Hypothecation, health taxation and hysterisis

The latest manifestation of the semi-permanent financial crisis experienced by the National Health Service (NHS) since its creation in 1948 has led the prime minister to promise a ‘multi-year’ funding plan for the health service. It is intended to get away from annual cash ‘top-ups’ to a long-term sustainable budget that allows the NHS to plan for the future. Ministers are publicly debating whether a hypothecated tax should be used to increase funding for the NHS. The Health secretary has said that he favours the idea, but the chancellor and chief secretary oppose this approach.

Nothing new under the sun

Hypothecated taxes are levies whose revenues are earmarked to be spent on a particular programme or use. The traditional approach to public finance is to pay tax revenues into a single pot (the Consolidated Fund) from which public expenditure is then funded. Hypothecated taxes have been around for a long time. In ancient Greece, for example, the individual citizen was not taxed directly. Instead, the (city) state taxed some form of commercial activity such as a road, a bridge, sea-lanes or a harbour. Those who used the facilities provided paid the taxes. But hypothecated taxes haven’t always worked well. For example, Charles I’s revival and extension of the medieval hypothecated ‘ship levy’ on coastal towns for the upkeep of the Royal Navy in peacetime fuelled the discontent that led to the English Civil War and his own execution.

Not quite 57 varieties, but

There are different types and degrees of hypothecation, ranging from strong and wide to weak and narrow. They can be classified as follows:

• Strong (or strict), wide earmarking: A whole spending programme e.g. education could be funded by its own dedicated tax. According to Julian Le Grand’s view (2003) ‘for a hypothecated tax to succeed, the hypothecation must be strong. The revenue must be genuinely circumscribed. That is, it must be the sole source of funds for the service concerned; and it must be used only for funding that service. A surplus in the hypothecated fund should not be used to fill in gaps in the funding of other services. Rather, it should be carried over for use in the next year’;
• Strong, narrow earmarking: A particular category of expenditure could be financed by a specific tax e.g. the BBC license fee;
• Weak, wide earmarking: e.g. National Insurance contributions (NICs);
• Weak narrow earmarking: The tax raises part of the revenue used to fund part of the expenditure on a particular spending programme e.g. the 1% increase in NICs from 2003 to increase spending on the NHS. Economic commentator Sir Sam Brittan (2002) said that ‘Earmarking is worse than useless if earmarked taxes do not finance the whole of the service in question’. He characterised such proposals as ‘tokenism’ and designed to ‘give the impression that a big improvement [..] could be had on the cheap’.

A case study

The National Insurance system is a weak and wide earmarked tax. It is used to fund the payment of contributory benefits, and specified certain other expenditures. All NICs receipts are credited to the National Insurance Fund (NIF). The NIF is an accounting concept, rather than a pool of assets. A claimant’s entitlement to benefits is established by reference to his or her history of NICs payments. Those who claim contributory benefits regard their previous contributions as having paid for their current benefits (‘I have paid my stamps’). The NICs system is a pay-as-you-go scheme. It uses current contributions to fund the payment of benefits, the entitlement to which has accrued in the past. Some people characterise it as a gigantic Ponzi scheme!

Surpluses on the NIF cannot be transferred to the Consolidated Fund. They can only be used for spending on the NIF’s specified purposes. The government borrows surpluses on the NIF to fund public services, and they reduce the public sector borrowing requirement. The Consolidated Fund (into which all tax revenues are paid) underwrites the NIF. General tax revenues are used to pay for contributory benefits when the NIF is in deficit. The earmarking of NICs to pay for contributory benefits is weak and nominal.

Whether hypothecation is good or bad

There is nothing inherently wrong about hypothecating tax revenues. But it is often alleged that economists in general, and the Treasury in particular are hostile to the introduction of hypothecated taxes. It is worth revisiting the pros and cons of earmarking tax revenues for a particular purpose, and the different approaches to hypothecation.

Advantages of hypothecation

• Reconnection, increased accountability and trust: Taxpayers find it hard to understand the link between the taxes they pay and the public services they consume. Hypothecation can reconnect tax payments and services. Taxpayers seem less hostile to paying NICs than income tax, because they know that NICs pay for their future entitlement to a state pension;
• Transparency: Most taxpayers find the tax system and public expenditure decisions totally opaque. Opinion polls regularly show that the public is keen to have increased public spending, particularly on the NHS, and lower taxation. Creating an explicit link between the public services provided and the taxes necessary for their provision promotes transparency about the value and cost of public spending and the choices faced by the government in funding it. Sam Brittan (1994) said that earmarked taxes would make voters put their money where their mouths are when they support higher expenditure on the NHS. ‘Eventually even belligerent uniformed nurses who turn up at every televised [..event] with their demands for ‘more, more, more’ will begin to be heard more critically’. Taxpayers would realise how meeting their demands would directly result in tax increases that affect them personally;
• Public support: It is difficult to meet public demand for new or additional services, without reducing spending in other areas. If there is robust evidence of public support for increased expenditure on particular measures, hypothecation can create public support for increased taxation specifically to fund it;
• Ring-fencing resources and revenues: Depending on the nature or the degree of hypothecation, earmarked taxes protect the revenues from being spent on anything other than the specified purpose for which it was raised. Public resources and revenues are allocated through a competitive and complex process. Government departments don’t always succeed in securing funding, even for measures that are popular with the public because general tax revenues have to fund all government activities. It is necessary for ministers to make tough collective decisions about national priorities, and to spend substantial amounts of revenue on measures that don’t figure high in the minds of electors e.g. Trident.

Disadvantages of hypothecation

• Inflexibility: The strict hypothecation of taxes that yield substantial revenues would restrict the Treasury’s ability to respond flexibly to economic cycles, shocks and other events by changes to taxes and public spending;
• Waste: Government spending financed through hypothecated tax revenues is not subject to the same detailed scrutiny, review and challenge as expenditures financed through inter-departmental bidding for a share of general tax revenues as part of a competitive public expenditure process;
• Getting revenues and needs ‘back-to-front’: Spending should be determined by needs and priority, rather than how much revenue a hypothecated tax raises. The government’s commitment to spend 0.7% of annual GDP on foreign aid could be compared to an earmarked tax. The policy has been heavily criticised for being wasteful. The Department of International Development has found it impossible to identify enough high-quality aid projects every year to use up its massive automatic funding. Some of its decisions have created controversy at a time when public expenditure on domestic public services has been constricted;
• Linking funding to requirements overtaken by circumstances: The level of hypothecated taxes are linked to the needs and requirements of the purpose for which they are earmarked at the time when they were introduced. These needs and requirements are likely to change over time. There are difficulties associated with changing hypothecated tax arrangements and legislation. It could result in the level of revenues remaining largely unchanged for many years. For example, the television license fee was introduced at a time when television sets were relatively expensive, and the BBC was a monopoly public broadcaster. The license fee survived the introduction of commercial television broadcasting funded by advertising revenues. But it is now increasingly under threat. The advent of digital and other disruptive technologies such as smart phones and live streaming has resulted in fewer people watching terrestrial broadcasting or purchasing television sets;
• Undermining funding for essential but less prominent areas of government spending: It is impracticable to hypothecate every item of government expenditure. It would undermine the established tax and public expenditure process if hypothecation taxation of revenues and ring fencing were used to fund ever- larger areas of public policy. The removal of the need to compete for a share of national tax revenues would result in a loss of central scrutiny and control of taxes by the Treasury, and deprive it of important policy instruments. Key areas of need that are unfashionable or not in the public eye would increasingly find it difficult to secure any, let alone adequate, funding.

Expenditure on health

Britain spends less on health, as a share of gross domestic product than other countries with a similar income level. The British government is directly responsible only for the health service in England. The governments of Scotland, Wales and Northern Ireland are directly responsible for their own health services. Most of the devolved administrations’ spending is financed by block grants made under the ‘Barnett formula’. (The Scottish Government can finance increased public spending on its devolved policy responsibilities through its ability to raise the Scottish rate of income tax. From April 2019, the Welsh Government will also be able to vary the rates of income tax payable by Welsh taxpayers.)

For many of those who campaign for higher spending on the NHS, the obvious solution is higher income taxes on the ‘rich’. They overlook that almost half of Britons (23m people) pay no income tax and that the top 1% of taxpayers (300,000 people) already pay 27.5% of all income tax. If the government continues to raise personal allowances at the same rate as they have been increased since 2010, it is possible that a minority of taxpayers could end up funding a service that is available free to the entire population. Those keen on increasing spending on the NHS often also suggest raising vast sums by stopping tax avoidance. But the UK already has extremely tight anti-avoidance rules and avoidance accounts for a fraction of the tax gap. Others argue that it would be possible to increase spending on the NHS if Britain followed the example of countries e.g. Germany that have dedicated funding for their health service.

Mythical Treasury hostility

Paul Johnson, the head of the IFS, recently noted that it seems obligatory to begin any discussion about the introduction of hypothecated taxation in Britain by referring to the Treasury’s alleged hostility to such taxes. It’s almost as if the introduction of a hypothecated health tax requires the sacrifice of Treasury orthodoxy at the altar of the NHS ‘the closest thing the English have to a religion’ (Nigel Lawson). The Treasury is in fact a pragmatic institution. It is guided by classical public finance principles, particularly avoiding waste and inefficiency, securing value for public money, and insisting on the determination of public spending decisions by policy decisions and by reference to governmental priorities (not favouring whatever is in the public eye).

The Treasury’s caution is about hypothecation is understandable, given its wider responsibilities for managing the economy and tax strategy. The widespread use of hypothecation would introduce inflexibility into the fiscal system, and limit the government’s ability to use fiscal policy to regulate the economy. But the Treasury isn’t doctrinaire about hypothecation. Lord (Nick) Macpherson, who was permanent secretary to the Treasury from 2005 to 2016, has been an influential contributor to the recent debate on a health tax. He has suggested that a specific tax rise (with NICs being the ‘strongest candidate’) might be the best way of raising additional funding for the NHS.

Increasing government spending on the NHS

The NHS is predominantly funded from general taxation. Any increase in government spending on health service would entail a corresponding reduction in other areas or higher taxation (borrowing would involve transferring the burden of current spending to future generations). The Tory party’s manifesto for the 2017 general election omitted its former pledge not to increase income tax, NICs or VAT (the so-called ‘tax triple-lock’). It promised instead to stick to its plans for corporation tax, keep VAT at the same rate, and simplify the system. Assuming it could get the necessary legislation through parliament, the government would have the flexibility to increase income tax and NICs. It could either increase the rates of income tax or NICs or widen the basis on which they are charged, e.g. by taxing all forms of work in the same way. Alternatively, it could raise more money through VAT by exempting fewer items, while keeping the rates unchanged, or introduce a brand new health tax. Any changes would have political consequences because they would create losers.

Carts before horses

It is undeniable that the health service is under intense pressure. A public service that provides us all with protection against the uncertainties of life is under siege for many reasons. The NHS is faced with coping with growing demand because of immigration and an ageing population, the diversion of patients to hospital A&E departments from primary care because of reduced GPs services, low morale, shortage of skilled staffing exacerbated by Brexit, and the ballooning cost of drugs, medical technology and compensation payments for medical errors (which have all tended to rise faster than the rate of inflation). But the spending pressures on the NHS are only one side of the equation. There needs to be a public debate about how much people are prepared to pay for the kind of service they say they want. And how that cost should be divided between different classes of taxpayers.

Reasons to spend more on health

The NHS is a creaky, 70 year old institution. There are doubts about whether it can afford the cost of rapid advances in medical science over the next decade, including a myriad of advanced stem cell and gene therapies, the revolution in robotics infused with artificial intelligence, and exploiting the potential from a human ‘Internet of Things’. The NHS needs massive and sustained injections of funding to continue providing a modern service.

The British health service is built on the principle that the funding of the service should be based on the ability to pay, but that the access to it should be based on need, and free at the point of use. Higher expenditure on health entails choosing where to spend less or higher taxation. At some point taxpayers would find that the loss of utility from lower government spending on other pubic services or having less net income outweighs the benefits they receive from increased spending on health.

Unlike most other areas of public sector, which have suffered cuts in spending, the NHS has received real-term increases in funding. Additional spending wouldn’t necessarily buy more health services. The new funding could easily be swallowed up by higher pay settlements. The NHS is the largest employer in the world after the Red Army and the Indian Railways. There are arguably other areas where is an equally pressing need for greater public expenditure e.g. preparing to deal with the unemployment of millions of workers through the wide-spread diffusion of robots infused with artificial intelligence.

As John Appleby (2017) has noted, funding on the NHS has to be increased to keep pace with demographic changes just to stand still. The population is increasing rapidly; it’s forecast to be 12 per cent larger over the next 20 years. It is also getting older; the number of people aged over 75 will be 73 per cent higher by 2036. (The average 85-year-old costs the NHS around eight times more than someone who is in his mid twenties.)

Health care is highly labour and skill intensive. It isn’t practicable to require the NHS to speed up medical treatment. The NHS cannot make do with less people or deploy untrained staff to generate and sustain improvements in productivity. That would be akin to asking the players in an orchestra to play their instruments faster or asking the conductor to sack the percussionists or violinists. Some taxpayers might choose to spend on private health care if adequate care wasn’t available on the NHS; those who cannot afford to do so would be denied care.  

A multi-year plan to fund the NHS

The NHS cannot plan and manage effectively if it continually has to seek annual top-ups; it needs a budget based on its projected long-term needs. The government has a variety of options for raising substantial revenues to fund current and future spending on the NHS, including:

• More of the same e.g. proposing a 1 or 2% increase in NICs, income tax or VAT rate and pledging the additional yield to the NHS
• Channelling the bulk of the revenue from a highly visible existing tax into the NHS e.g. NICs or VAT
• Introducing a brand new ‘health tax’

More of the same

It would be relatively easy for the government to increase the rate of NIC, income tax or VAT by a marginal amount and promise to allocate the additional revenue to the NHS. (Increasing the VAT rate would mean breaking its manifesto pledge not to increase it.) For the policy to be credible as meeting the long-term needs of the NHS, it would need to commit to future increases in the tax rate to ensure that revenues kept pace with growing expenditures.

The choice of the tax selected would depend on the level of desired revenue, their respective collateral impacts and the speed with which the government wished to implement the change. A 1% increase in these taxes would raise the following additional revenue:

Source: HMRC Direct effects of illustrative tax changes (January 2018)

If the government decides to increase income tax or VAT, it can formally announce the proposed changes at Budget in November 2018, and include the necessary legislative changes in finance bill 2019, which is expected to be enacted in March 2019. The Treasury does not have to consult about changes to tax rates. (Preparatory work may already be secretly underway.)

Any changes to NICs rates instead would have to be implemented through a programme bill. It could be introduced immediately after the Budget and expedited through parliament. Increases in employer NICs would be highly unpopular with employers (especially to small businesses) coming on top of the Apprentice levy, and the phased increases to the National Living Wage.  Small businesses and their employees have had to make increased contributions to workplace pension schemes from 5 April 2018. Their contributions are due to increase further from April 2019. These additional costs of employment would almost certainly result in lower employment and annual wage growth for those who keep their jobs. 

If the government decides to increase VAT, it would add to inflationary pressures and dampen consumer spending. The Treasury might face demands to hand over half of the additional Scottish receipts of any increase in the standard rate of VAT to the Scottish government, under the terms of the agreement struck between the British and Scottish government following the recommendations of the Smith Commission’s report in 2014. (The British government’s block grants to Scotland, Wales and Northern Ireland would be automatically increased under the ‘Barnett formula’ to adjust the amount of funding made available to them in line with increased expenditure on public services in England.)

Proposing a specific increase in the rate of income tax, NICs or VAT to fund increased spending on the NHS would represent a weak and narrow form of hypothecation (see above). The policy would lack transparency, arguably the most attractive aspect of hypothecation. Even if the government were to announce the policy from the rooftop of the Treasury, it would not necessarily raise the public’s awareness of their personal financial contribution to the increased spending on the NHS.

Depending on the tax selected, the total income tax or NICs paid after the increase was implemented would be reflected on wage slips; the total VAT paid would appear on sales slips. But it would be practically impossible for workers or consumers to know how much more they were paying specifically for the health service. The government could change the legislation governing what has to be legally included on pay slips or sales slips to require the identification of the element of the tax, NICs or VAT paid that would be going toward funding the NHS. (It would have to provide employers and retail outlets with a formula for isolating the NHS component of the total tax, NICs or VAT paid.)

Channelling the bulk of revenue from an existing tax to health

Sam Brittan (2002) suggested that it would ‘be best to channel the bulk of the revenue from some clearly visible existing tax’ into the NHS. The new money pledged in the Autumn Budget will take the total Department of Health budget to £124.7bn this year (2017/18), £126.4bn next year (2018/19), and £127.2bn in 2019/20. The projected receipts from the three largest sources of government revenue for the same period are:

Source: Office of Budget Responsibility (March 2018)

It would be courageous (as in the Yes Minister sense of ‘it will cost you the next election!’) for the government to pledge all or the bulk of income tax, its single most important source of revenue, to expenditure on the health service. Income tax has to fund many other aspects of the public finances.

NICs may appear to be particularly suitable for conversion into a payroll levy to cover expenditure on health.  Employers already pay a large proportion of NICs receipts (although the incidence of the charge would fall on their employees). Dedicating NICs receipts to health would leave the government without a mechanism for establishing entitlement to contributory benefits. This would undermine the ‘contributory principle’. NICs are a uniquely tolerable, if not popular, form of taxation. Taxpayers understand the quid pro quo they currently receive for their payment of NICs, particularly the state pension.

The NICs system redistributes income between high and low earners and between prosperous and poorer areas. It sustains the payment of a uniform state pension and contributory benefits throughout the UK. If the British government wished to devote future NICs receipts solely to funding the health service in England, it would need to consult the Scottish and Welsh governments. They might demand the creation of distinct National Insurance Funds for Scotland and Wales, and compensatory payments for the maintenance of a uniform level of contributory benefits.

On the face of it, VAT is a strong candidate to have the bulk or all of its receipts channelled into funding the NHS. VAT is a robust and buoyant tax. Its receipts have outgrown NHS expenditure in real terms, leaving some headroom for growth in NHS spending. But VAT is a regressive indirect tax. Its adoption as the NHS tax would be incompatible with the ‘equity’ objective of funding health according to ability to pay. Whether the government decided to divert NICs or VAT totally to financing the health service, it would have to commit to increasing the rates at which it is charged to ensure their receipts keep pace with the future funding needs of the health service. ‘In future, nothing could make clearer the cost of increasing NHS care than an automatic increase in the VAT rate or the extension of the tax to products now exempted’ (Sam Brittan, 2002).

On a brand new ‘health tax’

It is theoretically possible for the government to introduce a brand new ‘Health tax’ whose yield was legally earmarked to funding the health service. To do so, it would first have to identify a suitable base for the new tax. Earnings and consumption, the most productive areas for taxation, are already comprehensively fiscally harvested by income tax, NICs and VAT. The tax would require its own computation and assessment rules, and possibly a bespoke collection mechanism. The proposals would have be the subject of detailed consultation, before it was possible to prepare draft legislation (which would also be subject to consultation).

The operational work, especially the IT systems, necessary to implement the new tax could not be commenced until the second reading of the finance bill containing the draft legislation. The measure would consume scarce policy capacity in the Treasury and HMRC at a time when their resources are fully committed to Brexit, and distract HMRC’s IT and operational teams from delivering previously announced measures, including the new Customs Declaration Service. It would probably take at least three years before a brand new tax was fully operational. We would be in the run-up to the next general election by the time the tax was introduced.

The proposed tax would probably be dismissed as a ‘gimmick’. It would not be able to raise revenues remotely rivalling those collected through income tax, NICs or VAT. The yield expected from the new tax could be collected far more easily and cheaply by simply marginally increasing the rate of income tax or NICs or VAT.

Social care

There would be a strong case for re-labelling whatever approach is selected as being intended to fund Health and Social Care. A significant proportion of the costs of the health service are devoted to addressing the medical needs of the rapidly growing number of older members of the population. The demands of health care for the elderly are intertwined with the challenges of social care, and as Lord (Richard) Layard (2017) recently noted, ‘the present dichotomy between …health and social care is not working’. This is reflected in the prime minister’s decision to make the Health Secretary also responsible for social care.

Andrew Dilnot, the chairman of the Independent Commission on Social Care (2011) and Support, whose thoughtful recommendations were shelved in favour of the ‘death tax’ announced in the 2017 Conservative Party manifesto, has recently suggested that the million plus workers over the state pension age who are still employed or self-employed should cease to be exempt from NICs.

The logic behind the exemption of those over state pension age from NICs is supposed to be that they can ‘leave the club’ once they become entitled to draw their state retirement pension. They have finished paying for it. But this logic is flawed. The NICs system is funded on a pay-as-you-go basis. A significant proportion of NICs receipts are used to fund expenditures other than contributory benefits. Older workers, who haven’t started drawing their state pension yet e.g. to take advantage of the financial inducements available to those who defer drawing their state pension, can, in certain circumstances, acquire additional entitlement to a higher pension by paying voluntary Class 3 NICs.

There is principled case for removing the age exemption to NICs being charged on the earnings of those who choose to continue working past the state pension age. They are the biggest users of health and social care services. More than 40% of pensioners are now in the top half of income distribution. Many don’t need the full range of non-means tested benefits available to pensioners. It would help to address the intergenerational dislocation created by the rapidly rising increase in the dependency of older citizens on the working age population. (The ONS estimates that the number of people over the age of 65 years will increase by 7 million by 2046, while the number of working age workers will rise by only 3 million.) But the influence of the ‘grey vote’ might make the government wary of risking a row with its own backbenchers over the perfectly sensible change recommended by Andrew Dilnot.

Hysteresis

In economics, ‘hysteresis’ refers to effects that persist after the initial causes giving rise to the effects are removed. (The word is derived from a Greek verb meaning, ‘that which comes later’.) The funding problems of the NHS are likely, not just to recur, but get worse if the government fails to come up with a credible mechanism to keep funding contemporaneously apace with demand for NHS services. If it is necessary for NHS funding to be continually re-addressed whenever its crisis becomes too serious to ignore, the amount of the top-up required the next time would be greater.

The NHS’s chronic funding crisis is likely to continue until the electorate faces up to some key facts regarding the public provision of health services:

• Taxes will have to be raised substantially to pay for demographic changes and the kind of services the public expects when they need the NHS;
•  All consumers of NHS services who have taxable income or gains will have to bear a share of the additional costs through higher taxes or a reduction in other government services (probably both);
• Higher taxes on the rich or curbing avoidance cannot pay for all the NHS’s future needs;
• There is waste and inefficiency in the system but public service unions and professional bodies are resistant to changes that threaten their own status and pay and working conditions;
• The scope for greater efficiency, innovation and new practices in the NHS to secure better value for public money cannot be realised if any proposed changes are immediately condemned as attempts to ‘privatise the NHS’.