Long Term Prospects for Health Spending

<>Wellington School of Medicine Seminar Series: 22 November, 2013. [1]

Keywords: Governance; Health;


The Public Finance Act requires the Treasury to produce a statement on the Crown’s Long Term Fiscal Position every four years. It involves projecting 40 and more years out based on current policy assumptions with some alternative policy assumptions to assist assessment of the significance of possible policy responses.


The latest projections were released last July 2013 following extensive public consultation. I was one of the members of an advisory panel that the Treasury consulted with while preparing the projections.


The Treasury was keen that their report did not disappear but would contribute to an ongoing discussion of the issues the exercise high-lighted. What I am about to say, then, is not the Treasury’s view but my own. Indeed not all of these reflections on the implications for the health sector were discussed in the panel deliberations.


It is not entirely clear why such projections should be required. As a matter of course the Treasury produces as a part of the annual fiscal statement – popularly known as ‘The Budget’ – projections up to about a decade out. Apparently Treasury has been worried about the demographic impacts on the economy since the 1990s. They did not show up much in 10 year projections, so they began to project longer.


The Fiscal Challenge the Health Sector Faces.


However I learned lessons other than the population implications from the exercise. I’m only really going to talk about one – the implications for the health sector. I knew before I went onto the advisory panel that there was a problem with the growth of health care expenditure. It was not until the fiscal projections, however, that I began to think about it systematically. The issue is this: as the Long Term Fiscal Projections show, the expected rising demands for public funding of health care are beyond the capacity of the existing fiscal framework to finance them in the long run.


The Aging Population


This health funding challenge is independent of population change, although it is exacerbated by the aging population. So a word about that first. We all know that longevity is rising steadily and that it is expected to continue to rise. We also know – although it is a bit more complicated – that the fertility transition from roughly two daughters per woman to one has intensified the population imbalance. With the rising longevity, the fertility transition exacerbates our aging population because the labour force is not rising as fast.


Less obvious – although curiously enough I discussed it, albeit in a clumsy way, in a paper published in 1979 – the initial effect of the dynamics is a so-called ‘population dividend’ in which the falling proportion of children in the population eases economic and fiscal pressures. The salutary conclusion is that we have been benefiting from this demographic dividend but it ended in about 2011. From now on population change is going to make it more difficult to manage the economy and the fiscal position. We all know this, except we may be surprised that the turnaround occurred so recently.


The History of Public Spending on Health Care


As far as public health spending is concerned the aging population is a factor in the rising pressure. The elderly require more spending than those in working ages. Children are also health-spending intensive, but the fall in their proportion of the population does not offset the rising proportion of the elderly. I can illustrate the general problem with a chart of some work I did for the Treasury and also for the history of New Zealand I am writing. [Please contact author for graph.]


On its vertical axis is the spending of the Ministry of Health (and its earlier agencies) measured as a proportion of annual GDP going back to 1862. This is not all government spending on health care as we shall see later. It is the part-funded from general taxation. It certainly does not cover all spending on health. I’ll come back to the private spending later.


There are two lines on the graph. Actual spending is in blue and red, the former for when governments of the right were in power, the latter for when there were governments of the left. The track illustrates that –aside from the expansion of spending of the First Labour Government, it is hard to argue the thesis of differential spending behaviour by political preference. [2]


The black line is the spending if the composition of the population had been constant. You’ll see that it hardly varies from the blue and red line until the 1980s, after which population aging meant that spending began to rise faster than the black line. (The demographic dividend comes from the opposite trend happening to public spending on educational services.)


The graph provides a brief history of central government spending on health care. Until 1939 it was less than 1 percent of annual GDP. Most of it was on population-based health care. There was some additional public spending from rates levied by hospital boards which went into personal healthcare, but most was funded by private fees and charity. We dont have any figures for total spending on health care but it was probably low.


In 1938 the First Labour Government introduced a major expansion of public spending on health care, roughly trebling the relative level. Part was a transfer from private spending to public spending funded out of general taxation, part would have been giving care to those who had been missing out because of their inability to pay but some almost certainly was for a response to advance in healthcare. While the 2 percentage points of annual GDP increase must have seemed a huge lift at the time it is not such a big leap compared to what subsequently happened and what is expected to happen.


Suppose we take it that by 1950 the Labour Government’s transition had finished. Over the next 60 years spending rose relatively smoothly from about 3 percent of annual GDP to 7 to 8 percent of annual GDP (depending on how you want to treat the change in population composition). There is a bit of a swelling in the late 1970s, explained by public sector wages rising faster than usual. But, other than that, the post-war rise in spending share is surprisingly smooth.


Note, however, the choice of measure may mislead as to just how big the spending rise was. Actual public spending on health care rose over five to six times in the sixty years above the general rise in prices.


Why the Creep after 1950?


Why did it rise an additional 4 to 5 percentage points of GDP? We can see that about a percentage point was due to the aging population, but the rest is a bit of a mystery.


Normally the quantitative economist would split the increase into that to be attributes to price rises and that attributed to volume increases, but we do not have useful price and volume indexes, certainly not since 1938, nor recently either.


Statistics New Zealand is currently engaged in constructing price indexes to reflect changes in prices of inputs into health care; a good thing but limited. Let me explain with a simple example.


Once upon a time the ultimate treatment for persistent peptic ulcers was surgery. Then the Australian Barry Marshall identified Helicobacter pylori as the cause which can be effectively treated by medication. This means that a patient with a peptic ulcer now gets a better quality of care, but also that fewer resources are used to treat her or him. The index constructors will treat the reduction as lower resource outlays. They are unable to measure the better health outcome. So the indexes do not capture the productivity gain from the innovation, even suggesting a deterioration in health care when, in fact, there has been an improvement.


If carefully interpreted, the Statistics New Zealand exercise is progress. But it does not provide a comprehensive account because it cannot address changes in the quality of life as a result of improved care, one of the most important issues in health economics, one of the hardest to analyse and, as a result, one of the most overlooked.


Given this lack of quantification, I am going to have to do a bit of hand-waving to tell you what seems to be driving the rising spending on healthcare in economic terms. In summary they are the demand effect, the supply effect and the Baumol effect.


The Demand Effect


The demand effect is that as we become more affluent, our demand for health care rises even faster. This is true for the individual’s demand for personal health care but it is also true for the public’s demand for government provided health care.


You can see this happening very explicitly by the Pharmac threshold for cost effectiveness of pharmaceuticals creeping up over the years. The threshold is a valuation on the cost of adding to the quality of life by the use of a drug. The higher it is, the more drugs the public sector is willing to fund. Very expensive drugs in terms of their improvement to quality of life are not – as a rule – provided by the public sector. The threshold is a rationing device.


Pharmac is a bit vague as to what the threshold is, and it insists – quite properly – that it uses other criteria before making a final decision. What is evident is that the threshold has been creeping up, which means that as we have got more affluent, we have been willing to pay for more expensive drugs. The rationing device is being relaxed with affluence. That is true for other treatments too.


There is nothing wrong with such rising demand. It is perfectly rational for a person to wish to spend more on health care for a higher quality of life as their income rises – indeed it may be saner than demanding more and more trivial goods. In aggregate the community has similar demands. However the provision of that care by the public sector breaks the connection between an individual’s income and the payment for the treatment. This is a persistent problem in health economics, although the problem of the broken linkage is equally true for private insurance.


The Supply Effect


By a supply effect an economist means that the cost of supplying a treatment changes – in this case falls, so that people want more of it. The most important case in health care is when a new technology is introduced – say MRI or a new more effective drug. It may seem very expensive at the time but it is better than not being able to do the treatment. So there is a continued pressure to introduce new medical technologies, although economists query whether they are cost effective, that is whether they improve the quality of life sufficiently to justify the outlay.

As time goes by the innovation may become cheaper. Nobody in 1938 could have envisaged the full range of treatments that would become available in the next 75 years; we are equally ignorant of the future possibilities.


The Baumol Effect


The Baumol effect is named after an American economist William J Baumol, who pointed out that productivity rises faster in some sectors than others but that wages have to rise at roughly the same rate in all sectors. Those services experiencing low productivity rises find that their costs are rising faster than general inflation. Let me give a simple illustration from music.


Beethoven famously remarked that putting on a concert costs an awful lot of money. Suppose you repeated one of Beethoven’s concert today. You would have to pay a lot more for the players – perhaps about 16 times as much in real terms because of the general rise in productivity since his day. But the audience could not be any bigger since the size of a concert auditorium is constrained by the laws of acoustics. Costs up, volumes not. That means higher prices or subsidies.


Baumol’s law applies in medicine which is also labour intensive. Where person-to-person care is necessary, productivity gains will be limited, yet the pay of the health worker has to rise in line with increases elsewhere in the economy.


A Fourth Effect?


There may be a fourth effect which is raising spending on health care, although it may just be a special example of the demand effect. We seem to be extending the notion of what constitutes health care. One example will do. I am comfortable with the introduction of informed consent – indeed it is a human right. However, because it takes time, it has added to the costliness of the health care system.



In summary, there are considerable and understandable pressures increasing health spending as share of GDP. Moreover it seems likely that the pressures will continue; perhaps faster, perhaps slower. For instance Obamacare may increase international cost pressures as it absorbs more resources for those Americans with little current health care cover. (I know, I know, it promised to reduce them, but given the state of the American Congress it seems unlikely that the measures necessary to do that will be taken.)


On the other hand, many expensive pharmaceuticals are coming to the end of their patent life and the generics are much cheaper. However that has been happening in the past, with each generation of patented drugs coming out of patent to be replaced by a further generation. But there may be fewer blockbusters in the future.


Projecting Future Government Health Care Spending


There are all sorts of future possibilities, but it would be prudent to project the future trend in health care spending trending as it has in the past, subject to a margin of forecasting error.


That is what the Treasury Long Term Fiscal Projection does, concluding that the health care share will rise 4 percentage points of annual GDP in the next 50 years, slightly faster than it has in the previous 60 years, because of the aging of the population. [3]


Public Health Spending







% of Nominal GDP







In the long run, whatever the base, there is no other item which grows as much as health care in absolute terms (although there may be smaller items such as environmental spending and culture, heritage and recreation which may grow as fast in relative terms). That is the reason why the Treasury projection gave a lot of attention to health care spending; why am I here today.


Can the Treasury fund this sort of increase? The projections assume that government revenue – most of it from taxes – broadly remains a fixed proportion of GDP – at about 31 percent. We can have a long argument about whether that proportion can be retained. Many of the arguments are not very compelling, but the one which is important – I think – is that capital and skilled labour is increasingly mobile in a globalising world and that mobility is sensitive to international differences in tax rates.


So let us accept, for the interim, that the tax take is fixed as a proportion of GDP. (I skip the discussion about the borrowing – it is long tedious but it concludes that there is not a lot of flexibility here in the long run.)


Reducing Fiscal Pressures


Is there any other way we can avoid the fiscal pressures? In particular we have to pay attention to health-care spending because of the size of its increase. We cant simply let health care funding rip, because that would put severe fiscal pressure elsewhere. So what can be done?


Some of the obvious responses dont help. For instance, greater preventive care need not reduce spending on health care. A couple of colleagues in Australia have calculated that eliminating smoking would increase health-care spending in the long run, because people would live longer and the increased cost of their residential care is greater than the clinical reductions.That is not a case for abandoning anti-smoking campaigns – they are about increasing the quality of life. But we must not think that prevention necessarily leads to reduced spending.


Another obvious response is that we should increase the productivity of the provision of health care. Of course we should. We have been doing so as long as I can remember. The assumption of further productivity gains is already built into the projections. The question is whether we can accelerate the productivity gains in the future relative to the past. A good idea were it feasible, but it would not be a good idea to assume we can.


Restraining Government Spending: The Age of Retirement


A third option is to restrain public spending elsewhere making room for more health spending in the total. Given that other parts of the public system are also demanding increases in their share of the public cake advocating such restraint is easier said than done


The one area where there seems considerable agreement – the Prime Minister excepted – is to raise the age of eligibility for New Zealand Superannuation. My view is that this is little to do with fiscal stress but is about how we should think about increasing longevity. The age of 65 was chosen in 1898 when life expectation for 65-year-olds was 13 years. Today it is 20 years. Suppose it was 35 years so that most 65-year-olds lived past 100. Would an age of eligibility of 65 make any sense?


As long as longevity continues to creep up, the age of eligibility will have to rise. What we must avoid is the raising of the age in a hurry as is happening in European countries in fiscal crisis. It needs to be signalled so people can prepare for it and there needs to be an extension of the principle that those below that age who are struggling are entitled to a benefit under the sickness and invalids provisions.


While the basic case for raising the age of eligibility above 65 is the rising longevity, doing so would also relieve fiscal stress. Let me put it the other way around. If we dont raise the age, the growing numbers of very elderly with their increasing health care demands we may find that their care has to be restrained. The point needs to be made in the public debate; if we dont get the age up they could all have a more miserable old age when they really need government support. Perhaps we could throw in some sweeteners – such as set maximum waiting times for hip and cataract operations. Can we also reduce the financial pressures on families who are under 65 but need to look after their elderly parents?


However the Long Term Fiscal Projections show that the single policy of raising the age of eligibility, would insufficient to eliminate all the fiscal stress. We must look for other measures.


Restraining Government Spending on Health Care


What about restraining spending on health care? It would be easy to misunderstand what is being proposed here, by mixing up Ministry of Health spending with total spending on health care. Here is the latest year available share of each.


Share of Total Health Care Expenditure

            by Source of Funds 2009/10

Source Share
Ministry of Health




Other Government Agencies


Local Authorities


Total Public Funding


Private Household


Health Insurance


Not-for-Profit Organisations


Total Private Funding


Source: Health Expenditure Trends in New Zealand 2000-2010


So the Long Term Fiscal Projection was looking at less than three-quarters of total health care spending. Add in ACC whose funds come from a separate levy, and some other minor public funding and five-sixths of the total spending on health care comes from the public purse. Conversely one-sixth comes from private funding; mainly from households directly and the health insurance they pay.


The spending by households was in the following broad categories:

– insurance                                                                   $1000m

– medical goods dispensed to outpatients [4]             $641m

– outpatient dental care                                               $400m

– curative and rehabilitative care (inpatient)               $361m

– all other specialised health care                                $284m

– all other outpatient care                                            $129m

– long term nursing care                                              $109m

– basic medical and diagnostic services [5]                 $  84m

– curative and rehabilitative care (outpatient)             $  34m

(For an order of magnitude, $2000m represents 1 percent of annual GDP).


Some of the the spending is on treatments of little medical worth (placebo and attitudinal effects aside). Subtract those with unproven effectiveness and some sorts of cosmetic surgery, and the share of public provision for health care would be even higher than five-sixths. However for international comparisons this is the ratio we have to use.


Share of Public Funding in Total Health Spending

(2010 or nearest)

Place OECD Country Public Share %
1 Netherlands


2 Denmark


3 Britain


4 Norway


5 Czech Republic


6 Luxembourg


7 Iceland


8 Sweden


9 New Zealand


10 Japan




26 Australia


32 United States



In 2009 New Zealand comes about tenth out of the 35 OECD countries for the share of public provision of health care. It was behind the Netherlands, the Scandinavians, Britain, the Czech Republic and Luxembourg; New Zealand ratio is well above the weighted OECD average of 71 percent.


The Fiscal Projection considers restricting public health-care provision to 9 percent of annual GDP in 2060. Suppose, however, total outlays – public and private – remained the same at, say, 12.5 percent with the gap made up by direct household payments, greater private insurance and even employer-based medical insurance. The public contribution to total health care falls to 72 percent (from 84%) and New Zealand government share would fall to about the middle of the 35.

Please do not get paranoiac. The Treasury was exploring the implications of restraining public outlays on health. I dont think it shows much enthusiasm for such a structural change, certainly far less than there was during the National Government’s health services redisorganisation of the early 1990s. Presumably we have learned from the experience; the Americanisation of the health care system looks even less attractive, despite the improvements of Obamacare. It is well to remember that informed US opinion admires the unified funding system which New Zealand, among others, has.


Alternative Funding Arrangements?


Instead of the current funding arrangements, one might favour a compulsory system in which individuals pay into a single national fund at a rate which reflects their insurable medical needs. This would not involve competing (private) insurers, which would end the principle of unified funding.


The merits of such a change from the existing system are not evident. It would be complicated to administer, there would be problems about calculating the appropriate insurable rates and individuals would fall between the cracks. It would certainly not take health funding out of politics, one of the dreams of many in the health sector. There would still be rationing, there would still be supply failures, politicians would still have to set the overall levies (as they do for ACC) and aggrieved individuals would still appeal to politicians.


The only advantage of switching over to a national health funding system based on insurance that I can discern is that it would be a surreptitious means of increasing taxation, since a compulsory levy amounts to a tax.


Raising Taxes


In my view there is the possibility of raising tax levels in the long run. Earlier I said that the strongest case for keeping tax levels down is to be competitive in a globalising world. But the rest of the world is facing the same pressures for additional health care spending as New Zealand does. This will put pressure on their fiscal systems too, which in part could be relieved by higher tax rates. We can follow, or even cautiously lead.


The resulting balance of the tax increases is an open matter. It may mean higher rates for existing taxes, there may also be possibilities for new taxes and extending the existing tax base.


The issue was discussed briefly by the panel, but it would have been outside the terms of reference – the implications of the existing fiscal stance – to have devoted too much time to them. That is ongoing business.



I am not too pessimistic about the future of public health spending. However it will not be easy. Of course the easy solution for those in the health sector to say that funding is not their business, they are just going to spend it. If they take that approach, they face the possibility a repeat of the early 1990s, in which some who knew nothing about health care – or even health economics – tried to redisorganise it according to a formulaic ideology.


What is the alternative? The first thing is to press ahead with improving the productivity of the sector as much as we can.


Second, are there any activities which can be better provided by the private sector? Before you say ‘nonsense’, I have long been an advocate of increasing the range of over-the-counter medicine – prudently of course. We have done some of that recently.


My pragmatic position is that if we could find a decent solution by transferring the health sector or a part of it to private provision, we should do so. But sadly, despite much pondering, I have never found one.


Reflecting, I’ve realised that much government spending is because there is no satisfactory private solution. That is why the Americans – who have sought to use market solutions more than any other rich country – have ended up with such a goddam awful health care system – inefficient, inequitable and expensive. It is also true for other spending areas, such as the environment and heritage parts of culture, recreation and infrastructure. I am also not much impressed by market solutions to the provision of education although they seem to work better for training.


So I dont think that the only reason for government intervention is the income distribution; too often we lapse into the distributional excuse because it is easier than thinking.


The difficulty is that the all spending areas I have just listed as requiring susbtantial government provision have increasing demand with affluence. The rising demand for the services creates fiscal stress which has to be met by rising tax levels with all the complications they generate. Health care is only the most prominent example because it is the largest. (Some growth sectors are better provided for by the market; fortunately personal transport, from cars to flights, requires much less government involvement.)


Recognising the problem is the most important thing that those in the health sector can do in public policy terms. There is no simple solution and we need to respect those who struggle with the fiscal pressure. By working with them, we are likely to end up with as good a health system as we deserve. Not working with them and New Zealanders will suffer an inferior quality of life.




[1] I am grateful for contributions to this paper by Rob Bowie, Elizabeth Caffin, Geoff Fougere, Alan Gray, Diane Owenga and Paul Rodwell.

[2] More generally it is hard to see fundamental differences in expenditure patterns between right and left governments, with one salient exception. Governments of the right tend to be much more restrictive than governments of the left when it comes to social security.

[3] There is always a problem with the start-up assumptions in the any long-term projection. These assume that the government’s announced spending plans to 2014/5. As it happens the government says it plans to reduce public spending on health care as a proportion of GDP, in part because there is an allowance in the forecasts for future spending which is not allocated by sector. The effect is to depress the base from which the public health spending is projected, possibly by as much as 1 percent of GDP. I shall be very surprised if, with an election looming, that the government keeps to the health spending track set out in the budget.

[4] ‘Medical goods dispensed to outpatients’ include self-medication and other goods consumed by households (which may not be on prescription). This ranges from over-the-counter pharmaceuticals and bandages to alternative medicines.

[5] Includes private payments to GPs.