Take the lead! An experimental analysis of leadership in team production
description
Transcript of Take the lead! An experimental analysis of leadership in team production
Take the lead!An experimental analysis of leadership in team production
Enrique FatasLINEEXUniversitat de València
Sara GodoyLINEEXUniversitat de València
Motivation
Leadership in Team Production Leaders may alleviate free riding under some conditions
Leaders have the chance to exclude free riders, Guth et al. (2004) Leaders distribute earnings, Potters et al. (2006) They move first on the basis of additional information, Meidinger
and Villeval (2002) Leaders are endogenously determined and move first, Gächter and
Renner (2005), Arbak and Villeval (2007) Take the lead!
Endogenous leadership: anyone can be the leader Horizontal game: Leaders and followers make the very same
decisions Leaders cannot sanction followers: No sanctions, no punishment,
no informal peer pressure = low powered incentive system Leading by example is endogenous, costless but strategically risky
Motivation (II)
Take the lead! Endogenous leadership
Anyone can be the leader, in any period Horizontal game
Leaders and followers make the very same decisions Leaders cannot sanction followers
No sanctions, no punishment, no informal peer pressure = low powered incentive system
Leading by example Example is based on available information
Becoming a leader is costless but strategically risky A strong test for leading by example
Experimental Design
The game A standard linear public goods game (VCM)
MPCR=.5; group size =4; 10+10 rounds; 50 tokens; partners
Croson, Fatas and Neugebauer (EL, 2005) Treatments differ on the available information set
Follow the good leader treatment (FGL) Available information: highest contribution
Follow the bad leader treatment (FBL) Available information: lowest contribution
Natural baseline CFN Available information: vector of contributions, no trace
Standard experimental procedures
Results (I) Good leaders matter
Results (I) Good leaders matter
Results (II)
Endogenous leadershipGood leaders contribute more
Bad leaders do notVCM = 100
0
100
Highest Lowest
FGLFBL
Results (III)
The role of information
Table 3: Conditional cooperation patternRandom effects regression results; Dependent Variable: Individual contributions
FGL R01-10 R11-20 FBL R01-10 R11-20
Constant 23.63***
(3.58) 26.36***
(3.51) 16.92***
(4.356)*
11.66***
(1.84) 12.00***
(1.95) 9.38***
(2.28)
Restart -5.49***
(1.42) -1.63***
(0.57) Period -1.07***
(0.20)-1.27***
(0.23)-0.90***
(0.32)-0.62***
(0.16)-0.65***
(0.20)-0.63**
(0.26) LagHigh 0.09*
(0.05) 0.05
(0.07) 0.10(0.07)
LagLow 0.79***
(0.15) 0.75*** (0.17)
0.93***
(0.09) Nº Obs 648 324 324 648 324 324
Subjects do not follow good leaders, they pay attention to the bad ones
Concluding remarks Leadership significantly increases effort levels
(contribution) The usual PGG decline holds Good leaders contribute significantly more,
while bad leaders are not ashamed of their low contribution levels
Subjects follow bad leaders (instead of good ones)
“Don’t follow the leader, watch your parkimeter!” (Bob Dylan)