Lionel Page Profile picture
Sep 27, 2019 17 tweets 16 min read
Here is the explanation of the ‘homo economicus’ test!

The opportunity for a thread about what economists mean with the notion of economic ‘rationality’.

First, the test results:



(Thread)
Many seem to have interpreted it as a test of the ability to maximise expected value. The most chosen options are A & C in the poll 1 and B & D in poll 2 which maximise expected value.

Maximising expected value is rational, but you can also be rational with other choices.
Since the St Petersburg Paradox (Bernoulli, 1738), we know that valuing a gamble at its expected value can’t be the only acceptable solution.

Faced with a gamble with an infinite expected value (it’s possible!) people refuse to pay much to play it.

en.m.wikipedia.org/wiki/St._Peter…
Following Bernoulli, economists only assume that people maximise expected ‘utility’. An idea formalised by von Neumann and Morgenstern (1947).

This theory makes it totally fine to be risk averse and choose B (don’t play) in poll 1, for instance.
But, the theory imposes constraints on choices for them to be consistent. One of the most surprising constraints:

In the poll 1, if you don’t take the single gamble A then, you *cannot* pick the 100 gambles C.

It seems very counter-intuitive since C offers excellent odds.
This point was made by Samuelson in a 1963 paper which contained poll 1’s question as an example.

casact.org/pubs/forum/94s…
The reason is that avoiding the single gamble A and picking the 100 gambles C would violate the independence axiom, one of the key principles behind expected utility.

Here is an explanation of why for the economists around:
link.springer.com/content/pdf/10…
So, if you choose to gamble only once but not 100 times, A&D (5%), or not once but a 100 times, B&C (29%), you violated the independence axiom, you were not a ‘homo economicus’.

Only those who always gambled or never gambled respected the independence axiom.
The second poll, is the most famous violation of the independence axiom, it is one of the questions from the ‘Allais paradoxes’.
In 1952, Allais asked leading economists to answer a few questions at a conference in Paris. Many of them violated the independence axiom themselves!
The violation of the independence axiom is easier to see in poll 2.
A & B both contain 89% chances of getting $500,000.
C & D both contain 89% chances of getting $0.
The independence axiom says that the choice between A&B and C&D should not depend on this common element.
So only A&C and B&D respect the axiom of independence.

If you picked A&D (27%) or B&C (9%) you were not a homo economicus.
But were you ‘wrong’?

The most interesting aspect of the Allais paradox, is that the great Savage himself made the ‘wrong’ decision as he discusses in his second edition of The Foundations of Statistics. He then states that he made an ‘error’ and should change his mind.
Furthermore experiments show that people often do not decide to change their mind when they are told that they violate the independence axiom.

So if the best economists violate the independence axiom and if people don’t care about it, in what way should we consider it ‘right’?
Violations of the independence axiom have led to the development of non-Expected Utility theory, such as as Prospect Theory (Kahneman and Tversky, 1979).
Quiggin (1982) gave a rigorous theoretical underpinning to such models.
These models drop the independence axiom and explain our choices better. For many modern economists it does not mean we are wrong/irrational.

So if you were not a homo economicus in the test, don’t worry: that’s just a matter of preferences!

(End)

• • •

Missing some Tweet in this thread? You can try to force a refresh
 

Keep Current with Lionel Page

Lionel Page Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @page_eco

Oct 14, 2021
There has been a recent debate on Twitter on the notion of “utility” in economics. I’ll try to give some perspective to help understand the disagreements. 🧵
One reason why the discussion about utility can be confusing is that two types of interpretation co-exist in economics. On the one hand, economic students are typically told in introductory microeconomics that utility is just a number to reflect the order of preferences.
On the other hand, students will be presented either explicitly or implicitly with interpretations of utility as measuring some kind of satisfaction in other courses.
Read 18 tweets
May 10, 2021
This is an amazing story. In 1912, Greek soldiers took an island from the Ottoman Empire. The soldiers were thinking that the inhabitants were Greek, but the local kids said "No, we are Romans".
And they were Romans... from the former Roman Empire. 1/6
In the West, we date the end of the Roman Empire to around 476BC. But the "Eastern" Roman Empire persisted and its people *were* Romans. This Roman identity persisted even after the fall of Constantinople in 1453. 2/6
This is the extent of the Roman Empire in 555BC. It still included Rome! And it continued to include the city until the 8th century.
Rome had not been the capital of the Empire for a long time and losing Rome did not change their self-identification as "Romans". 3/6
Read 7 tweets
Mar 20, 2021
Everybody should know about Berkson’s paradox.

👉 When two positive traits are (spuriously) negatively correlated, in a population *selected on these traits*.

Examples: 1/
“Handsome men are jerks.”
via @anecdatally
“High standardised test scores before entering university do not predict university grades.”

People accepted with low test scores likely have other qualities helping them to succeed at university.
brilliant.org/wiki/berksons-…
“Good books make bad movies.”

Suppose the movies you can think of are either good movies or movies from good books: it will look like good books end up in worst movies!

by @FryRsquared
Read 8 tweets
Nov 1, 2020
In Italy, fascist groups used violence against political opponents to deter them and prevent them from seating in Parliament.
from Ebner's amazon.com/Ordinary-Viole…
In France, militarised leagues went close to what seems a coup d'Etat in 1934, with riots generated by conspiracy theories around a corruption affair.
from Kalman's amazon.com.au/Extreme-Right-…
Eventually, Germany's and Italy's democracy decayed into dictatorships.
There is definitely a lesson from history: democratic institutions can fail and we should not take them for granted.
Read 6 tweets
Oct 13, 2020
The Nobel Prize in economics went to theorists who made major contributions to our understanding of how to design auctions.
If it seems far from you, it should not. Auction theory is one of the major successes in economics which had very large real world implications. (Thread)
Economists are known to like markets. Markets are a way to allocate resources. But they are not the only one. Auctions are another way to organise who get what, at what price. You are surely familiar with art auctions, housing auctions, or eBay auctions.
But really the notion of auction is very broad. Contests such as sporting matches, legal battles, political races, grant applications can also be conceived as auctions where contestants bid resources in the hope of securing a prize.
link.springer.com/article/10.100…
Read 8 tweets
Oct 11, 2020
Suppose you could improve Twitter as a public platform, what feature would you add (or remove)?
I would suggest for Twitter to give users the freedom to adjust the parameters of the recommendation algorithm. Something like:

Emotions:
😱—————|——🤗

Seriousness:
🧐—|—————— 😂

Political content:
🗳—————|——❌

Outrage:
🤬——————|—😎
Why? People may be enticed to click on the posts promoted by the default algorithm. But they may regret wasting their time that way, and they could *prefer* different content.
Read 5 tweets

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Don't want to be a Premium member but still want to support us?

Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal

Or Donate anonymously using crypto!

Ethereum

0xfe58350B80634f60Fa6Dc149a72b4DFbc17D341E copy

Bitcoin

3ATGMxNzCUFzxpMCHL5sWSt4DVtS8UqXpi copy

Thank you for your support!

Follow Us on Twitter!

:(