There have been many studies of genetic risks for smoking, including GWAS (Genome-Wide Association Study) investigations. GWAS studies are ‘hypothesis-free’, that is, they take a given phenotype (e.g., initiation of smoking) and ask which of thousands of markers for alleles are associated with that phenotype. The results of GWAS research are frequently disappointing. They typically reveal few if any previously unknown genetic predictors of the phenotype. There are at least two reasons reasons for this. First, when you examine thousands of alleles that may be associated with a phenotype, many apparent associations will occur simply by chance. Second, the effect of a single allele is almost invariably small.

Caspi and his colleagues attacked this problem in two ways. First, they sought to overcome the “chance association” problem by using only markers for alleles that have been associated with smoking in multiple previous GWAS studies. Second, they dealt with the “small effect” problem by creating a polygenic risk score that combined information about many alleles into a single score. (The score was simply the count of the number of smoking risk markers.) The researchers then looked to see if the risk score predicted the development of smoking in a longitudinal data set.

The researchers have followed a cohort of more than 1000 New Zealanders from birth to age 38. From 12 interviews with each participant over these years, they have extensive data on their smoking, including how they started and how much they have smoked over the years. They conceptualize the development of a smoking habit as having three stages: initiation, conversion to daily smoking, and development of heavy nicotine dependence. They also have extensive genetic data which they used to calculate a polygenic smoking risk score for each study participant. They then looked at whether the genetic risk score predicted the smoking trajectories of the participants.

What they found was that polygenic risk had no association with whether the participant started smoking, but participants with a higher risk score ( = more risk alleles) were more likely to become heavy smokers if they did start.

In Panel A of the Figure, the bar graph shows the distribution of the risk score (shown on the horizontal axis) in the cohort. Notice that there are very few people with very high or very low risks. This bell curve pattern is likely to characterize many genetic risk patterns (because your genetic risk score is, in effect, the result of many nearly independent coin tosses). The superimposed scatterplot and regression line shows the relationship between the risk score and the total smoking by the participant up to age 38 (shown on the right-hand vertical axis). This results demonstrate that there is a clear dose-response effect: the more genetic risk factors you have the more cigarettes you are likely to consume. However, it is only a correlation of r = 0.12, so polygenic risk explains about 1% of the variance in cigarette consumption in the cohort.  Panel B shows that persons with high polygenic risk participants were moderately more likely to become nicotine dependent if they start smoking. Panel C shows that high risk participants who stopped smoking were slightly more likely to relapse.

The researchers were fair and judicious in their comments about what this means. On the one hand, they have evidence that this set of alleles affects mechanisms that lead to nicotine dependence. On the other hand, they acknowledge that the size of the association isn’t large enough to be useful for public health. In particular, I do not think it would make sense to routinely sequence the genes of adolescents to counsel them about their risk of becoming nicotine dependent. Some New Zealanders with no risk alleles smoked anyway and quite a few with all the alleles never became dependent. These gene risk factors just do not predict enough about smoking behaviour.

OK, but perhaps that is just the current state of the game in polygenic risk assessment? Could be, but I am skeptical. First, the effects of individual alleles are typically small, so finding a few more markers is unlikely to matter too much. Second, there is a diminishing returns process for the value of additional information, so you might have to quadruple the number of markers to double the predictive power. My belief is that this pattern will be typical for polygenic measures designed to predict individual behaviour. Genes will matter, just not very much.

However, it might be the case that we could combine polygenic information with phenotypic information to produce a substantially more powerful predictive score. For example, adolescents’ self-reports about whether their peers smoke help predict whether the adolescents will smoke. So a risk score that combined genetic information with data on a kid’s behavioural history might give us a powerful predictive tool.

Here again, though, the obstacles are daunting. Developing predictive tools combining multiple genetic and behavioural factors will require massive data sets. You can imagine mining this information out of very large electronic health record data sets. However, these data sets do not currently have data on smoking or other risk behaviours of sufficient quality to support instrument development; and it will be a long time before they do. We are a very long way from being able to predict individual behavioural trajectories from someone’s genes.

Do people not care about inequality, or do they not know the facts?

To test this hypothesis, Ilyana Kuziemko recently conducted an online experiment using members of the Amazon Mechanical Turk community (essentially an online labor market where individuals complete short computer-based tasks for negotiated wages). 

The treatment group received an “information shock” where they were told where they fall in the current income distribution, and what their household income might be if income inequality had not increased between 1980 and 2010. (Recall from a previous post that Americans underestimate the extent of income inequality).

The control group did not receive an information shock. The treatment and control groups then answered several questions about how important a social problem inequality is, and preferences for redistributive tax and transfer policies.

The main result is that across political affiliations and income profiles, people exposed to the information shock express much greater concern about income inequality (40 percent greater), but generally do not support more redistributive policies. Liberal-conservative differences on issues like raising taxes on millionaires hardly budge. They also do not have a huge effect on support for anti-poverty policies (EITC, food stamps, minimum wage).

The one main exception was that support for continuing the estate tax (which almost exclusively falls on very wealthy households), was greatly increased in the treatment group. Here’s one intriguing hypothesis offered by the authors:

“Correcting wildly misinformed views on welfare was not sufficient to change respondents’ support, though perhaps the lack of elasticity is due to the racial stereotypes the word “welfare” bring to mind (Gilens, 1996). It may be the case that the estate tax is one of a few issues on which voters are highly misinformed but is not linked to racial or other stereotypes.”

The other important thing to note is that trust for government was relatively low in both the treatment and control groups – but paradoxically trust for government decreases in the control group even as they profess greater approval for the government taking a more active role in addressing income inequality.

“In general, providing information about the growth of inequality and the ability of the government to raise taxes and redistribute have complicated effects on views of government. It appears to make them simultaneously see more areas of society where government intervention may be needed but simultaneously make them trust government less.”

What do we make of these findings? It suggests that working on anti-poverty policies may be a harder slog than anti-inequality policies. It also could suggest a vicious cycle: people feel less empowered and more distrustful in the face of growing inequality, even as they express an abstract desire for the government to do something about the problem.

Although it’s not ironclad evidence, the authors show that willingness to send an email to a U.S. Senator addressing the estate tax does increase substantially (although they do not have evidence that sending the email actually takes place). Giving people “voice” is in some respects harder in the current era, in which government feels very distant and unapproachable to most citizens. Finding small, but meaningful ways, to get people involved in the conversation on inequality is an important direction to pursue for future politics. Rather than concluding that the public will never support redistribution in policy, those who seek to reduce inequality might try to help voters to see a path that links their political participation with more redistributive policies (and conversely to punish elected officials that do not listen to constituent preferences on redistributive policies).

Timely as ever, I thought I’d finally get around to writing something about this Reinhart & Rogoff business. If you’re reading this blog, there’s a good chance you’ll be familiar with the story – a while ago, two high profile economists, Carmen Reinhart and Ken Rogoff, announced a finding with big implications for the austerity debate. They claimed that there was a strong negative relationship between national debt and economic growth. And further, that there was a pronounced ‘cliff’ at a 90% debt-to-GDP ratio – if a country crossed over this debt threshold, its prospects for economic growth were suddenly pretty dismal.

This finding was obviously of great interest to policy makers. If a high debt-to-GDP ratio strangles growth, then this is a big strike against the case for borrowing to stimulate the economy. Reinhart and Rogoff were invited to discuss their results with senators and congressmen, and were cited approvingly by politicians in the US and Europe.

That was until recently, when the story broke that they’d made a mistake in their calculations. Thomas Herndon, a graduate student at the University of Massachusetts, Amherst, when attempting to replicate their results for a class assignment, found that he couldn’t. It turned out that Reinhart and Rogoff made a basic data error in calculating average debt and growth figures for the countries they were analysing. Instead of selecting 20 countries, they’d accidentally omitted the first five countries alphabetically. As well as this, Herndon and his professors questioned the way they’d weighted various country’s data in the analysis. Basically, in Reinhart and Rogoff’s original analysis, any occurrence of negative growth in a country counted for as much as any other. So an economic contraction that lasted one year counted for as much as one that lasted for ten or twenty.  An alternative weighting meant a much weaker relationship between debt and growth and, crucially, no 90% ‘cliff’.

After all this came out, a lot of stories focused on how Reinhart & Rogoff’s original paper was never actually published in a peer-reviewed journal. A recent story on the always excellent More or Less emphasised the fact that politicians shouldn’t rely on unpublished work that hasn’t had a chance to be subjected to proper academic scrutiny. This is certainly true, and it’s worrying (though not at all surprising) that politicians weren’t quite smart enough to ask how country’s data were weighed in this very simple analysis, but were easily morally bankrupt enough to characterize a single unpublished study as a “widely acknowledged” fact, as did (EU economics commissioner) Ollie Rehn. However, what also worries me is that ‘proper academic scrutiny’ might not have solved this problem either.

In a generous mood I’d grant that peer-review, or subsequent academic criticism might have caught the weighting issue. This is the kind of thing that tends to be noticed by academics who disagree with someone else’s conclusions. However, what Herndon did was something that almost no-one does any more – he tried to replicate an existing finding. Despite the fact that one of the basic principles of the scientific method is that your results should be replicable, it’s very rare that an academic will go to the trouble of trying to reproduce an existing result using exactly the same materials and methods. This is true even in economics which often relies on existing, publicly available data – which should make replication pretty straightforward.

The main reason for this parlous state of affairs is that, if you’re an academic, replicating other people’s work doesn’t get you anything. You’d struggle to get anyone to give you money to repeat something that’s already been done, and no peer-reviewed journal would publish your results even if you did. Career-wise, replicating an existing finding will almost always be a complete waste of time – time that could be spent doing your own original work.

A second, maybe slightly less important reason we don’t see more replication is that academics don’t share. At least they don’t share as much as they should. In most studies in economics (and other fields that rely on existing data, like sociology and epidemiology), all it would take to replicate someone else’s findings would be access to their original ‘spreadsheet’ (the original data, along with a list of what calculations they did). The only reason Herndon found what he found was that Reinhart and Rogoff were kind enough to send their spreadsheet to him. A lot of academics wouldn’t do that. Why send this data, data you’ve likely spent months compiling, to someone who can just use it for their own research? Who’s to say they won’t use it to beat you to some interesting finding, setting back your own research agenda and putting you that much closer to being out of a job? On top of that, why show them your workings as well? When handing in, say a finished report at work, would you want someone else to see every step you went through to get to that final report? What if there’s an embarrassing mistake in there?

Reinhart and Rogoff’s embarrassing mistake got caught (if you listen to them, it didn’t change their results much anyway). But hundreds of mistakes like this slip through the academic net every day. They make it through into papers that may be cited hundreds or thousands of times; sometimes by policy makers or doctors with the ability to do real harm to real people. So here’s hoping the Reinhart and Rogoff case prompts a new enthusiasm for replication. But I wouldn’t hold my breath.

According to the website, the Oregon Health Study “is the first randomized controlled experiment to examine the causal effects of having some type of insurance coverage versus having no insurance at all.”  The findings, released a few days ago, have unleashed a storm of commentary on what the investigators did and did not find in terms of links between coverage and health outcomes.  Writing  over at The Incidental Economist, Harold Pollack quotes Joseph Newhouse for the notion that the “Oregon Medicaid experiment ‘is a Rorschach test of people’s views on the ACA.’”  I am going to try to defend that claim, although likely not in the way that good readers of Inequalities Blog might expect. 

So here’s the funny thing: even though I am an attorney, an historian, and a bioethicist who researches health inequalities, stigma, and social justice, I actually am less of an expert on the delivery of health care services than virtually every blogger here, and likely a goodly portion of the readership, too.  When interviewing for a job as a law professor some years ago, I was asked for my opinion on the fate of the ACA (then still in Congress), and I had to shrug and say that I really was not up to date on all of the provisions of the bill nor of its likely passage, nor of its potential impact.  (No, I did not get the job!).

This is not because of anti-intellectualism, I believe (and hope!).  This is rather because of my engagement with the overwhelming evidence that access to health care services is simply not a prime determinant of health and its distribution in human populations.  In a seminal 2007 essay in Health Affairs, Paula Lantz, Richard Lichtenstein, and the good Dr. Pollack himself note that “Lack of access to health care is not the fundamental cause of health vulnerability or social disparities in health” (p. 1256).*  The authors go on to warn of the limits of medicalizing health policy, and suggest that if we want to use laws and policies to improve overall population health and compress health inequities, we need to go way beyond simply expanding access to basic health care services.

I think this is an important point, and it is one I teach, speak, and write on in a variety of ways.  In fact, it is in some way a point of departure for virtually all of my work in public health ethics and public health law.  We have strong evidence, both historically and over the last half century, that access to health care services is only a relatively minor determinant of health and its distribution in human populations.  This evidence begins with the McKeown Thesis (in which the single largest recorded gain in life expectancy in Western history had little to do with organized medicine) and goes on through the Whitehall Studies (in which all subjects had access to basic health care services and stark social gradients in health were observed) and more than a few others.  I discuss some of the implications of this evidence here.

Ok, there is some debate on this subject; some argue that mortality declines in the past half-century are mostly attributable to medical innovation delivered through health care services.  I am dubious of these claims for a variety of methodological reasons (First, I think the studies tend to lump together public health interventions with clinical medical services.  Given that majority opinion on the McKeown Thesis notes that much of the mortality gain during the health transition is attributable to organized public health activity, we have every reason to believe such a lumping is unjustified.  Second, I think such studies do not, perhaps cannot, adequately control for the intergenerational effects of social and economic conditions over the life course).

But I am comfortable suggesting that the general consensus is that upstream social and economic conditions are by a fair margin more significant determinants of health and its distribution than access to medical services.

So, where does this leave us with regard to the Oregon Health Study? I have no idea, actually.  I am no expert on the bowels of Medicaid, and as I live and work in an extremely underserved region with devastating health inequities, far be it from me to claim that access to Medicaid in insignificant.  In fact, as I argue at length, we need to be very careful in the implications we draw from the above evidence.  First, it does not imply the moral insignificance of expanding access to health care services.  No matter how we structure society, people will fall sick, and I have no hesitation in saying that we ought to care for them.  But we should not confuse this ethical proposition with an empirical counterfactual, viz., that in a world in which we did expand such access we would see significant improvements in overall population health. Second, beware the false choice fallacy.  There is no contradiction between collective action on the SDOH and expanding access to basic health care services.

Does the Oregon Health Study support my claim above? Again, I have no idea.  And really, it is besides the point I am making here.  Even if it is the case that some kinds of access to health care has a positive impact on health outcomes — a conclusion I would of course never deny globally or even in any specific case – the weight of the evidence strongly suggests that if we wish to have a substantial impact in overall population health and in the compression of health inequalities, we will have to take collective and social action far, far beyond the provision of health care services.

The Oregon Health Study: definitely a Rorschach test.

* The choice of language of “fundamental cause” is important inasmuch as it invokes Link and Phelan’s fundamental cause theory of disease.  More on this in another post.

Thomas Insel, a psychiatrist, is the director of the National Institute of Mental Health. Through its funding of university research and its intramural labs, the NIMH is the leading engine of mental health research in the world. So Insel’s views matter. As it happens, he keeps a blog. You might expect that a blog post from a high government official would be sanitized and boring, but Insel’s post on 2013-04-29 was breathtaking.

Insel regrets the lack of scientific ambition shown in the “tweaking” and “modest alterations” of the DSM-5. What he is really concerned about, however, is not the details of individual diagnoses, but the architecture of the taxonomy itself. The DSM presents a world of discrete diagnoses such as “Attention Deficit Hyperactivity Disorder” or “schizophrenia”.  In the world as DSM sees it, you either have one of these disorders or you do not. Insel doesn’t buy it:

The weakness [of each of the editions of DSM] is its lack of validity. Unlike our definitions of ischemic heart disease, lymphoma, or AIDS, the DSM diagnoses are based on a consensus about clusters of clinical symptoms, not any objective laboratory measure. In the rest of medicine, this would be equivalent to creating diagnostic systems based on the nature of chest pain or the quality of fever. Indeed, symptom-based diagnosis, once common in other areas of medicine, has been largely replaced in the past half century as we have understood that symptoms alone rarely indicate the best choice of treatment.

These problems have been debated for decades and were recognized by the intellectual leaders of the DSM-5 project, including its chair (and my old boss), David Kupfer. Kupfer had hoped that the DSM would be able to shift to an ontology of mental illness characterized by continuous dimensions that describe the functioning of primary mental systems governing cognition, mood, and behaviour.  Instead of falling into one or more diagnostic categories, each of us would be a point in a space of mental functioning defined by these dimensions. The dimensions would be grounded in biomarkers and quantitative assessments of cognition and mood. The NIMH has also been interested in developing a dimensional system, called the Research Domain Criteria (RDoC) project.

In my view, there are many scientific and clinical reasons to believe that mental health assessment should be dimensional. However, neither the DSM-5 nor the RDoC projects have come up with a dimensional system that is ready for clinical use. Insel writes of RDoC that

It became immediately clear that we cannot design a system based on biomarkers or cognitive performance because we lack the data.

Similarly, the plan to develop a dimensional system for DSM-5 was a moonshot: a high-risk, high-reward, and high-cost effort. Unfortunately, it blew up on the launch pad. Not only did the projects require more data than could be collected in available time, but we may also need revolutionary advances in the many fields of neuroscience and psychology on which psychiatry rests.

Neither Kupfer nor Insel are giving up on the ambition to create a dimensional system. But what is the status of the about-to-published DSM-5, which retains the categorical system? Insel writes:

In this sense, RDoC is a framework for collecting the data needed for a new nosology. But it is critical to realize that we cannot succeed if we use DSM categories as the “gold standard.” The diagnostic system has to be based on the emerging research data, not on the current symptom-based categories. [Emphasis added.]

The quote in bold is extraordinary. “Gold standard” is a term of art in medicine. A gold-standard measurement system is supposed to be the reference used to anchor discussion and DSM has for decades been treated as the gold standard in American mental health research. Insel has withdrawn the NIMH’s scientific imprimatur from the DSM.

The NIMH’s views will not have a large effect on clinical practice. Clinicians currently using the DSM are unlikely abandon it without an alternative. The obvious current alternative, the World Health Organization’s ICD-10, is also categorical.

Insel’s is being candid about a debacle for the American Psychiatric Association. But this isn’t a scandal. Rather, it is progress for mental health research. We need to place mental health assessment on new foundations, but that does not imply that mental health research has no foundations. Science is just hard: you can’t easily test a theory without well-developed instruments, but how do you build a good instrument when you don’t yet have a solid theory? With NIMH’s support, investigators will have new freedom to shape research questions that can lead to progress.

I’ve written before about how I think a lot of people’s antipathy towards the benefits system comes from their ideas about the sort of people benefits claimants are. That they are a special, different sort of person that is unworthy of help. There’s a horrible sort of circularity to it – being the kind of person who claims benefits makes you exactly the kind of person who doesn’t deserve them.

Anyway, in reading around this subject for my day job, I came across some really fascinating work on something called infrahumanization. Infrahumanization is a very similar idea to the more familiar concept of dehumanization. In fact, it is explicitly framed as a milder form of this process. Infrahumanization means not going as far as denying someone’s essential humanity, but still denying them the full, subtle, complex humanity that ‘we’ possess. Groups that are ‘infrahumanized’ are not felt to have the same range of emotional experiences as other people. Specifically, while people find it easy to imagine them feeling ‘basic’ emotions like anger, pleasure or sadness; they have trouble picturing them experiencing more complex feelings like awe, hope, mournfulness, or admiration. This is a very important idea – if people are denied the same rich, subtle inner lives that we have, then it is that much easier to see them hurt and do nothing. After all, they don’t really feel it the same way we do.

This is a relatively new area of research (begun by two researchers at the University of Louvain; Jacque-Philippe Leyens and Stephanie Demoulin), and so far it has focused on people’s feelings about outgroups in general – mostly using ethnic or national groups. However, what struck me was how closely these ideas fit with the stereotypes people seem to hold about benefit claimants and ‘chavs’ (for non British readers unfamiliar with the term, a quick Google search will provide a solid, if unpleasant, education). They are widely pictured as grunting, shouting, semi-human oafs. In mental terms mostly only of aggression and the sort of low cunning required to cheat the benefits system.

You can try it for yourself. Imagine the most stereotypical ‘chav’ you can. Imagine their clothes, their surroundings, their posture, their attitude. Now imagine them feeling surprise, anger, or fear. Easy right? Well now imagine them experiencing reverence, melancholy, or fascination. If you found that just as easy, congratulations. But I’d bet for a few of you it was just that bit harder. I’m ashamed to admit it was for me.

So far, there hasn’t been any direct research looking specifically at whether people find it difficult to attribute complex emotions to benefit claimants (in the UK or elsewhere). However, the research that has been done is quite suggestive. Some existing studies have found that outgroup infrahumanization is stronger when the outgroup is of lower status (see here and here – journal subscriptions required). Most strikingly, there has also been mixed neuro and social science work at Princeton showing that the groups people find it most difficult to imagine experiencing complex emotions are those low status groups who are both disliked and disrespected (such as drug addicts and the homeless). This is a label that clearly applies to benefit claimants.

 While we should wait for more direct evidence, I still think that infrahumanization is something we should bear in mind when thinking about people’s seemingly bottomless appetite for policies that harm benefit claimants. If people really don’t think of claimants as experiencing life with quite the same richness that they do, then combating such policies is going to be very, very hard indeed.  

In a new article in Health Affairs, a group of sociologists and epidemiologists make a strong argument for adding an independent oversight panel to the DSM process. The panel would monitor social and environmental factors that shape diagnosis and treatment of mental disorders, and make recommendations for tweaking diagnostic guidelines or for initiating future research on the determinants of mental illness.

These recommendations are offered as a counterweight to some troubling trends. Diagnosis rates for some mental disorders have been soaring since the last revision of the DSM in 2001 – and the prevalence of these disorders has been unequal across social groups in the United States. I recently discussed the growth of ADHD diagnosis as an important example, where I argued that fundamental issues of causation and even prevalence are still unknown. It is known, however, that the risk of diagnosis is very different in the United States than in the United Kingdom, where the diagnostic criteria are generally more stringent. Diagnostic expansion is an important way in which the labeled and treated population grows over time.

Social context also matters. The DSM is fundamentally concerned with the neurobiological bases of mental disorders, and the authors of the DSM are mainly academic clinicians. Independent of biology, there are a multitude of factors that receive much less emphasis in the classification of mental disorders: cultural understandings of mental illness, changing social exposures (such as experience with combat that could lead to a proliferation of post-traumatic stress disorder), and the influence of major corporate interests, including the pharmaceutical industry. Anybody who watches television in the direct-to-consumer era realizes what a big business treating clinical depression has become for drug makers.

To what extent is the DSM to blame for these problems? And is there a better alternative to the current framework? 

Problems: Diagnosis or treatment?

There are good reasons to worry about the influence of the DSM. After the release of DSM-IV, drug makers created checklists of the new ADHD symptoms for elementary school teachers, which encouraged them to identify children with the disorder. Teachers are the main pathway into a doctor’s office, but they are not particularly good judges of ADHD symptoms (nor probably other behavioral disorders).

On the other hand, it’s not clear that the changing guidelines are to blame, so much as how the guidelines are applied by clinicians. Primary care physicians, rather than psychiatrists and psychologists, are now on the frontlines of mental health treatment in the United States. While most primary care doctors have a general familiarity with the symptoms of common mental disorders such as unipolar depression, ADHD, and anxiety, they do not have the resources or time to fully and effectively evaluate mental disorders. Again, sticking with the example of ADHD, the DSM-IV advises that children must be assessed across multiple domains (such as school, home, etc.). A thorough ADHD evaluation should include observation-based work, paired with validated rating scales given to parents, teachers, and older children. These are simply not tasks that pediatricians are set-up to administer, and so the default method of diagnosis is often a short interview with a symptom checklist.

These are problems rooted in limited clinical capacity, so we need to ask how much a more socially contextualized perspective on mental illness would be able to help.

What’s the role of an oversight panel?

Granting that diagnostic guidelines do contribute in some way (and it may be more indirect, but no less important), what could an oversight panel accomplish?

In addition to suggesting new directions for research, of the most ambitious roles such an oversight panel could fulfill would be in recommending revisions to the DSM – which the taskforce would be required to consider – based on the appearance of over- or under-diagnosis of disorders within subgroups in the population.

They would also mediate controversies about the cross-cultural issues in psychiatry: “By attending to the impact of social environments on biology in ways that can cause local variations in the incidence of mental disorders, the review body would help develop diagnostic criteria that would be useful internationally and cross-culturally.”

My impression is that the mental health treatment community is looking to embrace alternatives to a purely neurobiological model of mental disorders. Even the DSM-IV made some progress in that arena. But they are likely to resist the incursion of “outsiders” such as epidemiologists that would like to rein in the development of new diagnoses based on the social and cultural ramifications of those diagnoses. 

This debate is going to rage on for some time, because at its core it asks the question whether the classification of mental disorders should try to accommodate all the “messiness” of how clinicians actually practice in diverse, multicultural settings under considerable pressures from special interests. Getting the theory of diagnosis to align with the practice of diagnosing is the next frontier of psychiatry.

But does it matter how we create this thing called ‘public opinion’? Or put another way, does public opinion about public opinion matter – should we care about what people think about what other people think?

Public and private opinion

One possibility is that our impressions of public opinion really do matter, because our sensitivity to social influence changes what we ourselves think (or are prepared to say). But equally, the views of distant others may be completely irrelevant to our own views of how the world should be.

This has been tested in a lovely experimental study that my Kent colleague Trude Sundberg passed on to me, by Sonck & Loosveldt 2010 using an online panel survey in the Flemish part of Belgium. S&L asked for their opinions on a series of deliberately-selected issues – the independence of Flanders, the adequacy of retirement income, whether immigrants from poor countries should be allowed to live in the country, whether living standards were going down, and about solidarity between healthy and sick people in the healthcare system. Crucially, a random subset of the sample were given actual information on the public attitudes of their fellow-citizens.  For example, on the Flanders independence issue:

• Some people were given the following information: ‘An opinion poll commissioned by VRT and De Standaard revealed in November that 12% of the Flemish people favour the independence of Flanders’’
• Everyone ranked on a five-point agree-disagree scale whether ‘Flanders should be independent’
• And everyone answered, ‘In your opinion, what percentage of Flemish people favors the independence of Flanders (0-100 percent)?’

The results are intriguing. For the Flanders question, people on average thought that about 45% of people in Flanders favoured independence – far more than the figure of 12% revealed in the survey. Among the group who saw the poll information, they reduced their estimate of other people’s opinion by about 4.5 percentage points (so they on average thought about 41% of other people favoured independence). And amazingly there was still a measurable difference in attitudes three months later when people were asked again (still at about 2 percentage points).

Yet when it comes to people’s own opinions on independence, hearing about the poll made no difference whatsoever, even at the time that people heard the information. This is the same across all these different areas – S&L consistently found that hearing about polls changes perceived public opinion in both the short-term and medium-term, but had no impact at all on people’s private views.

This isn’t a nail in the coffin for the idea that perceived public opinion affects private opinion – this is only a single study, and a meta-analysis (£) of this kind of study suggests that we are more likely to find an effect on new issues that people haven’t yet formed an opinion of. And sometimes effects are more complex than the simple form that can be tested in this kind of experiment. But it seems reasonable to assume that these effects aren’t that big.

Policymaking elites and public opinion

But if perceived public opinion doesn’t matter much, then why is it such a battleground?  Why are the newspapers full of polls that they have commissioned that show that most people agree with the political slant of that paper?  For example, in the recent debate about whether benefits should be increased at less than the level of inflation, right-wing people cited polls showing this was popular, left-wing people cited polls showing that the public were split (see my chapter in this).

Partly it might be because people enjoy feeling in the majority, even if this doesn’t change their actual views. (Given the intense slant of all British newspapers – many to the right, but some slanted to the left – you might think that the main job of British newspapers is to pander to their readers’ views and make them feel morally superior. Clearly their role isn’t to create a more truthful public debate – see this, this & this…).

But my feeling is that the main impact of public opinion is in policymakers’ perceptions about what will win them votes, and that THIS is what all of these polls are speaking to (rather than a direct effect on their readers). This is perhaps not such a surprise given all the discussions of focus group politics et al, but this is a necessary piece of ground-cleaning for considering how public perceptions of deservingness, fairness and inequality matter for actual outcomes – which I would argue that they do.

It’s an old complaint. So old it’s basically a cliché. And this is a big part of the problem. We’re so accustomed to the idea that politicians ‘lie with statistics’ that we’re no longer shocked by even the most egregious examples of the form. We just accept it as par for the course, when there’s absolutely no reason we should.

I’m going to pick on a single example here today, but know that I didn’t exactly have to struggle to find one. The present government’s courageous willingness to back up policy decisions with meaningless numbers means there’s a lot of low-hanging fruit for the amateur political fact-checker. This means I’ve developed quite a thick skin when it comes to dodgy statistics; but even I was stung when reading some recent comments from Conservative party chairman Grant Shapps. Defending the government’s welfare policy changes, here’s what he had to say: “Welfare makes up a third of this country’s spending, so it’s our job to make sure it’s getting to the people who really need it [my emphasis]”.

This number is not wrong –as in it’s not factually inaccurate. Welfare does indeed make up about third of total government spending. But crucially this is only true if you include a lot of things people probably don’t think of when you say ‘welfare’. Things like child benefit (for which any family with children with a partner earning less than £50,000 is eligible), tax credits for working people, and (the big one – around half the benefits bill) the state pension. With apologies for repeating a previous post, only about 2% of total government spending goes on welfare for unemployed working-age adults. Even if you include everyone on the Disability Living Allowance, this number only goes up to about 4.5%.

You might say this is just standard political practice – using any vaguely relevant number to back up a point. But does that make it any better? Given the context of the quote (the changes to welfare for unemployed and disabled people), Mr Shapps must have known how the figure would be interpreted. He must have known people would think that the “third” referred to unemployed people. He’s not a complete idiot – he intended for people to come away with this impression. Because welfare cuts are easier to defend when people think we spend vast amounts of money on a group they think of as ‘scroungers’. This is not ‘cynical’ or ‘a dubious use of statistics’. It’s flat out intentional deception. It’s a lie, basically, and should be treated as such. Just because you use a ‘real’ number to do it doesn’t make it any less morally bankrupt.

The point is that politicians should be afraid to get caught doing this, but afraid they most certainly are not. This is a real problem for those of us who do research with an eye towards policy. What we do, essentially, is create numbers; facts (or as close as we can get) that we hope will make public policy better. But what is the point of this effort if this is how facts are treated by the people who make the policies – their strategy is “Policy first, then just stick in some numbers to make the point, no matter if they fit or not”?

I don’t know what we do about this exactly. Except that we should make a fuss. As many people as possible should write them letters, go to their surgeries, and generally give them a hard time. Basically, we should try our best to make “Welfare makes up a third of this country’s spending” as much of a scandal as “I did not have sexual relations with that woman”.

Update: After writing this post, I noticed that Declan Gaffney and Jonathan Portes have written something making broadly the same point on Guardian’s Comment is Free. Their example of number abuse is from the government’s claims about the ‘rush’ of benefit claimants trying to avoid the new, more stringent assessments for disability benefits. What did I say about low-hanging fruit?

“I think it’s important to understand it does not give anything.  It allows people access to the legal immigration system… The only thing you are earning here is an opportunity to apply for temporary status, and ultimately, potentially to apply for a green card, the way everybody else does.”

What people do not get is access to public programs. No food stamps (SNAP), no Medicaid nor Medicare, no social security. This continues federal policy established under the 1996 welfare reform law, which restricted public benefits to citizens and lawful permanent residents. In practice, the Senate proposal means that people who are undocumented right now, could not receive any public benefits for over a decade (the likely amount of time required to meet all of the requirements established under the new law).

I have very mixed feelings about this as a matter of moral values and policy.

The “wait your turn in line” argument has a powerful sway on how Americans think about immigration policy: undocumented immigrants should not have the opportunity to jump to the front of the line because they defied border control or entered the country legally and overstayed their visas. As a matter of basic fairness, I can see the deep appeal of this argument.

The reality is not so clear-cut, however. Parts of the American labor market, including meat processing and agriculture, have operated for decades with large populations of undocumented immigrants. These industries could not exist in their current form without the presence of cheap labor, a fact that is known by both employers and immigration authorities. If society has been taking advantage of the labor of undocumented immigrants, the line jumping argument becomes less persuasive. With a whisper and a wink, undocumented immigrants have always been whisked into the line. The implicit message has been that Americans have been happy to accept the benefits of undocumented labor, but not accept any of the burdens.

The rewards of American society for undocumented immigrants have been modest at best. Beyond minimum wage or less compensation in the labor market, immigrants rarely receive pensions or other private benefits, nor do they qualify for public benefits. In fact, most undocumented immigrants do not intend to stay in the United States for a long period of time. As Doug Massey and Jorge Durand explain, however, the presence of long-term undocumented populations in the United States is partially the perverse consequence of tighter border security. Rather than risking apprehension on a future visit, would-be seasonal migrants have made the calculation that it is safer to stay in the United States.

As the undocumented adults have stayed longer, they have also had families. This is a big challenge in the debate over undocumented immigrants. Any policy that impacts adult undocumented immigrants also impacts their children, many of whom are citizens by virtue of their birth in the United States. Getting legal status for parents will have a huge and beneficial impact on bringing those families out of the shadows – giving them better access to public insurance and other programs (U.S. born Latino children remain the most likely to be uninsured demographics). But this is just a first step. If we want to improve the wellbeing for children of immigrants, we need to consider the needs of their parents too. This could mean making a tough choice to expand public benefits to undocumented immigrant adults in order to increase their health and nutrition, and by extension to help their children.

There is yet another dimension to the debate over public benefits for undocumented immigrants. Because Congress steadfastly refuses to expand federal health and welfare benefits for the undocumented, states and localities are left to pick up the burden. For example, the Affordable Care Act categorically restricts access to subsidies and public insurance to permanent legal residents and citizens. Public hospitals and clinics, largely supported by local communities, will be left to provide coverage for undocumented immigrants when they ultimately need medical care. This problem has become deeper because the ACA also cuts subsidies to safety net hospitals. Hospitals in areas with large undocumented populations will be in great financial danger unless the federal government provides supplemental funds to help those community hospitals survive.

It is quite possible that the ACA could never have passed with health care benefits for the undocumented, but t leaving immigrants out of public health benefits also creates severe political and social consequences. Randy Capps and Michael Fix describe some band-aid solutions to the problem in a recent article, including greater access to community clinics. This may be all that we can hope for within the current political climate, at least for now. In the long run, bringing more immigrants into the legalization tent could lay the foundation for expanded public benefits, especially as the argument that they have paid their fair share becomes harder and harder to deny.