## Sunday, January 30, 2011

### Risk and uncertainty (revisited)

I have previously posted about the difference between risk and uncertainty based on economist Frank Knight's famous differentiation between risk and uncertainty:
Uncertainty must be taken in a sense radically distinct from the familiar notion of Risk, from which it has never been properly separated. ... The essential fact is that 'risk' means in some cases a quantity susceptible of measurement, while at other times it is something distinctly not of this character; and there are far-reaching and crucial differences in the bearings of the phenomena depending on which of the two is really present and operating.... It will appear that a measurable uncertainty, or 'risk' proper, as we shall use the term, is so far different from an unmeasurable one that it is not in effect an uncertainty at all.
As I noted, not perhaps the clearest distinction.

In the passage above, Knight expresses that difference as risk being “a quantity susceptible of measurement”, uncertainty as being where that was not possible. I would put it a little differently, I would say that ordinary risk is where the expected dangers are sufficiently structured that a pattern of expected risks can be derived from it (so that, even if specific values cannot be calculated, general rankings and ranges can be reasonably derived, even if only of the x is greater than y form) and uncertainty is where there is insufficient confidence in knowledge of how the dangers [likely outcomes] are structured so as to frustrate calculation, even in general terms. That the dangers [likely outcomes] are not able to be expressed mathematically in any useful sense, taking mathematics to be the science of pattern and structure, because there is insufficient pattern or structure within which likely outcomes can be assessed.

Expressed that way, we can think of uncertainty as a realm of possible outcomes across which people have little or no confidence in forming expectations whereas, with risk, people do have confidence in forming expectations. Hence, risk is where the factors driving outcomes are felt to be sufficiently patterned that expectations sufficiently specific to act on can be reasonably formed, uncertainty as where that is not so.

If we think of uncertainty as the range of possible events over which people do not believe they can form sufficiently confident expectations to act on, we can see the inhibiting effect on economic activity [negative] uncertainty is likely to have: it will encourage more holding of money and other liquid assets as “buffers” against adverse outcomes and/or resources to take advantage of opportunities that may present themselves.

Even if the uncertainty is arising out of some change felt to be positive, it would have to be confidently bounded as a positive in all aspects not to retain a “buffer” against adverse and change and, even then, there would be reason to hold resources to use when such positive opportunities present themselves. Either way, increasing one’s holding of money as a store of value rather than using it as a medium of exchange would be sensible, with contractionary effects due to people engaging in fewer transactions.

Though uncertainty due to negative factors will naturally tend to have greater contractionary effect than that due to positive factors. Not merely because the need for a safety “buffer” is a more direct response to fear of loss but also due to the fear of loss being generally greater than hope of gain. A well known, and rational, tendency: for, while both fear of loss and hope of gain are directed to what might happen, what you already do have has far more existential power than what you might have. The gain does not yet exist, and has not been experienced: that which one might lose already does and has been. So, fear of loss naturally tends to be cognitively stronger than hope of gain. Uncertainty has more fearful power in period of contracting economic activity than of expanding economic activity since the possibility of loss looms greater than any hope of gain.

Reducing uncertainty – i.e. increasing the ambit of matters over which expectations can be reasonably formed – will tend to promote economic activity, particularly economic actions with delayed pay-offs (such as creating and benefiting from capital). [Even uncertainty that is read positively is likely to be very unstable, to be easily subject to reversal – reading what Keynes called 'animal spirits' as how uncertainty is currently being framed – since it exists on the absence of a basis in which to frame expectations and so easily reversed by new information.] So business will often prefer policy clarity – even if the policy is hostile or otherwise problematic – to policy uncertainty, since the former gives some structure within which to calculate likely results from actions over time (particularly investment). Creating and sustaining a “bazaar” economy of transactions that are immediate swaps is easy and historically common. Creating and sustaining an economy where transactions across time, notably the production of capital, is encouraged, and so becomes extensive, is harder, and historically rarer.

Hence the long-term economic benefits of the rule of law. It encourages the creating and utilising of capital because it lessens uncertainty about whether one will continue to benefit from the capital one creates, moving beyond what people call ‘sovereign risk’: not the possibility of public debt default, but more general cases of official actions seriously undermining the value of assets (such as confiscation).

But the value of the rule of law extends beyond restraining officials. Contracts, for example, can be seen, not merely as ways of reducing risk, but of lessening uncertainty: but only if they can be enforced. Just as having well-defined and enforceable property rights does much to move economic (and other social actions) out of the realm of uncertainty and into that of mere risk. So people can form expectations with a reasonable degree of confidence, and act upon them.

It is impossible to completely abolish uncertainty, just as it is impossible to completely abolish risk. But public policy that seeks to sustain a stable and prosperous society should aim to decrease uncertainty, and avoid actions that increase it.