In his excellent book, Thinking, Fast and Slow, the psychologist, and Nobel prize winner, Daniel Kahneman discusses various habits of mind that, when applied to problem solving, can lead us astray. One chapter is devoted to problem substitution. When we are faced with a particularly difficult problem to solve, we often substitute a much simpler problem, and solve this instead. And awkwardly, we remain unaware this is what we’ve done. We believe that what we have ended up with is a reasonable response to the difficult problem.
An example he provides relates to the business of choosing political leaders. Knowing which of a number of candidates would best fit the role requires an awful lot of specialist knowledge. Ideally we’d know what sorts of decisions these leaders are likely to have to make, and under what conditions. We would need to know their style of leadership, the nature of their key working relationships, how they perform under pressure, where their priorities lie when policies need to be sacrificed, what motivates them, what scares them, how quickly they assimilate new knowledge and any other number of factors, and that’s before we even get to evaluating the policies they are peddling. A tricky task indeed, and one for which very few of us are well equipped. And yet we do form political preferences, and in many cases feel quite convinced that we are making the correct choice. How do we do it?
Kahneman suggests we substitute out the difficult question, choosing instead a far simpler one. Do I like this person? Are they the sort I’d like to have a beer with? Most of us, on the evidence of a television interview say, can get a handle on this aspect of personal compatibility quite quickly. Political surveys suggest the beer test is indeed a good predictor of ultimate popularity at the polls. It’s also slightly nuts. If I think about my friends, the people I choose to spend time with, they’re not necessarily the people I want running the country (and I count this in their favour). The two problems appear to be only tenuously linked, and yet the substitution happens without us even noticing.
I thought of this earlier in the week when The NZ Green Party floated the idea of using quantitative easing (we used it call it expansionary monetary policy) to fund the rebuild of the Christchurch earthquake and, if I read them correctly, to at the same time put downward pressure on the exchange rate. I’m not interested here in examining whether or not that’s a smart move, but rather in the methods that have been used by politicians and political commentators to assess its validity. What I think has gone on is a classic Kahneman substitution.
First, consider why this is such a difficult thing to evaluate. I have a degree in economics, I was usually at the top of my class at university or thereabouts, so presumably I should be reasonably well equipped to evaluate the policy, and yet I have no idea. The reason I have no idea is that the effectiveness of economic policy is always contextual. It’s never the case of do A, and B happens. Rather, it’s a case of do A, and depending upon the prevailing conditions, and the reactions of the key players, any number of things B through to Z might happen. So, to properly evaluate this proposal, we’d need to know how the money freed up by the move (the money otherwise committed to the rebuild) is then going to be used. Is it for debt, tax cuts or other spending programmes? We’d need to know the level of capacity in key sectors, how much extra demand can be absorbed before there are upward price pressures. We’d need to know how various markets are likely to react to the move. Will they buy it’s a one-off, or will they factor in further rounds of easing down the track? How will it affect The Reserve Bank’s targeting of inflation, what sort of accommodation might be reached? How will the lead unions react? What will it do to inflation expectations in the street? How will our financial sector respond to changes in reserves? How will foreign investors assess the move… the list continues for a good while. Some of the answers are always going to be guesses, but there’s also a huge amount of empirical evidence that can be gathered in order to make those guesses more accurate. I don’t work in the field, I don’t have that data, and so I can’t judge the proposal.
But, if we are to judge by the reactions of political leaders, editorial writers, columnists and voices from the public, others have had no trouble at all reaching a conclusion, and swiftly. If we look at what the opponents in particular are saying, because they’ve been the most vocal, they appear to have substituted the complex problem, what are the likely outcomes, for a much simpler problem, how do we feel about runaway inflation? The answer to the second question is easy, we don’t love it. Trouble is, it’s not at all clear that Green’s proposal will be especially inflationary. As I say, that depends upon the subtleties of context. But, the oversimplifiers have no time for subtlety. Instead the old favourites of 1930’s Germany, contemporary Zimbabwe and 1980s Argentina are wheeled out. Now, if the Greens were indeed proposing a Mugabe style suite of farmland repatriations, cronyistic spending sprees and international defiance, then sure, that might be a reasonable thing to bring up. Clearly they’re not. Our Prime Minister unhelpfully asked people to consider what might happen if the Government printed off nice big bags of money to give everybody for Christmas, substituting again a tremendously difficult concept for the very thought experiment economics teachers use to introduce thirteen year olds to the value of money. The link between the two is, however, far from clear.
It is absolutely true that if there is a one on one relationship between narrow money supply and price level, then the simplified analogies are valid. But even a cursory glance at the world economy past and present tells us the truth is much more complicated than that (the current US price response to easing is surely instructive in this). The often politically motivated pontifications of the oversimplifiers are based upon a false substitution. They are trying to decide the issue without doing any of the hard work, and as such are potentially cutting us off from a perfectly valid policy lever, albeit one that, in this country at least, is deeply unfashionable.
So, perhaps next time we come to vote, we might at least consider a new simplification. Don’t ask, would I enjoy sharing a beer with them, but ask instead, are they smart enough to avoid oversimplifying?