AnExperimentalStudyontheImportanceofGroupCompositionforLinearPublicGoodsGames Partners,StrangersandFreeRiders

(1)

Partners, Strangers and Free Riders

An Experimental Study on the Importance of Group Composition for Linear Public Goods Games

Matheus Albergaria de Magalhães*

Edvan Soares de Oliveira*

*FUCAPE Business School

Quinto Encontro de Economia do Espírito Santo (V EEES) FUCAPE Business School, Vitória (ES)

November 3-4, 2014

Matheus Albergaria de Magalhães

(2)

S ECTIONS

Motivation

Experiment

Evidence

Conclusions

References

(3)

M OTIVATION

I

Pure Public Goods: non-rival and non-exclusive .

I

A few examples:

I

Lighthouses (Classic Example).

I

National Defense.

I

Organizational Knowledge.

(4)

M OTIVATION

Examples of Public Goods

Lighthouse National Security

Source:Google Images(https://www.google.com/imghp)

.

(5)

M OTIVATION

Main Challenges:

I

Will people take part in collective actions involving public goods?

I

Will individual actions lead to socially efficient results?

(6)

M OTIVATION

Andreoni (1988) reported three consistent results for public-goods experiments:

1. No significant evidence of free riding in single-shot games.

2. In experiments involving repeated play, subjects’

provisions for public goods tended to decay.

3. Exact free riding was seldom realized.

(7)

M OTIVATION

Table 1

Marwell and Ames’ (1981) Summary of 12 Different Experiments

Source: Marwell and Ames (1981, Table 2, p.307)

.

(8)

M OTIVATION My goals today:

I

Describe a simple experiment, related to the provision of public goods.

I

Evaluate the importance of random rematching for linear public goods games.

I

Propose future research agenda for related themes.

(9)

E XPERIMENT The Experiment:

1. We ran five-round experiments in a Business School

(Accounting, Business, Economics and MBA students) . 2. Subjects filled a form deciding how to divide R$ 100

between a private and a public good.

3. For each R$ 1.00 invested in the private good, subjects would receive R$ 1.00.

4. For each R$ 1.00 invested in the public good, subjects would receive R$ 0.50.

5. Subjects were divided in two groups: (i) ’Partners’ (fixed composition); (ii) ’Strangers’ (random composition).

(10)

E XPERIMENT

A Linear Public Goods Game

Individuals were given a budget (m), which could be invested either in a private (x) or public good (g), with x + g = m.

Individual payoffs (P

_i

) were determined by the following formula:

P

_i

= x

_i

+ α

n

X

j=1

g

_j

(1)

where n was the number of group members. The parameter α

was chosen such that 0 < α < 1.

(11)

E XPERIMENT

A Linear Public Goods Game

Given the above game, we may have the following payoffs:

I

Investing R$ 1.00 in the public good has a private return of R$ 0.50..

I

..while it has a social return of R$ 2.50.

I

It is Pareto efficient for subjects to invest all of their money in the public good..

I

..but, since the private good’s return exceeds the return from the public good, the Nash equilibrium of this game is to invest zero in the public good (free ride).

(12)

E XPERIMENT

A Linear Public Goods Game

Two main hypotheses related to Random Rematching:

1. ‘Learning Hypothesis’: subjects may not immediately understand the incentives of the game, but after a few rounds, they start learning (free riding behavior increases).

2. ‘Strategies Hypothesis’: subjects believe all other subjects behave rationally in an incomplete information version of the Prisoner’s Dilemma.

Under the ‘Strategies Hypothesis’, we expect that giving by

‘partners’ will be greater than giving by ‘strangers’.

On the other hand, as the game approaches the end, we can

expect both ‘partners’ and ‘strangers’ to free ride.

(13)

E VIDENCE

Table 2 Descriptive Statistics

(14)

E VIDENCE

Table 3

Free Rider Index (FRI), 5-Round Experiments

(15)

E VIDENCE

Graph 1

Free Rider Index (FRI), 5-Round Experiments

(16)

E VIDENCE

Table 4

Percentage of Free Riders in Each Round

(17)

E VIDENCE

Graph 2

Percentage of Free Riders (Partners)

(18)

E VIDENCE

Graph 3

Percentage of Free Riders (Strangers)

(19)

E VIDENCE

Graph 4

Contributions to Public Goods Provision, 5-Round Experiments

(20)

E VIDENCE

Table 5 Econometric Estimations

(21)

C ONCLUSIONS

I

Main result: ’Strangers’ tended to cooperate more often than

’Partners’.

I

Other authors reported the same result (Andreoni and Croson 2008) .

I

Our results suggest that random rematching can play an important role in explaining the evolution of cooperation.

(22)

C ONCLUSIONS

Future research should concentrate on experiments with the following features:

I

Longer periods (10 or 20 rounds) (Andreoni 1988) .

I

Restarting dates (Andreoni and Croson 2008) .

I

No communication among subjects.

I

Complementarities between the laboratory and the field

(Fehr and Leibbrandt 2011) .

(23)

R EFERENCES

ANDREONI, J. Why free ride? Strategies and learning in public goods experiments. Journal of Public Economics, v.37, n.3, p. 291-304, 1988 .

ANDREONI, J.; CROSON, R. Partners versus strangers: random rematching in public goods experiments. In: PLOTT, C.R.; SMITH, V.L. (Eds.). Handbook of Experimental Economics Results, v.1. Amsterdam: North-Holland, 2008, p.776-783 .

FEHR, E.; LEIBBRANDT, A. A field study on cooperativeness and impatience in the Tragedy of the Commons. Journal of Public Economics, v.95, n.9-10, p.1144-1155, 2011 .