Globalisation and Propensity To Co-operate: Results From Multi-country Field Experiments

Globalisation and Propensity To Co-operate:

Results From Multi-country Field

Experiments

Gianluca GrimaldaCSGR – Warwick University

Joint with Nancy Buchan (South Carolina), Marilynn Brewer (Ohio), Enrique Fatas (Valencia), Margaret

Foddy (Carleton), Rick Wilson (Rice)

Motivation: Globalisation as a New Form of Connectivity

Previous research shows the importance of the structure and scale of individual connectivity on individual behaviour:

Henrich et al. (2004) show that the level of a society’s market integration in small-scale societies is positively related with pro-social experimental behaviour.

The social capital literature emphasises the importance of individual engagement in associations for the development of propensity to co-operate and civic-minded attitudes (Putnam, 1993).

This research looks at individual globalisation as a new form of connectivity, and at its influence on co-operative behaviour.

What do we Mean by Globalisation?

Globalisation may be conceptualised as the ‘de-territorialisation of human relations’ (Scholte, 2005; Robertson, 1992; Giddens, 1991). It leads to the creation of a network of relations among people on a global scale.

In accordance to the literature, such relations encompass the economic, social, cultural, and political sphere.

Individual globalisation is measured in this research as the extent to which an individual has access to this network and participates in it, that is, the extent of her ‘connectedness’ with other people on a (potentially) global scale.

Country-level globalisation is measured by the CSGR index.

Research strategy

6 countries spanning the globalisation spectrum: United States (Columbus, Ohio), Italy (Milan), Russia (Tatarastan), Argentina (Buenos Aires), South Africa (Johannesburg), Iran (Tehran)

Standard methodological controls for comparative research have been adopted (Roth, et al., 1991; Buchan, and Croson, 2006)

Sample of around 200 people in each country stratified according to age, gender, and socio-economic status, equivalent across countries

Strategy: (a) To Measure the Individual Propensity to Co-operate at the local, national, and international level through 3 experimental decisions; (b) To measure the individual degree of globalisation through responses to a questionnaire

Experimental Protocol Decision 1: Public Goods Game (PGG) at the LOCAL

level Decision 2: Nested PGG at the LOCAL/NATIONAL level Decision 3: Nested PGG at the LOCAL/GLOBAL level All decisions were independent and payoffs were not

communicated until the end of the three decisions People were randomly matched with people from

their own local area, other areas in their country, and other countries

An individual index of globalisation has been derived from the scores to around 70 items included in a questionnaire that people completed at the end of the experimental decisions. Items measure the frequency and the scale of connectedness of an individual across the social, cultural, economic, and political dimensions.

An Example from the Questionnaire

1. How often do you normally use the following products or services? Check one

option.

I own/have access to this product or service, and I use it:

I do not own/have access to this product or

service. Every

day Every week

Less often

Never

a. Landline phone 1 2 3 4 5

b. Mobile phone 1 2 3 4 5

c. Internet 1 2 3 4 5

d. Email 1 2 3 4 5

e. Credit card 1 2 3 4 5

f. Postal mail 1 2 3 4 5

g. A TV not connected to a satellite/digital/cable network

1 2 3 4 5

h. A TV connected to a satellite/digital/cable network

1 2 3 4 5

i. Radio 1 2 3 4 5

j. Fax machine 1 2 3 4 5

Overview of the Results: Mean Co-operation Rates per Decision

0

1

2

3

4

5

6

7

8

9

US Italy Russia Argentina South Africa Iran

Co

-op

erat

ion

rat

es

Local 1

C2

C3

Significant differences across countries

More globalised countries seem to experience a higher level of mean co-operation

Cooperation in Decision 3 is significantly lower than in Decision 2. This is the case (a) pooling all observations together (p-value= 0.0082 ); In Argentina (p-value=0.0144), the US (p-value= 0.0657), and South Africa (p-value=0.0220).

Regressions on Total Co-operation Rates in Decision 3

Strong Predictive Power of Country-Level and Individual Level Globalisation on Total Co-operation in Decision 3

Positive Relation

Dependent variable: Co-operation in Decision 3 ( 1 ) ( 2 ) ( 3 ) ( 4 ) Income 0.0414 0.0690*** 0.0148 0.0357 [0.031] [0.020] [0.034] [0.024] Education 0.105** 0.0361 0.0353 -0.026 [0.044] [0.033] [0.044] [0.038] Gender 0.0164 0.054 0.0253 0.0563 [0.094] [0.11] [0.097] [0.11] Year of Birth -0.00243 -0.00224 -0.00451 -0.00478 [0.0083] [0.0087] [0.0078] [0.0080] Country-level Globalisation Index 2.148*** 1.896*** [0.43] [0.37] Individual Globalisation Index 2.303*** 2.555*** [0.51] [0.52]

Observations 1029 1029 1029 1029 Pseudo R2 0.0216 0.018 0.0242 0.0216

Ordered logit model with robust standard errors clustered for country Country Dummies included in regressions (1) and (3) Robust standard errors in brackets *** p<0.01, ** p<0.05, * p<0.1

Regressions on Total Co-operation Rates in Decisions 2 and 1

Again, strong positive relation between globalisation (macro and micro) on co-operation rates

Dependent variable: Dependent variable: Co-operation in Decision 2 Co-operation in Decision 1 ( 1 ) ( 2 ) ( 3 ) ( 4 ) ( 5 ) ( 6 ) Income 0.0774*** -0.00088 0.0420** 0.0506*** -0.00472 0.0226 [0.019] [0.027] [0.019] [0.019] [0.032] [0.021] Education 0.0126 0.0742** -0.0499 0.0288 0.0486 -0.0264 [0.033] [0.030] [0.039] [0.067] [0.051] [0.071] Gender 0.0867 0.0289 0.0792 -0.133 -0.169 -0.134 [0.11] [0.12] [0.11] [0.12] [0.11] [0.12] Year of Birth -0.00084 -0.0029 -0.00363 -0.00019 -0.00218 -0.00253 [0.0070] [0.0058] [0.0059] [0.0062] [0.0056] [0.0057] Country-level Globalisation Index 2.355*** 2.137*** 1.921*** 1.711*** [0.64] [0.55] [0.54] [0.48] Individual Globalisation Index 1.947*** 2.562*** 1.931*** 2.199*** [0.45] [0.47] [0.23] [0.10]

Observations 1030 1030 1030 1030 1030 1030 Pseudo R2 0.0215 0.0327 0.0252 0.0142 0.0223 0.0169

Ordered logit model with robust standard errors clustered for country Country Dummies included in regressions (2) and (5) Robust standard errors in brackets: *** p<0.01, ** p<0.05, * p<0.1

Conclusions and Further Analysis

Significant Macro and Micro effects of globalisation, going hand-in-hand

Mediating effect of social identity? What are the relations with social capital and

generalised trust?

Appendix: The nested PGG

Local Group 2 from Country B

Local group 1 from Country A

I

Local group 3 from Country C

World Group .

Appendix: Relation Between Globalisation Index and Mean Propensity to Co-operate

US

Italy

RussiaArgentina

South Africa

Iran

56

78

Con

trib

utio

n to

Loc

al 1

.2 .4 .6 .8 1CSGR globalisation scale

Polynomial fitted line

Decision 1Relation between CSGR scale and mean co-operation rates

US

ItalyRussiaArgentina

South Africa

Iran

56

78

Sum

of

Con

trib

utio

n to

Loc

al 2

and

Nat

iona

l 2

.2 .4 .6 .8 1CSGR globalisation scale

Polynomial fitted line

Decision 2 - Total Co-operationRelation between CSGR scale and Mean Co-operation rates

US

ItalyRussia

Argentina

South Africa

Iran5.5

66.

57

7.5

8

Sum

of

Con

trib

utio

n to

Loc

al 3

and

Wor

ld 3

.2 .4 .6 .8 1CSGR globalisation scale

Polynomial fitted line

Decision 3 - Total Co-operationRelation between CSGR scale and Mean Co-operation rates

US

Italy

RussiaArgentina

South Africa

Iran

56

78

10000 20000 30000 40000GDP per capita, PPP (current international $)

Polynomial Fitted Line

Appendix: Co-Operation Rates with the Higher-Order Group per Decision

0

0.5

1

1.5

2

2.5

3

3.5

4

US Italy Russia Argentina South Africa Iran

N2-L2

W3-L3

Appendix: Relation Between Globalisation Index and Propensity to Co-operate with the Higher-Level Group

US

Italy

Russia

ArgentinaSouth Africa

Iran

01

23

4

.2 .4 .6 .8 1Glob Index

Polynomial Fitted Line

US

Italy

Russia

Argentina

South Africa

Iran

01

23

4

.2 .4 .6 .8 1Glob Index

Polynomial Fitted Line