8 The Match: Strong Medicine for New DoctorsRead more at location 1756
Note: Un tipico fallimento di mercato: il mercato diluito. Guarda al far west e alla regola "chi prima arriva meglio si accomoda": quanto spreco ha realizzato? Con un asta globale si sarebbero risparmiate molte risorse... Il mercato degli "interni" di medicine: gli ospedali, per ottimizzare le assunzioni, hanno gradualmente anticipato le proposte. I neolaureati nn avevano così un quadro completo tra cui scegliere... Exploding offer: offerta a cui rispondere subito (senza un quadro complessivo). Abbondano sui mercati diluiti in cui nn si può xdere tampo con un solo candidato compratore se nn si vuole travare accasati gli altri qualora vada male... Problemi (di congestione): mercati sempre + brevi => accordi mancati + accordi rotti... L'algoritmo fallito: si creavano due liste (una coi cluster di studenti a cura degli ospedali) e si abbinavano gli 1-1 poi si passa al secondo cluster di studenti privilegiando le prime scelte (2-1) poi si tornava al primo cluster privilegiando le seconde scelte (1-2) eccetera. Problema: non conveniva dichiarare la preferenza autentica. Inoltre l'esito nn è stabile: esistono alternative in cui tutti sono più contenti... L'algoritmo stabile (ad accettazione differita): Gli ospedali fanno un'offerta ai loro pupilli che possono accettare con riserva. Quando un'offerta viene rigettata si propone ai pupilli di seconda fascia e così via finchè nn ci sono più rigetti. In realtà questo processo è svolto in tempo reale da un PC a cui vengono sottoposte le luste di preferenze... Shapley e Roth vincono il Nobel dimostrando che l'Algo AA è stabile. In realtà l'argomento è semplice: nn possono esistere bloking pair poichè tutte le coppie sono state assortite e se una coppia nn si è formata significa che uno dei due membri ha preferito un altro partner... Edit
THE SOLUTIONS TO problems in market design are sometimes invented, sometimes discovered, and often a bit of both.Read more at location 1758
we can sometimes discover a solution to a new market failure in a design pioneered in another market.Read more at location 1759
nowadays, we can improve on our natural defenses with antibiotics. The first really important one was penicillin. Penicillin wasn’t invented, it was discovered, by the Scottish immunologist Alexander Fleming in 1928. Fleming noticed that the Penicillium mold found on bread produced a substance that killed bacteria.Read more at location 1763
Since about 1900, the first job American doctors take after graduating from medical school is called an internship or residency, in which they’re supervised by more senior docs.Read more at location 1775
there’s a lot of pressure on both sides to make good matches—on medical students to get good first jobs and on hospital residency programs to hire good young docs.Read more at location 1779
their competition for scarce medical school graduates, hospitals began to try to hire interns a little earlier than competing hospitals.Read more at location 1781
Students also were often forced to consider offers from one hospital at a time, without ever knowing what their prospects might be at other hospitals.Read more at location 1783
it was risky to hire a medical student two years before graduation.Read more at location 1786
it was also hard for medical students to tell what kinds of jobs they might want two years hence.Read more at location 1788
Although early hiring was pretty bad for both students and residency programs, we’ve already seen that unraveling isn’t solved by self-control. Only in 1945, when a third party—medical schools—agreed not to release information about students before a specified date, was the timing of offers controlled.Read more at location 1792
Hospitals now found that if some of the first offers they made were rejected after a period of deliberation, the candidates to whom they wished to make their next offers often had already accepted other positions.Read more at location 1797
This, of course, led hospitals to begin making exploding offers. Now candidates had to reply immediately, even before they could learn what other offers might be available. That in turn led to a chaotic market that shortened in duration from year to year and resulted not only in missed agreements but also in broken ones. In other words, the market suffered from congestion:Read more at location 1800
Instead of a completely decentralized market, they proposed to organize the last stage of the market through a centralized marketplace, a kind of clearinghouse.Read more at location 1805
students would submit to the clearinghouse a rank order listRead more at location 1809
residency programs would submit a rank order list of students.Read more at location 1810
each side could form well-considered preferences regarding the other. Note that they would do this in advance,Read more at location 1812
In the initial proposal, students were asked to rank individual residency programs, while residency programs ranked students in clusters, with 1 being reserved for the most preferredRead more at location 1817
The proposed algorithm first matched all residency programs and students that were each other’s first choices (1-1 rankings). After that, residency programs matched students in their second group if those students had ranked the residency program first (2-1 rankings), followed by matches of residency programs’ first choices with students’ second choices (1-2 rankings), and so forth (2-2, 3-1, 3-2, 1-3, 2-3 . . .).Read more at location 1819
The intention appears to have been to give an advantage to the students, since when the preferences conflicted, students’ first choices were considered earlierRead more at location 1822
after a trial run, students realized that it wasn’t safe for them to confide their true preferences to the clearinghouse. Much as with the school choice systems decades later in New York and Boston, a student who listed as his first choice a residency program that didn’t list him first might miss the chance of getting his second choiceRead more at location 1824
When I studied this and other successful labor market clearinghouses, I discovered one of the secrets of their success. It was that they produced outcomes that were stable, in the sense that no applicant and residency program not matched with each other preferred each other to their assigned matches.Read more at location 1843
notion of stability wasn’t clearly formulated until ten years later, in a 1962 article by David Gale and Lloyd Shapley with the intriguing title “College Admissions and the Stability of Marriage.”Read more at location 1870
Applicants and employers privately submit preferences to a clearinghouse in the form of rank order lists.Read more at location 1881
Each employer offers jobs to its top-choice candidates, up to the number of its available positions.Read more at location 1882
tentatively accepts the best one (the one highest on her preference list), and rejects any othersRead more at location 1883
Step n. Each employer that had a job offer rejected in the previous step offers that job to its next choice,Read more at location 1885
Note: N LE OFFERTE RIFIUTATE VENGONO PROPOSTE AL SUCCESSIVO CHE ACCETTERÀ O MENO IN VIA PROVVISORIA E VIA COSÌ FINO ALL ALLOCAZIONE COMPLETAEdit
Each applicant considers the offer he or she has been holding together with his or her new offer(s) and tentatively acceptsRead more at location 1886
The algorithm ends when no offer is rejected,Read more at location 1889
the final matching is always stable with respect to the preferences submitted by the employers and applicants, whatever those preferences happen to be.Read more at location 1892
The medical Match worked smoothly for decades. But it ran into trouble when women started to enter medical school in significant numbers.Read more at location 1915
