Tag Archives: rationality

Against Behavioural Economics and Irrationality

Prais­ing Maurice Allais as the fath­er of beha­vi­our­al eco­nom­ics rather than Kahne­man and Tver­sky,  John Kay intro­duces us to some of Allais’ ideas while sim­ul­tan­eously provid­ing one of the finest argu­ments against the simplist­ic view of beha­vi­our­al eco­nom­ics as the study of irra­tion­al­ity:

The skill of piecing togeth­er sense from frag­men­ted and inac­cur­ate inform­a­tion is a cent­ral attrib­ute of human intel­li­gence. Lit­er­al inter­pret­a­tion, and insens­it­iv­ity to con­text, are not marks of ration­al­ity but men­tal disorders. […]

The [beha­vi­our­al eco­nom­ics] exper­i­menter­’s trick is to con­struct an arti­fi­cial situ­ation in which nor­mally sens­ible beha­viour gives what he thinks is the wrong res­ult. The “mis­take” is detec­ted in a mean­ing­less prob­lem designed solely to eli­cit the “mis­take”. […]

Allais was less con­cerned to show that our beha­viour was irra­tion­al than to argue that the premises of ration­al­ity itself were irrational. […]

Allais’ most fam­ous exper­i­ment showed that we often treat very high prob­ab­il­it­ies very dif­fer­ently from cer­tain­ties, although “ration­al” indi­vidu­als would regard them as almost the same thing. But very high prob­ab­il­it­ies often are dif­fer­ent from cer­tain­ties: very high prob­ab­il­it­ies are usu­ally derived from cal­cu­la­tions whose rel­ev­ance and valid­ity are them­selves uncer­tain. […]

Irra­tion­al­ity lies not in fail­ing to con­form to some pre­con­ceived notion of how we should behave, but in per­sist­ing with a course of action that does not work. Some­times in mod­ern eco­nom­ics and polit­ic­al life, there is a big dif­fer­ence.

The example Kay uses is a bit glib but does serve its purpose.That last para­graph, how­ever, is the crux of it all. As you may have guessed, this is the Allais that designed the Allais para­dox – an exper­i­ment in beha­vi­our­al eco­nom­ics that shows the above won­der­fully.

The Virtues of Rationality

The name Eliez­er Yudkowsky imme­di­ately con­jours in my mind the word ration­al­ity (thanks to his addict­ive piece of fan fiction, Harry Pot­ter and the Meth­ods of Ration­al­ity). On a recent vis­it to his site, this con­nec­tion has now be strengthened after I saw his excel­lent essay on the twelve vir­tues of ration­al­ity:

  1. Curi­os­ity: A burn­ing itch to know is high­er than a sol­emn vow to pur­sue truth.
  2. Relin­quish­ment: Do not flinch from exper­i­ences that might des­troy your beliefs.
  3. Light­ness: Sur­render to the truth as quickly as you can.
  4. Even­ness: You are not a hypo­thes­is, you are the judge. There­fore do not seek to argue for one side or anoth­er.
  5. Argu­ment: In argu­ment strive for exact hon­esty, for the sake of oth­ers and also your­self […]Do not think that fair­ness to all sides means bal­an­cing your­self evenly between pos­i­tions; truth is not handed out in equal por­tions before the start of a debate.
  6. Empir­i­cism: Always know which dif­fer­ence of exper­i­ence you argue about.
  7. Sim­pli­city: When you pro­fess a huge belief with many details, each addi­tion­al detail is anoth­er chance for the belief to be wrong.
  8. Humil­ity: To be humble is to take spe­cif­ic actions in anti­cip­a­tion of your own errors.
  9. Per­fec­tion­ism: The more errors you cor­rect in your­self, the more you notice.
  10. Pre­ci­sion: More can be said about a single apple than about all the apples in the world. The nar­row­est state­ments slice deep­est.
  11. Schol­ar­ship: Each field that you con­sume makes you lar­ger.
  12. The Void

I believe that the ninth vir­tue, per­fec­tion­ism, is the most eleg­ant and I implore you to read the full essay if only to read that descrip­tion in full (and, I guess, to dis­cov­er what The Void is). How­ever the elev­enth vir­tue of ration­al­ity, schol­ar­ship, almost per­fectly describes why I write here and may go some way to explain­ing my diverse read­ing habits:

Study many sci­ences and absorb their power as your own. Each field that you con­sume makes you lar­ger. If you swal­low enough sci­ences the gaps between them will dimin­ish and your know­ledge will become a uni­fied whole. If you are glut­ton­ous you will become vaster than moun­tains. It is espe­cially import­ant to eat math and sci­ence which impinges upon ration­al­ity: Evol­u­tion­ary psy­cho­logy, heur­ist­ics and biases, social psy­cho­logy, prob­ab­il­ity the­ory, decision the­ory. But these can­not be the only fields you study. The Art must have a pur­pose oth­er than itself, or it col­lapses into infin­ite recur­sion.

Harry Potter and the Methods of Rationality

The First Law of Fan­fic­tion states that every change which strengthens the prot­ag­on­ists requires a cor­res­pond­ing worsen­ing of their chal­lenges. […] stor­ies are about con­flict; a hero too strong for their con­flict is no longer in tense, heart-pound­ing dif­fi­culty. […]

The Ration­al­ist Fan­fic­tion Prin­ciple states that ration­al­ity is not magic; being ration­al does not require magic­al poten­tial or roy­al blood­lines or even amaz­ing gad­gets, and the prin­ciples of ration­al­ity work for under­stand­able reas­ons.

That’s Eliez­er Yudkowsky in an intro­duc­tion to his acclaimed Harry Pot­ter fan fiction, Harry Pot­ter and the Meth­ods of Ration­al­ity.

The piece of “seri­al fic­tion” looks at cog­nit­ive sci­ence and ration­al­ity in a Harry Pot­ter-type world where Harry, hav­ing been raised by a sci­ent­ist step­fath­er, is a ration­al­ist, enter­ing the wiz­ard­ing world “armed with Enlight­en­ment ideals and the exper­i­ment­al spir­it.”

Cur­rently 63 chapters long–including chapters such as A Day of Very Low Prob­ab­il­ity, The Stan­ford Pris­on Exper­i­ment, The Unknown and the Unknow­able and Title Redac­ted, Part I–the Meth­ods is a fant­ast­ic read.

There’s a “book-style” PDF avail­able, ePUB and MOBI ver­sions for those on e‑readers, and a great TV Tropes entry. You can find more resources on the unof­fi­cial homepage, hpmor.com.

Although listen to Eliez­er when he says “This fic is widely con­sidered to have really hit its stride start­ing at around Chapter 5. If you still don’t like it after Chapter 10, give up”.

via Hack­er News

Anchoring Our Beliefs

The psy­cho­lo­gic­al prin­ciple of anchor­ing is most com­monly dis­cussed in terms of our irra­tion­al decision mak­ing when pur­chas­ing items. How­ever, Jonah Lehr­er stresses that anchor­ing is more wide-ran­ging than this and is in fact “a fun­da­ment­al flaw of human decision mak­ing”.

As such, Lehr­er believes that anchor­ing also effects our beliefs, such that our first reac­tion to an event ‘anchors’ our sub­sequent thoughts and decisions, even in light of more accur­ate evid­ence.

Con­sider the ash cloud: After the cloud began drift­ing south, into the crowded air­space of West­ern Europe, offi­cials did the prudent thing and can­celed all flights. They wanted to avoid a repeat of the near crash of a Boe­ing 747 in 1989. […]

Giv­en the lim­ited amount of inform­a­tion, anchor­ing to this pre­vi­ous event (and try­ing to avoid a worst case scen­ario) was the only reas­on­able reac­tion. The prob­lems began, how­ever, when these ini­tial beliefs about the risk of the ash cloud proved res­ist­ant to sub­sequent updates. […]

My point is abso­lutely not that the ash cloud was­n’t dan­ger­ous, or that the avi­ation agen­cies were wrong to can­cel thou­sands of flights, at least ini­tially. […] Instead, I think we simply need to be more aware that our ini­tial beliefs about a crisis – those opin­ions that are most shrouded in ignor­ance and uncer­tainty – will exert an irra­tion­al influ­ence on our sub­sequent actions, even after we have more (and more reli­able) inform­a­tion. The end res­ult is a kind of epi­stem­ic stub­born­ness, in which we’re irra­tion­ally anchored to an out­moded assump­tion.

The same thing happened with the BP oil spill.

Optimism as Incentive

Much has been writ­ten on the (ir)rationality of pur­chas­ing lot­tery tick­ets (Eliez­er Yudkowsky’s view­point is par­tic­u­larly fine), but little has been said on applic­a­tions of these biases that could improve the fin­ances of all of those who buy a tick­et.

Now beha­vi­our­al eco­nom­ists are attempt­ing to boost the his­tor­ic­ally poor house­hold sav­ings rate by using our lot­tery-like optim­ism as an incent­ive to save:

Psy­cho­lo­gists have long known that people tend to over­es­tim­ate the odds of rare events. Apply­ing that beha­vi­or­al insight, fin­ance pro­fess­or Peter Tufano of Har­vard Busi­ness School has devised a clev­er pro­gram called “Save to Win.” Launched earli­er this year for mem­bers of eight cred­it uni­ons in Michigan, it is a cross between a cer­ti­fic­ate of depos­it and a raffle tick­et. Mem­bers who put $25 or more into a Save to Win one-year CD* are entered into a monthly “sav­ings raffle” for prizes up to $400, plus one annu­al draw­ing for a $100,000 jack­pot. […]

In 25 weeks, the pro­gram has attrac­ted about $3.1 mil­lion in new depos­its, often from people who have nev­er been able to set money aside.

via Tech­dirt

* CD = Cer­ti­fic­ate of Depos­it (sim­il­ar to a sav­ings account).