Liberal economist Larry Summers says government must get bigger – Larry Summers is probably wrong on this. A brief genuflection: Summers is, as with Paul Krugman, both a very clever man and an excellent economist. He's also a political operative (Krugman is a political columnist) and excellence from the academic field doesn't quite ooze over into the murky world of spending everyone else's money (which is what politics is).

His arguments bear examination – the first being that the federal government largely takes care of the old, and there're more old people around today, thus the government needs more money. Except, we're supposed to be funding all that Social Security and Medicare out of those trust funds and lockboxes that we've all been paying into for a lifetime. What do you mean there's not enough cash there for us all to get what we've been promised?

Could it be that politicians have been promising us more than they've been charging? You know, the basic complaint from the Right about this welfare state for the Baby Boomers over these past few decades?

It could be just that. What Summers is saying now is exactly the counterpart to what President George W. Bush was saying about wanting to move Social Security over into private savings, at least in part — something he was shouted down for arguing, that for a couple of decades we weren't paying in enough for what we were being promised. It's even possible that Summers was doing some of that shouting.

Summers' second argument is that inequality is increasing and that, if a function of the federal government is to reduce inequality, then they need more money to do so. More tax and redistribution, obviously.

Yet, this is to fall foul of something called Worstall's Fallacy – we cannot measure or judge what we should be doing about something until we first take account of what we already try to do about that problem. Sure, we could say that the government should be reducing inequality (I don't, but that's another matter). Should we be looking at the inequality of market incomes, as Summers is appealing to? Surely not, we should be looking at what actually matters, the inequality of consumption after all of the things we do to reduce the inequality of consumption.

Part of this is being done by the digital revolution – you me and that homeless guy using the library computers all have exactly the same access to Facebook at exactly the same price as Bill Gates. We might regard this as a trivial reduction in inequality, but it is still a big reduction from when we all had to use the Postal Service to send cat pictures to each other.

The United States is a remarkably equal society in terms of access to much of life. Sure, some can have Aston Matins while others are the fifth owner of a Honda Civic, but there are few who cannot have a car at all. Almost everyone has just as much access to Big Macs and Cherry Coke as Warren Buffett does.

What inequality there is in consumption today is largely about positional goods — there are only so many Renoir paintings, houses in the Hamptons, and due to their limited supply, the prices of those things are quite high, thus their distribution is indeed highly unequal. In general, there's not so much equality in terms of roofs over heads, clothes on backs, and calories in stomachs. These past 50-80 years are probably the first time any society on the planet has been able to say that.

There's also the entirely true argument that we already do a number of things to reduce inequality.

The Census poverty numbers just released, as an example, don't in fact show the number of people living in poverty — They show the number of people who would be in poverty if we weren't doing something about it. The U.S. is unique in the way that it measures poverty – everyone else does it after the effects of taxes and welfare.

We know very well that millions are lifted out of poverty, as they should be, thanks to Medicaid, the earned income tax credit, food stamps, Section 8 housing, and all the rest of the hundreds of billions spent. We pretty much don't give poor people money these days — instead, we give them things, vouchers, tax rebates, and services. But we don't count the effect all that work has when calculating the number of people in poverty.

When we do count all of this, the roughly 20 percent of children living in poverty falls to about 2 or 3 percent — as I've noted before, that's pretty good for government work.

Which is where Worstall's Fallacy comes in — if we want to decide whether we should do more or less poverty alleviation, we need to know how much poverty is left after what we already do? Summers' mistake is that he's starting from market income inequality and demanding that we must do more. But what we really want to know about is consumption inequality, and we want to know about that after the effects of the tax and redistribution we already do.

Summers' fourth argument is the need for more defense spending, and even I'm not going to dip a toe into that nest of vipers.

It's his third argument which really fails, and fails hard to my mind. He points out that what government buys (primarily education and healthcare) has a higher inflation rate than the general economy. Thus, if the government is going to continue to buy those things, they need more money.

Yet, it's the twin strands of William Baumol's research which show us the answer here (strands which Summers is intimately familiar with, by the way).

Yes, education and healthcare are services largely composed of pure human labor, and as the society gets richer these will become more expensive relative to manufactured goods. It's easier to increase productivity in manufacturing than services — we can stick a new machine on the line, but we can't get a string quartet to simply play faster to increase their output. A richer society means more expensive labor — that's a definition of being richer, that we all get paid more. All labor becomes more expensive, the amount of labor in a service drops less than that in a manufactured good, and services become more expensive than manufactured goods.

This is the Cost Disease.

Yet, the other major strand of Baumol's work was, "How do we actually improve innovation?" The answer being that invention, the creation of entirely new things, can be done by both government and business. But innovation, or increasing productivity, is something that planned, government systems just cannot manage in any manner comparable to what markets achieve.

Putting these two strands together, we find that precisely because something is a service, where the Cost Disease applies, we should be using markets in order to do that more difficult job of increasing productivity. Or, to run this the other way around, because medicine and education are places where it's difficult to increase productivity (therefore they have higher inflation rates) the government shouldn't be running them.

And if we didn't have government running those two to a large extent, then we wouldn't have this problem that required government to get more money, would we?

As I've said, Larry Summers is both a clever man and a very good economist. But when he's talking politics, he's as fallible as any other.

Tim Worstall (@worstall) is a contributor to the Washington Examiner's Beltway Confidential blog. He is a senior fellow at the Adam Smith Institute.

If you would like to write an op-ed for the Washington Examiner, please read our guidelines on submissions here.