The G4S fiasco has led some to question - not necessarily consistently - whether the private sector is any more efficient than the public.

For me, the question makes little sense as an ideological one. Hierarchical organizations are prone to failures arising from adverse incentives, limited knowledge and bounded rationality (planning fallacy, anyone?).What matters is how organizations mitigate these dangers; whether they are in the public or private sector is not often the issue.

There is, however, a common claim - one I think I've made myself - I want to challenge. It's the idea that market forces mean that the failure of private sector organizations is less catastrophic than public sector ones.

Now, in many cases, this is true. When Woolworths collapsed, there was no generalized crisis; people simply bought their pic 'n' mix elsewhere. By contrast, when the "Border Force" screws up, people cannot use other means to get through immigration checks quickly.

What interests me, though, is that there are cases where private sector failure does have costs - and I don't just mean the fleeting political embarrassment that arises from G4S's failure.

The biggest and most obvious example of this was, of course, the banking crisis. It showed us that the collapse of a handful of organizations can have colossal macroeconomic effects.

I'd suggest three general ways in which private sector failures can have macroeconomic effects:

1, Common failures. If every organization is pursuing the same strategy, then any change which causes the strategy to lose money might kill off all organizations in an industry. Markets are like ecosystems (pdf).If there's insufficient biodiversity, environmental change will wipe out entire ecosystems.This is the story of banks.

2. Market blockages.Ashwin says: "When the incumbent fails, there must be a sufficient diversity of small and new entrants who are in a position to take advantage." But this is not assured.For it to be true we need (among other things) an entrepreneurial culture, a lack of barriers to entry and expansion and finance for entry and expansion. None of these are assured.

3. Being big enough to matter.As Xavier Gabaix has pointed out, some firms are sufficiently large that they matter for aggregate growth. The failure of one or two mega firms might then generate the sort of adverse productivity shock that RBC theorists fret about.

I reckon there are (at least) two implications here.

First, the standard macro-micro division of labour in economics can be dangerous, as it can blind us to the fact that micro failures can have macro effects.

Secondly, the question of whether private sector firms are well or badly managed cannot be left to the market, and nor can they be rectified by macroeconomic policies. The question of how private sector firms are managed is a legitimate political one. And this raises issues which the non-Marxist left has tended to ignore for too long.