Diane Coyle asks: why is profit-sharing not more common, given the evidence (pdf) that it increases productivity (pdf) and profitability (and well-being)?

To some extent, the premise behind the question is wrong: there are many more worker-owned firms than there are ones quoted on the stock market. Shared ownership is the norm for law, accountancy and medical practices - although they become less egalitarian as they get bigger. And in other cases, substantial worker-ownership is not feasible; I'm not sure workers should own their own oil rigs, for example.

Let's, though, grant the premise - which I think we should. There are several possible answers.

One is ambiguity aversion. The cost of profit-sharing is obvious and immediate, but the benefits are less salient and longer-term; more motivated workers, less staff turnover and so on. Given the cautiousness of many owners, this militates against shared ownership.

In other cases, such ownership is unnecessary. Yes, ownership motivates workers to either work hard or keep their hand out the till. But there is an alternative way of achieving this - direct oversight. And power biased technical change means this is more feasible in many work-places: containerization, tachographs, CCTV and computerization allow capitalists to control workers directly without the incentives of a profit share. It might be no accident that interest in shared capitalism peaked in the 70s and early 80s, before such technology became so available.

In yet others cases, there's simply no desire for it. On the one hand, some firms are happy to pay peanuts and get monkeys: just compare service standards at John Lewis to those in quoted retailers. And on the other hand, many workers don't want it. If you want a job where you can skive off for a sly cigarette, you're ill-advised to work at John Lewis where everyone is a boss.

In this context, Diane hints at an overlooked cultural change. She notes, rightly, that a pioneer of paying good wages to improve corporate profits was Henry Ford.What she doesn't say, however, is that he had to do so to reduce horrific staff turnover. The great Harry Braverman quotes Keith Sward:

The turnover of his working force had run, [Ford] was to write, to 380% for the year 1913 alone. So great was labour's distaste for the new machine system that toward the close of 1913 every time the company wanted to add 100 men to its factory personnel, it was necessary to hire 963. (Labor and Monopoly Capital, p 149)

Since then, however labour has accommodated itself to drudge work, so there's less need to buy off discontent. And learned helplessness - or a desire for an easy life - means there's little demand for worker control. The fact that preferences are in part endogenous is a truth that mainstream politics cannot acknowledge.

Yet another reason for the lack of worker ownership or control is path dependency. Hierarchical capitalism made sense when companies had huge capital requirements and were staffed by illiterates - as was the case in the early phase of the industrial revolution. It makes less sense when capital is plentiful but when workers are highly skilled, as Luigi Zingales pointed out (pdf). However, it's often the case that ideas persist after their material justifications have disappeared. Hierarchical capitalism might therefore be an example of an idea that's outlived its usefulness.

Finally, one reason for a lack of worker ownership or control is simply capital constraints. Workers cannot afford to take a stake in their workplaces, and the first rule of investing - don't put all your eggs into one basket - urges them not to. The John Lewis Partnership was formed not by workers, but by a gift of shares from John Spedan Lewis. Few employers are so generous.

It might therefore be that the prevalence of hierarchical capitalism is a form of market failure.

Another thing: you might think my support for worker ownership and control is a Marxian idea. I'm not sure. It's also founded in a Hayekian scepticism about the efficiency of centralized knowledge and control, economics' agency theory and the mathematical diversity trumps ability theorem.