Philip Cross, National Post, Nov. 5, 2015
Reinstating the 2016 mandatory long form census, announced today, may be problematic. You cannot turn back the clock and simply assume the census will pick up where it left off in 2006. The census has become politicized, with defenders of a mandatory census aligned on the left and opponents on the right. More than one million Canadians did not participate in the last census in 2006. Based on comments from people saying they won’t fill in the long form, participating in the census risks becoming a statement of one’s political affiliation and not a civic obligation.
The reinstatement of the census is an occasion to reflect on the controversy surrounding its replacement by the voluntary National Household Survey in 2011 (the short form census remained mandatory). The whole episode reflects poor communications by everyone involved, which is usually the case for public policy train wrecks.
The Harper government never adequately explained why it replaced the mandatory long form census. Publicly, it cited privacy concerns, substantiated by little evidence. Cutting government spending, the motivation for the Mulroney government’s proposed cancellation of the 1986 census, was not a factor, since the NHS actually cost more (the 1986 census was reinstated, on the condition Statcan absorb all the costs). However, the memory of the outraged response among academic and policymaking elites to the 1986 cancellation likely played a large factor in the decision to make the 2011 census voluntary.
You cannot turn back the clock and simply assume the census will pick up where it left off in 2006.
Statistics Canada was not blameless. Despite its experience in 1986, it was unprepared for the 2011 fiasco. Unlike all other subject matter areas, Statcan had no external advisory committee for the census, denying it an ear to the ground to detect rumblings about the intrusiveness or necessity of a census just as the proliferation of the Internet and social media was raising public sensitivity and awareness of such issues.
Infighting at Statcan complicated its position on the census. The call of former Chief Statistician Ivan Fellegi for the incumbent Chief Statistician Munir Sheikh to resign in protest over the census was an unseemly meddling in internal politics, something Sheikh pointedly promised never to repeat with his successor. The only comparable example of such behaviour in the federal civil service is the lame sniping by Kevin Page, the former Parliamentary Budget Officer, aimed at his successor, Jean-Denis Fréchette, which has only damaged Page’s own standing.
Nor was Statcan’s stance on the NHS entirely coherent. Earlier this year it challenged critics of the NHS to provide concrete examples of problems in the NHS, professing confidence in the data. But now Statcan rushes to reinstate the census, implying it had unarticulated concerns about the NHS data. This mirrors the government’s inability to state clearly its objections to the census.
Outside of government, many critics of the NHS were more interested in scoring political points than expressing real concerns about data quality. This was particularly true of former Bank of Canada Governor Mark Carney, who voiced concern about the impact of changing the long form census on the major economic data when in fact there was no connection. That Carney would express an uninformed opinion suggests he was interested in reinforcing the anti-government narrative about the census rather than informing the public debate, incredibly poor judgement for the head of a prominent agency that needs to be impartial to preserve its independence.
It was not just the Harper government that politicized the census. Many of the opponents of the voluntary NHS made shrill arguments based purely on ideology and the desire to oppose the Harper government, not an articulate defense of the need for good data. The census became Exhibit A in over-the-top arguments the government was waging a ‘war on science.’ Having responded so vociferously, partisans of the census cannot be surprised opponents also will use it as a political instrument. The CD Howe Institute, where I am a research Fellow, was one of the few participants that tried in 2010 to lower the temperature of the census debate by brokering a compromise between the government and Statcan. Egged on by the media storm surrounding the census, everyone else escalated the rhetoric, which only damaged the universal support a census must have.
In such a charged environment, it is naïve to think there will not be an impact on response rates and that non-participation will not skew the sample, the same criticism levelled at the NHS. As I mentioned, over one million Canadians did not participate in the last mandatory long form census in 2006, when the response rate was 94%. If the response rate drops to something below 90%, this would leave Statcan in the position of having to fine millions of people (although fines up to $500 a person would easily pay for the census, giving cost recovery at Statcan a whole new dimension). Would it be willing to do that, risking its good relationship with Canadians? Doubtful, but it is hard to see how else Statcan could maintain the nearly universal response rate which is integral to the integrity of census data. Clearly, Statcan faces some interesting decisions.
In such a charged environment, it is naïve to think there will not be an impact on response rates and that non-participation will not skew the sample
More broadly, people who think that rolling back the many policy changes made by the Harper government will be as simple as hitting the reset button are in for a surprise. The census is a good example; simply reinstating the mandatory long form ignores how this country has changed over the past decade. Some things are more political and polarized, including the census, due to both the government’s actions and the vehemence of its opponents. Not that the Harper government invented partisan politics; one of the worst examples was the night of the Quebec referendum 20 years ago, when Prime Minister Chretien pre-empted Conservative leader Jean Charest’s own speech to the victorious No forces.
Philip Cross is the former Chief Economic Analyst at Statistics Canada