Consider four issues:

1. John McDonnell objects thusly to the EU-India Free Trade Agreement:

If multi-brand retailing is suddenly and without safeguard opened up to EU retailers such as Carrefour, Metro and Tesco, 1.8 million jobs may be created, but at the cost of up to 5.7 million people working as street vendors.

But there are reasons to agree with Peter Mandelson that those 5.7 million are only 'short-term losers'. Those 1.8 million new jobs will be (relatively) high productivity an high wage ones. As those workers spend their higher incomes, they'll create new demand which will create other, maybe better, opportunities for those displaced street vendors. It's not at all clear that trade liberalization leads to lasting unemployment and poverty (big pdf).

2. One justification for a benefits cap is that the squeeze on housing benefit will force landlords to reduce rents. But in the process, some people might become homeless.

3. Jonathan Portes says it is possible that immigration raises unemployment in the short term, but that in the longer-term, joblessness depends other, macro, factors.

4. A return to national currencies within Europe might be superior to monetary union. But the cost of making the return - a huge banking crisis - is huge.

These look like four different issues. But they have a similar structure. In all cases, we have a problem of adjustment costs. Free trade, limited housing benefit, free migration or floating exchange rates might be superior to the alternatives, but there are costs of making the move.

So far, so straightforward. But here's a quirk. It is, in theory, quite reasonable to say 'state B is superior to state A, but the costs of moving are prohibitively high. So let's stick with A.' But, I suspect, few people take this position. It's far more common to conflate the adjustment costs with the argument that B is actually inferior to A even as a steady state.

Here's another quirk. Our first three cases above all have something else in common. People who think that markets operate reasonably smoothly will believe the adjustments are small, and so in all three we have a debate between market optimists and pessimists. But the issue of adjustment costs needn't always take this form. Consider a fifth case, thus:

A 'John Lewis economy' would be superior to one with external shareholders, as workers will be more productive, there'll be more effective oversight of management, and there'll be less inequality. However, a transition to a John Lewis economy requires that shareholders be expropriated and their stakes transferred to workers.

(I could add a sixth - the idea of cancelling debts).

I suspect that attitudes to this will be the opposite to that in our first three cases. The folk who think adjustment costs tolerable in those will think them intolerable in this. They might be right; if the state can violate property rights once, they can do so again, so such an expropriation might scare off future investors and thus depress economic activity. But this is arguable.

Which brings me to my concern. Could it be that popular debate about adjustment costs consists of a lot of fact-free hand waving?

I'm tempted to add - as Jonathan says - that it is overly influenced by cognitive biases such as the availability heuristic that causes us to focus too much upon lively anecdotes to the detriment of equilibrium thinking. But the complication is that the anecdotes are about real people, whereas equilibrium thinking is not.

I fear the issues here are trickier than are generally realized.