Measuring Norm s of Cooperation in Different Societies Simon - - PowerPoint PPT Presentation

Measuring Norm s of Cooperation in Different Societies

Simon Gächter (University of St. Gallen) Benedikt Herrmann (University of Göttingen) Sapporo, 26 June 2003

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

1

Contents

• Background & Motivation
• The experimental design and procedure
• The subject pools
• Expectations, decisions, attitudes and emotions
• A cross-generational & urban-rural comparison
• Conclusions

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

2

Background I : The cooperation problem

• Many social and economic problems are problems of

voluntary cooperation with free rider incentives.

• Groups of four subjects. Each subject is endowed with

y= 20 tokens. Subjects have to decide how many tokens to keep privately and how many tokens to invest in a group project.

• For each token invested in the project, each subject in

the group receives 0.4 tokens, i.e., the group together earns 1.6 tokens.

⇒

Group as a whole benefits from a contribution.

⇒

Yet, each contributor loses 0.6 tokens.

⇒

Purely self-interested subjects will never contribute.

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

3

Typical results

2 4 6 8 10 12 14 16 18 20 1 2 3 4 5 6 7 8 9 10 Period m ean cooperation

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

4

Cooperation in the presence of a punishment opportunity

• Modification: After contribution decisions each

member is informed about the contribution vector and can assign punishment points to each of the

• ther members.
• For every point assigned the punisher has costs of 1

and the punished player has costs of 3.

• Self-interest hypothesis predicts zero punishment and
• n contribution levels.
• Existence of reciprocal types predicts punishment and

hence an impact on cooperation.

• Important predecessors: Yamagishi JPSP 1986;

Ostrom et al. APSR 1992

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

5

Screenshot Punishm ent stage

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

6

Punishm ent solves the cooperation problem

Fehr & Gächter AER 2000; Nature 2002

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

7 2 4 6 8 10 12 14 16 18 20 1 2 3 4 5 6 7 8 9 10 11 12 Period Mean contribution

with punishment without punishment Replicated by

• Fehr & Gächter

2003a,b

• Page et al. 2002a,b;

2003

• Masclet et al. 2003
• Carpenter 2002, 2003
• Sefton et al. 2002
• Noussair&Tucker 2002
• …

Punishm ent

• f free riders

7.8 32.9 26.7 22 7.5 3.1 4.4 31.4 40.3 14.8 4.9 4.2 5.5 31.9 39.4 13 4.2 6

1 2 3 4 5 6 7 8 9 10

[-20,-14) [-14,-8) [-8,-2) [-2,2] (2,8] (8,14] Deviation from the average cooperation level of the other group members Mean expenditure by punishing group members

period1-4 period5-6 period6

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

8

Background I I : Cross-cultural experim ents

• International comparisons of cooperation and trust

e.g., Yamagishi 1986; Kachelmeier & Shehata 1995; Buchan et al. 2002; Ashraf, et al. 2003

• Henrich et al. 2001; 2002: How universal is behavior that

has been observed in affluent university students? Go to remote tribes to test. (15 small-scale societies).

• We go to Russia. Poor areas; some of them remote from

Western influence. Soviet “spirit” still alive.

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

9

Background I I I : Cooperation as a m easure of social capital

“Social capital generally refers to trust, concern for one’s associates, a willingness to live by the norms of one’s community and to punish those who do not.” Bowles and Gintis (2002, p. F419) Current literature focuses more on trust than on cooperation and norm enforcement. Cooperation and norm enforcement an important element

• f „social capital“.

The focus on trust alone neglects the problems of incentives for free riders to cooperate.

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

10

Experim ental research on cooperation and sanctioning behavior m ay yield a better understanding of social capital.

Why Russia?

• Results of several surveys show different patterns of trust

between Russia and Western Europe e.g.:

• Rose (2000), calls Russia a “antimodern“ society.
• Hjollund, Paldam and Svendsen (2001), formulate the hypothesis
• f negative social capital as a consequence of dictatorships.
• Collectivist and authoritarian ideology and practice.
• Lack of successful voluntary cooperation accounts for

many development and transistion problems.

• Woolcock (1998).
• Campos N.F. and Coricelli F. C. (2002).

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

11

Our research question

Are there different patterns of cooperation and sanctioning behavior in the former Soviet Union and Western Europe?

Test instrument: run exactly the same cooperation

game in different societies.

Identical incentives. Differences in behavior reflect cross-societal

differences.

Cross-societal comparison with 926 students in six

places (Zürich, St. Gallen, Goettingen, Minsk, Belgorod and Jekaterienburg).

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

12

Design overview

1 st one-shot Experim ent 2 nd one-shot Experim ent N-P experim ents No punishment (N) With punishment (P) P-N experim ents With punishment (P) No punishment (N)

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

13

Sequence of

• ne-shot experim ent ( N-P)

1 . Contribution decision ( N) Expectation and confidence about others‘ contribution? 2 . Contribution decision ( P) Expectation and confidence about others‘ contribution? 3 . Punishm ent decision ( P)

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

14

Expectation about others‘ punishment?

• 4. Post-experimental questionnaire

Methods

• Instructions with detailed explanations and control questions

were translated from German into Russian, forward and backwards, to ensure that texts are identical.

• In all places the same software (Z-tree) and the same displays

were used.

• The experimenter was in all places the same person.
• Experiment only continues when control questions are correctly

answered.

• Instructions and procedures orally summarized according to a

script.

• To avoid currency effects “Guilders” were used as experimental

currency units.

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

15

Geography of experim ents

West:

Goettingen

• St. Gallen

Zürich

East:

Minsk Belgorod Jekaterienburg Kursk Ust-Kinel

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

16

Characteristics of the subject pools

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

17

Russians and Belorussians Swiss and Germans Average share of females: 31.09% 35.04% Average age in years: 20.08 21.25 Average number of known participants: 1.63 1.33 Percentage of economists: 21.94% 31.39% Average income in experiment: 3.68 € 23.87 € Average monthly budget: 75.72 € 398.38 € Percentage of monthly budget earned in the experiment: 4.9% 5.3%

Results

1 . Expectations concerning cooperation

a.in N in the N-P experiments b.in P in the P-N experiments measure the first guess people have about cooperation of others

2 . Actual contributions

a.in N in the N-P experiments b.in P in the P-N experiments measure actual cooperation

• 3. Measure confidence in expectations
• 4. Look at the change in incentives in an N-P
• r P-N experiment respectively.

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

18

1 . Expected cooperation

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

19

Russia & Belorussia – N vs. P: t=1.05 Germany & Switzerland – N vs. P: t=2.02

5 6 7 8 9 10 11 12 13 14 15 mean expected contribution

Russia & Belorussia (n=288) Russia & Belorussia (n=216) Germany & Switzerland (n=252) Germany & Switzerland (n=159)

No Punishment (N) With Punishment (P)

t = 1.16 t = 1.89

2 . Actual cooperation

5 6 7 8 9 10 11 12 13 14 15 mean actual contribution

Russia & Belorussia (n=288) Russia & Belorussia (n=216) Germany & Switzerland (n=252) Germany & Switzerland (n=159)

No Punishment (N) With Punishment (P)

t = 4.96 t = 2.63

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

20

Russia & Belorussia – N vs. P: t = 1.27 Germany & Switzerland – N vs. P: t = 3.61

3 . Expected and actual received punishm ent

1 2 3 4 5 6 7 8 [-20,-14) [-14,-8) [-8,-2) [-2,2] (2,8] (8,14] (14,20] [-20,-14) [-14,-8) [-8,-2) [-2,2] (2,8] (8,14] (14,20]

Deviation from mean contribution of other group members Mean punishment Germany & Switzerland Russia & Belorussia

Expected punishment Actual punishment

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

21

4 . Confidence in one‘s expectation …

(1= no confidence; 10= full confidence)

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

22

4 4.5 5 5.5 6 6.5 7

M e a n c o n f i d e n c e

Russia & Belorussia (n=288) Russia & Belorussia (n=216) Germany & Switzerland (n=252) Germany & Switzerland (n=159)

No Punishment (N) With Punishment (P)

t = 4.16 t = 4.22 4 4.5 5 5.5 6 6.5 7 M e a n c o n f i d e n c e Russia & Belorussia (n=288) Germany & Switzerland (n=216) t = 0.12

… concerning cooperation … concerning punishment

5 . Expected and actual reactions to changed incentives

5 6 7 8 9 10 11 12 13 p=.160 p=.414 p=.064 p=.033 p=.053 p=.106 p=.000 p=.009

Mean cooperation level

Belgorod, Jekaterienburg, Minsk: N=504 Zürich, St. Gallen, Goettingen: N=411

expected actual expected actual

N P

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

23

6 . Expected and actual punishm ent in the N-P experim ents ( 2 nd sequence)

1 2 3 4 5 6 7 8 [-20,-14) [-14,-8) [-8,-2) [-2,2] (2,8] (8,14] (14,20] [-20,-14) [-14,-8) [-8,-2) [-2,2] (2,8] (8,14] (14,20]

Deviation from mean contribution of other group members Mean punishment Russia & Belorussia Germany & Switzerland

Expected punishment Actual punishment

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

24

Tapping m otivations and em otions

• Questionnaires on motives in the N-experiments:

“I believe, the other group members are mainly interested in maximizing their own income.“ “If someone has invested a lot in the one-stage experiment, it is his own fault if he is exploited.“

• Questionnaires on motives in the P-experiments:

“I believe that I will receive deductions points from the

• ther group members, if I contribute less than they do.

To avoid this, I decided to contribute the amount I thought the others would spend.” “I suppose that the deduction points will be used arbitrarily so I can‘t influence their distribution. That‘s why the deduction points had no impact on my contribution decision.”

• Emotions

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

25

“I believe, the other group m em bers are m ainly interested in m axim izing their ow n incom e“

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

26

0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50% agree strongly agree indifferent disagree disagree strongly

Russian & Belorussians Swiss and Germans

“I f som eone has invested a lot in the one-stage experim ent, it is his ow n fault if he is exploited“

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

27

0% 5% 10% 15% 20% 25% 30% 35% agree strongly agree indifferent disagree disagree strongly

Russians & Belorussians Swiss and Germans

Motives for the contribution in the P-experim ents I

Statement I: “I believe that I will receive deductions points from the other group members, if I contribute less than they

• do. To avoid this, I decided to contribute the amount I

thought the others would spend.” Percentage of agreement:

• Russians and Belorussians: 8.97%
• Swiss and Germans: 26.57%

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

28

Motives for the contribution in the P-experim ents I I

Statement II: “I suppose that the deduction points will be used arbitrarily so I can‘t influence their distribution. That‘s why the deduction points had no impact on my contribution decision.” Percentage of agreement:

• Russians and Belorussians: 19.31%
• Swiss and Germans: 5.16%

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

29

Em otions

• Self-reported emotions.
• Bosman & van Winden 2002.
• Emotions questionnaire used in psychology.

Contribution group member 1 X points Contribution group member 2 Y points

Sympathy

None o o o o o o very much None o o o o o o very much

Anger

None o o o o o o very much None o o o o o o very much

Contempt

None o o o o o o very much None o o o o o o very much

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

30

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

31

1 2 3 4 5 6 7

[-20,-14): [-14,-8): [-8,-2): [-2,2]: (2,8]: (8,14]: (14,20]:

The subject who is the target of emotion deviates from own contribution by ... Mean score Russians & Belorussians Swiss and Germans

Em otions - Sym pathy

Em otions - Anger

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

32

1 2 3 4 5 6 7

[-20,-14) [-14,-8) [-8,-2) [-2,2] (2,8] (8,14] (14,20]

The subject who is the target of emotion deviates from own contribution by ... Mean score Russians & Belorussians Swiss and Germans

Em otions - Contem pt

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

33

1 2 3 4 5 6 7

[-20,-14) [-14,-8) [-8,-2) [-2,2] (2,8] (8,14] (14,20]

The subject who is the target of emotion deviates from own contribution by ... Mean score Russians & Belorussians Swiss and Germans

Sum m ary of the cross-societal com parison

1. The students subject pools in East and West show different expectations concerning the impact of a punishment opportunity on cooperation behavior. 2. Russian and Belorussian students respond in a different way to the presence of the punishment option than the Swiss and German subjects. 3. Eastern and Western students have different moral judgments towards cooperation. 4. The measured emotions reveal similar positive feelings in both subject pools, but reduced intensities of negative feelings in the Eastern subject pool.

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

34

Are there intergenerational differences? Norm s of cooperation am ong urban and rural dw ellers

Gächter & Herrmann 2003b Can we find differences in the cooperation behavior between a.) non-students who lived the most part of their life in a collectivist society and students who were socialized in the post-socialist era? b) people from areas that are still more characterized by a a “Soviet life style“ like the rural areas and dwellers of urban centers?

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

35

Characteristics of our subject pools

I n Rubles I n Dollars Urban Students 126 20.1 18.3 915 30.5 Rural Students 111 20.4 28 667 22.23 Urban non- students 156 42.8 54 1775 59.16 Rural non- students 105 38.3 50.5 1330 44.33 Age in years ( means) I ncome ( mean/ month) Subject pool: Number

• f

Subjects ( n= 4 9 8 ) Gender ( % Female)

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

36

Contribution rates in NP and PN one-shot experim ents

13

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

37

5 6 7 8 9 10 11 12 p=.131 p=.393 p=.270 p=.166 p=.100 p=.123 p=.934 p=.840

Mean cooperation

NP PN

Urban students (n=126) Rural students (n=111) Urban Non-students (n=156) Rural Non- Students (n=105)

p<0.001 p=0.043 p<0.001 p=0.010 p=0.008 p=0.002 p=0.109 p=0.03

level

1 2 3 4 5 Deviation from the punishers‘ contribution

Mean punishment

Urban Students (N = 126) Rural Students (N = 111) Urban Non-Students (N = 156) Rural Non-Students (N = 105)

Punishm ent behavior

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

38

„ I f som eone has invested a lot in the one-stage experim ent, it is his ow n fault if he is exploited“

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

39

0% 10% 20% 30% 40% 50% 60%

agree strongly agree indifferent disagree disagree strongly agree strongly agree indifferent disagree disagree strongly agree strongly agree indifferent disagree disagree strongly agree strongly agree indifferent disagree disagree strongly

Urban students (n=126) Rural students (n=111) Urban Non-students (n=156) Rural Non- students (n=105)

• The non-student subject pools show higher contribution

rates than the urban students.

• The students have at least in the P-sequence of PN a

higher contribution rate than in the N-sequence of NP revealing a (insignificant) sensibility against the punishment

• ption.
• The punishment pattern of both the non-student subject

pool and the rural students differs strongly from students in Western Europe. = > It looks like the experiences of the Soviet past have shaped the norms of cooperation and attitudes towards free riders and cooperators.

Conclusion

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

40

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

41

The dynam ics of cooperation in the presence and absence of punishm ent opportunities

Gächter, Herrmann & Thöni 2003

2 4 6 8 10 12 14 16 18 20 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Period Average Contributions

Zürich Minsk Samara

No punishment Punishment

Kruskal-Wallis test: p=0.503 Kruskal-Wallis test: p=0.017

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

42

Punishm ent behavior

0.5 1 1.5 2 2.5 3

[-20, -15] [-14, -9] [-8, -3] [-2, 2] [3, 8] [9, 14] [15, 20]

Deviation from the punisher's contribution Average punishment points Zürich Samara Minsk

Concluding discussion

• Are there cultural differences in psychological

functionings?

No change in cooperation in Russia, despite higher levels of

expected punishment. Opposite result in the West.

Different attitudes and expressed emotions toward cooperators

and free riders.

• What is the role of institutions and life experiences in

shaping norms of cooperation?

• Experiments, in combination with standard empirical

methods seem to be good tools.

People react to the same incentive structure.

Very high degree of control.

(c) Simon Gächter, University of St. Gallen (FEW-HSG)

43