What ‘warfare versus welfare’ gets wrong about real-life economics

Lord Robertson’s claim that the UK cannot defend itself with an “ever-expanding” welfare budget has resonated loudly, given his previous positions as a Nato secretary-general and UK defence secretary. Following up on the UK’s 2025 strategic defence review, which he led, Robertson warned that low investment is leaving UK security “in peril”.

The comments have instant appeal in one sense. Defence is indeed awarded a far smaller share of the pie than social protection: 6.5% of total managed expenditure for 2026/27 against 28%, according to estimates.

The UK’s budget deficit is adding to already high public debt, and the IMF has forecast that Britain will be hit harder than other countries by the economic effects of the Iran hostilities. The government is already seeking savings from other departments as it tries to raise defence spending to 2.5% of GDP by 2027.

But the idea of a simple trade-off, with more weapons requiring less welfare, confuses two very different types of public spending.

Defence is part of “final” public expenditure, funding armed forces’ pay and the weapons and equipment they work with. This takes up money that can’t be assigned elsewhere in the budget, and consumes a share of national output when the government spends it.

In contrast, the welfare budget consists mainly of “transfer payments” that shift income between households. Some transfers are made according to assessed need, others also depend on past national insurance contributions. All represent a redistribution of income without any exchange of goods or services, leaving recipients to decide what to do with the money. This allows prices to steer spending away from scarce resources, while some is used to repay debts or clawed back in tax.

Demands on the public purse

As the government’s overall budget is in deficit (to the tune of around 4.5% of national income in 2025/26), it is true that welfare payments compete with other demands on the public purse. But the boost to recipients’ income is still largely offset by taxes collected from better-off households.

In principle, a country could raise its welfare budget to 100% of its GDP, by collecting all the money generated by production as tax and then paying it out to households. It would compromise efficiency, as happened in Europe’s “state socialist” countries before 1989. But such an economy could still function.

In contrast, raising the defence budget even to 3% of GDP – the UK’s target for the next parliament – will cause political and economic strain. This is due to the trade-off against other final expenditures, including healthcare, education and policing – all equally vital for national survival and security.

The UK and other countries with large welfare systems have reformed them with the aim of adding at least as much to output as to demand. Transfer payments are increasingly designed to keep people economically active, moving into new and more productive work. This matching of extra income to extra production keeps the inflation risk low, even if the government is “printing money” to fund some of its transfer payments.

Extra defence spending carries greater inflation risks. Paying for more weapons and military training generates new income and demand for consumer products. At the same time it can divert workers and materials away from civilian production, into military hardware that is intended never to be used.

Stronger defence could boost production as much as consumption if, as many advocates claim, it stimulates investment and innovations that other industries can adopt. The Manhattan Project remains a standout example of “mission-oriented” military spending that sped the arrival of new technologies and methods of organisation.

Studies confirm a pick-up in innovation and growth after major increases in military spending. But these tend to focus on the US and trace the improvement to increased research and development (R&D). Growth might be stimulated equally well, making more weapons and more welfare an affordable option, if greater sums went into R&D without a link to war preparations.

Of course, defence can be counted as an even more productive investment if, through effective deterrence, it prevents costly wars that would devastate civil production.

But again, there is an important difference between investing in military hardware and in social protection. The welfare bill is hard to forecast, as it varies with the state of the economy and trends in income and employment. But when transfer payments enable people to recover their health or acquire new skills and return to work – or when they keep pensioners out of poverty – the government gets a rapid return on its investment and reduces longer-term costs.

Investment in more soldiers and equipment may be easier to control in the short term. But it commits the government to maintenance and upgrades over the long term, without which the fighting capacity can soon become non-operational. The UK has a history of cost overshoots and delays keeping tanks and ships out of service. That’s why a Treasury set on cost-effectiveness will always choose butter over guns.

Alan Shipman has previously received funding from the British Academy/Leverhulme Trust and the Harry Ransom Center, University of Texas at Austin.