Group Identity and Competitiveness

Transcription

1 Group Identity and Competitiveness Marie-Pierre Dargnies Paris School of Economics, Université Paris 1 Panthéon-Sorbonne October Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

2 Motivations Main conclusions of the previous paper: Women do not enter the tournament significantly more often when it is team-based. High-performing men choose to compete less often when it is team-based unless they know they will be matched with a teammate of the same level. Idea: Belonging to a group and having to compete against members of another group may make women more willing to compete and high-performing men may accept to risk being matched with a low-performing teammate when he or she is a fellow group member. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

3 Motivations (2) Standard economic theory assumes economic agents act as isolated individuals. Akerlof and Kranton have been the first to emphasize how social identity may affect economic outcomes (Akerlof and Kranton 2000, 2002, 2005) Recent experimental results show that subjetcs make more socially-oriented decisions when they deal with fellow group members rather than with random participants or outgroup members (Eckel and Grossman (2005), Charness et al (2007), Chen and Li (2009), Sutter (forthcoming)). Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

4 Experimental Design Experimental design as close as possible to the previous one. Only change: When entering a tournament, a participant will compete: against opponent(s) of the other group with a teammate, in the case of team tournament, of the same group Compare the results to those of the previous experiment. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

5 Experimental Design: Identity sessions Participants come to the lab and sit randomly. They are told that the computer randomly assigned them to one of two groups and that it is important that they do not find out who is in each of the two groups. Group-identity-building activities Participants from the Identity sessions go through the same 8 tasks as participants from the Benchmark sessions. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

6 Experimental Design: Group-identity-building activities Choose a group name. Quizz: 4 general knowledge questions. For both activities, participants can use an instant message system which allows them to talk with members of their own group but not with members of the other group. Participants are reminded that anonymity is important and that they should be careful not to give information (including their gender) that could help others identify themselves. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

7 Differences between the Benchmark and Identity sessions 39 men and 37 women took part in the Benchmark sessions. 52 men and 38 women took part in the Identity sessions. This could explain some of the results. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

8 The effect of group identity on performance and confidence For both men and women, the Task 1 (piece rate) and Task 2 (mandatory tournament) performances are lower among participants of Identity sessions. The difference is significant only for the Task 2 performance of men (p=0.05) 78% of women in the Benchmark sessions and 58% in the Identity sessions think they are likely to win the individual tournament (p=0.06). Could be caused by the higher proportion of men in the Identity sessions. Respectively 85% and 83% (p=0.81) of men in the Benchmark sessions and the Identity sessions hold beliefs consistent with winning the individual tournament. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

9 The effect of group identity on entry in the individual tournament In the Benchmark sessions, 51% of women and 85% of men chose to enter the tournament (p=0.00). In the Identity sessions, 32% of women and 75% of men (p=0.00) made such a choice. Women from the Identity sessions choose marginally significantly (p=0.10) less often the tournament than women from the Benchmark sessions. This is not the case for men (p=0.31). Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

10 Proportion of men and women from Benchmark and Identity sessions entering the individual tournament conditionnal on performance level Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

11 The effect of social identity on entry in the team tournament In the Benchmark sessions, 62% of women and 59% of men (p=0.82) chose to enter the team tournament. In the Identity sessions, 42% of women and 69% of men (p=0.02) made such a choice. In both treatments, women choose the tournament more often when it is team based, even if not significantly so (p=0.48 both times). While, in the Benchmark treatment, men entered the tournament significantly less often when they were part of a team rather than alone (p=0.02), in the Identity treatment, they chose the tournament almost as often whether it was individual or team-based (p=0.66). Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

12 Proportion of men and women from Benchmark sessions and Identity sessions entering the individual tournament (IT) and team tournament (TT) Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

13 Logit of Decision to Enter the Team Tournament (Task 4) Regressors Benchmark Identity All (1) (2) (3) (1) (2) (3) (1) (2) (3) Id*Female (0.06) (0.14) (0.22) Id 0.16 (0.36) 0.17 (0.45) 0.19 (0.46) Female 0.04 (0.90) 0.05 (0.82) 0.08 (0.81) (0.01) (0.02) (0.10) 0.08 (0.90) 0.10 (0.91) 0.14 (0.98) Prob (0.09) (0.19) (0.19) (0.28) (0.51) (0.52) (0.05) (0.15) (0.15) Guesswin 0.21 (0.14) 0.11 (0.95) 0.09 (0.32) 0.09 (0.46) 0.04 (0.63) (0.75) 0.10 (0.22) 0.05 (0.45) 0.01 (0.68) Submit (0.00) (0.00) (0.03) (0.04) (0.00) (0.00) Idpartn (0.87) (0.03) (0.05) Observations Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

14 Entry rates for each of the 3 tournaments by gender and performance level Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

15 Logit of Low-Performing Participants Tournament Entry Decision (Task 3 and 4) Regressors Men Women (1) (2) (3) (4) (1) (2) (3) (4) Id*Team (0.26) (0.22) (0.17) (0.11) 0.12 (0.18) 0.06 (0.31) 0.04 (0.19) 0.05 (0.10) Id 0.30 (0.38) 0.36 (0.29) 0.45 (0.24) 0.45 (0.80) (0.04) (0.08) (0.05) (0.06) Team 0.30 (0.26) 0.41 (0.12) 0.45 (0.16) 0.50 (0.09) 0.22 (0.70) 0.26 (0.53) 0.27 (0.88) 0.34 (0.88) Prob (0.56) (0.14) (0.11) (0.02) 0.03 (0.96) (0.79) 0.01 (0.99) 0.51 (0.38) Guesswin 0.22 (0.01) 0.27 (0.01) 0.26 (0.03) 0.10 (0.44) (0.51) 0.06 (0.93) Submit (0.00) (0.01) (0.29) (0.54) Idpartn (0.00) (0.00) Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

16 Logit of High-Performing Participants Tournament Entry Decision (Task 3 and 4) Regressors Men Women (1) (2) (3) (4) (1) (2) (3) (4) Id*Team 0.13 (0.02) 0.14 (0.03) 0.10 (0.06) 0.12 (0.08) (0.27) (0.26) (0.28) (0.28) Id (0.09) (0.13) 0.01 (0.18) 0.06 (0.24) (0.56) (0.58) 0.02 (0.93) 0.04 (0.96) Team (0.00) (0.01) (0.01) (0.02) 0.24 (0.28) 0.26 (0.29) 0.33 (0.32) 0.35 (0.33) Prob (0.44) (0.31) (0.26) (0.36) 0.49 (0.27) 0.47 (0.30) 0.44 (0.24) 0.39 (0.30) Guesswin 0.55 (0.00) 0.48 (0.02) 0.40 (0.03) 0.07 (0.56) (0.94) (0.97) Submit (0.00) (0.02) (0.00) (0.01) Idpartn (0.21) (0.46) Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

17 To sum up Men s behavior seems to have shifted towards decisions more in the favor of the group: Low-performing men enter the team tournament less often unless they are matched with a teammate of the same (low) level as their own. High-performing men do not shy away anymore from the team competition when they risk being matched with a less able teammate. Women become less competitive and act a bit more selfishy: Low-performing women prefer entering the team tournament when they do not know for sure they will be matched with a teammate of the same (low) level as their own. High-performing women enter the team tournament (a bit) more often when their teammate is of the same (high) level as their own. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

18 Why do women become less competitive when belonging to a social group? It may have nothing to do with group identity per se. It could be due to the higher proportion of men in the Identity sessions. (1) The group-identity-building activities may not have had the intended effect on women (Cadsby and Maynes (1998), Brown-Kruse and Hummels (1993)). (2) Women may react differently to the group identity: may feel more comfortable taking advantage of a fellow group member than of a random participant... (3) Cannot discriminate between hypothesis 2 and 3. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

19 Exploring hypothesis 1 10 men and 10 women participated in each of two additional sessions and they were told they belonged to a group composed of themselves + the other women or men. No group-identity-building activities. Participants went through the same 8 tasks and whenever they had to choose whether to enter a tournament they knew their opponent(s) would be of the other gender while their teammate, in the case of team competition, would be of the same gender as their own. If the higher proportion of men in the Identity sessions than in the benchmark sessions caused less women to enter the competition, we would expect even less women to enter tournaments in the "Gender-identity sessions" where they know for sure their opponent(s) are men. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

20 Exploring hypothesis 1 Women from the gender-identity sessions are found to be more confident in their chances of winning (75% hold beliefs consistent with winning the individual tournament) than their counterparts from the Identity sessions (58%) and about as confident as women from the Benchmark sessions (78%). Women from the gender-identity sessions enter the individual tournament at a slightly higher rate than women in the Identity sessions (40% vs 32%) and a little bit more often than their counterparts from the Benchmark sessions of which 51% prefer the individual tournament to the piece rate. Due to the scarce data available (only two Gender identity sessions were run), those differences do not reach significance. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

21 Men and women willingness to compete as part of a team Men (and especially, high-performing men) are less willing to compete as part of a team than alone when they do not know the performance level or anything else about their teammate. Knowing that their teammate will be of the same level as their own or sharing a group membership is sufficient to have high-performing men self-select into the team competition. Women are not more prone to enter a competition when it is team-based. If anything, the fact that their teammate is a fellow group member makes them even more reluctant to enter competitions. Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22

22 Future research The idea of team used in these papers is a rather minimal one: Participants do not know their teammate before deciding whether to engage in the competition. Since the team does not exist previous to the decision to compete, there is no discussion between team members as to whether to enter the competition. It may be interesting to study how some people may be prone to being talked into entering a competition. Does a male have more chances than a female to convince a teammate to enter the competition? Marie-Pierre Dargnies (Univ Paris 1) Group Identity and Competitiveness October / 22