to be taken are major changes and we really cannot expect in the short term and without a great effort to make changes.

Q93 Dr Stoate: Is there anything that the Department of Health and NHS have managed to achieve to bear down on inequalities? Or, again, is that difficult to answer?

Ms Dixon: I would like to come back on the point about QuOf, if I may. Obviously there have been high levels of performance and at least the analysis that has been done is showing that there are not major differences in performance between those practices serving more deprived populations. However, given that there is just so little variation, it does suggest that perhaps if the standards were increased - which I believe is the intention, now we have GPs meeting this first set of standards, to increase the level of performance that is required - we may start to see more variation appearing in meeting higher standards. On some particular indicators there are areas where some practices are underperforming. There have been some spearhead PCTs which have made big inroads in terms of preventable mortality through very aggressive statins and hypertensive prescribing and so on, so I think there is potentially more that could be done around the contribution of primary care, both for things that are within QuOf and some of the things that are outside of QuOf, in terms of ensuring that their contribution to reducing health inequalities is maximised. But I think we do not know enough yet about what GPs need to be doing on a routine basis outside of what is measured annually in the QuOf. On your last question, as to where there have been achievements, I would say that there have been some achievements in that area. I would be less able to comment on achievements on the wider determinants.

Q94 Dr Stoate: Is there anything else the others would like to add?

Professor Graham: Could I come back on what you were saying initially about the Programme for Action and this kind of anomaly that there is progress on a number of the indicators - and I think you mentioned 75 ----

Q95 Dr Stoate: Yes, 75 out of 82 was our information.

Professor Graham: And yet the targets do not appear to be met. I think the point that Margaret was making is really important, about time lags. Infant mortality and life expectancy are outcomes that are influenced by conditions - over one's life, obviously, with life expectancy, and, with infant morality, very much by the health of the mother, which is in turn influenced by the social conditions in her childhood. We would expect long time lags between interventions and effects but also the indicators, the 75 of the 82 successful indicators, the basket of indicators, came from the Treasury-led review of strategies, the Programme for Action strategy to tackle inequalities, and by and large it was an attempt to map existing initiatives against the targets; namely, given that this is what we are doing, how might they contribute to reducing inequalities? I think there is an opportunity to turn the question around the other way and say, "If we want to reduce inequalities in health, what should we be doing in policy terms?" I think if you did that, then the interventions may be covering some of the bases but not all of the bases, and the problem with not covering all the bases is that some of the bases you should be covering may be moving in the other direction. Just to give you an example, if you take the child poverty target, which is a very important platform for the Programme for Action, we know from evaluations of the moves towards that target that the actions the Government have taken have been very successful in moving families who are just below the poverty line across the poverty threshold into better circumstances, but other policies have continued to lift incomes at the top end of the income distribution so that the overall target has not been met - not because the policy itself is unsuccessful but because there are other countervailing forces at work. When we are looking at the Programme of Action actions, we need to be asking ourselves: Would we expect them to be doing something about inequalities in health if other policies are moving in the wrong direction?

Q96 Dr Stoate: So it is fair to say that the gaol posts have shifted rather than things getting worse.

Professor Graham: It could well be that the interventions themselves would be more effective if they had a groundswell of other policies behind them. The interventions, by and large, are additional initiatives to mainstream policies rather than changes in mainstream policy.

Dr Stoate: That is helpful. Thank you very much.

Q97 Charlotte Atkins: Do you think that the progress in the spearhead authorities has been disappointing or do you think they have made progress?

Ms Dixon: From some analysis that we have done, the identification of the spearhead local authorities has not meant a great deal on the ground. Obviously it was an important way of measuring against the target in terms of life expectancy but, particularly in those local authorities that have been designated spearheads, there has not really been a sense - and we did some interviews with directors of public health and others working in local government - that spearhead status had really made much difference in their focus on health and health inequalities, which is clearly quite disappointing. The other disappointment was that there were additional monies supposed to be allocated as part of the Choosing Health White Paper but these were not earmarked funds and there is clear evidence that quite a lot of that money was spent on other priorities and was not spent on public health. Both in terms of the resources that were allocated to spearhead PCTs and also the status of spearhead, I do not think it has made a great deal of difference. In terms of their performance, it does seem amongst the spearheads - and there were a number of criteria which meant that spearheads were given that status, to do with their levels of deprivation and other indicators - that those who have performed best were already the ones that were least deprived. It does seem that those spearheads which met all five of the criteria to get that status have been the really toughest ones in term of their improvement. I think there still begs the question for those who are dealing with very, very difficult and deprived and socially excluded communities as to whether there has been sufficient focus and resource given to public health.

Q98 Charlotte Atkins: So the spearheads have had extra money but you seem to be saying that there has been lack of leadership within those spearhead authorities to focus on things like public health and that is largely because the funds were not targeted or ring-fenced for that particular purpose.

Ms Dixon: I am just saying that in relation to the designation "spearhead", particularly within the local authorities - and we did a small sample of interviews, so I cannot say in terms of the national picture - they felt that that particular label did not necessarily ... That is not to say that they are not giving focus through other partnership arrangements to health - which I am sure many of them are.

Professor Whitehead: I think it varies greatly around the country. In the North West, for example, the spearhead designation really has brought the public health and the local authority people together to work effectively on this. I see that they are working in an integrated way, so I would hold that out as a positive example.

Q99 Charlotte Atkins: When you say the North West, which ones are you speaking about?

Professor Whitehead: Merseyside and Greater Manchester, for example. There is a high concentration of spearhead local authorities in those two areas and they are working well together, but I would emphasise that they are working in the most difficult conditions. These area-based initiatives are really struggling against deteriorating conditions, so it may be that the best it can do is to stop things getting worse rather than to try to make things better. If you are trying to measure what they are doing, their ability to stop things getting worse may be a great achievement.

Q100 Charlotte Atkins: Where there is partnership between the PCT and the local authority, is that crucial to success? Are you indicating that where that partnership does not exist properly, progress is not made to such an extent?

Professor Whitehead: I am saying that when it works well - for example, in Merseyside I have seen it working very well - then more can be done, more multi-sectoral work can be done to try to tackle this, but it is a long-term issue. There are many area-based initiatives that are introduced one after the other. They are only given a few years to prove themselves and then, when they cannot prove themselves within that short period, they are stopped and something new comes along. There is a continual procession of area-based initiatives and that in itself is quite disruptive. Nothing is given time to really bed in and function.

Q101 Charlotte Atkins: We are going back to the issue you raised earlier about the lag in terms of the results coming through.

Professor Whitehead: Yes.

Q102 Charlotte Atkins: Clearly we have seen with things like Sure Start, for instance, which obviously has had the biggest response in terms of research studies, that, yet again, the issue is that it will take time for it to come through.

Professor Whitehead: Yes. That is a good example of something that I think has great potential and the evidence base is very strong for it, but it needs that time and space to be able to make a difference.

Q103 Charlotte Atkins: What advice would you be giving to the Department of Health in terms of making sure that these spearhead areas are performing to the best possible level they can?

Ms Dixon: One issue is to do with areas where there has been PCT reorganisation. The reorganisation has brought more PCTs in to be coterminous with their local authorities - which one would hope would be beneficial in terms of supporting some of this partnership working - so I guess one thing would be not to reorganise the primary care trusts again, to allow some of those relationships to develop. I suppose the other issue around the partnerships is to do with the local area agreements and the new comprehensive area assessment that is being introduced as part of the Local Government Act: to give them time, in terms of allowing those partnerships to establish and to identify what their priorities are within the local area agreements and to build, I hope, on previous partnerships that have existed in those areas.

Q104 Charlotte Atkins: What impact do you think that population mobility has on the problems faced by spearhead authorities?

Professor Graham: That was something that was raised in the Status Report, the question that we are not looking at stable populations when we are measuring things like life expectancy over time. The lives that we are saving may be those of people who no longer live in the areas where the intervention was targeted. I think it is a big issue, particularly for poorer areas. The evidence we have suggests that mobility, moving in and out of areas, is related both to your social circumstances and to your health; namely, if your circumstances improve and you are in better health you are more likely to leave a poor area and move into a more affluent area. Conversely, if you live in an affluent area and your conditions deteriorate and your health deteriorates you are more likely to leave the affluent area and move into a poor area. If you think of an intervention that has been introduced into a spearhead community, and let us say it has bedded in well and become effective, then one could envisage a situation where people whose social circumstances have benefited, say, from a New Deal for Communities initiative and whose health is improved through more intensive primary care, those very individuals would have an increased propensity to leave the area.

Q105 Charlotte Atkins: Especially if you have, as you do in Stoke on Trent, a huge programme of housing regeneration. Some, of course, will be resettled into the area; others will be coming into my constituency and the adjoining Staffordshire Moorlands. How much do the statistics really mean in terms of the progress we are making in spearhead authorities? Is the analysis robust enough to demonstrate whether spearhead authorities are really making progress or not?

Professor Graham: I think the evidence has to be based at the area level because that is the way in which the spearheads are defined. All the research community can say is to caution against simple assumptions that if there is not progress it means that the interventions that have been put in place are not effective, partly for the population mobility problems that you have raised, which I think are important, but also for the issues that Margaret was discussing; namely that we would not expect interventions which really are trying to improve people's trajectories over time to be manifest in large improvements in their health in the short term, particularly when, as Anna has mentioned, a lot of these communities have had poor health and poor conditions for generations. Analysis has been done looking at the 19^th century which suggests that the spearhead areas are very similar to the areas in Victorian Britain which were displaying multiple social disadvantages and poor health, so these are very long-seated problems. As time moves on, people become more and more impatient to see results, and it may simply be a question that time is needed to yield any information that is reliable about whether or not they are effective.

Q106 Sandra Gidley: Perhaps I could pick up on something Professor Whitehead said. You said that in Merseyside and Greater Manchester it is working well. I am not quite sure what you mean by working well. Do you just mean they are going to hit the target? Because only 40% of areas are.

Professor Whitehead: I meant, in terms of leadership, the fact that the people concerned are getting together and organising and working together in an effective way to tackle and to plan the initiatives.

Q107 Sandra Gidley: What is being effective? The fact that people are engaged by this and having a lot of meetings does not necessarily affect outcomes - although one would hope it would. Are they one of the spearhead areas on track to deliver against the target, because only 40% are?

Professor Whitehead: No.

Q108 Sandra Gidley: So it is not really working well.

Professor Whitehead: Well, it depends how you look at it in that respect, whether being designated as a spearhead PCT area allows you to do something new, something effective and to plan a programme that might have an effect, and whether it is worth being a spearhead area in that respect. I was saying that it is variable. In some area like Merseyside and Greater Manchester that seems to have been a positive impetus and facilitator for bringing people together to work on long-term programmes but those programmes have to be integrated and long-term and, coming back to the issue of whether you will actually see results in the short-term, you may not see results in the short-term.

Q109 Sandra Gidley: So working well is not connected to outcomes.

Professor Whitehead: No.

Charlotte Atkins: But the Department of Health is saying that the spearhead authorities in the North West are all off track.

Sandra Gidley: That has been acknowledged.

Charlotte Atkins: I do not understand exactly why you were saying that from your point of view they were doing well.

Sandra Gidley: Because they have a lot of meetings, apparently. I am sorry.

Charlotte Atkins: I know, but the Department of Health consistently are saying that all of them are off track. You were saying, of course, there is a different approach and everything else, but for them all consistently to be off track seems rather odd.

Stephen Hesford: Could I tell my colleagues that I live in this area and I know they are doing a very good job.

Charlotte Atkins: Yes, they may well be.

Stephen Hesford: Hold on, I live there and this academic discussion about what this might mean ----

Chairman: Let us not have a debate. We are talking here about how we measure health inequalities and what the term is.

Stephen Hesford: I understand what Professor Whitehead is saying.

Chairman: All right. Let us not have a debate. Sandra, do you have another quick question and then I want to move back to the agenda.

Q110 Sandra Gidley: I have a more general question because I have looked at the proportion of spend in all PCTs, particularly spearheads, that was allocated to public health and there seemed to be absolutely no correlation between being designated a spearhead PCT or area and the amount of money that was spent. In fact, where the extra money was received, in some of those authorities or trusts the spending has gone down on public health over time. That may have been the financial pressures. Has anybody done any work to show whether it is just about money or whether the outcomes would be better affected by the attitudes of people working together? Do we get too obsessed with the money side of things?

Ms Dixon: On resource allocations I think there are two issues. One is that historically the resource allocation formula has gradually brought these spearhead trusts closer to what is called their target capitation (the capitation they would receive if it was truly distributed on the basis of the need of their population) but historically they have had a significant under-allocation and even in the current year they are slightly below the target that would be expected if they were given a full needs-based allocation. We are talking here of a historic under-spend in general on health and therefore on health services in a population that we would probably assume is more needy and that to reach them effectively with health care might cost more money. The data you are referring to, I think, is the programme budgeting data, where you can get a breakdown by PCT of how they are spending money on mental health, cancer, coronary heart disease and so on, and obviously there is a public health component to that. From our other analysis, certainly, there seems to be very little correlation between what PCTs are spending on particular areas of health care and the needs of the population. So there is very wide and probably unjustified variation in spending and I am sure that is true also of public health specifically.

Q111 Dr Naysmith: I cannot help but speculate a little bit that the discussions we have been having in the last ten minutes illustrate something about academic research, which is that it is much easier to demonstrate that something does not work than that it does work. In the case of Sure Start it is clear that some academics have jumped in too quickly and tried to assess something which is going to take a long time. Some of the things you have been talking about just now suggest that if you do not make the observations at the right time, then the effects you have had will have disappeared and moved on and gone somewhere else. As an ex-academic myself, it just illustrates why it is really frustrating doing research sometimes. But I do not really want to talk about that, I want to ask Professor Graham a question. When you started off you talked about how in the 1950s and 1960s and the 1970s there was an increase in economic wealth and so on, and that was accompanied by an improvement in health but there was not an increasing gap. Now, in the 1980s and 1990s and currently, we are obsessed with this difference in the gap there is that is developing. When we were in the Netherlands last week it was suggested to us that, perhaps, rather than focusing on narrowing the health gap between social classes, as the current Government is attempting to do with its targets, a more sensible approach would be to raise the levels of those with the lowest health status and concentrate on that and not constantly get worried about the gap between those with the highest and lowest status. What do you think of that, Professor Graham?

Professor Graham: The White Paper objectives are to do the first as a way of getting to the second in England - and there are exactly the same goals across the UK.

Q112 Dr Naysmith: People keep reporting the gaps widening, while people at the lower end are getting healthier all the time.

Professor Graham: Yes. The pattern we have in England - which is repeated across other rich countries - is that the health of the poorest groups are improving but they are improving at a slower rate than the groups above them and the better-off groups: while their health is improving, the guys up here are improving quicker, so the gap is widening, but it is not that the health of the poorer groups is deteriorating. Their health is improving.

Q113 Dr Naysmith: Is not the important thing the improvement in health in the poorer groups, and if we were to concentrate on that that would be much more useful than worrying about the gap.

Professor Graham: That is the strategy of the Government at the moment. The concentration of effort is on those in the poorest circumstances and in the poorest health and spearhead is a way of designating that status. The ambition is that, by increasing the rate at which health improves in those groups, the rate will increase at a similar or quicker rate than better off groups and therefore the gap will narrow I think the answer to the Netherlands is that in some sense that is what we are doing. We are concentrating on those in the poorest circumstances, seeking to improve their health over time - which historically we have been achieving - but, additionally, seeking to improve their health over time at a quicker rate and, therefore, enabling those groups to start to catch up with the health of the better‑off groups.

Q114 Dr Naysmith: Why do you think the gap was not widening in the 1950s and 1960s and 1970s? Presumably something was happening to everyone.

Professor Graham: When I was giving the overview in the response to the Chairman at the beginning, I was saying that in the 1940s, 1950s and 1960s income inequalities were not widening. We have had a trend towards widening health inequalities over a longer period, although there have been periods where that has been less intense. I think the Government is trying to rectify a long-running historical trend; namely, of improving health in poorer groups but at a rate that leaves them increasingly detached in health terms from the mass of the population. If we were to take the Netherlands approach and simply say, "All we want to do is to improve the health of the poorest group," we could all go home. We are doing it. We have improving health in the poorest groups, and historically we have had that, but we have a much quicker rate of improvement in the better-off groups. The real challenge is to enable everybody to reach the kinds of levels of health that children in the best-off circumstances can look forward to over their lifetime. I feel the answer to the Netherlands is not, "Don't do what we are doing," but, "We do it in order to achieve something more ambitious," which is to begin to close the gap.

Q115 Dr Naysmith: Do you have any views on this point, Professor Whitehead?

Professor Whitehead: No, I think that is a very clear explanation.

Ms Dixon: No, thank you.

Q116 Dr Taylor: Professor Graham, you are very clear that the major determinants of health lie really beyond the NHS. In these next few questions we want to concentrate on NHS solutions. Do you see really that the NHS can do much about reducing inequalities, or should everything be aimed at the solutions outside the NHS?

Professor Graham: I think it has to be a dual strategy. The major drivers of health over people's lifetimes are not ones on which the NHS can necessarily have a direct impact, but that does not mean that the NHS does not have an absolutely vital role to play. One of the difficulties about having indicators based on, in effect, death - which is death at the end of the first year for the infant mortality target, and life expectancy is really a measure of whether you die young, because we are looking at deaths over 65 - so we have very death-focused targets - is that health is much more broad than that. If you begin to think about what health inequalities means across people's lives, I think you begin to get the sense of the capacity of the NHS really to do an awful lot. If I were to say, which we know from evidence, that children come into the world in poorer households, on average, with a lower birth weight and they are more vulnerable to low birth weight - which is a highly powerful marker of child health. We know that through their childhood they are more likely to develop slowly physically, to have intellectual difficulties, to have physical disabilities. As we move into adolescence we know that those children from poorer backgrounds are more likely to take up smoking and they are more likely to be heavy smokers across their lives. Anybody here who is a doctor can begin to see the moments at which these problems are walking in through the surgery. When we get into adulthood we know what is called co-morbidity (the clustering of more than one health problem) are socio-economically patterned; they are more likely in poorer groups. Psychological problems as a backdrop to physical problems are more likely. We know from a number of studies that the ageing process sets in about ten years earlier in poorer groups than it does in the most advantaged groups. We know that onset of chronic disease comes younger and we know you die younger. There are so many points at which people's lives are crossing in and out of surgeries and hospitals, so I think to say the NHS has no role to play would be really to abrogate its major stewardship role in terms of looking at the health of all communities and, particularly, those which need it most, which are the poorer groups.

Q117 Dr Taylor: Could I check on your evidence about infant mortality. It appears that we are only looking at certain groups and excluding, I think you say, babies born to other groups of parents, including those solely registered by the mother and jointly registered to parents where the father is long-term unemployed. These are missed out. Have I understood this correctly?

Professor Graham: Yes. The measure of infant mortality, as far as the target is concerned, is based on children born to parents where an occupation can be registered as routine and manual. In order to register the occupation, you need either for the parents to be married, so the occupation of the father can go down, or registrations where the father is present at the registration, where the father's occupation goes down. Where a mother registers as a solo parent, that baby falls into another category which lies outside the target. The sole registered births have higher infant mortality rates than those jointly registered babies born to fathers who are in the manual and routine occupations.

Q118 Dr Taylor: Effectively, the figures are probably worse than the ones we get.

Professor Graham: They are worse, yes.

Q119 Dr Taylor: What can the NHS do at this stage of life?

Professor Graham: One of the things that the NHS fully understands is that there are vulnerable groups who lie outside the target, in particular this group that I have mentioned: babies born to mothers who register the child on their own. Their children are particularly vulnerable to a range of health problems, so I do not think we can assume that the NHS in some sense abandoned that group. We know very much that that is not the case, but it is one of the anomalies of the target that, in order to have robust data, the target is based on parents where there is an occupation to register.

Q120 Dr Taylor: Is this well recognised?

Professor Graham: Yes.

Q121 Dr Taylor: Sufficiently so that PCTs will instruct their staff to target these sorts of groups?

Professor Graham: I cannot comment on PCT level but I think it is certainly well recognised at policy level that the target is an indicator of child health and that what it is doing is alerting communities and those who serve them to the particular vulnerabilities of disadvantaged groups, but the group on whom the target is based only represents one of those disadvantaged groups and there are others who are not included in the target but who will be included in the interventions in those areas.

Dr Taylor: Right. I will come back to Professor Whitehead with a later question because I know she has some suggestions.

Q122 Stephen Hesford: I am asked to direct this question at Anna initially. Social marketing, the Five-a-Day sort of campaign and other interventions, do they work?

Ms Dixon: When compared to the earlier attempts at health education and health promotion campaigns, I think there is some evidence to suggest that some of the more sophisticated approaches that are understood by the term social marketing - which is not just about an information or advertising campaign but is a whole array of tailored and targeted information, and there may be promotional incentive schemes or other types of ways of getting people to change their behaviour - can be more effective in getting behaviour change than simply the information, commercial type of approach that was adopted in the past. Professor Graham very eloquently described, in terms of smoking, some of the trends, where, in the early period, when there was greater education about the dangers of smoking that obviously had a differential impact in terms of, particularly, educational level. There is evidence to suggest that people who are higher educated will respond to just the more simple knowledge or education messages but to get people to change behaviour you have to do more to lead them to action. Just simply knowing that something is not good for your health or, in the case of Five-a-Day, that something is good for your health is not sufficient and it does need a concerted effort. I think the Five-a-Day has been successful because they have engaged supermarkets: there has been labelling as well as the advertising and there have also been some supermarkets doing incentive schemes around additional points and so on. They have tried to look at ways of doing it in a multi-level approach and I think there is some evidence that that can be more successful in getting behaviour change than just a simple message.

Professor Whitehead: Social marketing is obviously a form of health education. The evidence shows that health education on its own is not effective; it needs the support of other programmes to make the environment more health enhancing, et cetera, making the behavioural choices easier to make, et cetera. It needs the support of structural programmes.

Q123 Stephen Hesford: In relation to aspects of what other support there might be or other buttresses, is there a role for financial incentives to change individual behaviour?

Ms Dixon: We produced recently a short review of some of the literature around using financial incentives to change health behaviours. The evidence is pretty limited and it is fairly focused on getting people to enrol in a programme. Where a financial incentive could be, for example, to sign up for a Weight Watchers programme or to join a Quit Smoking programme it has been quite successful, but it has no bearing on the success of then changing the behaviour. Where they have been applied, they have generally been applied for some very short-term and specific behaviour but, in terms of maintaining more complex behaviour changes over time, there is at the moment a lack of evidence to suggest that financial incentives can be used in that way. Although there are experiments: some insurance companies are trying to model ongoing financial incentives linked to their products, for example, around gym usage and so on, and they are showing that they can to some extent maintain those behaviours, but clearly that is with a very select and slightly wealthier group of people.

Professor Graham: The UK evidence I do not know - and Anna has clearly spoken on that and I am sure has done studies - but there is something called conditional cash transfer programmes which have been introduced, particularly in Latin American countries, in Mexico, Chile, Guatemala and countries like that. Those schemes are targeted at poor families, usually on a village basis, and the family receives quite substantial sums for agreeing to enrol on particular health programmes and sending their children to school and that sort of thing. There have been some evaluations of those programmes which are now coming out. They do support what Anna has just suggested, which is that they obviously sign up to the services that they are linked to - because it is mandatory: you do not get the cash unless you go the services - but the recent work I have seen which has come out of Mexico suggests that the effect they have on child health is coming through the fact that families have more money. It is the cash that is important. When we are looking at cash incentives, we do have to ask whether the behaviour is changing, if it is changing, and if child health is changing because of the service intervention or because families have more money to spend on their children and feel less stressed and therefore the family environment improves in those ways.

Q124 Stephen Hesford: I understood you to say before about income inequality. Does that emphasise the role of income inequality in health inequality?

Professor Graham: Yes, I would say it does. The evidence that we have suggests that if you increase income in poor families then that income disproportionately goes on aspects that improve the conditions for children: food, fuel and items that support their lifestyle.

Q125 Stephen Hesford: Dr Taylor introduced this section about NHS solutions within that policy area. What can we say about penalties? If there is a financial incentive, what about penalties? For example, if you have potential heart problems, your doctor says, "If you don't stop smoking you are going to get x, y or z, and if you don't stop smoking you can't have your heart operation because we told you." What about penalties to influence behaviour?

Ms Dixon: I am not an expert on the psychology literature but I have been briefly looking at it in the context of this work we have been doing. I think there is a lot to suggest that when things are framed negatively it is not usually conducive to positive behaviour change. Things that are around positive goal setting, about motivation and helping people to live their lives in a more autonomous way, are generally much more successful. The use of penalties would probably fit in the negative framing of health behaviour which I think has been shown to be not particularly conducive to positive behaviour. That is a separate discussion from whether in principle you would want to deny care on the basis of people's health behaviours, which I think gets into a whole different ethical discussion about what it means to have a universal service and what clinical decisions can be made. I think everybody would agree that doctors should be advising clinically, given people's risk, about weight and smoking status for particular health treatments, but whether there should be some sort of use of it as a penalty to drive behaviour change I do not think is a sensible policy direction.

Q126 Dr Stoate: Some people have said that access to services makes a relatively small difference to health inequalities. Do you agree with that?

Ms Dixon: I think there is a lot of evidence to suggest that health care, particularly now in a developed society like our own, can make a very big difference to people's health status. Clearly these wider determinants are extremely important, but when we come to look at things like premature mortality there is a lot that medical care can do and there has been good research to try to quantify the sorts of levels of mortality, perhaps less so around morbidity, that are amenable to medical care. If we accept that fact and then take a look at the healthcare interventions and healthcare services that can make a difference there, we do see inequalities in access persisting, particularly once we go beyond primary care. I think this country in international comparisons has a pretty good record in terms of equity of access to general practice and primary care but both aggregated studies and a lot of the smaller studies of particular NHS services clearly demonstrate that there continue to be socio-economic inequalities in access when accounting for differences in need, and I think that addressing that would still be an important part of addressing some of the inequalities.

Q127 Dr Stoate: You are talking more about secondary care there. What can you do to improve access to secondary care for these particular groups? How would you go about that?

Ms Dixon: There have been some studies which have tried to look at the reasons for the inequalities. Most of the studies just measure the inequalities in access and then tentatively suggest a few ideas, but there has been less done to investigate. It seems that there are issues around car ownership and public transport, the lower socio-economic groups finding it difficult to get to out-patient appointments, and often around some of the out-patient systems, where letters are either not received because people are more mobile or perhaps not in permanent housing, or literacy issues about receiving a letter and meeting that appointment, putting it in your diary and getting to it, so there are some practical issues about why there are higher "did not attend" rates. Another issue is to do with clearly late presentation in primary care, and there do appear to be some potential differences in the propensity to onward refer, either because of the presentation of symptoms in primary care or the communication that goes on between patients and their general practitioners; so that, say, chest pain in one person is not expressed as such a great problem and therefore may be left longer and a referral not be made. These are all things you will know from practice.

Q128 Dr Stoate: Yes, I accept that some people do present later. Is there any mileage in improving access to primary care? Do you think that is a worry? Or do you think the access to primary care is already adequate?

Ms Dixon: I think it is what happens to people when they are in primary care.

Q129 Dr Stoate: I am talking particularly about access. The question is about whether improving access is a way to go. Is that a cost-effective method of improving inequalities?

Ms Dixon: There is evidence that there are some areas, particularly deprived areas of the country, where GP access, the actual presence of general practice and the use of general practice, does remain a problem. There is an issue about getting GP or primary care services more generally available in those areas, but ensuring the quality of those services - which is back to the discussion on whether QuOf is sufficient - and ensuring that people are getting the appropriate services that will benefit their health. It is at several levels: they have to have the primary care; they have to be able to access it; and when they get it they have to receive high quality primary care and get appropriate onward referral and treatment where necessary. I think there is a whole chain of different points at which inequalities in access arise, and we do not understand where the key points are and what we can do at each point to ensure there are fewer inequalities in access right the way across the chain, I suppose, or the patient pathway.

Q130 Dr Taylor: Perhaps I could turn to Professor Whitehead and explore your written evidence because you have given us seven main fronts on which health should be tackled. In paragraph 5 of your Executive Summary you say, "We need to monitor and identify which policies and interventions are making things worse, but also recognise and preserve the equitable systems that we do have, and guard against erosion by countervailing forces." Could you give us some idea of policies and interventions which could be making things worse? What do you mean there?

Professor Whitehead: As far as equitable systems were concerned, I was, for example, thinking of the NHS itself as a model of an equitable system. If anybody were trying to design an equitable healthcare system they could hardly do better than the foundations on which the NHS is based, from the financing of it, progressive taxation, to free at the point of use, to a comprehensive integrated service, et cetera. That was a prime example of an equitable system which should be protected. I do think there are some threats to it.

Q131 Dr Taylor: What are the threats?

Professor Whitehead: There are external threats to this sort of universal system - indeed, many of the universal systems in Europe - from, for example, the World Trade Organisation Directives. There is a continual pressure to bring universal healthcare services of different countries within the World Trade Organisation agreements so that healthcare system would be just like another commercial commodity. I think that is a great threat. Internally, I think we have to be very careful about different health sector reforms to make sure that they are not damaging some of the major principles. Some of the proposals and some of the reforms appear for example to be fragmenting the service. We have a comprehensive integrated service able to plan services in relation to need, but some threats are coming from fragmenting that service, from commercialising it, and privatising some sections of it, and the long-term threats to that are great in terms of whittling away at the universal service provided in relation to need rather than ability to pay, for example.

Q132 Dr Taylor: Thank you very much, that is music to my ears, but I am not sure to everybody's, and that is great. Can you just go through your seven main fronts and would you pick out one or two really important ones and expand on them?

Professor Whitehead: The main point I was making here is that although inequalities in health are not, in the main, generated by inadequate access to health care in this country, there is a lot that I think the health sector can do, and I have listed these things. I think that the Health Service can recognise the severe gradient in health across the social spectrum and make extra provision for the extra need that is evident in different groups in different areas. It can try very hard at matching services to need. It can help, as I say, alleviate the health damage caused by the wider determinants of health. In the paper I give the example of unemployment. People who lose their jobs are at great risk of a decline in mental health, for example. Maybe the Health Service cannot do anything about getting them a job, but it can anticipate that decline in mental health and provide extra support and extra services to try and prevent the decline in mental health that groups that are unemployed might experience. Again, I think the issue of boosting preventive and health promotion programmes with an equity orientation is tremendously important. As Professor Graham has said, there is a very steep gradient in risk factors like smoking and dietary risk factors, etcetera, that the Health Service can do a lot about. The final one I think is perhaps the most important, that the Health Service can motivate and facilitate the sort of multi-sectoral actions that know to take place outside the Health Service. They need to work with colleagues to actually stimulate and motivate action beyond the Health Service. I would say that if the Health Service is not going to take a lead on this, then who will? What sector will actually take a lead on inequalities in health, will monitor, will make the inequalities in health visible, et cetera, and will measure the health impact of policies outside the Health Service? Who will do it if the Health Service does not?

Q133 Dr Taylor: I take your point. Finally to all of you, what impact will the patient choice programme have on health inequalities? Do not spare your punches if you wish to or, if it is excellent, tell us.

Ms Dixon: I think initially there was a lot of concern that patient choice would worsen inequities in access to care, with pushier more vocal, middle-class patients exercising choices. With the pilots that were implemented both for cardiac patients and for patients in London, the data showed that there was no difference by socio-economic background in patients choosing, but the way that those programmes were implemented was quite different from the implementation of choice that there will be from 1 April, which is free choice at the point of referral. The choices were supported by a patient care adviser, travel costs were paid for to the more distant provider if that was what was chosen, and even travel and accommodation was organised for carers in some of the cardiac pilots. I think the reason why we did not see inequalities there may be partly the way the policy was implemented. We do not have the evidence yet around choice at point of referral in general practice as to whether there are differences by patient. There are certainly very big differences by GP in their propensity to offer choice, but what we do not know is whether patients are more demanding of being given that choice or indeed exercising it, and it is something that we are going to be doing some research on.

Q134 Dr Taylor: Have you anything to add?

Professor Whitehead: Just that I suppose patient choice puts patients competing with one another potentially, and that is a problem. If you get patients having to compete for the fastest or best service or whatever, then the more articulate and the more pushy will naturally win in that race.

Q135 Chairman: Have they not always done that, Margaret?

Professor Whitehead: Yes, they have always.

Q136 Chairman: That is why we have disparity in the share of monies for health as well.

Professor Whitehead: That is absolutely true that the more articulate have been able to find their way around the system. This patient choice agenda adds to that and augments that process, and if you get the more articulate, et cetera, getting quicker service, hogging the best service if you like ---

Q137 Charlotte Atkins: If you have a choice agenda, is it not more explicit whereas in the past it was less explicit and there the middle classes would have been able to navigate the system much more? At least now the choice agenda is much more in your face and therefore surely people are aware that they have the ability to choose whereas perhaps a few decades ago, the people who were more articulate and more capable of navigating the system were the ones in fact who won out? Has there been any research done on that or is it too early?

Ms Dixon: It is too early, but I do agree that in theory a process by which a choice is offered more explicitly and more routinely and is supported with information, actually has the potential to enable people who may previously not have asked specifically to go somewhere else to be able to benefit from faster access or going to a particular provider which offers a higher quality service. Many GPs would say that before the 1990s and the introduction of contracting they were pretty much able to send patients anywhere in the country anyway, and in a sense they had a choice, and some GPs would presumably have had patients who would have said, "Actually I have heard this is a good hospital and I would like to go there," and that would be something that I think you would expect a very small number of people to do whereas now if it is more routinely offered you could expect people from lower socio-economic backgrounds to be willing. We know that they are willing from the patient choice pilots and from surveys as well and they are interested in exercising choice, and I do not think there is anything to suggest that people in lower socio-economic groups for whatever reason are less capable or willing. I think the only issue is travel. It is a cost to go to a more distant provider and we know that that is a barrier to access and it is likely to be a barrier to access for these groups more than for middle-class patients.

Dr Taylor: It is very good to hear that there is somebody who remembers that choice is not something entirely new and that it did exist a long time ago.

Chairman: With the profession, but we will not go there. Let us move on to Sandra.

Q138 Sandra Gidley: Some questions for Anna. You alluded to the QOF earlier. Would it be fair to say that your interpretation is that the QOF really does not do anything very much to address health inequalities at the moment?

Ms Dixon: I think the fact that pretty much all practices are attaining such a high level means that it is not possible to say that it is directly doing anything to address health inequalities.

Q139 Sandra Gidley: You alluded to the fact that it has been revised. What would you like to see included in the QOF? What do you think would make a real difference to longer-term health outcomes?

Ms Dixon: I think that what needs to happen is there needs to be further work done to identify more clinical indicators where we know there is good evidence that if people receive that care that it will have a significant impact on improving their health and potentially preventing premature mortality. I think there needs to be a lot around the area of secondary prevention and what probably needs to be reviewed is some of the thresholds that are set, for example for diabetes management and so on, to ensure that they are being consistently met and that they are clinically stretching and sufficiently clinically appropriate thresholds to ensure good control by diabetics across the course of a whole year for example, so I think it is more about some addition and, as I said, I would give emphasis to clinical indicators. I think there were a lot of process indicators, which was important to have in at the beginning, for example setting up disease registers and so on, and these provide the basis for doing better population management and are a very important first step, but I think now we need to move it on to say we need to judge primary care professionals on their ability with that registered population of people with particular diseases, or with co-morbidities, to manage effectively the health of those people over the course of a whole year. There may be some additions and I think there may be cause for more stretching thresholds within some of the standards.

Q140 Sandra Gidley: Is there any way, for example, that primary care could be incentivised to tackle the teenage pregnancy rate or do you think that is not a role for general practice?

Ms Dixon: I would not want to comment specifically on that in that respect. I think the other point, though, to note is there is a lot of focus on QOF, and it has obviously been a very powerful incentive within the contract, but there are other aspects of the financing of primary care that perhaps also need to be looked at in terms of addressing health inequalities. At the moment there is no mechanism for allocating money to practices, and obviously with practice-based commissioning this is more and more important on the basis of the needs profile of the registered population, and so if we really want to reward practices who are serving deprived and least healthy populations and ensure that they have sufficient resources to maximise and improve the health at this quicker rate that we are wanting to happen, then we need to make sure that the resources are appropriately allocated to the level of practices.

Q141 Sandra Gidley: That is very much a carrot approach. Is that the best approach or should there be penalties for people whose services do not hit certain targets?

Ms Dixon: I think that there needs to be better quality, I would hate to use the phrase performance management because I do not think it is that, but it would be good to have more quality information about the care. I think one of the great things about QOF is that it has generated for the first time systemic data on practices' quality of care, and I think we could go much further in being transparent and actually help practices within practice-based commissioning consortia and with PCTs to analyse how are they doing and why their populations are having not such good outcomes as others. There are other things that they can learn and having that data not to punish practices but more having it for a level of transparency can promote quality improvement in primary care.

Q142 Charlotte Atkins: Can I come in on that. It is interesting that you are talking about QOF, but if you are talking about dental practices, in fact dental practices can be incentivised to look after a more deprived community by increasing the price of the unit of dental activity. For instance, in my area one dental practice is going to get an enhanced level of UDA simply because they are going to be asked to target particularly disadvantaged areas. Do you think that is something which could be more widely used to try to reduce health inequalities?

Ms Dixon: That sounds like an imaginative use of the contract and I would envisage that PCTs may wish to be looking at ways of using some of their primary care contracting mechanisms in an imaginative way in order to address health inequalities.

Q143 Charlotte Atkins: You think that some PCTs are not as au fait with their commissioning role that they would use it in such an imaginative way?

Ms Dixon: I do not think there is a lot of active commissioning of primary care because the main focus around commissioning has been on commissioning secondary care.

Q144 Chairman: You talked about clinical indicators that are brought up through QOF and we know a lot more now, and most of those indicators are about ill-health; should one of the issues not be about how we measure health within these communities that cover primary care centres or general practices in general terms so that we are able to make good comparators? What could we do which would incentivise ill-health prevention/pro-health as opposed to measuring ill-health?

Ms Dixon: I will take that briefly. I think the level at which that needs to occur is primary care trusts and I think that in the future primary care trusts' targets and the way that their performance will be measured will be more and more on the health of the population, and that will partly come through the local area agreements in terms of how they discharge that by working in partnership with local authorities and others to do health promotion and prevention. I think that primary care trusts, however, will need a longer time horizon. The problem is that at the moment we want to give them one year's money and be able to measure whether they have done something worthwhile within a year. That is probably not sufficient to do some of the more proactive health improvement for people already with chronic conditions and in terms of long-term conditions management, and it is certainly not a sufficient time horizon over which to performance-manage in relation to some of these more intransient health determinants. I think the focus for that and the responsibility and accountability should firmly lie with primary care trusts in terms of responsibility for the health of their population, but what measures over what period you can use to hold them to account for that is more difficult in the area of health promotion.

Q145 Chairman: Are patients more likely to take more notice of their doctor than their local MP? Is the issue around health inequalities and ill-health or potential ill-health not that they might be more effective in terms of being able to perhaps change somebody's lifestyle to the good so they do not end up on these clinical indicators on QOFs, and should not they or somebody who works for the team in primary care have some measured responsibility for promoting pro-health? You do not think - not now but in the future - that is something we should look at?

Ms Dixon: I would hope that it already is. There is a lot of what is called 'opportunistic' health promotion that goes on and I think that it would be very good if that was more systematised. There is good evidence about some of the skills that individual professionals, whether a primary care practitioner or a specialist nurse working in a heart clinic at a hospital, need and there are techniques that they can use that are more effective than just advice and admonishing people and saying "you really should lose weight". Motivational interviewing and brief negotiation are techniques that we could be training our health care professionals in, and they could form part of the curricula for health professionals and encouraging professionals to see that very much as part of their role to be effective in those opportunistic times, because when people have a diagnosis or are facing an intervention they are often more motivated or at least open to the possibility of changing, for example giving up smoking, changing diet, exercise, or alcohol intake.

Chairman: Let us move on, thank you for that anyway. Doug?

Q146 Dr Naysmith: There has been quite a lot of talk about smoking this morning and its effect on health and health inequalities. Professor Graham, the Financial Times' medical correspondent Nick Timmins has recently argued that something like £0.5 billion has been spent by primary care trusts on smoking cessation activities. He is quite a well-respected journalist and he reckons that it has had virtually no impact on smoking levels. Do you agree with him?

Professor Graham: If he is talking about smoking prevalence ---

Q147 Dr Naysmith: In adults particularly.

Professor Graham: --- In adults, then the prevalence rates between the early 1990s and 2002/2001 hardly changed at all. They hovered around 27/28% and since then they have dropped by 3% from 27% to 24%, so prevalence has actually started to fall again after a period of stability where we saw very little change.

Q148 Dr Naysmith: There is not much evidence of the PCT programmes have helped much.

Professor Graham: I would challenge that. When we are thinking about prevalence we have to remember that two factors drive prevalence, once is the rate of uptake of new smokers coming into the smoking population and the rate of cessation, and however effective a cessation service is, it is not going to stop the new recruiters coming in. It is trying to target the groups that are already smoking. We know that the baseline smoking rate at which people in the population would give up is about 1% of smokers a year, it is not very much, so we are losing smokers at about 1% a year out of smoking cessation and that is called the baseline cessation rate.

Q149 Dr Naysmith: People talk about a matter of weeks giving up, it is not a hugely long period in which to measure success?

Professor Graham: If you have friends and family who are ex-smokers, you will know that they live on what is called a 'revolving door', that you give up, you resume, you give up, you resume and eventually, God willing, you give up and stay given up. We know that about half of all smokers in any one year will try to give up but actually we know we are only losing 1% of smokers to cessation a year, and that is the baseline rate. In the work that has been done on smoking cessation services in England, which was work undertaken by Professor Ken Judge and colleagues, what they estimated is that the one-year abstinent rate for people using those services is about 15% which is, extraordinarily, much higher than the baseline rate, so I would challenge the fact that these cessation services are ineffective.

Q150 Dr Naysmith: Even though £0.5 billion is a lot of money that might be spent more usefully doing something else?

Professor Graham: I think it is very difficult to argue if on the one hand we know we have a very, very high burden of smoking-related disease in the UK (which we do) and we know that half of the smoking population every year is trying to give up at least once and then turn round to smokers and say, "We are not offering you any smoking cessation services." I think there is something slightly inconsistent about that policy. We know those smokers, if they continue, are going to shorten their lives appreciably and make very heavy demands on health services. I do not know where the costing came from but if it is £0.5 billion then this is a integral part of what a Health Service should be to encourage the kind of healthy living that Anna was discussing just now. I do think that what is important to realise is that the smoking cessation service on its own will battle to make changes in long-standing and habitual behaviours like smoking, particularly if what happens is the ex-smoker, newly reformed and clean of cigarettes, walks out of the smoking cessation services and returns to an environment which is largely unchanged. Other things have to be happening in their life to sustain that. I think smoking cessation services are a really important component but only one component of a much wider tobacco control strategy, so I would say it is expensive but worth it.

Ms Dixon: I would agree with that. It is about resilience and maintenance and we know that is very hard. These quit programmes generally are measured at six-week quit rates so if you do a longer term follow-up you will see a drop-off but that does not mean to say that they are not cost-effective.

Dr Naysmith: It was on my fifth or sixth attempt over about ten years that I actually succeeded. Thank you.

Q151 Mr Syms: A final question to you all: what are the three key evidence-based interventions that you would recommend to tackle health inequalities, starting with Professor Whitehead?

Professor Whitehead: I think the first one has to do with tobacco control because in this country we do have such a steep social gradient. It needs to be tobacco control in a wide sense; universal population level interventions coupled with extra intensive interventions for the worst-off groups, and also doing something about the wider social inequalities, so a broad comprehensive tobacco control. Certainly in the long term we have to do something to give our children the very best start in life, so I would continue the child poverty policies, the early education initiatives, Surestart, et cetera. It is very, very important to give our children the best start in life. Thirdly, I would just make a plea to give promising initiatives time to work and not cut them off when they are just getting started.

Professor Graham: I am going to echo some of the points that Margaret has mentioned. I think the first two would come from the recommendations of the Acheson inquiry, and one of its three core recommendations was about reducing inequalities in living standards, and it goes back to the very first issue that we discussed about what has been happening to inequalities over the last 50 years. If we go back 50 years and just before it, we had a period associated with the Second World War where we had a very rapid rise in life expectancy, the most rapid rise we have seen in life expectancy across a 50/60-year period, and we saw a reduction in health inequalities. The poorer groups' health improved relative to the richer groups and what that was associated with was a rapid narrowing of inequalities in living standards, so we do have some historical evidence that that is an effective approach and, as I say, it was one that was in the Acheson inquiry. The second point was also in the Acheson inquiry and was one that Margaret has mentioned which is about thinking about policies in the early years, and it would be a plea not to lose heart in Surestart. I think the most recent evaluation that has come out has begun to indicate the sorts of improvements that were expected. The programme on which it is based in the US was more or less abandoned because they did not think it was having any effect, until they started to follow up the children long term and saw that it did have these what are called 'sleeper' effects, meaning you can see nothing for some time and then you can get evidence that it has made a difference to their trajectories, so my second suggestion would be not to lose heart with Surestart. The third is one that again Margaret has mentioned, which is the need for a comprehensive tobacco control strategy. There are now quite good sources of evidence that suggest when tobacco control policies are weak then we see particular effects on poorer groups, either that prevalence will rise or that rates will fall less quickly, so they do seem particularly sensitive to the wider policy environment, and smoke-free is part of that, smoking cessation is part of that, ad bans and sport sponsorships are part of that, and restricting sales to children are part of that. I think we are slowly moving towards that comprehensive strategy and that will be a very, very important moment in public health policy.

Ms Dixon: I would probably focus on what the NHS can do, although I would agree with those as wider government policies. In terms of what the NHS could do, I think that in terms of addressing health inequalities there is still a need to ensure that those with the poorest health get the most care and the highest quality of care. That would stretch from both primary and secondary prevention and making sure that primary care in those areas of greatest health need particularly have the resources and deliver that care, and also ensuring reduction in inequalities in access in treatment and management of chronic illness and co-morbidity. I think that is fundamental and we still have not fully addressed it within the NHS. I do think there is a role for the NHS to work closely with local authorities to create a health promoting local environment for the population, and so I would like to see the appropriate use of local area agreements to really deliver on a local health agenda, for which the primary care trusts should be held partly accountable. Finally, I think we will probably need more research before this can truly happen, but I think PCTs need to look at commissioning effective behaviour change support for those in poorest health. We know very little about what really works for these people who have the most complex social needs and the worst risks of future health. I think we need to know a bit more about that, but as soon as we do we need PCTs to be spending their money and commissioning that sort of support.

Chairman: Okay. Could I thank all three of you very much indeed for coming along and helping us this morning with our inquiry.

Witnesses: Professor Kay-Tee Khaw, Professor of Clinical Gerontology, University of Cambridge, Professor Richard Wilkinson, Professor of Social Epidemiology, University of Nottingham, and Professor Sir Michael Marmot, Professor of Epidemiology and Public Health, University College London, and Chairman, Commission on Social Determinants of Health, gave evidence.

Q152 Chairman: Welcome. Could I thank you very much. I am sorry about the slight delay we have had at the beginning of this second part of our second evidence session on our inquiry into health inequalities. Could I ask you to introduce yourselves and the position you hold for the sake of the record?

Professor Wilkinson: Richard Wilkinson. I am Professor of Social Epidemiology in the Department of Epidemiology and Public Health at Nottingham, and I have worked basically on health inequalities and social determinants of health for about 30 years.

Professor Sir Michael Marmot: I am Michael Marmot. I am Professor of Epidemiology and Public Health at University College London. Also relevant to this morning, I chair the Commission on Social Determinants of Health set up by the World Health Organisation basically to look at what we can do globally on inequalities in health. I also chair the Department of Health's Scientific Reference Group on Health Inequalities that produced the Status Report on health inequalities ten days ago.

Professor Khaw: Kay-Tee Khaw, Professor of Clinical Gerontology, University of Cambridge.

Q153 Chairman: Thank you. I did not say this to the previous witnesses but thanks also for the written evidence that you have sent in for this inquiry. I suppose really, Michael, this is headed to you in view of what you have said in your 2007 Status Report, published, coincidentally I am quite sure, on 13 March, which was the first evidence session that we took here on health inequalities. It indicates that 75 out of the 82 government commitments in Programme for Action have been wholly or substantially met. There is also, obviously, evidence that QOF payments are similar across deprived and less deprived areas, so why have health inequalities continued to increase at this stage? Are we doing the wrong things or are we just measuring the wrong things? What do you think?

Professor Sir Michael Marmot: As I said in the introduction to the report, I think the headline that health inequalities continue to increase does not quite capture the reality of the data. There are two key aims. One is to improve health for everybody and the second is to reduce the inequalities. What we showed in the report is that life expectancy for the worst off had continued to improve, so much so that, taking the Spearhead areas, which was the way the data were tabulated, life expectancy for the worst off group at the latest figures was marginally higher than the average for England eight years earlier, so that the worst off group today had pretty well the same life expectancy as the average eight years earlier. Health was improving for everybody across the board, so saying that health inequalities continue to increase does not quite capture that very welcome improvement of the worst off. The second was, has the gap narrowed, and the answer is no, it has not narrowed. I think it is reasonable to predict that if you declare your intention to set a set of activities in motion in 2002 and they begin, let us say, in 2003, they could all be met and have no impact at all on health inequalities in the next two or three years when the data are collected because these are long term issues. In fact, in my Commission on Social Determinants of Health we are, not to let the secrets out of the bag, talking about closing the health gap in a generation. We think that is what it takes, at least a generation to close the health gap. That said, within that context of saying not much has changed to narrow the gap - and we are all disappointed. It is not as if we would not have wanted the gap to be narrowed - there were one or two encouraging signs. Looking at mortality from cardiovascular disease, looking at mortality from cancer, the absolute differences between the worst off and the best off, or the worst off and the average, showed some signs of narrowing, so that is movement in the right direction. We could have a slightly arcane discussion about whether it was right to set the targets in terms of relative differences or whether they should have been set in terms of absolute differences. They were set in terms of relative differences, so the Status Report did look at relative differences, but we also looked at absolute differences and there are these glimmers of encouraging signs with mortality from cardiovascular disease and cancer.

Q154 Chairman: We heard earlier today about the length of time that somebody is having it, and it seems to me that we are talking about cultural changes in many instances. Of course, these government targets were set for 2010, which is not going to change the culture of South Yorkshire in terms of its liking for beer and fish and chips, and smoking, sadly, to some extent. What is the exercise about if we are setting targets for 2010 and yet we recognise the fact that these are lifestyle and cultural issues that may be generational? Do you think that we are looking at the right things?

Professor Sir Michael Marmot: First, I think it was incredibly important to do it, to set the targets. I was also a member of the Acheson inquiry and it was all swirling about at that time: should there be quantitative targets for reduction of health inequalities, and, being a typical academic, I said, "On the one hand ... on the other hand". On the one hand, I said, we do not know what to do. We cannot quantitate how much money it will take to make how much change in the right inputs and what the effect of that would be. I said it might be that the best thing we could do to reduce social inequalities and coronary heart disease 50 years from now would be early child development, but to quantitate that effect is close to impossible, so that would argue against having targets. How can you have targets if you do not know what the input should be and you cannot measure the output properly? On the other hand, if we have targets it is aspirational. We set them up there. They ought to be sufficiently challenging, and some people said that 10% reduction was not challenging, but, as you can see from the data, if the average is moving up, 10% is incredibly challenging because the worst off have got to improve more rapidly and that rarely happens. They have got to be sufficiently challenging but somehow appear amenable to action. Then, if we do not meet the targets, everybody runs around saying, "We have to work harder. What should we do?", and I think that is absolutely the right thing to do. We all pay attention to what is it we need to be doing to meet these targets. My own view is that a lot of the things we have put in place will reap benefits, but the last thing we want is to stop the average going up because we want everybody to get healthier. The real challenge is, can we get the worst off improving more rapidly?

Q155 Chairman: I accept that, and I accept that we are getting both wealthier and healthier as a nation and that is a very good thing, but we do set these targets and they are commented on from time to time. It just seems to me that, given that we have 2010 targets, we now say that it is going to take time; it is cultural. It could be a decade, it could be two decades in some instances before you see a change. Let me go on a different tack: how we measure. I have looked through the Status Report. If we were doing a health promotion in Spearhead areas, in terms of trying to change the diet and the habits of people we measure mostly normally in terms of outcomes, and health or ill health might be a better measurement of that. If you are doing public health campaigns, have we ever looked at communities that have been running these, about what they consume or do not consume, what is shifting off the shelves in local supermarkets or if tobacco is decreasing, or alcohol. We now know the top five retailers, if you have a loyalty card particularly, know where you shop, know what you get. You go into different stores and there are different things on the shelves. It is different here in London from what it is in South Yorkshire. Do we ever use any of that evidence to be able to say whether any interventions were taking place over time (and it could be a decade to change the culture or the diet of the community), whereas at the moment we measure in the short term, not just looking at individuals and their outcomes but at other matters, in terms of what is happening in these areas? Have we ever done that?

Professor Sir Michael Marmot: There has been a little bit of pilot work using retail sales as a tool for health monitoring. I do not think it is used in any major way, and in fact, with another hat on, I have been attacked by the food industry more than once and I have said to them, "You must have an enormous amount of information and knowledge about what promotes behaviour. Why do you not use that for the public good instead of attacking health people who are trying to improve public health?", and they remain very silent on that issue. They must have an enormous amount of knowledge and they say it is commercially sensitive and they are unwilling to share it, but I have said to them, "You will make as much profit if people eat healthy food as if they eat unhealthy food. Why not join with us in trying to promote the public health?". I think there is a repository of knowledge in there that we cannot get access to, but it would be quite helpful if, as we scientists do; we share our knowledge, they were more willing to share. I would make another point which relates to, you hit all the targets of action, then why are things not getting better? Professor Jerry Morris, I think after his 90^th birthday, calculated the minimum income for healthy living for a pensioner and he did it by consensus. He went round to the various experts and said, "How much does it cost to eat a healthy diet?", and, "Is it reasonable to expect people to buy presents for their grandchildren and make visits to friends and so on? How much would all that cost?", and he summed it up. Then he looked at what a single pensioner gets with the state pension and there is a huge gap. People who rely on the state pension who are pensioners do not have enough money to lead a healthy life. That is the clear judgment and it is the same for a couple. They do not have enough money to live a healthy life. We can give all the health education we like. If people cannot actually afford to do the things they need to do to remain healthy then they are not going to be healthy. That has to be a key issue in inequalities and we have not solved that one.

Q156 Chairman: Why do poor people take the wrong decisions then in terms of smoking more than richer people do and having unhealthier diets? Why is that?

Professor Sir Michael Marmot: Hilary Graham, who has just left, is the world's expert on this issue, and what her research shows is that smoking is not a key issue for people living in relative poverty when they have a number of other key issues that concern them more immediately, and when those statements are being made at times they have been taken as pouring scorn on the smoking issue. Neither Hilary Graham nor I would do that. Smoking is of vital importance. If you look at Washington DC, young black men have a life expectancy of 57. Young black men also have a one third probability of being incarcerated for drug dealing between the ages of 18 and 24, so they are either going to die early or they are going to be put in prison. You go to those young men and say, "You know, you really shouldn't smoke because you might get lung cancer when you are 60". I will not repeat the language that they are likely to use if you say that, but I do not think you would get a very welcome reception. That is an extreme case but I think some of that goes on if people have multiple problems and smoking does not rank so highly on their list of problems that they are willing to do something about it.

Q157 Chairman: I am sorry about this conversation. Do you have any views on what you have just heard?

Professor Khaw: Professor Graham earlier talked about the changing gap in some of the chronic diseases, many of which are related to lifestyle factors, as you indicated, such as smoking and dietary intake. A lot of interest has been focused on which interventions are effective and why some people respond more to messages than others, and I think there is increasing evidence that some interventions work very much on individual choice and some people will have more capacity to make those lifestyle choices than others, whereas some other interventions are much more environmental, such as anti-smoking legislation, pricing of cigarettes, and there is increasing evidence that people with different levels of education or social class may respond differently to different interventions, so I think we have to be much more careful and specific about which interventions may increase health equalities. For example, if it is in terms of personal education or health promotion messages, it is quite likely that better educated people will take these messages on and change their diets and stop smoking, and then you get this increase in health inequalities, whereas interventions that are applied much more across the board, which involve much less individual choice, such as environmental measures, seem to affect all classes much more equally. On the one hand we have to examine which interventions may affect everybody more and at maybe much more societal levels, like legislation and pricing and passive smoking and environment. The second issue which has been raised is that we are not very good at knowing what the effects of these different interventions are, whether they are individual smoking cessation clinics or local level policies or national policies. Some research which has been done, for example, suggests that individual social class, educational status and local residential deprivation independently predict your likelihood of having a healthy diet or not smoking, and they are all separate, so in fact the effects of living in a deprived area influence you independently of whether you are well educated or whether you are low or high social class. That suggests that local area interventions may also have an impact on health behaviours, apart from your individual educational status fix, and that suggests an area you can target. I think where we have failed is that the NHS as well as local authorities are a huge test bed for all these different policies that are being implemented all the time. Research is not just a passive thing of, "We generate research and here you implement it and this is where it is". We are continually generating the evidence all the time, every time we change policy, every time the PCT decides to put out incentives or not put out incentives. We have a huge opportunity with the NHS with different primary care trusts working in different ways with local authorities to monitor the impact of different interventions, both the long term, which is the mortality, as Michael has indicated, which is what we are interested in, and the intermediate measures, such as the health behaviours in terms of smoking prevalence in different local areas, in terms of fruit and vegetable intake. I think the limitation is that we do not really have the capacity to do that because there has been a huge decline in public health activity, in academic public health activity, and if the Health Service is not the focus of public health activity that is generating the impetus to ask local authorities to evaluate the impact of their changes in policy, to evaluate what they are doing to monitor the health impact, and to report this back and say, "It works in this place; it does not work in this place; this is what we should be trying; this is the effect of the national policy on smoking prevalence. This is the effect of local policy", I think we should be encouraging the NHS to turn some attention to it, but it is increasingly difficult to do that in an atmosphere where people are entirely focused on very specific targets and there are huge incentives. I agree that targets are important, both aspirational and used to inform behaviour, but the minute you tie rewards to the targets you get counter, perverse incentives, such as with the infant mortality where people target the thing they are measuring and not the thing that they are really interested in, which is the single parent mortality. It is important to put it on the agenda as something that is important to be aware of, but also to be very careful about tying rewards to that.

Professor Wilkinson: Can I address two of the points you raised with Sir Michael Marmot? First is the time lags issue. I think it is important to distinguish between, if you like, cultural lags and biological lags. Although I think it is reasonable not to expect results to come through in the short term, just as Michael Marmot was saying, we do know that in the Second World War the changes then were big enough for death rates in different age groups from quite different causes of death to drop from about 1941, quite substantial drops, so we know it is possible to have improvements in the short term, and, of course, the experience of eastern Europe and Russia in the transition shows that you can have a worsening in the short term as well. Whether we only expect effects in the longer term because we are not doing anything very dramatic or whether there are more fundamental, maybe biological reasons, it seems to me that perhaps there are not the biological reasons. Even things to do with cardiovascular disease where we know there are long histories, to some extent seem additive processes and maybe you get a fall in death rates whenever you stop that additive process. I do not think, even with diseases that have long histories, one should rule out the possibility of short-term effects. On the health behaviours, I think you ask a very good question, why do poorer people behave unhealthily or badly or whatever. The fact that nearly all health-related behaviours have similar social gradients, they are all more common lower down the social scale, I think is a real clue to the situation of how people experience their lives. In a way, a lot of them are props, whether we talk about eating for comfort or smoking, the sort of work that Hilary Graham did on smoking, or alcohol or drugs, or not taking exercise. In a way, health-related behaviour is all about resolutions to give up the things you do not want to give up and to do the things you do not want to do. You cannot do that, you cannot make the resolutions and stick to them, unless you are feeling on top of life. I think in that picture there is a clue to what is driving the social gradient in health-related behaviours and maybe also a clue to the kinds of pressures that we need to tackle in order to change them. I suspect a lot of the health promotion activities do not address what really lies behind the social gradients.

Q158 Chairman: We are going to explore some of those issues in the next few minutes. We were in the Netherlands recently and it was suggested to us that, rather than focusing on narrowing the health gap between social classes, as the current government targets do, a more sensible and practical approach, and you heard this in the earlier session, is that we should just target on the health inequalities in those areas where the health inequalities are worse. In this case, the way we measure it, that would be social classes C4 and C5. Obviously, with the targeting that we do at the moment you would meet a lot of raising people up, as it were, improving health without actually closing the gap. Do you think that would be a better way of measuring and meeting the 2010 targets or not?

Professor Sir Michael Marmot: It might be a better way of meeting targets but it would not address the problem of health inequalities. My signature tune, as it were, is the social gradient. We looked at this first in the Whitehall Study of British civil servants not far from here, and what we found in people who are not poor, so who by and large do not live in Spearhead areas, do not meet any of the usual criteria, are by and large not single mothers, not on welfare, people in stable employment, was that the lower the position in the hierarchy the higher the mortality, the greater the rate of illness. In fact, we published recently from the Whitehall II Study, if we have been talking about mortality and life expectancy, that if you look at functioning and the quality of health-related functioning, it was as if the high grades were 12 years younger than the low grades in terms of their health-related functioning. In other words, a 72-year old high-grade man was like a 60-year old low-grade man, so this is an enormous difference and it is a gradient; it is a finely graded phenomenon. If one targeted only the worst off one would miss the overwhelming majority of the inequality related problems because the smallest group at the bottom have high risk but they are a relatively small group. The people above the bottom are at higher risk than those at the top and they are a much larger group, so that the problem for society, for all of us (and most of us do not think we are at the top but think of ourselves as below the top), we are at higher risk than those above us, so it affects all of us. That is why I think that taking action on health inequalities is taking action about improving society as a whole, improving life for everybody, not just for those worst off. Somehow or other one has, I think, to combine the universalist principles of the NHS, of Beveridge, et cetera, with, in addition, progressive focusing on those worst off to bring them into the mainstream. I would not say that the right strategy is somehow just to define a particular sub-group and focus on those because it misses the problem.

Q159 Charlotte Atkins: This morning, I know, on recent years we have very much focused on socio-economic inequalities, but, if we take something like infant mortality, clearly there are huge differences between black and ethnic minority groups in terms of infant mortality, so are we missing out on analysing health inequalities beyond the issues just of income and class? Should we be looking at these wider issues as well?

Professor Sir Michael Marmot: Yes, we should. On my Commission on Social Determinants of Health, we think of inequalities in relation to a number of different dimensions: ethnic group; belonging to an indigenous group; and gender. In this country, as in most developed countries, women have a longer life expectancy than men, but that is not the case when you look globally. In country after country, you find that what you would expect on biological grounds, women to be healthier than men, you do not find, which is clear evidence of gender inequality, so I think it is vitally important, and most of my research has been on socio-economic differences, but that does not mean that we should ignore the others. There has been a lot of debate about this issue in the US, that black and ethnic minority differences are, to some extent, socio-economic, and, to the extent that we do not pick up all the black and ethnic minority differences by our conventional measures of socio-economic position, it might be that it is not that the differences are not socio-economic, it might be a problem with our measurement. For example, this has been well-studied in the US that, if you take income, within an income group, blacks have lower incomes than whites. Take education, within an educational group, blacks have less education than whites, so the fact that, within an educational group, blacks have higher mortality than whites might not be saying that there is something other than education that is important, it might just be saying that our way of measuring education does not quite capture all of it. My own view is that there is something else going on that is not just our conventional socio-economic measures, but I do not think they do a complete job in summarising the black and ethnic minority differences.

Q160 Charlotte Atkins: So does that read over to gender as well, given that women generally are going to be on lower incomes than men? Do you have an even greater widening in health inequality in terms of gender?

Professor Sir Michael Marmot: The gender issue in this country is a complicated one because, as in all rich countries, women live longer than men, but, for many morbidity measures, women have higher morbidity than men, so they live longer, but they seem to have more illness, so it is complicated as to why that is the case. There had been a view that socio-economic differences applied more to men than to women and that the social gradient was less in evidence in women than in men. That view is wrong, in my view. It is wrong because of the measurement issue. If you get the measurement right, you see the social gradient in women as you do in men, so I think behind your question there is some of this problem of measurement.

Q161 Charlotte Atkins: You said that the reasons were complicated for the difference between men and women, but what are some of the reasons?

Professor Sir Michael Marmot: Well, I do not think anybody has sorted this out really, why women have more morbidity in general and less major illness and longer lives. People have looked as to whether women are somehow more sensitive to bodily changes, what goes on and they pick up symptoms more, whether in fact the morbidity is related to the quality of women's lives as opposed to men's lives, whether it is a labelling issue, women are more likely to get labelled. I do not think anybody quite understands it. I, for one, do not and my reading of the literature is that this is not a well-understood issue.

Professor Khaw: The main reason for the big difference in female life expectancy and male life expectancy is due to two differences. One is there is an early life between 15 and 30 when young men die far more from violent death than women and in the past, 100 years ago, there was not this difference because young women died in childbirth and they are no longer dying in childbirth, so there is a big gender gap there and, largely, it accounts for the big difference in life expectancy because the earlier you die, the bigger the gap in future life expectancy there is. The second big gap is in coronary heart disease, not total cardiovascular disease, but just coronary heart disease where, in the 40s to 60s, men have far more coronary heart disease than women, so the main reasons for the big difference in the sex difference is actually violent death in younger men and coronary heart disease in middle-aged men. If we are interested in narrowing the gap, I think those are the two areas that we want to tackle, reducing coronary heart disease in middle-aged men and reducing violent death in younger men from alcohol, road accidents, that sort of thing.

Q162 Charlotte Atkins: Would you like to comment on that as well?

Professor Wilkinson: Only on the ethnicity issue, simply that I have not done any work on ethnicity because it seems to me it is, largely, the same issue and it is a mistake to think of it primarily in ethnic terms, and I think very often people have started looking at ethnic differences, thinking they might be genetic. Clearly, there are some differences in how socio-economic status and so on works round ethnicity in that you cannot escape from your ethnicity and you tend to get specific cultures round ethnicity, but I think it is very important that we regard it as the same ballgame as social determinants of health and health inequalities more generally.

Q163 Dr Stoate: I have just a quick comment to Professor Marmot. I think any GP can tell you why morbidity rates are different in men and women: you never see men between the age of 18 and 45 in general practice. It would be extremely difficult to measure morbidity because blokes get morbidity, they just do not ever present with it and, when they do present, it is much further down the illness event or they are much more likely to have something very nasty indeed when they do get to the practice, and this is out of all the morbidity studies. I am sure that is a huge amount to do with it, but we just do not have that data. The question I wanted to ask really was to Professor Wilkinson, and that is that you have claimed that unequal societies cause health inequality, but we live in a Western capitalist society and surely the very essence of Western capitalism is inequality because it would not work otherwise. Is that not in some ways unsurprising?

Professor Wilkinson: My work is about the differences in the amount of inequality in different market democracies and also different American states. I am not talking ever about any sort of perfect equality that does not exist. It seems to me that the research shows quite clearly that more equal societies, like the Scandinavian, the Nordic countries and Japan and also the more equal US states, have fewer of almost all of the problems that have social gradients, including ill-health. It is not about getting rid of inequality as a whole, and of course our kind of society will inevitably generate substantial inequalities whatever we do, but they are small in some countries than others and that matters.

Q164 Dr Stoate: Except that in Sweden, for example, health inequalities are still persisting, despite sort of upstream interventions and more rigid policies. They have still got the same inequalities that we have.

Professor Wilkinson: Yes, it is a complicated picture because it looks now as if more equal societies have better health overall. It does not just help the people at the bottom. It is not simply that there are fewer poor people in a more equal society. It is that, on whatever given level of income you are on, you do better in a more equal society. The effects of greater equality in most of the analyses, not all, seem to be bigger at the bottom than the top, but whether they reduce health inequalities, I think, is largely a matter of whether you express those relatively or absolutely. If you have got a little graph of health inequalities, even the people who are already the healthiest do a bit better and the others do better still, so the curve goes down from a steep one high up, but still the difference between one group and another may remain a two-fold difference, but, in terms of the absolute number of deaths per 1,000 or whatever, the difference in most analyses suggests that greater equality in a society as a whole makes those differences smaller.

Q165 Dr Stoate: So really you do believe then that, if we could just simply level our socio-economic inequalities, we genuinely would improve health inequalities?

Professor Wilkinson: Well, we have looked at this in not only the rich, developed societies, but, as I say, in a quite different test-bed, the American states, and it is not just our work, there have been close to 200 papers now looking at that issue. In the papers which measure inequality and health over large areas, like whole societies or whole states, 70 or 80% of them find support for this relationship. That is not true when you measure inequality within small, local areas because of course a deprived neighbourhood does not have bad health because of the inequality within it, it has bad health because it is deprived in relation to the rest of the society.

Q166 Dr Stoate: Just to follow on from that, is there any evidence then that in non-Western capitalist countries the picture is any different?

Professor Wilkinson: Margessen(?), a long time ago, did a paper looking at improvement in mortality over a ten- or 20-year period using data from the 1960s to some time in the mid- or late-1970s. He said it was surprising how well communist societies do. That, I think, was true until around 1970 actually when suddenly the communist societies, at least in Europe and the Soviet Union, had ceased to improve. However, if you are talking about Third-World societies, there are analyses looking at infant mortality and life expectancy which suggest that it makes a difference there too. The important factor which, I think, distinguishes between the problems in rich and poor countries is that the GNP per capita matters very much in poorer countries, whereas amongst the richest 30 or so countries it seems to make very little difference to which country does better than another in terms of mortality.

Q167 Dr Stoate: Just from the point of view of government policy, is there any actual evidence that things like income tax changes or child benefits actually can narrow these gaps? We all assume that, if we give out these things, it is going to help, but is there any actual evidence for that?

Professor Wilkinson: There have been one or two experiments with negative income tax in the States and they were a long time ago, but they did show improvements, one which particularly addressed early life and pregnancy and tried negative income tax for poor mothers. One is concerned, I think, not simply with income as allowing people to buy more of whatever they need, but what inequality does to the whole culture of a society, so a lot of the concern that we have been reading, and people have, about the changing culture in Britain, I see as the long-term results of the rises in inequality which took place, particularly in the 1980s and early-1990s, so people, even if you improve their income, are part of that same culture where there is a lot of status competition and so on. That is what one has to change and that means changing the income inequality in the whole of society.

Q168 Stephen Hesford: What I was going to ask, and I was going to ask Michael about this, touches on the conversation that has just been had, but in terms of reducing child poverty, if you attack child poverty, is it basically an income issue and, if it is, if you tackle the income issue, will that reduce health inequalities?

Professor Sir Michael Marmot: It is a very important question and perhaps I could, in answering your question, make a comment on the previous question about what constitutes evidence and the chain of evidence. Overwhelmingly, we are not dealing with, "We've done the experiment, we can look at the result on health inequalities" because we just do not have that sort of evidence. You ask the right question, but I cannot give you the right answer because it does not exist, by and large. What we do know, I think, is two clear things. One is that policy changes poverty levels and, secondly, poverty is not only about income. To illustrate both of those - and I sent in the presentation that I made to the Secretary of State's expert group, advising him on his health inequality review, so I sent in some of the data that I had put to the Secretary of State's group - if you look at child poverty, it relates across Europe and developed countries. It relates directly to the generosity of family policies, that in the Nordic countries family welfare policies are more generous either because they promote two-earner families, they allow for both parents to go out to work, or because they actually have a high level of ----

Q169 Stephen Hesford: You mean childcare issues?

Professor Sir Michael Marmot: Yes, but also the level of the childcare issues and child support, so the family policy, because it is more generous, leads to lower levels of poverty. I think it is an entirely reasonable chain of evidence to suggest that a reason for the very low infant mortality rates in the Nordic countries relates to the fact of fewer children living in poverty. We do not have that direct evidence, we do not have the trial evidence that, if you change the benefit strategy, you will drop infant mortality, but I think it is a reasonable chain of evidence that government policy can affect poverty, and poverty levels do relate to outcome of pregnancy, infant mortality and child mortality, ergo, that is a reasonable chain. Now, I said that poverty is not only about income, and it relates to what the previous panel said about the importance of early child development, early child development is absolutely key in this and we know that early child development is not only related to income, but it is harder to give children what they need if you are strapped by a lack of cash and the inability to do things properly. To take the Nordic example again, the data show very clearly that, if you look at a measure of outcome of education, the literacy levels of young people, the higher the education of parents, the better the literacy level of their offspring, and it is a gradient. In Britain, that gradient is quite steep. The lower the educational level of parents, the lower the literacy level of the offspring. In Sweden and Japan, that gradient is very shallow; the education of parents makes much less difference to the literacy levels of their offspring and, hence, the life chances of their offspring in Sweden and Japan than it does in this country, and Canada is somewhere in between the US and Sweden.

Q170 Stephen Hesford: Why?

Professor Wilkinson: I think that, when people have studied this, looking at literacy levels as an outcome, they see three key classes of influence: the family; the wider level of degrees of deprivation; and the school is the third in this. The family is very important, the level of deprivation, which is presumably some measure of the general social environment and culture, to use the Chair's term from earlier, and then the school. Now, we know that children with poorer early child development do worse when they get to school and so do some of the problems, and that is why Surestart, and its progenitor in the US, Headstart, was potentially so important because it is trying to help deal with some of these issues before children get to school. Parental involvement, that is what the evidence shows, is very important. I am not an old-style 'family first', but the evidence shows that the family is really very important. There have been key studies of looking at literacy levels related to parental involvement in children, how much they talk to children and the literacy level of their offspring in relation to all of that, so caring, for the parents, is really highly important. The level of deprivation of the area, I think, functions both as peer-group effect and, I think, also relates to what Richard said earlier, the degree to which people are on top of their lives, which is a graded phenomenon, and what they give their children, I think, is related to the degree to which they are on top of their lives, and then of course the schools. You have got a lot of issues to deal with, but I think the educational gradient in outcome, the educational performance by area of deprivation is an absolute scandal. We do terribly poorly in this country. The fact that you can predict how poorly children are going to do just by simply knowing where they live and knowing the level of deprivation, it does not have to be that way and it is an absolute scandal that it is and that is clearly going to relate to children's life chances as they become adults and it is clearly going to relate to their health and health inequalities.

Q171 Stephen Hesford: In terms of what I will describe as 'poverty of opportunity', education and circumstance versus being on top of one's life, being able to cope, what is the more important? If you had to pick, what is the more important aspect or the quicker-win aspect, if you had to tackle it?

Professor Sir Michael Marmot: I think they are very closely linked. I think being on top of your life relates to the circumstances in which you are born, grow, live, work and age. Circumstances matter enormously, so poverty of opportunity is clearly linked to poverty of having control over your life. You cannot have control over your life if you do not have the opportunity to exercise it, so I think they are really very closely linked.

Professor Wilkinson: It seems to me that one can separate out family, neighbourhood, society and so on, but family is shot through with the wider society. I do think that in the past we have not, and I do not like to use the term 'joined-up thinking', but I suspect that we are, as human beings, highly sensitive to social status. We have evolved to be attentive to social status because it is affected throughout our evolution, reproductive opportunities and also access to scarce necessities, resources. Also, I think, similarly, we are highly sensitive to issues to do with friendship which are also a very important protective of health. Parenting is partly a matter of preparing children for the kind of society they are going to have to deal with. Poor parents affected by low social status pass on an experience of adversity to their children, and we know that there are very important processes affecting stress responses that go on through maternal stress in pregnancy and in early childhood, affecting how people behave, their cognitive development and their stress responses. It is not an accident that we have that sensitive period in early life and what it is doing is preparing us for the kinds of social relations we are going to have to deal with, so the parents' experience is passed on to the child, and part of the question we were dealing with earlier to do with lags, you need both the parental experience of a different kind of society and early programming to change, so there are always going to be two stages in how a society responds. There are going to be the short-term effects as we pick up that we are in a different kind of social environment, that we have different sorts of opportunities, and we are going to have the longer-term effects as one generation comes through having had a different childhood experience. The issues to do with opportunities, what little data there is that allows us to compare social mobility in different societies, there is some consistent data from about eight different countries and it is highly related to income inequality. The more unequal societies have much less social mobility, particularly in Britain and the United States, whereas again the Nordic countries have high social mobility. If you have got bigger differences, the cultural differences, so the social distances are also larger, and the downward social prejudices, the class prejudices, I suspect, are greater, so social mobility opportunities are reduced.

Q172 Stephen Hesford: That is extremely helpful because that comes to my final supplementary. On income inequality, for the reasons that you have just suggested and I know, there is a little tension here about the importance, so how do you usually measure income inequality to make sense of this?

Professor Wilkinson: There are lots of different measures and they are highly correlated. The one that I have used most is the ratio of how much richer are the top 20% than the bottom 20% and in some societies, Japan and the Nordic countries, about four times as rich, and in Britain, the United States and Portugal, it is more like eight or close to nine times as rich, and this is after taxes and benefits. People also use the Genie Co-efficient, we use that in our American data whic