Wellington School of Medicine Summer School  22 February, 2019.

But, first, you may well ask, why should anyone – especially a health practitioner – care about health economics? When I was teaching health economics to sixth-year medical students, I wrote on the board so that it was there during the entire session:



My students were uncomfortable with this – as I intended them to be. They took the view that they were there to act in the interests of their patients; why should they have to be concerned about these other things?

The role of a health economist is to explain why resources have to be taken into account, although today I will be working on a wider canvas because I shall also be talking about population-based healthcare as well as healthcare for the individual: in my professional lifetime the ‘treatment’ with the biggest health gain has been from reducing tobacco consumption.

There are many dimensions to health economics, many topics on which I have worked.

Charging for Health Care

My first memories of a health economic topic arose over the question of whether visits to the doctor should be free. That had been the intention of the First Labour Government when they established in 1938 what we know today as the public health system. (Incidentally I have never been disadvantaged by learning about the history of any topic I am studying.) The Labour government wanted all health services free but doctors claimed the right to continue to charge their patients – as they generally do today in general practice and private secondary medicine.

But should they? Isn’t health care fundamental to life? Why should patients pay? Actually a lot of things are fundamental to life, but we pay for them. Consider food. The difference is that food needs are regular, but health care is erratic. Some people can go for years without the need to see their GP, others may have to visit weekly.

An answer might seem to be insurance. In effect the welfare state was seen to be a form of social insurance but, for reasons that belong to a full health economics course, private insurance proves to be much trickier.

A related problem is where do you draw the line? Should free care cover dental fees or physiotherapy fees or aspirins and other OTC medicines or even, say, quack treatments.

Observe that charges are a rationing devise. You go to a GP if you value the service you are going to pay for, although in practice there other factors which affect your decision. That you go because you value the service more than you pay is also true when you purchase food – economics has put in a lot of effort to explain the exact mechanics.

What is different for healthcare, besides erratic demand, is that you go to the doctor find out whether you need treatment. Supermarkets do not advise you on your food needs. Sometimes it is said that GP charges discourage frivolous visits, but they may also discourage necessary visits – say for early identification of a condition such as cancer. Prescription charges are even more puzzling. A qualified medical adviser says one needs the drugs; why should you pay for them?

The charges seem to have been introduced to eke out the available funding for public health care. The alternative would be to raise taxes; in fact a prescription charge is a tax, a tax on the sick.

Not surprisingly, some people go without. A recent Ministry of Health Survey shows that 7 percent of adults and 4 percent of children leave a prescription unfilled each year because of cost. This affects women more than men, Maori and Pasifika adults more than Pakeha, the most economically deprived neighbourhoods more than those from the least deprived neighbourhoods. I find it very hard to justify prescription charges.

(Incidentally, when Muldoon proposed to introduce prescription charges in the 1970s we argued that it made more sense to tax drugs  like alcohol and tobacco which damaged the public health, Muldoon switched tack and did exactly that. But rust never sleeps and some years later the Treasury got its way by convincing some craven politicians that there should be prescription charges.)

What about the opportunity to purchase secondary care from a specialist or private hospital? The option was always there but the establishment of Southern Cross Medical Insurance in 1961 intensified the issue. It made it even easier to jump the treatment queue, that is to enable the patient to be treated earlier than more urgent cases waiting in the public line, especially if specialists divert their effort from public treatment to private treatment. The outcome had shades of a multi-class provision of healthcare which the First Labour Government was opposed to.

I have already said that medical insurance is a big topic which requires complicated analysis. But sitting behind it is the question of the level of public healthcare funding. Just how much does the taxpayer, rather than the sick, pay? The answer is ‘a lot’  – the health budget is one of the government’s largest outlays. Many people, including myself, conclude though, it is not enough.

Why don’t we increase taxation and use the additional revenue for spending on public healthcare (and other worthy causes)? The short answer is that  in New Zealand there is a strong aversion towards raising taxes.

You might think that in the case of healthcare everyone would be willing since they would be receive better healthcare from the public sector. But the costs and benefits do not fall evenly. Those at the top of the income distribution would pay more tax than the healthcare they would receive on average. They judge that they personally can get a cheaper deal by purchasing treatment from the private sector when they need it. Those on high incomes, a little below the richest judge, may be able to get a cheaper deal by purchasing medical insurance which will provide treatment from the private sector.

Before finishing this rapid overview of charging and overall funding of healthcare, let me mention a couple of other issues.

First, in 1938 the implicit view was that there was a limit to the totality of possible treatments. This was before the postwar explosion of pharmaceuticals and other novel treatments. Such is human inventiveness that there does not seem to be any ending of the trend, with resulting upward pressures arising from new treatments.

Meanwhile, new demands for care have arisen. Longevity increases the demand for care although we do not have a very good understanding of by how much. It is ironic that improved healthcare which eliminates deaths probably costs the nation more in further care in the longer lives.

We know less about the causes of rising obesity, of rising mental health demands and of rising addiction. On the other hand we have made good progress by reducing tobacco use and we have done much the same reducing the misuse of alcohol – although there remain serious problems here;

Changes in treatment approaches have also chewed up resources. For instance, while I support procedures to obtain informed consent, I recognise that takes time and adds to treatment costs.

The big takeaway is that unlike the 1938 vision, there does not seem be any practical limit to the resources we can throw at the health system. Yet, there is going to have to be some constraint on what is available. That means waiting times and the non-provision of low-return treatments in the public system and the limits that come from affordability in the private system.

Second, I was not involved in the establishment of the no-fault ACC system. In health system terms the scheme is interesting because it provides comprehensive coverage for certain treatments arising from conditions caused by accidents as well as prevention, rehabilitation and compensation.

It is frequently argued that we should extend the entire health system so that is run on the same principles as ACC. The historic record is that ACC was fundable because it abolished a lot of expensive litigation. There are no similar hidden funding sources for extending the scheme to sickness. To do so would require some form of taxation (or compulsory levies). Practically, ACC can probably provide such a good health service because it is only a small part of the total treatment system. Politically, the scheme has been greatly contested, especially by those who pay the levies, especially in the 1970s.

In the case of ACC there was the sneaking back of the fault notion in regard to treatment injury – when a treatment goes wrong –  quite against the spirit of the 1967 Woodhouse Report.

I was involved in addressing the anomaly in the 2000s. The criterion is now, basically, that the injury as a result of the treatment has to be unusual; no longer fault has to be proved. Court decisions mean the notion of ‘unusual’ is being reviewed.

What we were doing here was realigning the incentive systems towards the needs of the patients over any alleged failure of practitioners. Meanwhile there remains a parallel system for dealing with incompetence in a health professional. In my view the design is still not quite satisfactory because it does not give all the encouragement it might to the institution to provide a quality safety culture.

To summarise, funding the health system has substantial impacts on how the system runs; the arrangements affect not only the cost to the taxpayer and patient (and to the insurance system) but the efficiency of the system and the care people receive.

This is nicely illustrated by the US health system which is the most expensive system in the world and has some of its finest clinicians; yet its performance in terms of the health of the average American is mediocre by international standards. There are two reasons: poor system design (the topic of the next section) and poor public health (the topic of the third section).

Organising the Health System

The structure of the health system has been in almost constant flux. I rarely use the term ‘reform’. The word ‘change’ is usually adequate and it does not have any connotations of improvement. Too often a ‘reform’ has proved – prospectively and retrospectively – to result in a reduction in performance. I am going to devote this section to the health redisorganisation of the early 1970s.

There was with the 1991 Mother of All Budgets a ‘Green and White’ paper proposing major changes to the public health system. (The reason why it was combined green – for discussion – and white – where government decisions had been made – was that even the National Cabinet could not agree to some of its more extreme proposals.)

I use the term ‘commercialisation’ to describe the redisorganisation. At its core was the idea that the best way to deliver health care was by businesses competing in the market. It would not be possible to do this completely. The aim would be to commercialise as much as possible.

The redisorganisation was sometimes described by the critics as an ‘Americanisation’ of the health system. New Zealand’s healthcare culture is different from America’s with its more commercial focus. America’s is probably near unique among rich country health systems. The proposers of the redisorganisation wanted an even purer – more commercial – form of the US system.

This was not evidence-based policy for there was no evidence that the redisorganisation would give a better health system. Indeed, insofar as it was moving toward, and then past, the American system, the evidence was that it would get worse. Those involved had absolutely no experience in the disciplines most knowledgeable about the design of health systems. Those in charge of the change specifically excluded those with expertise. That included not bothering to consult some leading world experts who happened to be visiting New Zealand at the time.

The ignorance of those planning the change soon became evident. Sometimes the errors were laughable. One advocate confused an intensive care unit with a post-operation recovery unit. One hopes that had he a heart attack he would not choose where he was to go.

The promise was that these changes would generate a 20 percent productivity increase. That, the advocates said, would mean that 20 percent more health care could be provided on the current budget, although perhaps some hoped to cut public funding by that amount.

One of those working on the redisorganisation explained that he had worked on other industries and always got a 20 percent productivity gain. What he did not understand was that his earlier businesses had been capital intensive – say 80 percent of costs were related to capital inputs – but the health sector was 80 percent labour intensive. Given this sort of misunderstanding, little attention was paid to the labour input. Not only were the specialists in health systems analysis ignored but so were the health professionals.

This was nicely illustrated by one chief executive of a Crown Health Enterprise (notice the last word) who had been recruited from the business sector. He had a dozen reporting directly to him, only one of whom was a health professional. Most of the appointed chief executives were from business and were more comfortable with business people. The governing boards of the new entities were also stacked with business people.

I remember a chief executive talking enthusiastically how he expected his clinicians to join him to achieve his CHE’s objective. He was politely asked what the objective was, and the businessman said it was the ‘bottom line’ and then sort of deflated as he realised the implications. The clinicians he was talking to, did not give a fig for profits but had someone in the room had a heart attack you would have been clear that their objective was the welfare of patients.

The theory was that just as any economist was able to pontificate on the health service, any manager could manage a health entity even though they had no knowledge of health issues. We call them ‘generic managers’ and their cult still dominates Wellington today, although not so much in the health system. You need professional expertise to run it.

Given the disaster of the health redisorganisation of the 1990s, you may also want to dismiss economists as having nothing to say about health systems. It is generic economists you need to dismiss.

Health practitioners were outraged by the proposed redisorganisation and actively opposed it. One of the most important lobbyists was the Coalition for Public Health. I worked as a technical adviser for them. They could see that the proposed changes were out of touch with reality but they did not have the framework to understand what was going on, nor make sense of what the other side was thinking. Thus we health economists made our contribution.

Did the public opposition to the commercialisation had any effect? It was such a stupid proposal that it would have broken down under its own weight anyway. Probably the opposition’s effect was to slow down the changes and bring them to an earlier end.

People suffered. There are documented cases of people dying during the turmoil because of managerial mistakes. The falling mortality rate slowed, which indicates that there were many more undocumented cases of death and discomfort from the cutting back of resources and the organisational turmoil.

Within a couple of years the government acknowledged there would be no significant productivity gains. Even so, the momentum of the redisorganisation continued; even today the current government struggles with its legacy despite much having been reversed. Destruction is easy; recovery and creation takes decades.

It is said that success has many parents but failure is an orphan. Nobody today mentions their active role in the redisorganisation; some have rewritten history to disguise their involvement. Some involved in the shambles were kept on advising or promoted in the health system area.

The government suffered. Jim Bolger, prime minister of the time, said that National almost lost the 1993 election because of its health policies. (Probably what protected National was that the Labour Party played such a small role in the opposition to them.) Those opposing the changes may take a little pleasure in this, although I do recall any sentiment in the Coalition for Public Health that it was trying to bring down the government. The focus was on stopping the stupidity.

Another curiosity is that those who led the opposition have never had public recognition. In a healthier political world its generals – Alan Gray, Peter Roberts, Alistair Scott – would have been given knighthoods.

Legal Drugs and the Public Health

I now switch to a quite different area of health economics: public health policy illustrated by the regulation of alcohol and tobacco. By an unfortunate or fortunate accident I worked first on alcohol policy. Tobacco policy is much easier so I shall deal with it first.

The health advice is that one should not smoke. The more one smokes the worse for one’s health. Moreover – and this distinguishes tobacco from alcohol – the damage seems to be proportional to the quantity consumed.

The health economist asks ‘what is the problem?’. Smokers know the risks – they do now anyway following heavy publicity campaigns – and they choose to take the long-term consequences for the short-term benefits of smoking. People do that in many other areas – skydiving or walking across the road.

One answer is that some of the costs arising from the consequences of smoking are borne by others. Economists call this the ‘externality problem’. Much of the analysis of how markets behave well assumes that making a decision in one’s own interests does not reduce the welfare of others. If it does, then the price mechanism lacks some of the properties which make it so effective.

Examples of externalities include passive smoking (I help develop the smoke-free policy) and the devastation to family and friends from a smoker’s early death. The big one is that the health system incurs additional treatment costs from a smoker’s sickness which are not paid by the smoker. .

The last statement involves a trickier analysis than it seems, because if one avoids death from smoking one lives longer and incurs extra costs to the public health system later in life. The tobacco companies commissioned consultants to show that these later costs were greater. (To do this the consultants had to admit that the tobacco caused sickness and deaths, which they have not been keen to do.) The logic of their analysis is exploding cigarettes which kill smokers on the day they retire from the workforce. The paradox arises because we need to take into consideration longevity and the quality of life associated with it, something I shall talk about in today’s final section.

There is another tricky issue here, which is that economics tends to assume that people make rational decisions in their own interest. By way of background, economics is an older discipline than scientific psychology and it had to invent the assumptions which underpin economic behaviour. That a person makes rational decisions which optimise their self-interest is a useful one with a number of analytic advantages. In my view the assumption is a useful first step, but there are economists who defend it fiercely as though it is unquestionably true.

Even when I began the work, we knew that most smokers exhibited a paradox for they say they would like to give up smoking but they do not. That is difficult to explain in terms of rational economic behaviour.

More recently, economists have incorporated the psychological literature which suggests that people do not behave as rationally as traditional economic theory assumes. Behavioural economics is a relatively new research area and we are fumbling our way to incorporating it into the central economic paradigm.

It shifts policy beyond making sure people understand the consequences of smoking and eliminating externalities and to pursue policies which actively discourage smoking – especially discouraging (young) people from taking it up.

It turns out that one of the most effective ways to reduce tobacco consumption and discourage taking it up is a high excise duty on tobacco; raising it is an integral part of today’s anti-tobacco policy. So economists have been integrally involved in the campaign.

Alcohol is a little different, because while one’s aggregate consumption affects one’s health, the intensity of consumption also has serious additional effects – binge drinking is neither good for you nor for the others you impact on. My reading of the literature is that modest low level drinking is not particularly bad for one – as I said, many things one does like skydiving and crossing the street are like that. In fairness I should say that the view I have just enunciated is contested but it is my best judgement – over the years it has changed as new evidence became available..

Because of the different impacts from tobacco and alcohol on health and society it is not so easy to regulate drinking. That there are numerous impacts means that there is a need for multiple policy interventions. For instance, there is no tobacco parallel to drink-driving rules.

Taxing alcohol consumption is trickier too. The ideal would be a tax rate which rose as your blood alcohol consumption rose. I have struggled with practical alternatives over the years. That is a big story, so let me just tell you a cheerful part of it.

When working on rates of excise duties on alcohol, I found an anomaly in duty levels which meant there were low tax rates – and hence prices – on light spirits which were taxed as though they were wine but whose alcohol characteristics were closer to full spirits. A bottle of light spirits cost about a hour’s wages, but its contents were sufficient to kill a drinker of a full bottle. Sadly, while I was working on the project, some youths did and died.

I took my findings to the Treasury who were appalled. Bugger the health effects, there was leak in their tax system. Very quickly the tax regime was changed; ALAC, for whom I was working, said it was the quickest tax increase they ever got.

One effect of the new tax regime was to raise the price of fortified wines. The newspapers gleefully pointed out that little old ladies would have to give up their night sherry, failing to mention light spirits which were the primary public health target. In fact, the impact on sherry consumption proved to be small, but light spirits consumption fell to zero. (The light-spirits manufacturers did not end up crying into their liquor. They provide RTDs; I know of no direct deaths from alcopops although of course they are a source of alcohol abuse.)

The other area involving alcohol and tobacco which I worked in was measuring the social costs of tobacco use and alcohol abuse. That fits nicely into the next section.

Economic Evaluation and Social Costs

Before I worked on economic evaluation and social costs in the health sector I had used cost benefit analysis in other areas such as the Major Projects/Think Big. The intellectual underpinnings are much the same. I shall confine myself to sketching in some of my eccentric experiences.

I cannot resist mentioning that a good colleague and friend Rob Bowie and I were approached to consider the introduction of MRI –magnetic resonating imaging – into NZ for assisting the diagnosis of multiple sclerosis.

Couple of lessons here. One was that the machines were being heavily promoted but we could not find any evidenced-based studies of their usefulness that we need for an economic evaluation – all there were was promises. Because of the data limitations our advice was to compare the numbers of MRIs in Australia.

The second lesson is that technological forecasting is very difficult. Economists are not too good at this. Somehow you have to get a balance between hyper-hysteria and over-caution. The best advice I can give is to be aware of the difficulties, to be humble about outguessing human ingenuity, to keep an open mind and to be ready to laugh at your blunders..

My big effort int the area occurred when I was on an international working party to set out the principles to calculate the social costs of drug abuse. Previously the exercise of measuring the social costs of a drug were all over the place. There was no rigorous underpinning, rather there were arbitrary procedures which were often meaningless. Eventually our standard was published by the WHO.

I’m proud of that effort. My principle contribution was to ensure the social-cost methodology was consistent with cost-benefit analysis, not only because the CBA was a well established methodology based on standard economic theory. It also meant that the social cost exercise could be converted to a full economic evaluation.

Afterwards, I calculated New Zealand social costs of tobacco use and alcohol abuse. It is too big a job to report today on my findings – the costs were huge – or on the underlying methodology.

However, I report that subsequently a generic economist was commissioned to update my work. It was very badly done as he did not seem to understand the issues (for instance,  he did not even refer to the WHO report, but made it all up as he went along). The report was vigorously attacked by another generic economist of a different political persuasion who was equally ignorant.

Neither referred to one of the most fundamental principles in economics that a ‘cost’ is always based on an ‘opportunity cost’ and that requires a counterfactual scenario to compare with the current situation. For instance, are you talking about what happens if all tobacco consumption ceases tomorrow but there is a hangover from past consumption, or are you assuming that tobacco was never used in New Zealand? Each will give a different total social cost so it is really important that one states what the alternative scenario is, otherwise the estimate is just a muddle.

This muddle is not confined to health applications. We often get commissioned reports which make outrageous claims about the social costs or their equivalents of some policy change. Unlike in the case of the major projects the studies are not checked by independent economics and they tend not to be very reliable. The next time you see a promise that this or that will generate or cost billions of dollars, yawn.

I want to end this discussion of economic evaluation by taking you through an issue which comes from health economics but has even wider implications.

I began in a discussion of how to choose between different road designs. At issue was how to value road safety. The convention was to include in the evaluation the number of lives saved which shifted the design choice towards safer roads – an eminently sensible thing to do. The particular issue was how to put a dollar value on lives saved.

There is a tricky ethical issue here which worried the international working party. What does it mean to put a dollar value on a life – is that the ‘value of life’? The working party was confronted with the fact that some religions would not tolerate putting any value on life. What we agreed was that we were not valuing life as such but identifying the amount that public policy would spend in order to avoid a death. I shall fudge the issue for the rest of this talk.

There is a bit of a problem with the roading approach because it does not cover losses from a lower quality of life – say from permanent injury or even from short-term accident recovery and trauma. Such issues are acute in the health sector because increasingly treatments are concerned with better quality of life – including for patients in a terminal condition – and not just preventing death. Would it not be crazy to conclude that a treatment was not economic because while it gave a better quality of life it did not prolong life?

Additionally, how do we deal with the prolongation of life? Should we not favour a treatment which gave two extra years of good quality life against one which gave only six months? For instance, tobacco kills more than alcohol, but the dead from alcohol abuse tend to be decades younger than those who die from tobacco-induced diseases. Should put in relatively more effort to eliminating alcohol abuse than tobacco use than the relative death rates suggest?

Another issue is whether we treat all people the same? It turns out that even the elderly think we should not put as much resources into prolonging their lives as we should for a mother with young children. Sounds right to me.

It is a complicated task answering all these questions. There has been progress, but it is limited by useful data. A particularly robust procedure is using a measure of quality-adjusted life years. You calculate the difference in QALYs, as they are called, between two scenarios and value them according to some procedure.

When I did this for tobacco use and alcohol abuse I found that the value of the total QALYs from the elimination of the use and abuse were enormous compared to the resource costs involved – by a factor of about ten. That meant that the impact of tobacco use and alcohol abuse on the quality of our lives was far larger than it was on GDP. I long puzzled over this. Here is my brief explanation.

Economists knew from the beginning that material output (say per capita GDP) was not the same as wellbeing. Sometimes we forget, sometimes we seem to forget – the public, especially the commentariat, are much worse.

The puzzling result suggests that material output makes a smaller contribution to wellbeing than other things – things which economists do not have a lot to say about. (I add that there is a lot of other evidence to underpin this conclusion.)

That is a sobering thought. Yes, economists can improve the human condition by promoting good quality material output, but do so humbly for there are things which are outside our competence. It is possible that those who are in the health sector make a bigger contribution to wellbeing than business does – even using business criteria. That is not the national rhetoric.

One can also see the current government grappling with related questions in their promise of a wellbeing budget this year – although I do not have high hopes it will make a lot of difference.


I have had to take you through some tricky economics, even if superficially. Dont worry about the details. My central messages are twofold.

The minor one is that I hope I have conveyed that being a health economist can be interesting and challenging and that it can even have a little influence on human wellbeing.

Most of you have no ambition to become a health economist. The message for you is that economic issues are going to influence your health practice. Today you have seen it in terms of

– pricing and funding, and the amount of resources available to you;

– the environment and structure of the system you work within;

– population-based health policies;

– the availability of treatments.

In each case economics has contributed to shaping outcomes, often for the better, sometimes for the worse I am afraid, although even in the case of the health redisorganisation health economists played their part in the resistance.

What does this mean for a health practitioner? One can ignore economics and live a professional life buffeted by economics in ways one cannot comprehend. The buffeted will probably rely on poor quality economic theories – even fallacious ones – without realising they are using them or knowing that the theories they use are not very good.

Alternatively, one can think, as I do when I obtain medical advice or treatment. I am in charge but I am aware of my ignorance. I expect the adviser to act in my interests but I try to understand what is going on. That includes being prepared with the basics, asking questions where I do not sufficiently comprehend and – yes I admit – consulting Dr Google and the like. Being prepared including keeping abreast with the latest developments; that requires an understanding of the relevant science.

The same broad principles apply to health economics. You should respect the expertise of the professionals and be knowledgeable enough to identify quacks. Keep the experts on tap not on top. One of the reasons this paper has asked lots of questions, is because I am unsure of the answers. But I hope it will help you when you are challenged by the issues I have covered.

If you do not go down this path, do not be surprised if you find yourself limited in a world over which you have little control.