Saturday, July 24

Census brouhaha

Since everyone else is talking about the census, I figured I would weigh in as well.

I consider myself to be somewhat libertarian, so I appreciate the argument that the census, at least to some extent, is an invasion of privacy. One has to admit that some of the questions might make some people a little uncomfortable to answer (even if they only have to answer it, on average, once every 25 years).

On the other side of the coin, however, replacing the long-form census with a voluntary survey generates data that is of much lower quality. The people who choose not to respond may tend to have similar characteristics, and if the survey no longer picks up their data, the results will be skewed. Even after the voluntary long-form survey is conducted, we won't be able to know how bad the problem is.

At the end of the day, however, life will go on. Researchers will probably continue to use the information from the voluntary survey, since they won't have much alternative. The data quality will be worse, meaning the conclusions researchers draw using the new data will be less convincing. But the data will still be usable to some extent.

What we will probably see, however, is that fewer researchers will choose to use Canadian data. I remember one of my professors telling our class how the norm in economics is to use American data. U.S. data is preferred because of the country's leadership position in the academic world, but also because Americans generally collect more data and make it cheaper to access. My prof explained that she usually has to have a compelling argument as to why she's using Canadian data instead of comparable U.S. data in order to be taken seriously by top economic journals.

If the long-form census becomes voluntary, the quality of Canadian data will go down. That means Canadian researchers, many of whom already choose to use foreign data for their research, will look outside of Canada more frequently for data. That's unfortunate, because conclusions about how things work in other countries may not necessarily apply to Canada. All things being equal, it would be better to have professors use Canadian data for research, since their findings could help us identify and fix problems in our own backyard.

Having a mandatory census does intrude on Canadians' privacy. Whether the intrusion is great enough to warrant moving to a voluntary long-form survey is a political question, and it will be fascinating to see how the controversy plays out in the coming days.