Monday, 9 December 2013

Finns teach Finns rather well. Can they teach us?

All Finnish primary school children seem to look particularly fresh-faced, healthy, and actively engaged, a bit like their teachers, actually. The head teacher tells me (don’t the head teachers look young), the pupils at Ressu comprehensive school in Helsinki are not from particularly privileged backgrounds. The catchment area has three groups: academics, artists, and working class. The head teacher is polite, well informed, generous with her time, and very indulgent, given that I am not the first to make a pilgrimage to pray at the shrine of Finnish education. Ever since the PISA (Programme of International Student Assessment) showed Finland’s 15 year olds scoring particularly well compared to other European countries on standard tests, people have been wondering how they do it.

The UK made its reputation as being a brainy country – a great many discoveries, wonderful creative arts, brilliant universities. Despite that, we don’t do very well on PISA tests. The US is no slouch when it comes to innovations and skills, either. Why do the Finns do so brilliantly and the British and Americans so badly on internationally standardised tests? It’s quite a neat natural experiment because the UK and Finland do things so differently.

PISA scores in Finland (blue) and the UK (red), 2012[1].


The current approach in the UK to improving education for our children, is for the relevant minister to take central control: abolishing meddlesome local authority control of schools, setting the national curriculum, setting out a programme of school visits by inspectors to make sure they are up to scratch (the schools not the inspectors – might be interesting if it were the other way round), setting the criteria for what children should be able to achieve at various stages of their school career, examine at each of these stages, publish league tables of school performance to name and shame schools; abolish minimum qualifications for teachers in order to stimulate creativity. Oh, and freeze pay of teachers and criticize them as obstructing progress. Throw into the mix repeated discussions of whether abolishing grammar schools in favour of comprehensives didn’t rather lead to mediocrity. One last thing: make everything that goes wrong the fault of the last set of politicians to be in charge, whoever they were.

Finland has followed pretty much the obverse. There is a national curriculum but teachers have a great deal of autonomy in deciding what they teach; schools are under the control of local authorities; there are no tests to see if children have reached the relevant competency at various ages; there is one national test at the end of nine years of basic education in comprehensive schools but the results are not published, not fed back to pupils, and are used by the schools for statistical purposes so they can see how they are doing compared to other schools; all teachers have a master’s degree – unthinkable to have less; teaching is highly prized and teaching jobs are much sought; teacher training is research-based so teachers are encouraged to develop an enquiring approach to sort out solutions for children who are having difficulties.

The Finnish system is based not on managerial control of teachers but on teachers taking responsibility and students taking responsibility. The lack of emphasis on national exams represents not only a judgement that they are not necessary for high levels of performance, witness the high PISA scores, but they represent too narrow an approach to education. Finland is interested in educating future citizens who know how to work with people from diverse backgrounds, are educated in music and culture, have learnt traditional Finnish skills such as working with wood and textiles, and know how to cook.

I put two counter-hypotheses to my hosts – the head teacher and an official from the teachers’ union – one is that school reduces the effect of social disadvantage on educational performance. The other was that the school amplifies the effects of social advantage and disadvantage because the kids from more stimulating backgrounds are better placed to take advantage of what schools have to offer. No question in the teachers’ minds: Finnish comprehensive schools reduce the effect of social disadvantage on educational performance. Children who are having trouble get special attention, up to 30% of pupils. Equality is the most important word in Finnish education.

I suggested that if educational performance follows a bell-shaped curve, their special attention to the children in trouble may have reduced the prevalence of dumb-bells, but it may have reduced elite performance, fewer Nobels. (I am grateful to Helena Cronin for this terminology). They were not having any of that. The teachers work not only at helping the slower kids, but at challenging and extending the more able.

Might it not be the case that Finland’s good performance educationally may have less to do with schools and more to do with a relatively homogeneous population with low levels of child poverty? Typical academic’s question, wanting to isolate relevant variables. They chided me gently. You cannot separate the performance of the schools from the society and culture in which they are embedded. Schools are influenced by culture and society, and their mission is to contribute to society and culture in a positive way.

[1] Creating using the PISA ‘compare your country’ tool at

Monday, 29 April 2013

Debating the Evidence of Capital Punishment

Capital punishment terminates a human life. Is there any case at all for a doctor being in favour of it? Hippocrates; first, do no harm; do we really need to spell it out?

At a recent meeting of the Council of the World Medical Association – I represent the BMA there – there was debate over a resolution deploring capital punishment and, specifically, medical involvement. Most national medical associations were in favour of such condemnation, but not all. One appeal was to justice: he shot others, the argument runs, surely he deserves to die. Another appeal was to protecting the public: he is a terrorist, he bombed and maimed, unless we execute him, he’ll kill others.

My contribution to the discussion, fully ready to be criticised by the lawyers in the room, was that there are three possible reasons for the criminal justice system: deterrence, revenge, and protecting the public.

Taking them in reverse order: protecting the public. If a dangerous killer or rapist has been caught, the public need to be protected. Locking up the convicted criminal up is a way of doing that. It is not necessary to execute them in order to protect the public. In fact, in the USA, those handed a sentence of execution, can spend twenty years or more behind bars, as the appeals go on, so the main way the public is protected is locking up the criminal – even if the criminal has been sentenced to death. After such a time, and with the public protected, why go through with the execution?

Revenge. No doubt, advocates for capital punishment would not call it revenge, but closure, satisfaction or something of that nature for the bereaved or for society as a whole. But what else is it but revenge? Does not killing a killer mean that we have descended to his level? Eye for an eye is primitive, surely beneath a civilised society. If the organs of society do not feel moral repugnance at becoming killers, then we have a long moral argument ahead of us.

Deterrence. It could be argued that if capital punishment deterred others from committing murder, killing (executing) one person could lead to a net saving of life. It is a means and ends argument. The vile means of executing someone would be justified by the end of saving lives. The proponents of capital punishment could therefore put themselves on the side of preserving human life.

Quite apart from the question of whether doctors should be complicit in executions, what does the empirical evidence show on whether there is a deterrent effect? A friend told me that there had been several reviews of the deterrent effect of capital punishment on homicide. Some found convincing deterrent effect; others found none. My friend told me that ALL the reviews that found deterrent effect were by economists; all the reviews that found none were by criminologists.


Economists believe in incentives and rational choice. Some young man is feeling violent; he weighs up the costs and benefits of committing murder; and where the curves cross he acts. If capital punishment is in force, he thinks the costs of committing murder are too high and desists. Can you believe that?

Unlikely as it may seem that real people function this way, this view of how the world works shapes the interpretation of the evidence – i.e whether the data support a deterrent effect or not.

I had a quick look at the literature and found an interesting article (2013 Columbia Law School) from Jeffrey Fagan, Professor of Law and Public Health at Columbia. He dates a particular debate from 1975:

"...when University of Buffalo Professor Isaac Ehrlich published an influential article asserting that during the 1950s and '60s, each execution "…saved eight innocent lives" by deterring murder. Inspired by an economic model of crime developed by Professor Gary Becker of the University of Chicago, Ehrlich theorized that would-be murderers would choose between illegal and legal behavior based on the threat of execution."

Gary Becker, Nobel Prize winning economist, apparently suggests that people do make rational choices based on threat of execution. The debate now takes on a  polarised tinge: Cass Sunstein (of nudge fame) and Becker are convinced. Jeffrey Fagan, presumable closer to the “criminologist” dichotomy is most certainly not. Fagan reviews more recent studies that the economists find convincing. He rejects them on methodological grounds, among which are that the analyses did not consider other causes of fluctuations in homicide rates or the possible deterrent effects of imprisonment, as distinct from homicide. A quick look at other literature show that there are questions over the assumptions that went into the choice of instrumental variables, much beloved of econometricians. Essentially, an instrument should be correlated with execution rates but not with the “outcome”, in this case homicide rates. In that way, the independent relation between execution and homicide can be determined.

This all sounds rather familiar. I have written with passion about how economists analyse the same data we do and find the causal arrow runs from health to wealth and not, as we conclude, that features connected to socio-economic position are causal of poor health. (See: A continued Affair with Science and Judgements. International Journal of Epidemiology 2009). Same data; opposite conclusions. Here we are again.

I remain interested in why people, who purport to be empirical scientists, take such opposite views of what the evidence shows. It has to be related to starting assumptions. The interesting question then is why views on the deterrent effect of capital punishment (and the health damaging effects of social conditions) vary by academic discipline. Predictably, I side with the Professor of Law and Public Health and against the rational choice theorists, in finding the deterrent effect of capital punishment unpersuasive. But then, as a doctor, I find participation in legal execution to run counter to what we do. I am hopelessly biased by concern for human life. We all have biases that colour our view of the evidence. At least my bias does not lead to my advocacy of killing people.

Thursday, 4 April 2013

Social Movement and Swedish Paradox

Back in Stockholm. Whatever for? I was asked. It is true that I was in Malmo three weeks ago for the launch of the Malmo Commission on a Socially Sustainable Malmo, and in Stockholm in January. The answer is simple: our social movement on SDH is alive and thriving in local government in Sweden, and they wanted my continued input.

I went first to the Parliament for lunch as the guest of Anders Jonsson, a paediatrician who is a Centre Party MP, and secretary of the Social Committee. Paediatricians are naturals for social determinants of health because the effects on children are so plain to see. Anders Jonsson had invited other doctors who are MPs or otherwise involved in politics. Barbro Westerholm, who is a senior advisor on our European Review, gave a refreshing account of what it means to be an MP. She had been head of the Swedish Board of Health. After retiring from there she became an MP. But, she said, after a while she recognised that she had run out of fresh ideas and so left parliament to work with organisations devoted to the elderly. Working with Civil Society, she developed a bucket full of ideas, and so came back to the Parliament. 

(Good heavens! Fancy quitting being a member of parliament because you had run out of ideas. I can think of one or two who would serve our country well if they took such a view.)

I joked to the Local Government people the next day, that the Parliament is a fact-free zone – my powerpoint would not load at the post lunch seminar that I was giving to the Parliament’s social committee. So I had to ad-lib it or, as I put it, talk ideas rather than data. There is a recognisable debate in Sweden across the political spectrum about the role of the state and the individual, but it is different to the UK. 

I have been told that it was Labour’s Clement Attlee (couldn’t track the quote on Google) who returned from a trip to the US and explained to his fellow Labour Parliamentarians: They have two political parties in the US. The Republicans are a lot like our Conservative Party; and the Democrats are a lot like our Conservative Party. 

If the right-left debate in the US is to the right end of the spectrum, the UK debate is further toward the centre, and the Swedish debate is further to the left. It was put to me that none of the major parties seriously question the Swedish welfare state. The Social Democrats may have to have given way to a Centre-Right Coalition for two elections in a row, but the legacy of decades of Social Democratic government is more or less intact. 

And Sweden looks pretty good – on life expectancy, low levels of child poverty, relatively high equality on UNICEF’s Report card on children’s living conditions. And, of course, their economy is doing well.

But, as I have previously reported, questions have been raised about the magnitude of health inequalities in Sweden. The graph above, from Finn Diderchsen, using data from Johann Mackenbach’s latest effort, shows that the countries of Central and Eastern Europe have a high Gini coefficient – although not much higher than the UK – and high educational inequality in mortality; the Nordic countries have low Gini and low inequality in mortality. BUT, and this is the so-called Swedish paradox, health inequalities in Sweden appear not to be narrower than in other West European countries with higher Gini coefficients, and less generous welfare states.

Two comments. As we learnt from Olle Lundberg and CHESS: health of the most disadvantaged has been improving in Sweden. This is a societal success. Whether due to the welfare state, or not, it is a major societal success. One criterion of societal success is precisely improvement in the lot of the worst off, and Sweden looks good. But inequalities are increasing – this a second challenge that must be faced.

Espin Dahl from Norway points out that if you look at self-reported health rather than mortality, the picture is different. Now, he sees that the more generous is a country’s spending on welfare the NARROWER are health inequalities by education. 

More to do on this agenda.

The Local government conference – SALAR, Swedish Association of Local Authorities and Regions – was inspiring. 300 representatives of at least twenty local areas came together to make their commitment to pursue local policies for health equity, very much based on Closing the Gap in a Generation, the report of the CSDH. This IS our social movement in action.