Stumbling and Mumbling

Author: chris dillow   |   Latest post: Tue, 12 Nov 2019, 1:55 PM


Why we are wrong

Author:   |    Publish date:   |  >> Read article in Blog website

Eric Lonergan asked a good question yesterday: "why does it take so long for economic beliefs which are repeatedly falsified to change?" I can think of several reasons.

A failure to appreciate that the institutional background has changed.

Take, for example, big bonuses paid to bankers and CEOs. The idea that we need these to incentivize people is (mostly*) wrong. Instead, they crowd out other motives (pdf) such as professional ethics and can encourage excessive risk-taking.

In truth, though, investment bankers' bonuses did not begin as an incentive at all. BITD when stockbrokers were owned by partners, they were instead a way of smoothing the partners' earnings: in bad years, they could protect their incomes by cutting bonuses. They weren't an incentive: having (often scary) partners looking over one's shoulder was incentive enough.

In not appreciating sufficiently that the world has changed, people continue to believe that big bonuses are efficient, even though they might not be.

I suspect there are other examples here. Ideas such as the Nairu or output gap might have made sense in closed economies when inflation expectations were volatile. But they are less valid in open economies with stable expectations. And the idea of expansionary fiscal contractions made sense when private sector capital spending was constrained by high real interest rates and low profit margins, but is nonsense now we're at the zero bound. And I suspect that some of those who champion the achievements of capitalism are insufficiently awake to the fact of stagnation now.

Bad ideas serve the rich.

There's another reason why so many believe in the efficacy of high pay and big bonuses; they suit the well-off. And there's never a shortage of people happy to defend their interests. As Adam Smith said:

The great mob of mankind are the admirers and worshippers, and, what may seem more extraordinary, most frequently the disinterested admirers and worshippers, of wealth and greatness

Bayesian conservatism.

It's perfectly common for all of us to fail (pdf) to update our beliefs sufficiently in light of new evidence: this is one reason we see momentum effects and post-earnings announcement drift in stock markets. And it's not unreasonable to err on the side of conservatism; new evidence can be unreliable and Burke was right to claim that there is the wisdom of ages. ThomasBayes

Sadly, though, this reasonable motive can be amplified by others. Ego involvement causes us to cleave to strongly to an idea merely because we've expressed it ourselves. And groupthink causes us to stick to bad ideas if our colleagues do.

Sunk costs.

A friend recently showed me a rejection letter for a paper he'd written. Reviewer 2 claimed not to understand a section because it was "purely descriptive." It was, of course, wholly comprehensible. What Reviewer 2 was doing was protecting his investment. He'd sunk his human capital into formal max U (pdf) modelling, and was damned if he was going to accept other ways of doing things**.

In this vein, if you've spent years solving DSGE models you'll be loath to abandon them in favour of more useful approaches.

Few ideas are wholly false.

Take the efficient market hypothesis. It's wrong, in the senses that defensive and momentum stocks do better than the EMH predicts, and that the aggregate market does sometimes over-react. But it is correct in predicting that active fund managers are mostly useless (or worse!). When faced with ideas of mixed truthiness, it's easy to exaggerate their validity: it's a form of halo effect - inferring from one quality that something possesses other ones.

The Duhem-Quine problem.

It's rare that an idea is cleanly and wholly falsified; even the best experiments test joint hypotheses and are subject to the problem of external validity. This is even more true in macroeconomics. Does the lack of significant inflation in the face of low unemployment mean the idea of a Phillips curve is wrong, or are there just some variables which have confounded the relationship between unemployment and inflation? Is the weakness of western economies in the face of low real interest rates a sign that conventional monetary policy is futile, or does it just tell us that its impact has been outweighed by the forces of secular stagnation and declining profits? And so on.

"The poison is the dose"

The minimum wage has had little impact on labour demand. The Low Pay Commission says:

None of the research that we have commissioned has shown strong evidence that minimum wages have led to falling employment...[There is] some evidence of falling hours in response to higher minimum wages, but again the findings are not consistent across different specifications, and the effects are generally small.

This means opponents of the NMW were wrong to predict large adverse effects. But it doesn't mean they'll always be wrong. If we jacked the NMW up to £50, they might well be right.

Our formative years matter

Our beliefs are disproportionately shaped by our formative years. Erin McGuire and Ulrike Malmendier (pdf) have separately shown that people who grew up in hard times own fewer equities than others even decades later (controlling for obvious confounders). Bad times in our impressionable years make us unduly worried about recessions.

I suspect this explains why the ECB has been too scared of inflation. Memories of the high and volatile inflation of the 70s and 80s have undue influence on the minds of 50- and 60-somethings.

We're creatures of history

It's not just our own personal histories that shape our beliefs. So too does the distant past. For example, David Fielding has shown (pdf) that, in the UK, people's opinion of

21st century immigrants is significantly more positive when the respondents live in a constituency that was home to a medieval Jewish immigrant community. On average, these respondents also express less authoritarian views about crime and civil rights, and show less support for far-right political parties.

In this light, isn't it plausible that German's hostility to inflation should be based not so much upon current evidence as upon memories of the 20s?

If our beliefs are not formed by current evidence, why should they be changed by it?

Now, in saying all this I am NOT pretending these factors apply only to other people. I suspect I'm as prone to them as others. It's just that it's easier to see others' errors than our own.

* I say mostly, because there is an exception. Big bonuses might deter bosses or bank traders from selling assets cheaply to rival firms.

** Anonymity cuts both ways. It enables you to dish out abuse, but also enables us to attribute the worst motives to you.

Share this

More articles on Stumbling and Mumbling >>

  Be the first to like this.


2063  2119  703  3405 

Top 10 Active Counters
 WINZL 67.12-0.96 
 TRX 1.125+0.35 
 IPOST 390.75-4.60 
 VAST 0.395-0.015 
 EUZ 0.028+0.002 
 BOU 0.024+0.001 
 PREM 0.138-0.002 
 SNT 0.16-0.055 
 INSP 0.063+0.028 
 SD12 0.020.00 
Partners & Brokers