Universal healthcare, based on there being no cost for consumers, was partially realised in New Zealand in 1938,1 but an unresolved dispute with the medical profession meant that primary care remained privately provided. Escalating costs and inequality of provision have subsequently triggered multiple reviews and reforms. However, starting with the establishment of Area Heath Boards in 1983, none of these largely structural changes have either lasted or been able to fundamentally address these cost and equity issues in a sustainable manner. Given we seem imminently destined for yet more structural change, it is important to ask why these reforms have not been successful.

Political economy underpins eight decades of ‘unsuccessful’ health reforms

Some of this failure is due to poor design and implementation.2 However, more fundamental forces are at play. Accommodating the ideals of “free” and “universal” inevitably requires rationing to be affordable. This is manifest as restrictions on entitlements (coverage) and on access to those entitlements (queuing).

Various public and private interests expressed through the political process determine the distribution of cost, entitlements and access, and hence the success of any reform. Changes in policy affect a range of stakeholders who have different interests and degrees of influence. This is the “political economy” of reform.3

The landscape of health stakeholder interests

Many stakeholders, such as those employed in the sector, have a concentrated interest in the outcome of any reform and are well organised to ensure their interests are protected and advanced. Some consumers of healthcare similarly enjoy organised representation. These are typically frequent users and their families, whose interests are often well supported by provider and advocacy groups.

However, for most people most of the time, healthcare is more akin to insurance, something they might have occasion to use episodically. They assume that it will be available when they really need it. Inevitably, their judgment about how good this will be when that time comes relies on the opinions of those in the health system and on social-media influencers.

While most consumers make small co-payments to help meet the cost of their healthcare, the vast majority of costs are met by taxpayers. Taxpayers only have a diffuse and poorly organised interest in influencing this cost. Additionally, most taxpayers are the episodic consumers of healthcare described above.

The other main stakeholders are those who also rely on government expenditure and whose interests would be compromised if health took a greater share of available revenues. These interests will have their own priorities, and ministers need to juggle budget bids from across publicly funded activities to remain within the agreed overall spending limits. Heath is not the only set of interests that are well-represented in the budgetary process.

Although taxpayers’ interests are weakly represented in deciding the healthcare budget, this interest gains weight when healthcare costs start to push total government expenditure up as a proportion of GDP. At that point, governments can no longer fund demand for increased spending from the growth in tax revenue that is “naturally” fuelled by economic growth; ministers have to consider new or higher tax rates for existing taxpayers, or running up debt for future taxpayers, and that is when taxpayers take more notice. Ultimately, it is ministers who must balance these competing interests during reform and associated annual budget processes, which underscores the point that healthcare reform is inherently political.

The stylised facts are consistent with this landscape

Given the distribution of influence described above, it is easy to see why healthcare is taxpayer funded and why it is so difficult to reallocate resources when that reallocation threatens the interests of well-organised workforces or consumer groups. This resistance to reallocation is magnified by “loss aversion,” the well-known tendency for people to fight harder to avoid a loss than to realise a gain of similar value.4 Politicians well understand and are sensitive to this aversion.

Less well understood are some of the other within-system forces that resist healthcare reform. Non-cooperative game theories help explain this resistance when health spending is tightly constrained. For example, the various “players” in mature health systems are argued to be unable to change their “strategies” unless the other “players” agree to change theirs—the transfer of privilege then, in its many forms, requires the support of the relatively advantaged. This condition is known as a Nash Equilibrium.5 An alternative insight is to consider any reallocation when health funding is constant or worse relative to health inflation as a “zero sum game,”6 which inevitably creates losers. By contrast, when health spending is increasing rapidly, it should be possible to increase equality of access without reducing spending on those services of most interest to well-organised workforces or consumer groups. However, a 2004 Treasury study looked at the drivers of health spending from 1951 and concluded that most of this growth was driven by “cost and coverage” factors and not by the size or health profile of the population.7 Typically, additional spending has been used to increase wages and incomes for existing providers, to improve coverage from existing providers (especially as technology changed) and/or to reduce wait times for existing services.

The same factors that influence the distribution of “new money” are likely to bear on the scope for, and distribution of, any savings generated by efficiency gains or other measures. The appetite for these will be constrained by the limited tolerance that concentrated and well-organised interests have for measures that could variously: increase their risks or costs or threaten their incomes; reduce access to the services they provide; and/or limit their professional discretion. Even then, concentrated and well-organised interests are best placed to ensure any savings are recycled into existing services delivered in existing ways.

Inequality of access and chronic under-servicing persist in large part because the groups whose healthcare is compromised lack the support of well-organised provider interests. This is not surprising. If simply delivering “more of the same” could address inequality and chronic under-servicing, then those issues would have been addressed long ago. Instead, success requires innovation in service mix and design and an integration of health, government and social services. The health system lacks incentives for innovation and integration and consequently lacks the community of providers who have a concentrated stake in delivering “novel” services and advocating for their funding. Sometimes the problems associated with under-servicing become impossible to ignore, and so additional resource will be allocated, as happens episodically with mental health services. These exceptions only help to prove the rule.

The observed pattern of health spending is also consistent with the view that taxpayers’ interests are relatively weak until a period of increased public health spending, relative to GDP, pressures total government spending to move in the same direction.7 The Treasury noted that decades of significant and above-average growth in health expenditure as a proportion of GDP had been followed by decades with much slower, or even negative, growth in this ratio. This pattern has continued through to 2019. The increase in core health spending relative to GDP following 2002 accelerated in 2005 with the “pay jolt” for nurses but this trend ended in 2011. This ratio eventually returned to 2007 levels by 2018. At some point, concern about escalating health costs has found effective political expression.

This landscape of healthcare interests is inherently unstable

Given these underlying political pressures, it is not surprising that governments often “lean against the wind” of rising healthcare demand in order to ensure that they can fund this demand without having to raise taxes or crowd out other competing interests for public funds.8 It is also not surprising that increased spending typically funds “more of the same” mix of health services provided by existing providers and accessed in existing ways. This ensures that the best-represented workforces and consumer interests in the system are never worse off and often better off. While this approach can work for some time, it builds tensions within the health sector that require attention and so is inherently unstable.

First, “leaning against the wind” leads public providers, especially district health boards (DHBs) and their predecessors, to delay spending that is not literally required for current service provision, such as capital investment and maintenance. Rather than save to spread the cost of capital works, the pattern is to spend more on existing services and run deficits. As more DHBs fund existing services from deficit spending, the problem becomes defined as sector-wide underfunding, rather than indicating that individual DHBs cannot manage their budgets. This pattern has been repeated over recent decades and inevitably ends with governments at least partially funding the deficits of public providers in order to close the metaphorical “second chequebook.”9

Second, resistance to reallocation of health spending ensures that existing services delivered by existing providers and accessed in existing ways continue to be given priority. Almost by definition, these services do not adapt to changing demands. Inequality of access persists and some consumers are chronically under-serviced. As the numbers in these groups grow, or as the consequences of this under-servicing become more obvious, governments face mounting pressure to do something. Mental health, disability and ethnic disparity are illustrations.

Finally, the unfettered introduction of new healthcare technologies that almost always add cost but sometimes add little value, along with population aging and the associated protracted period of poor health in old age, adds to underlying cost pressure in existing service delivery.10,11 That makes it ever harder for ministers to balance the various interests noted above without burdening either current or future taxpayers.

Governments eventually understand that health is only going to get more difficult and that they are likely to “limp” from one mini-funding or care crisis to another. That might help explain the ongoing desire to want to “reform” healthcare and why it seems a change in government ushers in a new reform process. However, until these exercises explicitly consider the political factors that continue to frustrate each attempt at reform, then they will neither change the political calculus that drives the decisions nor the status quo.

Changing the political calculus to improve equity

How do you change that calculus and create a health system that will address existing inequalities?

First, reformers need to figure out whether the increasing health demands of an aging population can be met from “natural” growth in tax revenue, because it makes no sense to borrow to fund what is essentially consumption spending.12 That will come down to assumptions about the sustainable rate of income growth the economy can produce. The impact of population aging on the budget has been well-known for some time and both major parties seem comfortable that they do not have to raise tax rates to deal with it.

Second, reformers also need to recognise that the most powerful interests in the system will simply not accept being worse off as a result of any reform and will be in a strong position to capture any additional health funding. That means that the incomes of the existing workforce and the quantum and quality of existing services delivered to existing consumers with concentrated and well-organised interests will need to be protected. This makes it almost impossible to reallocate a meaningful quantum of existing resources to meet the demand of those groups who are chronically under-serviced. Moreover, allocating extra funding to meet this demand is unlikely to be successful unless it is effectively targeted for this purpose (as ring-fencing funding for mental health services tries to do).

Third, reformers should not assume that increasing tax rates or introducing new taxes will solve the problem. As noted above, this is likely to trigger active taxpayer resistance that makes this course politically unattractive. Although this resistance might initially be overcome, our experience suggests that periods of strong growth in spending will be followed by periods of fiscal stringency,7 so any gains may prove short-lived. Our experience also suggests that politicians with an eye to re-election will use new money to deliver “more of the same” in order to, for example, reduce wait times. This is a major reason why inequity of unmet need has proved so enduring.

So the question becomes, how do we make more money available to service these groups without having to impose new taxes, raise tax rates or squeeze other government expenditure?

The answer to that question requires a detailed examination of the consequences for future taxpayers of continuing to under-service these groups (in terms of either prevention, detection, treatment or rehabilitation). Given the nature of these groups, it seems likely that not addressing at least some of their health needs will result in significant social expenditure in the future, as well as undermining their workforce engagement and productivity. When this turns out to be true, then improving the health outcomes for these groups should generate a dividend for future taxpayers that could justify an investment that is funded by borrowing against that expected dividend.

It is not enough to just assume that improving health today will reduce health costs tomorrow. Aneurin Bevan, one of the architects of the UK National Health Service, “… had hoped that as people’s health improved the cost of the NHS would fall. In reality, people demanded more services and the cost of providing those rose.”13 Given the landscape of interests and the political calculus described above, it is no surprise that we see the same relentless rise in demand and cost, or that marked inequality in health outcomes persist even in periods when health spending increases strongly.

Future reform efforts aimed at improving equality of access and outcome need to be focussed on three areas:

1. doing the actuarial work to discover the size of this dividend and, therefore, the quantum of borrowing that it can support
2. exploring the institutional arrangements that will produce the right incentives on providers and funders to discover the service mix and design that works best in improving health outcomes for specific underserved groups, and ensuring that extra funding is effectively targeted, so that this dividend will in fact be realised with reasonable certainty14
3. identifying the new system architecture required to secure this body of analysis and to support the implementation of policies grounded in that analysis.15,16

Realising this dividend is critical. It is no longer as easy as it once was to shift funding problems onto future taxpayers by continuously running fiscal deficits and increasing public debt. Compliance with the now legislated principles of fiscal responsibility require governments to be more transparent. Moreover, current and forecast borrowing following COVID-19 is much greater than what previous governments of various stripes have considered “prudent” and close to what has been seen as the upper limit for debt relative to GDP. The government will need to explain how it intends to reduce this debt to “prudent” levels, so any additional borrowing will need to show a credible future dividend.

Meeting some unmet need will generate a fiscal dividend but meeting others will not, so this approach is not the complete answer to addressing equity issues. Interventions that help increase and prolong workforce attachment and productivity are, for example, more likely to generate a dividend than those that help people live longer in retirement with multiple health conditions, which require expensive treatment. This does not mean the latter are less worthy. Rather, addressing these issues requires a different approach that is likely to be more challenging, given the prevailing political calculus.

Conclusions about healthcare reforms in New Zealand

Previous healthcare reforms have failed to change the underlying political calculus and consequently have disappointed, leading to demands for still more reforms. Necessary changes to this calculus are possible, although the path to a different outcome is narrow. However, an actuarial and future dividend focus for funding some elements of healthcare has a much greater chance of succeeding than previous efforts.