Impaired decision making: Why good census data matters for decision makers


Two years ago, the Canadian federal government made a unilateral decision to scrap the mandatory long-form questionnaire of the Canadian census, in favor of a voluntary survey sent to a sample of 4.5 million households.  This decision was met with considerable backlash from the academic, government, and economic community, including the resignation of the Chief Statistician of Statistics Canada (the federal agency that administers the census) in protest.

In the census year 2011, the long-form was replaced with the National Household Survey, which was completed by individuals on a voluntary basis.  The voluntary aspect of this survey was the driving force behind the flurry of protests in Canada, with those opposed to scrapping the long-form questionnaire citing low response rates and non-response bias associated with voluntary surveys.  In contrast, those who supported the change have cited privacy concerns and intrusion of individual liberties.

The fires have yet to cool from this heated debate.  This year marks the first release of data from the revised 2011 census, with topics on families, dwellings, marital status, and languages to be released in the coming months.  But what will this data tell us and why has the change from a mandatory to voluntary survey caused such an uproar across the country?

U.S. Census Bureau. Washington, DC: 2012.

Historically, the Canadian census has contained two components: a short-form, completed by 80% of households; and a long-form, completed by the remaining 20%.  The short-form contains 8 questions on demographics, and is used by government agencies to allocate resources across the country.  The long-form contains 53 additional questions on education, ethnicity, mobility, income, employment, and dwelling characteristics; and is used by an array of decision makers, market research groups, and private corporations to plan community services, allocate health resources, and decide where to locate new shopping centres.

Although the detailed questions on sociodemographic data will be replicated in the NHS, the switch to a voluntary survey may compromise the underlying reliability of the data and negate any comparisons with historical census data.  Survey researchers, for example, have long battled with low response rates and biases due to differential response among certain segments of the population (e.g., lower income groups, or ethnic minorities).

As with the majority of policy decisions, these tend to have an impact on the most vulnerable populations in society.  A report from the House of Commons, for example, called for the reinstatement of the mandatory long-form, citing substantial impact on gender equality and women’s rights.

You can see why academics, private corporations, and community groups are fighting hard to preserve every ounce of reliable data.  Data users from across the health, economic, and development spectrum have voiced concerns over the utility of the new census data, with fears that the biased data will negatively impact policy and decision making.

Privacy concerns are valid, especially in middle- and high-income countries with electronic records and detailed administrative databases.  However, census data often pales in comparison to the data collected by retail corporations or social networking sites.  Moreover, the data is anonymized and confidential, meaning that users of the census data cannot identify individual survey respondents.  And perhaps most importantly, one cannot ignore the public good that arises from the use of the data, such as equitable resource allocation and health research.

Of course, there are other sources of data that can be used for health and development research.  In Scandinavian countries, for example, a system of government administrative databases are used to track detailed data on individuals; whereas in North America, population-based data repositories (here and here) have been used extensively to conduct health services research and evaluation.

Nevertheless, more detailed data is often necessary for health and development studies, above that which can be provided by administrative data.  Any researcher worth their salt would be well aware of the social and behavioral determinants of health that exert their influence across the life course.  Take just one look at the census survey contents and you’ll see how these can fit into multiple levels of the ecological framework for injury and illness prevention.  This is why reliable, population-based data is necessary for decision makers, and why our society needs to fight tooth and nail to preserve our decision making capacity — so that we are not making decisions in the dark.  The call for reinstatement of the mandatory long-form census remains loud and clear.


Tags: , , , , , , , , , , , , , , , ,

One Response to “Impaired decision making: Why good census data matters for decision makers”

  1. chenjo Says:

    While I agree that census surveys are important for the many reasons states above, one of the things I personally like doing the least is filling out surveys, especially if they are long. I for one would consider filling a the short 8 question survey but would likely not fill out the 53 question one so I can understand the reasons why the long-form census was discontinued. There are so many databases to gather population information from and with the technology available out there today, perhaps even different “apps” for smartphones could do some of the data gathering!

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: