Thursday, 7 May 2015

The Machine that goes Ping

Despite the best efforts of Nigel "immigration" Farage, and irrational scaremongering about the Caledonian horde, the dominant theme of the general election campaign has been public spending. Unfortunately, the discussion has focused largely on quantity not quality. From deficit levels through welfare to increases in the tax-free allowance, the assumption is that numbers matters more than what they represent and that context is irrelevant. This gives rise to the paradox of a Conservative party that espouses individualism while advocating one-size-fits-all policies, such as the benefits cap and the spare bedroom tax. Whereas once their argument was means-testing, now they dabble with arbitrary cut-offs: no child benefit for larger families. But Labour are similarly conflicted, too often reducing an argument about values (what is right) to mere accountancy (what we can supposedly "afford"), and still prone to beasting the poor. Nowhere is this more obvious than in the case of the NHS.

According to The Guardian, we have only "6.8 computerised tomography scanners per million people, which is less than half the OECD average". This rather precious fact was culled from a report by the Economist Intelligence Unit that compared the NHS with other national health systems. Though the report looked at a number of dimensions (employing the "balanced scorecard" paradigm beloved of neoliberal thinktanks), the media coverage focused on resources, particularly staffing and hardware, with the general implication that these are inadequate. This material inadequacy was then linked to insufficient funding (the Guardian article echoed Labour's claim of a prospective £2bn NHS deficit), with the further implication that an increase in expenditure would lead to an improvement in at least some of these ratios.

As summarised by the BBC, "The UK is in the lower half of a league table of 30 nations for health staffing. In 2012, the UK had 2.8 doctors and 8.2 nurses per 10,000 people, compared with averages across the OECD of 3.2 and 8.9, respectively. The EIU argues that the starkest differences are apparent when it comes to physical resources ,with the UK sitting near the bottom of the OECD league table. The UK has just 2.8 hospital beds per 1,000 people against an OECD average of 4.8. For equipment such as CT and MRI scanners, availability is less than half the average". The problem is that this tells us nothing about the structural factors that influence the ratios (i.e. context), nor anything about the efficiency of capital utilisation (i.e. what these numbers represent).

The EIU report itself has this to say about resources: "Hospital beds, admittedly, are in decline nearly everywhere: many countries have been steadily closing wards and consolidating hospitals. This is partly to reduce inefficiency and fixed overheads, but also reflects the fact that modern medical techniques—such as keyhole surgery—have cut the length of hospital stays." In other words, the lower ratio of beds per capita could indicate a number of positive developments, from more effective preventative care through more efficient resource management. I'm not suggesting that this is necessarily the case (though there is plenty of evidence that the NHS is efficient in managing its resources relative to other health systems), but that trying to assess performance using isolated data like this is foolish.

The biggest structural factor influencing resource ratios is the extent of private enterprise. Though the traditional critique of centrally-managed institutions is that they create waste due to the dispersed knowledge problem, markets typically produce over-capacity. This is partly by design, the aim being to provide headroom for peaks in demand, but also partly the result of market churn: as providers enter and exit the market, some capacity will be temporarily duplicated. In theory, this overhead is offset as better, more efficient providers compete the weaker out of the market, however there are problems when this theory is applied to inelastic public goods. An example is the way that free schools open in areas that aren't short of school places. The same number of pupils is now spread over a larger capital base, so the capital/pupil ratio is higher, but you also have higher running costs because of fixed charges (e.g. building maintenance). In terms of capital allocation efficiency, the situation has got worse.

Similarly, the lower ratios of doctors and nurses could point to greater efficiency, not less. If we invest in preventative care and better social care for the elderly and disabled, we'll end up with fewer hospital beds. Hospital consolidations will also reduce staffing and equipment ratios. Though they often face local opposition, mergers are more likely to occur in a nationally-coordinated system than in a free market where you have to rely on provider exit. In terms of high-tech equipment (the modern equivalent of the "machine that goes ping"), the distribution of this is heavily influenced by commercialisation. Hospitals used by the rich tend to have more kit than they need (over-capacity to avoid waiting times and meet peak demand), while hospitals used by the poor tend to have less. The UK's greater equality of access, relative to the OECD norm, means that the "average" the UK falls short of will include a degree of structural waste that we have avoided.

Tellingly, the media coverage of the EIU report ignored the section on staff costs. This was because the NHS pays pretty low wages, relative to its international peers, with one notable exception: "Although figures are scarce, OECD data for 2011 suggested that self-employed GPs in the UK are the highest-paid in the OECD, earning 3.6 times the average wage. This includes most family doctors. The figures for salaried GPs (which may be employed by a GP practice) and hospital doctors are far more modest ... To solve the growing problems in A&E, for example, there needs to be better access to out-of-hours care elsewhere. Nevertheless, asking GPs to take on extra responsibilities fits poorly with efforts to restrain their wages." The significance of this is that GPs represent the largest (and longest-established) incursion of private business into the NHS.

The EIU report adds to a growing body of international health system comparisons, such as the US-based Commonwealth Fund (their latest ranking puts the NHS well out in front). Though there are variations in the way that rankings are arrived at, the broad consensus is that the NHS is the most cost efficient (because it is a nationwide system with minimal charging) and equitable (high levels of access and lower levels of inequality of treatment) but has limitations in terms of overall patient outcomes, though this last is more broadly the product of public health policy (i.e. including factors such as diet, poverty, housing, alcohol consumption etc).

In short, the challenge for the NHS is one of management (the efficient allocation of resources), not just an issue of more money to boost resource ratios, though I'm sure a few more computerised tomography scanners per million people would not come amiss, particularly if they go "ping". This raises difficult questions for those tired of "constant top-down reform", but it also highlights a further risk associated with privatisation and outsourcing, namely the fragmentation of strategy and the capacity inefficiencies inherent in markets.


  1. I'll read again, but thanks for addressing an issue I have thought about but not really decided on - on principle I don't care if NHS services are privately provided if free(ish), but what actually happens if services are privately provided?

    Obvious q I know, but less obvious answer

    1. It costs a hell of a lot more.

  2. How about a follow up post on the impending demise of the NHS?

    Or perhaps leave it a while, I'm depressed enough already.

  3. Today is a great day to join the Labour Party

    Standard Membership £3.88 per month