Correlation versus causation and the apparent external benefits of education

There is a widely reported positive correlation between education and the amount of volunteer time given. Many researchers assume this is a causal relationship, so increased volunteer work has been added to the list of 'external benefits' of education. These 'external benefits' make public subsidies to higher education seem more justifiable. This study uses data from a specially collected sample of New Zealand identical twins to test the relationship between education and volunteer labour supply, holding unobservable family effects constant. Multiple measurements of education levels were also collected so that any bias in the results due to measurement errors could be corrected. The results show that once family unobservables are controlled for, extra education significantly reduces the amount of volunteer time donated. This reversal in the results is a reminder of the lesson that correlation does not imply causation.