The 21 st thing is that a larger government actually makes economies more flexible: thus we should have larger governments in order to increase the necessary flexibility of the economy. And if the first part were true then the second part might indeed follow: only it isn't, at least not in the sense that Chang means it.
As part of his argument there's one thing that Chang does which is really very naughty indeed. He compares the growth rates of various European countries to that of the US. And he divides the growth rates into two periods: early 1950s to late 70s, late 70s to now (or what was now when he wrote a couple of years ago). He's right that the European economies, with their larger state sectors, grew more quickly than the US did in the earlier period. Hence the claim that larger government can mean more economic growth. Except, well, there's just one thing missing from this calculation: the Wermacht. As the perceptive will have noticed the German Army did have something of a European tour in the years immediately preceeding the early 50s. And the destruction of getting them to go home again was considerable: something which did not happen to the US at the time.
Just as we expect a developing country to have a higher growth rate than a developed one, given that copying is easier than being at the technological cutting edge, so we also expect an economy recovering from a total war being fought on its territory to grow faster than one which is not. So while the growth rates are true we cannot use them as proof that larger governments will create more economic growth.
The real problem with Chang's position though is that he confuses two entirely different things. He talks about employment inflexibility: the way in which it's difficult to get fired and thus the workers all feel secure. He also talks about the existence of a decent welfare state: unemployment pay, health care, retraining opportunities and so on. The problem is that he sees these as being equal: both increase the security felt by the workers and thus increase their flexibility. Which is untrue: they work in very different ways indeed.
It is true that a decent welfare state does lead to greater flexibility in the economy. The workers (and everyone else in fact) will be less stuck in their ways if they know that a change in the economy does not mean destitution. But job security works entirely the other way around: those who are too secure in their jobs won't accept any change at all. Thus reducing the necessary flexibility of the economy.
The reason that this becomes important is because Chang points to the Nordics as evidence of his assertion that you can still have decent economic growth with a large government sector: indeed that it increases growth to have a large such sector. But the very success of the Nordics argues against all of the other strictures about free markets and capitalism that he wants us to understand and adopt. For it is true that they do have large and generous welfare benefits: the unemployment pay, the retraining and so on. They also have decent economic growth. But what they don't have is the sort of regulation, planning and government intervention into the economy that Chang proposes. Look behind the tax numbers (necessary to pay for those benefits) in the economic freedom index  and look instead at everything else. They have less regulation of markets than we do, greater economic freedom than even the US, less intervention into capitalism than just about anyone other than Hong Kong. Which is what makes the places work of course: as Scott Sumner is fond of pointing out, Denmark might well be the most classically liberal economy on the planet  underneath that welfare state.
All of which I admit I find rather amusing. It is true that the Nordics are nice places to live, despite those crippling tax rates (almost all of which are tumbling down). It is indeed true that they've had very decent economic growth over the years and decades. It's entirely true that they have a lavish social insurance system. But those economies work precisely because they ignore, do absolutely the opposite of, everything else that Chang proposes a government should do to an economy. They're more free market and capitalist than even the US: which is why they work. Indeed, it's probably true to say that the only way in which you can have a social safety net like they have is if you allow capitalism and markets to let rip: how else can you possibly afford to pay for it all?