"the government steps in, briefly nationalizes them, sells off the parts (the shareholders get nothing, of course), and life continues"

That's the solution I would have preferred in 2008, but the political system has shown it can't commit to that solution. You lose any support for that resolution plan from conservatives at "briefly nationalizes", because they're afraid any nationalization won't turn out to be brief at all. So nationalization is off the table, and we have to stumble through with bailouts.

Per the article, size is correlated with problems but not the cause of them. It happens that large banks are the ones who are more likely to get bailed out, thus they are incentivized to take more risk, so they do. It is the potential for government bailouts/protections that is the root of the issue, thus the proposal is basically to stop doing that.

The problem is not the bailout, per se, the problem is that the bailout didn't wipe-out shareholder value (and executive compensation) in the process.

the problem is the bailout, just don't bail them out.

Your first argument rests on the idea that they will pass those economies of scale on to consumers, which I have never found to be the case. In my experience, smaller banks are consistently more competitive; they pay higher rates, charge lower rates, and have fewer fees.

If you do that you're still subsidizing banks. The simple act of telling creditors that if the banks gets into trouble the government will step in, wipe out shareholders and keep depositors and creditors whole, makes credit more available to them. Of course I'd rather lend to a bank with that kind of backstop than a small savings bank/credit union.

And then if I'm a bank manager, I'm going to borrow as much as possible, and do a lot of high-risk, high return type trading, if it works out I make a ton of money, if I lose it's the government's problem.

Finally, it's not that likely for a Treasury official to tell a TBTF bank they're taking it over - it's an administrative nightmare, and the banks have captured the regulators.

So if a bank gets weak, markets will keep extending credit as long as there's a government backstop, and it won't get taken over, and management will keep playing double or nothing with taxpayers' money.

Temporary nationalization of banks was suggested in 2008, but the idea is not politically viable on the right in the US.

We have a structural problem, when things are "too big to fail", the normal democracy does not work anymore, the bank gets too powerful on its own and goverments can't regulate it effectively.

Moreover, bigger companies are not necessarily more efficient. A smaller company (or country) has the agility of a shorter decision chain.

It costs a lot less to save a small bank than a big bank, and the risk is spread out over multiple organisations. If we had (for example) small banks that were more specialised, if the housing market falls through, those sorts of investment banks would disappear, but others should stay relatively untouched. That was the whole point of Glass-Steagal : if the investors mess up, at least make sure that we don't have to wipe out Grandma's savings.

The sums of debt some of these banks got into swamped their share value, and in some countries were a fair chunk of the GNP. But I certainly wish we had removed shareholders from any bailed out bank.

One issue is some banks were on the border on if they needed bailing out, and shareholders might prefer the risk of catastrophic failure than taking a bailout which would lose them their shares.

It's like asking why we don't let a structurally unsound skyscraper just go bust. Not a problem in isolation, but not really safe in the midst of a crowded city block.

>The large banks are (arguably) more efficient because of economies of scale, eg. less duplicated management

Someone has never worked in a large bank before.