The FT reports that thousands of cash point machines might close. One of the very few worthwhile financial innovations of my lifetime might therefore be about to go into reverse. This poses the question: shouldn't we take the possibility of technical regress more seriously?

Certainly, it was common before (pdf) the industrial revolution; plumbing, for example, was better in parts of India around 2000BC than it was until quite recently in the west. The dark ages weren't wholly misnamed. But we have also seen it in our lifetimes.

For example, it takes longer now to fly across the Atlantic than it did in the 80s when we had Concorde - especially if you consider longer check-in times and time spent getting through airport security. Equally, some train journeys take longer now than they did years ago. And of course, men are no longer travelling to the moon.

It's not just in transport that we've seen regress. Older readers might remember that we used to get two postal deliveries a day, one of them before lunch. Paul Romer has claimed (pdf) that macroeconomic research has seen intellectual regress. The fact that companies have for years built up cash piles suggests that banks are no longer as good at financial intermediation as they were - not that they ever (pdf) much improved (pdf) in the first place - perhaps because the rise of intangible assets deprives firms of collateral against which they can borrow. Newspapers used to keep readers well-informed with numbers of foreign correspondents and detailed parliamentary reports. (Brendan O' Carroll, for example, managed to track down his grandfather's killer by using newspaper reports at the time). Today, they prefer to flatter readers' prejudices. Most worryingly of all, antibiotics are becoming less effective which might cause a general regress in medicine.

You might object that we still know how to do most of these things, but it's just that the costs of doing so have risen prohibitively, For my purposes, this is irrelevant. BITD we spoke of the socio-technical conditions of production: we knew that we cannot separate technology from the social circumstances that facilitate or not its use.

Regress, then, is a real possibility.

This can help explain cyclical downturns. If we combine Xavier Gabaix's point (pdf) that recessions can start from large individual firms getting into trouble with Daron Acemoglu's description (pdf) of how network effects can amplify such fluctuations, we have a story of how regress can cause recessions. This, I think, was true of the 2008-09 recession. Banks' ability to produce credit declined. That technical regress caused a collapse the effects of which are still with us. What's daft about real business cycle theory is its conception of labour markets and use of representative agents, not its identification of the cause of downturns.

This poses the question. What possible mechanisms might give us longer-term regress. Here's a non-exhaustive list:

- Baumol's disease. Where productivity cannot improve, relative costs increase over time as wages rise. This might render some services unprofitable. This is probably the story of postal services.

- Intangibles. We used to think that one route through which technology improved was that failing firms' assets would be bought by better firms. This may well be true of physical assets. But is it so true of intangible ones? If a firm's know-how consists in organizational capital, it's demise might cause knowledge to be lost. And as Luigi Zingales has said (pdf), "a temporary shock such as financial distress may have very long-term consequences."

- Path-dependence. Brian Arthur has described (pdf) how a poor technology might gain a temporary advantage over a good one because of a lucky accident. Producers then continue piecemeal improvements in the inferior technology with the result that a potentially better technology is abandoned. This might be the story of steam-powered cars, or perhaps economic research.

- Incentives. If there are big incentives for rent-seeking rather than innovation, we'll get less of the latter. Worse still, fighting over shares of income can reduce that income. This is the story, in different ways, of the natural resource curse and John Kay's marvellous parable of the ox.

Perhaps a bigger danger, though, is simply the rise of anti-intellectual and anti-scientific attitudes - of the sort we see variously in opposition to vaccinations, the belief in a post-truth world and the attack on universities.

Progress is more fragile than we suppose. We must ask how we can shore up the material and intellectual preconditions of it.