Does Cognitive Imprecision Result in Under-Reaction or Over-Reaction?

That is the topic of a new paper by Michael Woodford, to be presented at this year’s AEA meetings.  Here at least is the abstract:

Men who score high on standardized IQ tests display forecast errors for inflation that are 50% lower than forecast errors for other men in a representative sample of Finnish households. High-IQ men, but not others, have consistent inflation expectations over time and their inflation perceptions align with past expectations. Only high-IQ men increase their consumption propensity when expecting higher inflation in line with the consumption Euler equation. High-IQ men are also twice as sensitive to interest-rate changes when making borrowing decisions. Heterogeneity in education, income, or financial constraints do not explain these results. Limited cognitive abilities are thus human frictions to the transmission and effectiveness of economic policy and inform research on heterogeneous agents in macroeconomics and finance.

How many papers are there in economics about psychological cognitive bias as a source of error, as opposed to low IQ as a source of error?  Might that be a bias of sorts and not one of IQ?


Fascinating - for two decades, the spectre of inflation has been haunting Europe, particularly among those who consider themselves possessing greater intelligence-

Does the finding suggest that the inflationistas are low IQ men?

Low IQ men = hand-to-mouse households. Got it.

Um. Mouth.

Daniel Kahneman's work would be more interesting if he identified which of his trick questions worked simply by fooling lower IQ people the most, and whether or not he could come up with Talebian trick questions that actually fooled higher IQ people more than lower IQ people.

Stupid is as stupid does.

Nothing to see here.

I recall two differently worded but identical probability questions being asked of doctors, who were pretty hapless at seeing the questions as identical. Prolly the low IQ doctors.

Finland scores high on global rankings for gender equality. So, I have to wonder why the conclusion was only about men with "hi-IQ". P-hacking is the cynical answer. Which leads to the question of whether or not this study ('s hypothesis) was pre-registered. The other question I have is why would men be equated to households in a country ranked 5th in women's rights? Do "hi-IQ" men make better economic decisions than average men? In Finland? (It's also a country with high "business friendly" rank as well as low inequality)

The author has probably used the same Finnish IQ dataset as other behavioural econ researchers (see for example Greenblatt & Keloharju). The data is from an IQ test taken by all military conscripts in the country - overwhelming majority of test takers are men because the conscription is only mandatory for them.

The efficient market hypothesis would predict that the various biases and misperceptions of various individuals and sub-groups should cancel out to an average that's approximately appropriate.

The abstract is not specific enough to indicate whether prediction errors were after the fact (e.g. a sample predicted 5% but only 3% occurred one year later) or whether prediction was matched to inflation expected by models or policy makers.

If the latter, then taking into account the EMH again, no one knows whether the "correct" reaction to _policy_ changes is actually "correct" w.r.t. the actually occurring future.

Shouldn't we expect low IQ expectations to be influenced by trend and herd behavior, rather than self-sorting into a random distribution?

The EMH implies that no one is assured to beat the market, regardless of knowledge or IQ. ANd current opinion in cognition research is that high IQ people are just better at rationalizing their biases - not that they're free of biases.

Nice support for the Elizabeth Warren campaign, because of course if financial ability is a birth lottery you have two choices, simplifying rules, or sheeps and wolves.

The AEA messed up the abstract of Woodford's paper. This is from the next paper in that section, entitled "Cognitive Abilities and Inflation Expectations".

Does this mean this post is by a bot and not Tyler? The abstract clearly doesn't fit the title, so you would expect him to pick up on that.

This reminds me of a thought I had all throughout grad school: high IQ, nearly autistic (mostly) men made models of... themselves, and then wondered why nobody else acted like them.

Never thought of that, but it would explain a lot.

"How many papers are there in economics about psychological cognitive bias as a source of error, as opposed to low IQ as a source of error? Might that be a bias of sorts and not one of IQ?"

I find these two sentences incoherent.

First, writers should not ask questions of their readers. We're here answers or speculation.

Second, here's my first attempt at a restatement of how I interpret the two sentences:

"There has been much research attempting to show low IQ as a source of error. However, I think that psychological cognitive bias may be behind these results, and I think it should be studied."

Third, psychological cognitive biases are listed on this Wikipedia page:

Comments for this post are closed