Sure.
Our study was based on behaviours rather than on the survey question directly asking individuals whether a tax credit was important to them. I agree that statistically it's a difficult thing to test.
The Canadian survey of giving, participating, and volunteering asks a specific question about whether tax credits are important to you and if you plan to claim for a tax credit. That wasn't the measure we used in our study. We used a regression analysis, where the price of donation was a variable. So we included a variable that represented the tax credit, and that was shown to be statistically significant.
Granted, it's a difficult variable to measure, and I'm not sure what the statistical expertise of the committee is, but we have to deal with endogeneity issues. You probably work and consult with Statistics Canada people as well, but it's a difficult variable to measure, and we use a proxy variable to measure it.
There is a little bit of uncertainty, but there is whenever you're dealing with survey data. The specific study we have done has been recently accepted for publication in a peer-reviewed economics journal, Applied Economics Letters, in the U.S., so it has gone through a peer-reviewed process as far as statistical technique.
In addition to that, in one of our papers we compared our results with some other results done in different areas. For example, a couple of studies done in the U.S. have also found.... And I understand their tax credit system is different. Instead of using a tax credit, they use a tax deduction. A study by Brooks in the U.S. from 2007 also found the tax incentive to be effective. The two former Canadian studies we found that go back a little ways—Kitchen and Dalton in 1990, and Kitchen in 1992—also found tax incentives to be a statistically significant variable.
Although there are some complications and difficulties in measuring such a variable, our results did appear to be in line with some other studies that were quite similar.