At least I did until I read this quite terrible argument. Matt’s trying to argue that government will inevitably get larger so we needn’t worry about how much of the economy is swallowed by the Congresscritters….it’s all part of a natural progression, you see?

Meanwhile, one needs to understand that, somewhat counterinuitively, when you have a very efficient economic sector what happens is that it tends to go away. Consider agriculture. Our modern-day agricultural technology is way better than what was available 200 years ago. But agricultural progress hasn”€™t meant that everyone goes to work in the super-charged high-tech agriculture of the future. It’s meant that more food than ever is grown with fewer person-hours of labor than ever. We should expect this to continue apace. For all the talk of trade’s impact on American manufacturing, the bigger issue has been automation and robots. But either way, even though people will continue to consume manufactured goods”€”just as we still eat”€”manufacturing will be a less-and-less important part of the economy. Not because manufacturing “€œisn”€™t important”€ but because it”€™ll get more efficient. And that’s how the whole private sector part of the economy will go. Markets, doing their work, will make those sectors more and more efficient leading them to shrink as a share of the overall economic pie.

There’s a certain amount of truth to that set up, yes. We do need to distinguish though between employment and contribution to the economy as a whole. Manufacturing employment, for example, has very definitely been falling, while manufacturing output has continued to rise. So we need to be careful when we talk about a sector shrinking….do we mean in employment or in output? And further, do we mean shrinking as a whole or shrinking relative to other parts of the economy? Farming is a smaller percentage of the economy than it was 100 years ago but total output is vastly higher….it’s just that the non farming sectors of the economy have grown even faster than farming has.

But where Yglesias goes horribly wrong is in the next part:

Markets, doing their work, will make those sectors more and more efficient leading them to shrink as a share of the overall economic pie. What will be left is big government. Or, rather, bigger and bigger government. Teaching kids. Taking care of the elderly. Patrolling the streets. Making the SUPERTRAINS run on time. And it’s going to be fine. Which isn”€™t to say we should crank spending up to 93 percent of GDP next year.

That’s an extraordinarily bizarre view of the economy, that there are three sectors, there’s agriculture, manufacturing and government? Whatever happened to services?

The vast majority of the economy is now indeed in services. This is everything, well, everything that isn’t agriculture or manufacturing really. And that everything does indeed include government, but just because government is a service does not mean that all services are best supplied by government. Cutting hair is a service and I’m certainly unconvinced, to put it mildly, that we want the nation’s barber shops run from DC. Insurance is a service and I’m absolutely certain that we don’t want, say, flood insurance, provided by the politicians. You may have noticed that said politicians are the only ones who are willing to insure houses built in flood plains, something that private insurers are too intelligent to do.

There are indeed some services which are better provided by government. I’m just fine with the idea that we’ve only one military: the European feudal days of anyone who could raise one having their own army were really not all that much fun. Similarly I think having just the one, rather than competing, criminal justice systems is just dandy. But to say that simply because things are services then they should or must be provided by government is simply nuts.

Educating kids for example: there are countries (Holland and Sweden come to mind) where any qualified group of teachers can set up a school and the government sends a check for every pupil they attract. Education vouchers in short. It simply doesn’t have to be government that does the educating. There are similarly other countries that have truly private education systems. Taking care of the elderly is something that’s been done not by government for millenia. We even have a phrase for the mechanism: a family.  Trains in the UK were run privately for 125 years of their 190 year history (roughly). American freight trains are still private, it’s only a few passenger services which are government run.

But to go from nuts to truly bizarre:

And it also does mean that it’s always important to find ways to make the public sector more efficient and more effective. It can be done. Public agencies are better-and-worse managed and offer better-and-worse performance. But it’s difficult to do and it doesn”€™t happen automatically the way it does in a well-functioning market.

There’s an idea called Baumol’s cost disease. Essentially, it will always be more difficult to increase productivity in services (the same thing as “making them more efficient”) than in agriculture or manufacturing. Baumol’s own example was an orchestra: you can’t make them more efficient or productive by getting them to play twice as fast. True, but we can increase the productivity, the efficiency by which music for us to listen to is provided, by changing the technology. Radio, records, CDs, iPods, all make an orchestra more efficient at what we actually want, which is to be able to listen to an orchestra.

The same is true of almost all other services: it might be more difficult to invent the gramophone (oddly, first marketed as a form of dictaphone while the telephone was first mooted as a way of listening to concerts at a distance) than it is to make an assembly line more efficient, but we can indeed increase productivity in services. But it’s precisely this difficulty that makes it so much more important that services, except those that are truly necessarily done by government, are indeed left to markets. For as we’ve already seen, even Matt thinks that markets increase efficiency and the more difficult such an increase is the more we need to use the best method we have of doing so.

As I say, I used to think Matt Yglesias was smart. In fact, I still think he is, he’s got a philosophy degree from Harvard, a university that wouldn’t, even if I tried really hard, allow me to sully their hallowed halls, not with my SATs. But I think it’s obvious that he’s markedly ill informed on this matter at least.