August 5, 2010
Mowat Centre features analysis by Arthur Sweetman and Herb Emery on the effectiveness of Canadian statistics and information gathering for decision-making bodies in Canada.
As the provincial premiers gather for the Council of the Federation meeting in Winnipeg, there is undoubtedly some discussion of the census long form controversy, even if it is not on the formal agenda. Both short and long-term issues are at play and there are important, but quite different, considerations in the two time frames.
In the very short run there is a need to find a compromise regarding the 2011 census before it goes to print in the very near future. Several sensible compromises have been advocated, but as far as we are aware no one has addressed the core disagreement. The government has taken the position that a voluntary long-form survey will produce information that is as reliable as a mandatory census, or perhaps more so since the government also proposes to increase the distribution of the long form from the mandatory 20 per cent, to a voluntary 30 per cent, of the population. The two most recent Chief Statisticians of Canada, and a host of others, have argued that this is a technical issue and that a voluntary survey is not reliable. In particular, many experts argue that voluntary surveys have proven to be ineffective for reaching certain groups, especially disadvantaged members of the population for whom data to inform program development and design are most critical. Interestingly, the Minister of Industry seems to understand some of the opposing issues. In his discussions with the CBC regarding why the Census of Agriculture remains mandatory (with jail penalties and fines) he said, “The argument obviously to farming associations and to farmers is, ‘You fill out the form, it’ll help the government help you in your farming activities.'”
Paradoxically, the sought-after compromise may also be the solution to this core disagreement. From its point of view the federal government could give a little, for 2011, and reduce rather than eliminate what it perceives to be a major intrusive burden on the Canadian population. This middle road would entail reducing the mandatory long form distribution not from 20 per cent to zero, but to 10 per cent, and at the same time continuing to distribute a voluntary version of the form to 20 per cent of the population. So the total distribution could remain as the government plans (or it could be lowered somewhat on the voluntary side to reduce costs), but it would be split into mandatory and voluntary versions. Responses to each could then be compared and an informed decision made for the 2016 census.
Continue reading
While an argument can be made that the very public political standoff will alter people’s behaviour regarding the 2011 census and make it difficult to predict outcomes for 2016, this incrementalist approach still provides useful information for decision-making at little or no cost.
MBreakout1ore broadly, this controversy also raises some long-term issues that are extremely important. We live in an information age and our economy depends on credible and reliable knowledge. The importance of agencies such as Statistics Canada is only increasing. As governments around the world discovered centuries ago regarding the Judiciary, and decades ago regarding Central Banks, it’s important for the sake of socioeconomic stability and growth to establish mechanisms by which these institutions are arms-length from political interference. It may be time to consider a new governance structure for Statistics Canada, making it a National Institution, as opposed to an agency of the Federal Ministry of Industry. Historically, Canadian politicians have had a more than admirable track record of providing Statistics Canada with the independence it needs to produce credible statistical information about which there is no taint of political interference. However, for example, recent issues around the mismeasurement (deliberate mis-statement?) of the Greek government’s financial accounts illustrate the importance of independent and credible social and economic information. Increasing the credibility of Canada’s statistical information will only help the nation going forward.
Additionally, given that we live in a federation with two levels of constitutionally recognized government, does it make sense for the nation’s statistical agency to be subject to only one? This misalignment is the reason why, for example, Canadian health data is managed partly by Statistics Canada, and partly by the Canadian Institute for Health Information (CIHI). The situation also implies that substantial portions of provincial government data services are subject to federal decision-making.
Altering an important organization’s governance, of course, needs to be done with much forethought, and not in the midst of a political standoff. It is an endeavour for which getting the details right matters. We recommend, therefore, that the First Ministers, as a group, establish a blue ribbon panel to examine the need for, and potential structure of, a new intergovernmental governance arrangement for Statistics Canada.
Two issues are of prime importance. First, it must be ensured that both domestic and international audiences in our knowledge-based economy and society believe that Statistics Canada continues to provide high-quality and unbiased data. Secondly, although some policy decisions are matters of principle and/or ideology where data does not matter, our society needs to ensure extremely good quality information for efficient and effective development and implementation of the many other decisions in our increasingly complex and information dependent world where good data does matter.
Author
Arthur Sweetman & Herb Emery
Release Date
August 5, 2010
