The statement was, at best, ambiguous. Its political consequences more profound. George Osborne said he will not hit his target for a budget surplus by 2020. Was this just a judgement that Brexit would slow deficit closure given his planned spending and tax policies? Or admitting the whole framework set out in the last Budget and previous spending reviews was to be overthrown, and more borrowing allowed? Conservative leadership candidates seemingly assumed the latter – really clutching onto the statement to wriggle free from ‘austerity’ and re-open the spending taps.

It’s not clear why Osborne needed to do this. Short-term uncertainty will undoubtedly hit the public finances. But a ‘Nike-tick’ recovery with catch-up growth in that scenario may well have allowed the target to still be hit. If he believes Brexit is a negative supply shock, then in an age of lower potential growth post-Brexit (and assuming the Government takes no supply-side measure to counteract), then more austerity would be needed, not less, in the medium term. Instead, the ‘leadership final three’ offered up more spending in general, and more infrastructure spending in particular. Amidst this backdrop, the new Prime Minister should acknowledge some hard truths.

Our budget deficit for 2016/17 was projected to be around three per cent of GDP, circa £60 billion, a full eight years after the financial crisis. This was down from around 9 per cent in 2010, when the Conservatives were elected, mainly driven by very modest real terms overall spending cuts (1.5 per cent overall through to 2015/16), tax rises and economic growth. For this parliament, overall real spending was expected to essentially be flat, with growth and rising tax revenues closing the deficit to a surplus.

Allowing this deficit to remain slightly higher relative to plans by not adjusting discretionary spending or tax rates in reaction to any slowing of growth is probably sensible. A catastrophic market reaction is unlikely. But increasing spending further deliberately – i.e. adopting a ‘stimulus’ programme – to run deficits of 4.5 per cent or 5 per cent of GDP would be unsustainable in anything but a year or two, raising the public debt-to-GDP ratio to over 100 per cent of GDP fairly quickly, creating very little room to deal with future shocks and running directly into the fiscal headwinds of an ageing population in the next decade. Add to this the potential impact of lower migration of working age people and these recent pronouncements could have a toxic effect on the public finance outlook.

In other words, to tolerate short-term deficit increases which reflect automatic stabilisers is one thing. To increase discretionary spending massively is another. Britain still needs a credible medium-term deficit reduction plan, which is robust to shocks. This means both a target and spending and tax plans to achieve it, such that the debt-to-GDP gets firmly on a downward path to historic peacetime levels. To throw out one fiscal framework in the last parliament was forgivable. To now throw out another is careless. May and her chancellor must sort this out, pronto.

In the age of deficit reduction, Keynesians have always appealed to the beguiling idea of ‘infrastructure spending’ as the answer to our woes – almost like a free lunch. And now Conservatives seem to appeal to this logic too (Stephen Crabb and Savid Javid in particular). With interest rates low and ‘weak demand’, the argument runs that now is the perfect time for government to borrow to invest. This wouldn’t just support demand now, we are told, but also improve the supply performance of the economy in the longer-term.

This can be true, of course. Good infrastructure, particularly transport infrastructure, can grease the wheels of economic activity. It does make sense to borrow to invest when the costs of doing so are low. If we can find projects that genuinely have high returns, enhance the supply-side, and are cost effective, who would object?

Here we begin to hit the difficulties. Treasury analysis has actually found that Crabb’s plan, to relax spending constraints by taking capital spending out of deficit targets, tends to lead not to much more infrastructure spending, but simply more current spending. It is for this reason that Osborne adopted a hard surplus rule. That’s not to mention that, as the removal of the post-crisis investment ‘stimulus’ showed, temporary spending creates permanent expectation – and much wailing as that spending is subsequently reduced.

But the most important critique of this agenda can actually be observed from the hugely diverse empirical literature: the link between public spending on investment and economic growth, as opposed to actual physical measures of infrastructure, tends to be weak. The literature on the link between different types of infrastructure and growth is varied – though, as the LSE growth commission has outlined, there are many studies that show that good infrastructure can enhance output and productivity in both the short and long-term.

This is intuitive. Whilst good investments can enhance the supply-side of the economy, government procedures, the politicisation of capital spending and lack of market signals can often lead to bad investments being made. Poor government oversight and perverse incentives created by the knowledge of taxpayer bail-outs can lead to significant cost overruns for state-led projects. And, as we observe with planning constraints, as with the government blocking of Heathrow and the expansion of toll roads, the state is often the roadblock to private sector investments where taxpayers would not have to fend the downside risk of failure.

We do need to do detailed regressions to see this. The US, which ranks worse than us than on the quality of its infrastructure, almost always grows more robustly. Japan spent tonnes of public money on big projects, and ended up with bridges to nowhere. Spain, likewise, was left with empty airports. The major investments in road infrastructure and motorways in the UK failed to deliver the promised benefits. On energy, governments grant subsidies to inefficient renewables producers which can actively undermine the productive potential of the economy. When the government does access large pots of money, it often blows them on huge prestige projects such as HS2, rather than much more cost-effective schemes with higher returns such as road improvement projects.

Whilst nobody is against good infrastructure, the liberal critique of public infrastructure spending is that the political process does not have the mechanisms to choose good projects. Indeed, politics can create actively perverse incentives for pork barrel and white elephant spending. What we need is institutions and mechanisms to harness as much private spending on infrastructure as possible, and to allow private actors to take the risks – just as they did in Victorian Britain.

A major barrier to this is the UK’s land use planning system. Indeed, it is precisely things such as this that means there are rarely ‘shovel-ready investment projects’ – which negates the effectiveness of Keynesian demand management in practice, even if you presume it works in theory.

My plea to Theresa May is therefore two-fold: first, maintain fiscal sanity. Set out a clear target and framework for getting debt-to-GDP back on a downward path. Second, think harder than just assuming more investment spending is good – consider barriers to private investment, and institutions which can depoliticise the process by which decisions are made.

This article was first published on Conservative Home.