Mul$ple Equilibria and Economic Theory Roberto Weber - - PowerPoint PPT Presentation

mul ple equilibria and economic theory
SMART_READER_LITE
LIVE PREVIEW

Mul$ple Equilibria and Economic Theory Roberto Weber - - PowerPoint PPT Presentation

Mul$ple Equilibria and Economic Theory Roberto Weber Department of Economics University of Zurich LATSIS Symposium September 2012 Lets start with


slide-1
SLIDE 1

Mul$ple ¡Equilibria ¡and ¡ ¡ Economic ¡Theory ¡

Roberto ¡Weber ¡ Department ¡of ¡Economics ¡ University ¡of ¡Zurich ¡ ¡ LATSIS ¡Symposium ¡ September ¡2012 ¡

slide-2
SLIDE 2

Let’s ¡start ¡with ¡an ¡exercise ¡. ¡. ¡. ¡

Choice ¡“Le*” ¡ Choice ¡“Right” ¡ Select ¡one ¡of ¡the ¡two ¡choices ¡– ¡if ¡you ¡select ¡the ¡most ¡common ¡ choice ¡in ¡the ¡room ¡you ¡win ¡CHF ¡10; ¡alterna:vely, ¡you ¡win ¡nothing ¡

slide-3
SLIDE 3

The ¡fundamental ¡problem ¡in ¡tacit ¡bargaining ¡is ¡that ¡of ¡coordina$on ¡. ¡. ¡. ¡ explicit ¡bargaining ¡requires, ¡for ¡an ¡ul$mate ¡agreement, ¡some ¡ coordina$on ¡of ¡the ¡par$cipants’ ¡expecta$ons. ¡

¡

The ¡final ¡outcome ¡must ¡be ¡a ¡point ¡from ¡which ¡neither ¡expects ¡the ¡other ¡ to ¡retreat: ¡yet ¡the ¡main ¡ingredient ¡of ¡this ¡expecta$on ¡is ¡what ¡one ¡thinks ¡ the ¡other ¡expects ¡the ¡first ¡to ¡expect, ¡and ¡so ¡on. ¡Somehow, ¡out ¡of ¡this ¡ fluid ¡and ¡indeterminate ¡situa$on ¡that ¡seemingly ¡provides ¡no ¡logical ¡ reason ¡for ¡anybody ¡to ¡expect ¡anything ¡except ¡what ¡he ¡expects ¡to ¡be ¡ expected ¡to ¡expect, ¡a ¡decision ¡is ¡reached. ¡These ¡infinitely ¡reflexive ¡ expecta$ons ¡must ¡somehow ¡converge ¡on ¡a ¡single ¡point, ¡at ¡which ¡each ¡ expects ¡the ¡other ¡not ¡to ¡expect ¡to ¡be ¡expected ¡to ¡retreat. ¡

¡

¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡Thomas ¡C. ¡Schelling, ¡1960 ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡ ¡The ¡Strategy ¡of ¡Conflict ¡

slide-4
SLIDE 4

. ¡. ¡. ¡professional ¡investment ¡may ¡be ¡likened ¡to ¡those ¡newspaper ¡ compe$$ons ¡in ¡which ¡the ¡compe$tors ¡have ¡to ¡pick ¡out ¡the ¡six ¡prePest ¡ faces ¡from ¡a ¡hundred ¡photographs, ¡the ¡prize ¡being ¡awarded ¡to ¡the ¡ compe$tor ¡whose ¡choice ¡most ¡nearly ¡corresponds ¡to ¡the ¡average ¡ preferences ¡of ¡the ¡compe$tors ¡as ¡a ¡whole; ¡so ¡that ¡each ¡compe$tor ¡has ¡ to ¡pick, ¡not ¡those ¡faces ¡which ¡he ¡himself ¡finds ¡prePest, ¡but ¡those ¡which ¡ he ¡thinks ¡likeliest ¡to ¡catch ¡the ¡fancy ¡of ¡the ¡other ¡compe$tors, ¡all ¡of ¡ whom ¡are ¡looking ¡at ¡the ¡problem ¡from ¡the ¡same ¡point ¡of ¡view. ¡It ¡is ¡not ¡ a ¡case ¡of ¡choosing ¡those ¡which, ¡to ¡the ¡best ¡of ¡one's ¡judgment, ¡are ¡really ¡ the ¡prePest, ¡nor ¡even ¡those ¡which ¡average ¡opinion ¡genuinely ¡thinks ¡the ¡

  • prePest. ¡We ¡have ¡reached ¡the ¡third ¡degree ¡where ¡we ¡devote ¡our ¡

intelligences ¡to ¡an$cipa$ng ¡what ¡average ¡opinion ¡expects ¡the ¡average ¡

  • pinion ¡to ¡be. ¡And ¡there ¡are ¡some, ¡I ¡believe, ¡who ¡prac$se ¡the ¡fourth, ¡

fiUh ¡and ¡higher ¡degrees. ¡

¡

¡ ¡ ¡John ¡Maynard ¡Keynes, ¡1936 ¡ ¡ ¡ ¡The ¡General ¡Theory ¡of ¡Employment, ¡Interest, ¡and ¡Money ¡

slide-5
SLIDE 5

Mul$plicity ¡and ¡economics ¡

  • The ¡preceding ¡example ¡is ¡a ¡simple ¡coordina:on ¡game ¡

with ¡mul$ple ¡equilibria ¡

  • Similar ¡to ¡“games” ¡described ¡by ¡many ¡prominent ¡

economists ¡(e.g., ¡Schelling, ¡Arrow ¡and ¡Keynes), ¡in ¡ discussing ¡important ¡economic ¡phenomena ¡

  • Applica$ons ¡to: ¡

– Firm ¡produc$on ¡(Camerer ¡& ¡Knez ¡1994) ¡ – Macroeconomic ¡ac$vity ¡such ¡as ¡investment ¡(Bryant ¡1983) ¡ – Bank ¡runs ¡(Diamond ¡& ¡Dybvig ¡1983) ¡ – Social ¡interac$on ¡(e.g., ¡language, ¡Selten ¡& ¡Warglien ¡2007) ¡ – Confidence ¡in ¡the ¡SNB, ¡EU, ¡Greece, ¡Italy, ¡Spain, ¡. ¡. ¡. ¡

slide-6
SLIDE 6

The ¡basic ¡game ¡ X ¡ Y ¡ Z ¡ X ¡ 1, ¡1 ¡ 0, ¡0 ¡ 0, ¡0 ¡ Y ¡ 0, ¡0 ¡ 1, ¡1 ¡ 0, ¡0 ¡ Z ¡ 0, ¡0 ¡ 0, ¡0 ¡ 1, ¡1 ¡

slide-7
SLIDE 7

Broad ¡ques$ons ¡

  • What ¡happens ¡in ¡these ¡kinds ¡of ¡situa$ons? ¡
  • How ¡good ¡is ¡economic ¡theory ¡for ¡predic$ng ¡

what ¡happens ¡in ¡these ¡kinds ¡of ¡situa$ons? ¡

  • How ¡can ¡research ¡outside ¡economics ¡provide ¡

us ¡with ¡insights ¡about ¡these ¡kinds ¡of ¡games? ¡

  • How ¡do ¡the ¡limita$ons ¡that ¡these ¡games ¡pose ¡

for ¡economic ¡theory ¡make ¡them ¡useful ¡tools ¡ for ¡economic ¡research? ¡

slide-8
SLIDE 8

Broad ¡ques$ons ¡

  • What ¡happens ¡in ¡these ¡kinds ¡of ¡situa3ons? ¡
  • How ¡good ¡is ¡economic ¡theory ¡for ¡predic$ng ¡

what ¡happens ¡in ¡these ¡kinds ¡of ¡situa$ons? ¡

  • How ¡can ¡research ¡outside ¡economics ¡provide ¡

us ¡with ¡insights ¡about ¡these ¡kinds ¡of ¡games? ¡

  • How ¡do ¡the ¡limita$ons ¡that ¡these ¡games ¡pose ¡

for ¡economic ¡theory ¡make ¡them ¡useful ¡tools ¡ for ¡economic ¡research? ¡

slide-9
SLIDE 9

The ¡basic ¡game ¡ X ¡ Y ¡ Z ¡ . ¡. ¡. ¡ X ¡ 1, ¡1 ¡ 0, ¡0 ¡ 0, ¡0 ¡ Y ¡ 0, ¡0 ¡ 1, ¡1 ¡ 0, ¡0 ¡ Z ¡ 0, ¡0 ¡ 0, ¡0 ¡ 1, ¡1 ¡ . ¡. ¡. ¡

slide-10
SLIDE 10

Schelling ¡(1960) ¡

You ¡are ¡to ¡meet ¡somebody ¡in ¡New ¡York ¡City. ¡You ¡have ¡ not ¡been ¡instructed ¡where ¡to ¡meet; ¡you ¡have ¡no ¡prior ¡ understanding ¡with ¡the ¡person ¡on ¡where ¡to ¡meet; ¡and ¡ you ¡cannot ¡communicate ¡with ¡each ¡other. ¡You ¡are ¡simply ¡ told ¡that ¡you ¡will ¡have ¡to ¡guess ¡where ¡to ¡meet ¡and ¡that ¡ he ¡is ¡being ¡told ¡the ¡same ¡thing ¡and ¡that ¡you ¡will ¡just ¡have ¡ to ¡try ¡to ¡make ¡your ¡guesses ¡coincide ¡. ¡. ¡. ¡the ¡two ¡of ¡you ¡ must ¡guess ¡the ¡exact ¡minute ¡of ¡the ¡day ¡for ¡mee$ng. ¡ ¡ Grand ¡Central ¡Sta:on ¡at ¡noon ¡

slide-11
SLIDE 11

Mehta, ¡Starmer ¡& ¡Sugden ¡ ¡

(1994 ¡American ¡Economic ¡Review) ¡

  • Name ¡any ¡flower: ¡

¡ ¡ ¡ ¡ ¡ ¡ ¡Rose ¡(67 ¡percent) ¡

  • Write ¡down ¡any ¡day ¡of ¡the ¡year: ¡

¡ ¡ ¡ ¡ ¡ ¡ ¡25 ¡December ¡(44 ¡percent) ¡

  • Write ¡down ¡any ¡colour: ¡

¡ ¡ ¡ ¡ ¡ ¡ ¡Red ¡(59 ¡percent) ¡

  • Write ¡down ¡any ¡boy’s ¡name: ¡

¡ ¡ ¡ ¡ ¡ ¡ ¡John ¡(50 ¡percent) ¡

slide-12
SLIDE 12

Crawford, ¡Gneezy ¡& ¡Rojenstreich ¡

(2008 ¡American ¡Economic ¡Review) ¡

“X” ¡ “Y” ¡ “X” ¡ 5, ¡5 ¡ 0, ¡0 ¡ “Y” ¡ 0, ¡0 ¡ 5, ¡5 ¡ “X” ¡ “Y” ¡ “X” ¡ 5, ¡5.1 ¡ 0, ¡0 ¡ “Y” ¡ 0, ¡0 ¡ 5.1, ¡5 ¡

68% ¡ coordina:on ¡ 38% ¡ coordina:on ¡

slide-13
SLIDE 13

Stag ¡Hunt ¡ A ¡ B ¡ A ¡ 1000 ¡ 0 ¡ B ¡ 800 ¡ 800 ¡ Minimum-­‑input ¡(weak-­‑link) ¡coordina3on ¡game ¡ Minimum ¡choice ¡ 7 ¡ 6 ¡ 5 ¡ 4 ¡ 3 ¡ 2 ¡ 1 ¡ 7 ¡ 130 ¡ 110 ¡ 90 ¡ 70 ¡ 50 ¡ 30 ¡ 10 ¡ 6 ¡ 120 ¡ 100 ¡ 80 ¡ 60 ¡ 40 ¡ 20 ¡ 5 ¡ 110 ¡ 90 ¡ 70 ¡ 50 ¡ 30 ¡ 4 ¡ 100 ¡ 80 ¡ 60 ¡ 40 ¡ 3 ¡ 90 ¡ 70 ¡ 50 ¡ 2 ¡ 80 ¡ 60 ¡ 1 ¡ 70 ¡

Coordina$on ¡games ¡with ¡Pareto-­‑ ranked ¡equilibria ¡

slide-14
SLIDE 14

Van ¡Huyck, ¡Bajalio ¡& ¡Beil ¡ ¡

(1990 ¡American ¡Economic ¡Review) ¡

1 ¡ 2 ¡ 3 ¡ 4 ¡ 5 ¡ 6 ¡ 7 ¡ 1 ¡ 2 ¡ 3 ¡ 4 ¡ 5 ¡ 6 ¡ 7 ¡ Mean ¡Minimum ¡Choice ¡ Period ¡ Large ¡Groups ¡(n ¡= ¡14 ¡or ¡16) ¡ Small ¡groups ¡(n=2) ¡

slide-15
SLIDE 15

Brandts, ¡Cooper ¡& ¡Weber ¡ ¡

(2011 ¡working ¡paper) ¡

10 20 30 40 6 7 8 9 10 11 12 13 14 15 16 17 18 Average Minimum Effort Period

Internal Appointed Leader Internal Elected Leader External Appointed Leader

slide-16
SLIDE 16

Kogan, ¡Kwasnica ¡& ¡Weber ¡ ¡

(2011 ¡American ¡Economic ¡Review) ¡

Group ¡1 ¡

(Market) ¡

Group ¡2 ¡

(Outsiders) ¡

Asset ¡Market ¡ trading ¡(futures) ¡ shares ¡based ¡on ¡ Group ¡1 ¡outcome ¡ ¡

slide-17
SLIDE 17

1.0 2.0 3.0 4.0 1 2 3 4 5 6 7 8 Mean Minimum Input Period

Figure 1C: Group Minimum Input (Small Groups)

Control (n=8) Market-H (n=6) Market-L (n=4) Market-H Outsiders (n=6) Market-L Outsiders (n=4)

slide-18
SLIDE 18

1.0 2.0 3.0 4.0 1 2 3 4 5 6 7 8 Mean Minimum Input Period

Figure 1D: Group Minimum Input (Large Groups)

Control (n=8) Market-H (n=6) Market-L (n=4) Market-H Outsiders (n=6) Market-L Outsiders (n=4)

slide-19
SLIDE 19

Equilibrium ¡Selec$on ¡and ¡ Economic ¡Theory ¡

  • Equilibrium ¡refinements ¡(Kohlberg ¡& ¡Mertens ¡1986; ¡

Harsanyi ¡& ¡Selten ¡1988) ¡

  • Theories ¡of ¡dynamic ¡adjustment ¡(Kandori, ¡Mailath ¡& ¡

Rob, ¡1993; ¡Crawford, ¡1995) ¡

  • Theories ¡of ¡payoff ¡uncertainty ¡and ¡noisy ¡behavior ¡

(Anderson, ¡Goeree ¡& ¡Holt ¡2001; ¡Carlsson ¡& ¡van ¡ Damme, ¡1993) ¡

slide-20
SLIDE 20

Payoffs ¡ Behavior ¡

slide-21
SLIDE 21

Stag ¡Hunt ¡ A ¡ B ¡ A ¡ 1000 ¡ 0 ¡ B ¡ 800 ¡ 800 ¡ Minimum-­‑input ¡(weak-­‑link) ¡coordina3on ¡game ¡ Minimum ¡choice ¡ 7 ¡ 6 ¡ 5 ¡ 4 ¡ 3 ¡ 2 ¡ 1 ¡ 7 ¡ 130 ¡ 110 ¡ 90 ¡ 70 ¡ 50 ¡ 30 ¡ 10 ¡ 6 ¡ 120 ¡ 100 ¡ 80 ¡ 60 ¡ 40 ¡ 20 ¡ 5 ¡ 110 ¡ 90 ¡ 70 ¡ 50 ¡ 30 ¡ 4 ¡ 100 ¡ 80 ¡ 60 ¡ 40 ¡ 3 ¡ 90 ¡ 70 ¡ 50 ¡ 2 ¡ 80 ¡ 60 ¡ 1 ¡ 70 ¡

Coordina$on ¡games ¡with ¡Pareto-­‑ ranked ¡equilibria ¡

slide-22
SLIDE 22

The ¡basic ¡game ¡ X ¡ Y ¡ Z ¡ X ¡ 1, ¡1 ¡ 0, ¡0 ¡ 0, ¡0 ¡ Y ¡ 0, ¡0 ¡ 1, ¡1 ¡ 0, ¡0 ¡ Z ¡ 0, ¡0 ¡ 0, ¡0 ¡ 1, ¡1 ¡

slide-23
SLIDE 23

The ¡basic ¡game ¡

X ¡ Y ¡

Z ¡

X ¡

1, ¡1 ¡ 0, ¡0 ¡ 0, ¡0 ¡

Y ¡

0, ¡0 ¡ 1, ¡1 ¡ 0, ¡0 ¡

Z ¡

0, ¡0 ¡ 0, ¡0 ¡ 1, ¡1 ¡

slide-24
SLIDE 24

Payoffs ¡ Behavior ¡

Uniqueness ¡ / ¡Cogni$on ¡

slide-25
SLIDE 25

Equilibrium ¡Selec$on ¡and ¡ Economic ¡Theory ¡

  • Equilibrium ¡refinements ¡(Kohlberg ¡& ¡Mertens ¡1986; ¡

Harsanyi ¡& ¡Selten ¡1988) ¡

  • Theories ¡of ¡dynamic ¡adjustment ¡(Kandori, ¡Mailath ¡& ¡

Rob ¡1993; ¡Crawford ¡1995) ¡

  • Theories ¡of ¡payoff ¡uncertainty ¡and ¡noisy ¡behavior ¡

(Anderson, ¡Goeree ¡& ¡Holt ¡2001; ¡Carlsson ¡& ¡van ¡ Damme ¡1993) ¡

  • Theories ¡of ¡“salience” ¡and ¡cogni$on ¡(Bacharach ¡and ¡

Bernasconi ¡1997; ¡Blume ¡& ¡Gneezy ¡2010) ¡

slide-26
SLIDE 26

The ¡basic ¡game ¡

X ¡ Y ¡

Z ¡

X ¡

1, ¡1 ¡ 0, ¡0 ¡ 0, ¡0 ¡

Y ¡

0, ¡0 ¡ 1, ¡1 ¡ 0, ¡0 ¡

Z ¡

0, ¡0 ¡ 0, ¡0 ¡ 1, ¡1 ¡

slide-27
SLIDE 27

The ¡basic ¡game ¡ 1, ¡1 ¡ 0, ¡0 ¡ 0, ¡0 ¡ 1, ¡1 ¡

slide-28
SLIDE 28

The ¡inherent ¡difficulty ¡of ¡equilibrium ¡ selec$on ¡for ¡economics ¡

  • “. ¡. ¡. ¡In ¡the ¡final ¡analysis ¡we ¡are ¡dealing ¡with ¡

imagina$on ¡as ¡much ¡as ¡with ¡logic ¡. ¡. ¡. ¡Poets ¡ may ¡do ¡bejer ¡than ¡logicians ¡at ¡this ¡ game ¡. ¡. ¡.” ¡(Schelling ¡1960) ¡

  • “. ¡. ¡. ¡predic$ng ¡which ¡of ¡the ¡many ¡equilibria ¡

will ¡be ¡selected ¡is ¡perhaps ¡the ¡most ¡difficult ¡ problem ¡in ¡game ¡theory” ¡(Camerer ¡2003) ¡

slide-29
SLIDE 29

Payoffs ¡ Behavior ¡

Uniqueness ¡ / ¡Cogni$on ¡ Social ¡/ ¡ cultural ¡ factors ¡

slide-30
SLIDE 30
  • Can ¡economics ¡as ¡a ¡scien:fic ¡discipline ¡that ¡

must ¡extricate ¡itself ¡from ¡its ¡current ¡ conceptual ¡crisis, ¡benefit ¡from ¡concepts, ¡ methods ¡and ¡insights ¡developed ¡in ¡other ¡ disciplines, ¡notably ¡the ¡natural ¡sciences? ¡

  • Yes ¡(but, ¡in ¡this ¡case, ¡not ¡easily) ¡

– Computa$onal ¡neuroscience ¡/ ¡ajen$on ¡/ ¡visual ¡ percep$on ¡ – Complex ¡systems ¡/ ¡computa$onal ¡sociology ¡ – Computer ¡science ¡/ ¡ar$ficial ¡intelligence ¡ – Conven$ons ¡in ¡philosophy, ¡linguis$cs ¡and ¡law ¡ – Poetry? ¡

slide-31
SLIDE 31

Payoffs ¡ Behavior ¡

Uniqueness ¡ / ¡Cogni$on ¡ Social ¡/ ¡ cultural ¡ factors ¡

slide-32
SLIDE 32

“good” ¡ “neutral” ¡ “bad” ¡ “good” ¡ 1, ¡1 ¡ 0, ¡0 ¡ 0, ¡0 ¡ “neutral” ¡ 0, ¡0 ¡ 1, ¡1 ¡ 0, ¡0 ¡ “bad” ¡ 0, ¡0 ¡ 0, ¡0 ¡ 1, ¡1 ¡

Telling ¡a ¡lie ¡

¡

Sharing ¡money ¡with ¡someone ¡in ¡need ¡

¡

Accep:ng ¡a ¡bribe ¡

¡

Arriving ¡late ¡for ¡work ¡ ¡ ¡

slide-33
SLIDE 33

The ¡value ¡of ¡coordina$on ¡games ¡. ¡. ¡. ¡

  • Behavior ¡in ¡coordina$on ¡games ¡can ¡also ¡tell ¡us ¡

(economists) ¡about ¡underlying ¡social ¡factors ¡that ¡may ¡

  • therwise ¡be ¡hard ¡to ¡iden$fy ¡
  • Examples ¡(all ¡using ¡pure-­‑matching ¡games): ¡

– Krupka ¡& ¡Weber ¡(forthcoming, ¡JEEA) ¡uses ¡simple ¡ coordina$on ¡games ¡to ¡iden$fy ¡and ¡measure ¡social ¡norms ¡ (which ¡then ¡predict ¡behavior) ¡ – Houser ¡& ¡Xiao ¡(2011 ¡Experimental ¡Economics) ¡use ¡ coordina$on ¡games ¡to ¡classify ¡the ¡content ¡of ¡natural ¡ language ¡messages ¡ – Von ¡Ahn ¡& ¡Dabbish ¡(2004 ¡SIGCHI ¡Conference ¡proceedings) ¡ use ¡coordina$on ¡games ¡to ¡label ¡images ¡on ¡the ¡internet ¡