In hard times, many a government has spent money it does not have on infrastructure projects it does not need.
Japan has spent trillions since 1990 on projects including an art museum with no paintings, an airport so empty it offers passengers money to use it, and a $70 million bridge to an almost deserted island.
At its peak, Japan spent more per year on public infrastructure than was spent building the entire Panama Canal.
And as you can see from the graph, it did not work.
We should be incredibly cautious about embarking on large taxpayer-funded infrastructure projects.
But there is a non-Keynesian, market based rationale for infrastructure investment during recession.
Whether we like it or not, large parts of our infrastructure are either provided by the state or by massively dominant monopolies that exist hand in glove with state regulators.
During recession the state can take advantage of the fact that capital investment can be done cheaply. However, this perspective demands that public infrastructure projects meet three requirements.
First, they must be short-term projects completed during the period of economic difficulty.
Second, each project must be designed to deliver a major, long-term economic boost.
Third, where possible these projects should be done at arm’s length from the state: facilitated rather than executed by government.
The Government’s flagship infrastructure scheme, High Speed 2, fails all these tests.
A £30 billion, 25 year long prestige project is precisely the sort of white elephant we must avoid.
To make infrastructure spending work for taxpayers in a recession, the money should be spent on smaller projects with a quicker return.
On the railways, that means improving existing lines, clearing bottlenecks, and building more conventional lines in congested areas.
And if you really want to give Britain a competitive advantage, then perhaps we should focus on a modern technology that has been handicapped by the lack of a real competitive market in its provision, namely our digital infrastructure.
With tiny fraction of what we would spend on HS2, we could create a broadband network which puts Britain ahead of the world in its digital infrastructure.
By 2013 South Korea will have one Gigabit broadband direct to every home – ten times the speed of the most ambitious British plans.
As it stands only 0.2% of UK households have a super-fast broadband connection.
Similar arguments could be used for unclogging the choke points on our roads, again with significant more private sector involvement.
The government has been thinking out loud about that. I would suggest that they stop musing and start acting, because if it is ever necessary, it is necessary now.
As for airports, well I will leave that to Boris.
But let us be clear, spending money that we don’t have for its own sake is a bad idea, sometimes promoted by businessmen with a vested interest of their own.
If the project does not stand up on its own economic benefits it should not happen.