I’ve received a promotion and been busy with the Coast Guard Auxiliary and graduate school in addition to work, so I apologize for my tardiness on getting a post up.  That said, I’d like to treat you to a quantitative analysis I recently conducted on the relationship between a country’s wealth and its culture.

The purpose of the analysis was to attempt a identify what characteristics of a culture contribute to a nation’s wealth.  I first looked at Hofstede’s dimensions and the data set on his website, however, I deemed it not usable since it lacked completeness.  Many countries did not have data across all five dimensions.  I then chose the GLOBE Project data set which is similar to Hofstede’s dimensions.  I downloaded the 2004 GLOBE Phase 2 data set for Society Cultural Scales.

I combined the GLOBE Project data (the Practice dimensions only, not the Values dimensions) with Polity IV Polity Index for the 57 countries in the data set.  I chose Polity because it was a good single value proxy for the variable I am really after, which whether the country has a government that functions on Weber’s rational-legal authority type.  A key control for whether the state is wealthy because it has good government, or whether the state is wealthy because it has a culture that makes it wealthy.  Thus I built my set of 10 independent variables (Polity plus nine dimensions from the GLOBE Project).

For the response variable I created a dichotomous variable from the World Bank’s estimates for states’ income.  The states classified as “high income” were coded 1 and the states with lower incomes were coded 0.

I then ran a logit regression in R.  results of which are below:


> summary(globe_glm_1)

>

The model reduces the deviance more than the NULL model. An analysis of variance of the model terms also shows the important coefficients.

> summary(globe_anova_1)
Df Deviance Resid. Df Resid. Dev
Min. :1 Min. : 0.1141 Min. :47.0 Min. :35.19
1st Qu.:1 1st Qu.: 0.5720 1st Qu.:49.5 1st Qu.:41.63
Median :1 Median : 2.7185 Median :52.0 Median :64.87
Mean :1 Mean : 4.5143 Mean :52.0 Mean :56.83
3rd Qu.:1 3rd Qu.: 7.5384 3rd Qu.:54.5 3rd Qu.:67.39
Max. :1 Max. :15.0461 Max. :57.0 Max. :80.34
NA's :1 NA's :1
Pr(>Chi)
Min. :0.0001049
1st Qu.:0.0122373
Median :0.1073930
Mean :0.2366209
3rd Qu.:0.4799755
Max. :0.7354827
NA's :1
>


The results indicate that the Humane Orientation Societal Practices and the Collectivism II Societal Practices (In-Group Collectivism) with both significant at the 0.05 level and negatively related to the odds of being wealthy society. What this indicates is that societies that are individualistic and competitive are more likely to be wealthy.

Furthermore, the Collectivism I Societal Practices (Institutional Collectivism) is significant at the 0.1 level and positively associated with the odds of being a wealthy society. What this means is that solidarity with employers is also associated with wealth–i.e., cultures like Germany, Japan and South Korea.

Interestingly, the Polity Index did not contribute more to the model. An investigation of outliers revealed that both Kuwait and Qatar, Gulf oil kingdoms, have negative polity scores–i.e., are authoritarian. All others are Polity 10’s, which is the highest score on the strength of democratic institutions and lack of authoritarian tendencies. Thus, there is an identity in this data set between wealth and liberal governance. Oil kingdoms appear to be a class all their own.