How do you change Australians’ views about aid? Earlier this year I wrote a blog post presenting my first cut analysis of three survey experiments in which I’d tested what information, if any, would shift Australians’ views about whether their country gave too much aid or not. A majority of Australians supported the 2015/16 aid cuts, and given there’s some evidence to suggest public opinion has an effect on aid policy, there were obvious reasons to want to see whether support could be increased.
I ran three separate experiments with three different groups of participants. In the first experiment, the treatment group was ‘treated’ by being told just how little aid Australia gives. In the second experiment, the treatment group was treated by being shown that Australian aid as a share of gross national income (GNI) was decreasing. In the third experiment, the treatment group was given information contrasting Australian aid cuts to aid increases in the United Kingdom.
As I wrote in the original blog post, I thought the first and second treatments were the most likely to be effective.
I was wrong. The only treatment to have a substantively meaningful (and clearly statistically significant) effect was treatment three: the comparison with the United Kingdom.
I’ve now published a Development Policy Centre Discussion Paper in which I: fully detail the experiments; conduct additional (more robust) tests of the treatments using logistic regressions; and link the findings to various literatures from psychology. The findings remain the same. Although it turns out that while the findings were surprising at first glance, they do fit well with what we know about human psychology. (Australians aren’t strange. People are.) If you want the details, read the paper. (If you just want the basics, the original blog post will still do.)