The Treasury’s Real Investment Blind Spot

Before Britain builds a railway, road or hospital, officials try to weigh its benefits against its costs. Labour thinks these evaluations have been too restrictive, deterring investment and economic growth.

As chancellor, Rachel Reeves revised Treasury guidance to give projects outside London a “fair hearing”, emphasising the importance of hard-to-price benefits and warning officials against vetoing schemes for weak benefit–cost ratios. Her successor, John Healey, wants a lower discount rate, making benefits decades away count for more in evaluation today. Andy Burnham has argued that the Treasury’s Green Book sends investment where the money already is. No 10 North is his bid to support projects beyond London.

But what if the Treasury’s biggest bias in evaluation goes the other way? There’s a strong case its appraisals framework flatters many projects, making them look more valuable than they really are.

Suppose a £1 billion bridge promises £1.2 billion in benefits. That’s a £200 million economic gain, right? Not necessarily. Raising the tax revenue to fund the project brings wider economic harm. By lowering the reward for work, investment, or trade, taxing doesn’t seamlessly transfer funds to the Treasury. It changes behaviour. People work less, rearrange their finances, or forgo moving due to tax disincentives.

Economists call these effects the deadweight loss of taxation. It means raising one extra pound of tax costs the economy more than a pound. Yet, as a recent newsletter by the economist Duncan McClements reminds us, the Treasury’s Green Book, while acknowledging such harms, instructs officials to ignore them for most appraisals. The result is overly rosy benefit-cost ratios.

The Treasury’s position has a certain logic. Ministers have already fixed departmental budgets. If a money pot must be spent, the tax distortions to raise it are irrelevant to project selections. Yet governments also choose how much to spend. We could do fewer projects, so trimming highly distortionary taxes.

This matters, because as the tax burden has risen to its highest share of GDP post-Second World War, governments have raised revenue in ever more complex, destructive ways. Earn between £100,000 and £125,140 and the withdrawal of your personal allowance takes 60p in income tax from each extra pound earned. Transaction taxes like stamp duty deter moving. Different taxes on salaries and dividends encourage company owners to change when they take income. All create large economic costs.

How much does it cost to raise an extra pound in revenue? In 2006, Henrik Kleven and Claus Kreiner, two economists, estimated around £1.26 for Britain under one set of assumptions about taxes on work and our willingness to supply labour. A European Commission study in 2013 put the cost of raising £1 through labour taxes here at £1.81. McClements estimates it at £1.80 to £2.20 using a broader measure of responses to tax and accounting for the fall in private investment when people work less.

The precise figures are debatable. Yet, the implication is that, once tax costs are considered, many approved projects destroy economic value, on net. The A1 upgrade from Dishforth to Leeming, for example, opened in 2012, promised only £1.40 of benefits per pound of cost (a later evaluation found just 90p). East West Rail’s western section scored £1.30 including wider economic benefits and Scotland’s Borders Railway £1.30 by including fuzzy “social inclusion” gains. If raising funds cost £1.40 or more, none would have cleared the bar.

Across the government’s Major Projects Portfolio, McClements identifies 16 schemes today worth £7.6 billion a year with benefit-to-cost ratios in the danger zone of one to two. This includes the New Hospital Programme (£1.26), the Affordable Homes Programme (£1.68) and the smart meter rollout (£1.35). The combined infrastructure pipeline aggregates a figure of just £1.30.

The Treasury gets accused of penny pinching. Yet proponents of more government investment should be grateful it refuses to account for the cost of the pennies it raises.