He showed (pdf) that, under certain conditions, it was possible for competitive markets to achieve an optimal allocation of risk. One of these conditions is that there exist securities which pay out in all possible states of the world, so that we have in effect complete insurance markets. But, he concluded:
Many of the economic institutions which would be needed to carry out the competitive allocation in the case of uncertainty are in fact lacking.
He wrote that in 1952. Not much has changed since then. Yes, we now have index-linked gilts which pay out for all different possible rates of inflation. But we don’t have GDP-linked bonds that pay out in different rates of GDP growth, nor securities that pay out at different levels of house prices, or which pay out at different levels of occupational employment and so would offer insurance against our jobs being taken by robots. As Robert Shiller wrote in 1993 – 24 years ago! - we just don’t have markets that insure us against big economic risks: his book, Macro Markets can be seen as an attempt to operationalize Arrow’s high theory.
Why, then, has so little changed since Arrow wrote those words?
It’s not because his ideas are flawed: many far worse ideas have been adopted. Nor is it because the welfare cost of business cycles is small as Robert Lucas argued (pdf): recessions hit real people, not representative agents. Nor is it because the state does such a good job of insuring against economic risks that there’s no role for markets. Nor is it because there’s anything very unusual about this particular case; there was a gap of over 30 years, for example, between Ricardo arguing against the Corn Laws and their abolition.
Instead, I’d suggest other reasons.
One is simply a status quo bias. This isn’t wholly a bad thing. Things often exist or not for good reasons. The status quo might embody the wisdom of the ages. But a healthy scepticism can spill over into antipathy for change – as in the “not invented here” syndrome.
Another is that it’s easier for the financial “services” industry to make money by rigging markets or flogging over-priced products than it is for them to undertake genuinely worthwhile financial innovation. As William Nordhaus has famously shown, the private rewards to innovation are often small.
This might be especially true in finance. The worth of financial assets often lies in their liquidity, the ease with which they can be traded. Liquid markets, however, don’t fall from heaven. They need to be created which requires hard work. Donald MacKenzie discusses how Leo Melamed, then chairman of the CBOE, cajoled traders into trading index futures thus creating a market:
A market, says Melamed, “is more than a bright idea. It takes planning, calculation, arm-twisting and tenacity to get a market up and going. Even when it’s chugging along, it has to be cranked and pushed.” (An Engine, not a Camera, p173)
Perhaps Arrovian securities are lacking because there’s been no Melamed figure to chivvy them into existence. Markets don't exist because of a market failure.
In this context, markets and government are complements not substitutes. If we want macro markets the government might need to create them and give them away; if it can create money out of nothing it can create other claims on wealth too.
One point that critics and supporters of markets are both apt to forget is that, when viewed from the perspective of Arrow-Debreu theory, the astonishing thing about markets is how under-developed they are.