Reflections from IPPR’s new report launch – Bismarck versus Beveridge Revisited

Capital concerns

I was in the room this week when IPPR launched Bismarck versus Beveridge Revisited, the most comprehensive comparative analysis of health system funding models produced in years. Health Secretary Wes Streeting gave the keynote response, addressing the increasing pressure the Government faces to switch the NHS to a social health insurance model – backed heavily by the new evidence base produced by the IPPR.

The analysis draws on data from 22 OECD countries across five performance dimensions: capacity, access, equity, quality, and efficiency. The headline finding is unambiguous. No funding model systematically outperforms another. Performance varies more within model types than between them. The best Scandinavian tax-funded systems outperform the best social insurance systems on most measures. Tax-funded systems hold structural advantages on equity – lower out-of-pocket costs, less reliance on voluntary insurance – and on administrative efficiency, spending 2.2% of health budgets on administration compared to 3.5% in social insurance systems. Were the NHS to match the social insurance average, it would need to redirect £7bn annually from patient services to bureaucracy.

Streeting endorsed the findings unreservedly and was explicit about the political stakes. Nigel Farage has stated publicly that he wants an insurance-based system and does not want the NHS funded through general taxation. Reform has released no health policy ahead of the Welsh elections. Streeting called this deliberate – “keeping their powder dry and counting on us to fail” – and challenged Reform to answer publicly what premiums patients would pay, what protections would exist for the uninsured, and who funds the additional administrative cost.

Settling the funding model debate, however, is not the same as knowing what to do next. And the more important – and largely unanswered – question is allocative, not structural.

Model misbehaviour

The IPPR data is sobering. The NHS records 71 treatable deaths per 100,000 – deaths from conditions where timely care should have prevented a fatal outcome – the second worst of the 22 countries analysed, ahead only of the US. Capital formation stands at 0.4% of GDP, roughly half the comparator average and less than half that of high-performing tax-funded peers such as Australia, Norway, and Denmark, all at 0.9%. The NHS carries 2.4 hospital beds per 1,000 population and 19 diagnostic machines per million – significantly below average on both.

These are not structural failings. They are the consequence of chronic underinvestment and, critically, of spending in the wrong places. Compared to high-performing peers, the UK invests substantially less in primary and community care, and dramatically less in long-term and social care. The Netherlands spends approximately three times the UK’s per capita social care spend. Denmark and Finland spend more than double.

The cost-effectiveness case for rebalancing is stark. Generating a quality-adjusted life year through public health budgets costs approximately £3,800. Through NHS acute spend, the same QALY costs approximately £13,500 – more than three times as much. CF’s analysis of unmet care gaps in cardiovascular disease and chronic kidney disease shows what this misallocation produces in practice. Across ICBs and PCNs, large proportions of patients with diagnosed conditions remain undertreated against NICE guidelines – only 37% of CVD patients have cholesterol under control, and only 28% of people with type 2 diabetes have optimal HbA1c. The downstream cost of that undertreatment, in avoidable hospitalisations and acute demand, is measurable and large.

The IPPR report identifies four priority investment areas: capital infrastructure, primary and community care, long-term and social care, and public health and prevention. Streeting endorsed all four and pointed to the neighbourhood NHS as the vehicle for the primary and community shift. On capital, he confirmed openness to private investment alongside public funding — explicitly distancing from PFI while citing the Waltham Forest model as a better template. On the Federated Data Platform, he was unequivocal: it is not going away, and his test is ethical governance, not provider identity.

The productivity conundrum

I asked from the floor how the government intends not just to increase investment in primary, community, and social care, but to ensure it delivers strong returns – given the known variation in what that investment actually produces. The answer was less convincing than the diagnosis.

This is the harder question. CF’s analysis shows an average return on investment of approximately 4x from spending in primary, community, and social care. But that average conceals variation from less than 1x to more than 8x depending on the intervention, the population, and the system context. The average is not the point. The distribution is. Increasing investment without the analytical infrastructure to distinguish high-return from low-return deployment will not close the performance gap. It will reproduce it at higher cost.

Streeting’s instinct – to name underperforming trusts publicly and drive consistency of improvement – points in the right direction. CF’s own productivity analysis shows that the gap between the highest and lowest performing NHS trusts on output per unit of resource is large and persistent. Closing it by raising the floor is achievable. But it requires commissioning decisions to be driven by evidence of return, not by formula and convention. The mechanism for translating investment into consistent returns is not yet fully articulated – and the Welsh elections notwithstanding, that is the question the Health Secretary will need to answer next.

The debate that now matters

The insurance model question has been answered. The political consensus has been re-established on firmer evidential ground than before, and IPPR’s contribution to that is significant. Lord Darzi’s foreword – drawing on his 2024 independent investigation – gives the report a weight that political opponents will struggle to dismiss.

The debate that now matters is less ideologically satisfying but more consequential. It is about whether the NHS can move from a system that spends reactively on acute demand to one that invests proactively where the returns are highest. It is about whether the shift to primary, community, and social care will be driven by analytical rigour or by good intentions. And it is about whether the data infrastructure now available to the NHS will be used to answer the allocative question, or primarily to improve operational reporting.

The funding model is not the constraint. The question of what the model is used for – and how investment decisions within it are made – is where the argument needs to go next.