I wrote this posting for my other blog, but I figured that I’d post it here too, given that (a) most of my readers probably don’t read my other blog and (b) I just had a little debate on the issue on my Facebook wall, which prompted me to share this info a little more widely (also, I’m adding in a few more things than my original posting):
A friend of mine just sent me a link to this news story and it’s gotten me quite livid:
Tories scrap mandatory long-form census
StatsCan says quality of data will suffer
Every five years, Canada conducts a census, with the next one scheduled for 2011. In the past, every household received the short census form, which contains just a few questions (like number of people in the home and their age and sex) and 1 in every 5 households received a mandatory long form. The long form contained questions about a variety of things, like income, education, and ethnicity, and provided a lot of really important information about the population of Canada. For example, I work in health care and we use census data all the time. Knowing the makeup of our population allows us to make informed decisions about providing health care to meet the needs of the people living in our region. All levels of governments (municipal, provincial, and federal), community agencies, and other organizations use the data from the census long form to develop evidence-informed policy.
But the ability to do that now at risk, as the federal government has, apparently without consulting anyone, decided to scrap the mandatory long form, citing that the long form represented “what most Canadians felt was an intrusion into their personal privacy in terms of answering the longer form” (Source) – though I haven’t seen from anything I’ve read thus far how they determined that this is how “most” Canadians feel. And given that Canada’s Privacy Commissioner has had only 3 privacy complaints in the last 10 years, “none of which have been upheld” (Source and Source), I’m doubting that “most” Canadians are up in arms about the intrusiveness of the census. Instead of the mandatory long form, they are replacing it with a “voluntary household survey” that will be sent to 1 in 3 households, the members of which can choose to complete – or not to complete – the survey. This raises very serious concerns about the quality of the data – the people who choose not to respond to the long survey may be different from those who choose to respond to it, which will result in skewed information. Which means we won’t have the data we need to make policies and provide appropriate services.
“Senior statisticians at Statistics Canada have conceded the change will make it more difficult to obtain reliable, detailed information.” (Source)
All of this brings up questions about politicians’ understanding of the importance of data and evidence-informed practice, not to mention their ignoring the scientific experts on the matter – in this case, the statisticians at Stats Canada (and any other statistician who has been asked). As David Eaves noted in his article, “Why you should care about the sudden demise of the mandatory long census form”:
“This is a direct attack on the ability of government to make smart decisions. It is an attack on evidence-based public policy. Moreover, it was a political decision – it came from the minister’s office and does not appear to reflect what Statistics Canada either wants or recommends. Of course, some governments prefer not to have information; all that data and evidence gets in the way of legislation and policies that are ineffective, costly and that reward vested interests (I’m looking at you, tough-on-crime agenda). ” [emphasis mine; Source]
In another news article on the topic I read:
Liberal MP Marlene Jennings “argued that Clement has shown in postings to the social media site Twitter that he does not understand how the mandatory nature of the long-form census allows Statistics Canada to properly weight the short form data. Clement debated sample size and data weighting with other posters, including an economist.
“(That’s) something Mr. Clement seemed not to understand when he was tweeting yesterday, so maybe he should take a stats course,” she said” (Source)
This lead me to check out Clement’s Twitter stream, where he referred to having a mandatory (as opposed to voluntary) long form as “state coercion” – I guess it’s fine to have the state “coerce” you to complete the short form – or, you know, obey any of our other laws – but they draw the line at the long form? And as for his understanding of statistics – well, he said that statisticians can fix the problems resulting from the bias that will result from a voluntary (instead of mandatory) form with “large sample sizes.” Now, I, my Official Statistician, and anyone who has passed a first year statistics course can assure you that this is not true. There is no way to create data from any groups that don’t respond to a voluntary census.
As I mentioned, I was engaged in a debate about this on my Facebook wall and my sister’s contribution to that was so awesome, I just have to share it with you. Essentially, the debate I was having involved the other person making claims like the mandatory long form of the census is “undemocratic” because it invades our privacy and that the census is “inaccurate” (then suggesting that market research, which would not be appropriate for the purpose of the census, would be better), and, of course, that the long census being mandatory is “state coercion” (i.e., Tony Clement’s words). My sister’s comment:
Does anyone else find it funny to read a conversation about protecting privacy and rights on Facebook? Why does everyone want to envision the government like something out of a science fiction novel. We are the government. We have health-care, public/secondary and post-secondary education at a fraction of its real cost. If my personal data on who does more house work in my home or if my child can read can help society as a whole, of course they can have it and should have it. Do we really feel safer with the illusion that we are protecting ourselves from exploitation because it is not mandatory? Is this before or after we have posted up our kids latest baseball game home movie on YouTube?
She has it exactly right. Statisticians, academics, social scientists and like need access to the robust data provided by a mandatory census in order to provide services to Canadians. The Conservative government and their supporters are making it sound like we are sitting around planning all sorts of nefarious things to do with census data – completely ignoring the fact that there are lots of safeguards on the privacy of census data.
Stuff you can do about this:
- The Liberals are demanding the Conservatives reverse this decision to scrap the mandatory long form and, if they don’t, to introduce legislation to protect it themselves. Personally, I’ll be writing to my own MP, Industry Minister Tony Clement, and the Prime Minister to inform them about why the long form is so vital. And I’ll be writing to some Liberals to suggest that they stick to their guns on this one and introduce legislation to protect the mandatory long form. Don’t know who your MP is? You can search here to find out!
- Sign the petition: http://www.gopetition.com/petitions/keep-the-canadian-census-long-form.html
- Spread the word!
Remember, even if you aren’t going to use the census data directly, this is an issue that affects you and the services you receive!