FEWO Committee Report

Introduction

The Standing Committee on the Status of Women passed the following motion at its meeting on October 5, 2010, calling for a study on the Census:

That the Committee undertake a study on the cancellation of the mandatory long form census and its impact on women’s equality in Canada; as well as the removal of questions on unpaid work—formerly listed under Question #33 in the 2006 mandatory long-form census—from the new voluntary National Household Survey; and

That the study takes into account the United Nations Beijing Declaration and Platform for Action to which Canada committed.

The Committee held three meetings on this subject in November 2010 with witnesses.

Background information

About the Census

The first Canadian Census was held in 1871, and asked 211 questions, although not all households answered all questions.[1] The Census was taken every 10 years until 1956, and has been taken every five years since. In 1941, detailed questions about housing were sent to 10% of households; this sample was increased, 10 years later, to one in five. Until 1971, all Censuses were taken by interview; since that time, respondents have been able to self-enumerate and complete and return their own Census forms.

Also in 1971, two questionnaires were used:

• The "short" form distributed to two-thirds of Canadian households covered the basic population questions and nine housing questions.
• The "long" form, distributed to the remaining third, contained the same questions as the short form with the addition of 20 housing questions and 30 socio-economic population questions.

In 2001, 80% of households received the “short questionnaire,” and 20% the long questionnaire, which included 52 additional questions.

The Census is mandatory because the Statistics Act requires persons to complete the Census, with penalties for refusing or neglecting to do so:

31. Every person who, without lawful excuse,

(a) refuses or neglects to answer, or wilfully answers falsely, any question requisite for obtaining any information sought in respect of the objects of this Act or pertinent thereto that has been asked of him by any person employed or deemed to be employed under this Act, or

(b) refuses or neglects to furnish any information or to fill in to the best of his knowledge and belief any schedule or form that the person has been required to fill in, and to return the same when and as required of him pursuant to this Act, or knowingly gives false or misleading information or practises any other deception thereunder is, for every refusal or neglect, or false answer or deception, guilty of an offence and liable on summary conviction to a fine not exceeding five hundred dollars or to imprisonment for a term not exceeding three months or to both. 1970-71-72, c. 15, s. 29.[2]

The same piece of legislation allows for the creation of voluntary surveys, other than the Census:

8. The Minister may, by order, authorize the obtaining, for a particular purpose, of information, other than information for a census of population or agriculture, on a voluntary basis, but where such information is requested section 31 does not apply in respect of a refusal or neglect to furnish the information. 1980-81-82-83, c. 47, s. 41.[3]

About unpaid activities

In an earlier report on economic security of women, the Committee identified “the economic costs incurred by women who decrease their labour-force attachment to take on a greater role in unpaid care giving.”[4] The Committee reported that witnesses had told them that:

[W]omen continue to assume a disproportionate share of non-standard work (part-time, part-year); [and] that their unpaid work limits their ability to maintain a stable attachment to the labour force during the times they provide care to children and other family members.[5]

In her written submission to the Committee, Kathleen Lahey provided a brief history of the collection of data on unpaid activities:

Beginning with the 1970 report of the Royal Commission on the Status of Women, developing and collecting high quality, continuous, and comprehensive sex-disaggregated data on all aspects of women’s lives—including data on women’s ‘invisible’ and unpaid activities—has been increasingly recognized as one of women’s fundamental human rights. The basic equality principles and the detailed evidence-based analysis used throughout the lengthy report to document the extent of women’s inequality at that time led to the clear conclusion that no society could expect to eradicate discrimination without high-quality statistical data on all aspects of women’s lives—especially on their paid and unpaid work lives.

The Royal Commission report has had tremendous impact. In the 1970s, courts and legislatures began to recognize women’s unpaid work in family property and business law, and Statistics Canada began developing comprehensive data on unpaid activities beginning in 1971.[6]

The question with respect to unpaid activities was added to the long form in 1996, and was asked in the 1996, 2001 and 2006 Censuses of the Population. The question in the 2006 Census was:

Last week, how many hours did this person spend doing the following activities:

• doing unpaid housework, yard work or home maintenance for members of this household, or others?
• looking after one or more of this person’s own children, or the children of others, without pay?
• providing unpaid care or assistance to one or more seniors?[7]

Statistics Canada collects other data on unpaid activities in voluntary surveys, including the General Social Survey (GSS). The time-use segment of the GSS is asked every five years, most recently in 2010. The sample size was more than doubled that year, from 10,000 to more than 20,000. The GSS is conducted by telephone and respondents are asked to keep a time diary for one week before responding to questions. Questions related to unpaid activities from 2010 are below:

Last week, how many hours did you spend looking after:

• one or more of the children living in your household, without pay?
• one or more children living outside your household, without pay?

Last week, how many hours did you spend doing:

• unpaid housework, yard work or home maintenance for your household?
• unpaid housework, yard work or home maintenance for persons who live outside your household?

Last week, how many hours did you spend providing:

• unpaid care or assistance to one or more seniors living in your household?
• unpaid care or assistance to one or more seniors living outside your household?[8]

The Committee notes two changes to the 2011 Census:

• the mandatory Census long-form questionnaire, distributed to 20% of the population, was to be replaced by a voluntary National Household Survey, distributed to 33% of the population;
• the question related to unpaid household activities was to be eliminated.

What the Committee heard

The Committee heard from witnesses and in written submissions about several issues related to the cancellation of the mandatory long-form Census and the removal of the unpaid activities question from both the Census and the National Household Survey.

Cancellation of the Mandatory Long-Form Census

The decision to cancel the mandatory long-form Census was made public on June 26, 2010 in the Canada Gazette,[9] where the date and questions for the 2011 Census were announced. As the Committee heard, the Government has the responsibility under the Statistics Act to determine the questions to be asked in the Census: “The Governor in Council shall, by order, prescribe the questions to be asked in any census taken by Statistics Canada...”[10]

However, a former Chief Statistician told the Committee that the means by which answers to those questions were collected was not appropriately a political decision:

When I spoke out about the voluntary long-form census, it was not about whether or not the government has the right to decide what should get in the census, but whether it is inappropriate for any government to tell Statistics Canada how to do a survey, because that is a technical issue, and there is a long tradition, including United Nations resolutions, about the technical side of statistics having to be free of government interference. That is the only issue on which I spoke out. I am in total agreement with the tradition and the law that governments determine the content of the census.[11]

Non-response rates and bias

There has been considerable public discussion and debate about whether the new survey would provide data that are sufficiently reliable for a variety of purposes, including for use by governments at all levels to assess the impact policies and programs may have had over time. Witnesses commented on this in their presentations and in response to questions from Committee members.

A Statistics Canada official told the Committee that “we still don't know what the repercussions will be on the quality. We already know from a few surveys we have done that some sub-populations are likely to have fewer respondents,” and that “we are putting more procedures in place at a number of levels to properly control the consequences of an under-response from some groups.”[12]Officials from Status of Women Canada said:

As data users, not producers, and at this early stage in the decision, we are not in a position to comment on the impact of eliminating the long-form census, and would defer to our colleagues at Statistics Canada, and other data experts, on this matter.[13]

Ivan Fellegi, a former Chief Statistician of Canada, said that a lower response rate to a voluntary survey:

...would not matter much if the lost responses were evenly distributed over all population groups, but we know this is not the case. Past experience from Canada and elsewhere shows that underprivileged groups, such as aboriginal people, new immigrants, visible minorities, and, generally, people with low incomes, will respond at a disproportionately low rate—and no extra sampling will compensate for this disproportion.[14]

While many witnesses made similar points, one submission to the Committee made the case that a higher response rate to compulsory surveys may not always be more accurate:

While it is no doubt true that compulsory surveys have a higher response rate, it is also likely that, where the subject matter is sensitive, they will generate a higher level of false responses. It seems probable that when people are forced to divulge information they would prefer to keep private, some will respond with evasions.[15]

The Committee also heard from a witness that corrections could be made for such responses:

Certain errors always come into a census, for many reasons. A small number of people may [...] report being Jedi, but the feeling of the census experts is that those numbers are relatively small. Overall, yes, there is theoretically bias, but it really doesn't affect, for example, the trends in the information we get.[16]

Usability of data

An additional theme addressed by many witnesses was whether the shift from a mandatory long-form Census to a voluntary National Household Survey would mean data could not be compared over time, and whether that could result in lack of continuity over time.

In explaining the efforts being made to correct for any under-representation that might result from a voluntary survey, a Statistics Canada official told the committee “we have high hopes that, for a number of uses, the data will be comparable and that we will still be able to conduct studies that will be very useful.”[17]

A second Statistics Canada official acknowledged the difficulty in comparing data, but reminded the Committee that it is too soon to know the scope of the difficulty. She added, “Statistics Canada is confident, however, that the survey will produce usable and useful information that will meet the needs of many users.”[18]

Mr. Fellegi commented:

Statistics Canada said—and I absolutely implicitly trust them—that it will meet many user needs. There's no question about that. The trouble is we won't know which ones and to what extent, because bias is unknowable.[19]

Mr. Fellegi continued:

[The doubt caused by unknowable bias] will shift the debate from the underlying issues to whether the data can be trusted for this purpose. That's what I'm really concerned about. The next five years will be spent debating the data as opposed to the underlying issues they are supposed to reveal.[20]

Kathleen Lahey wrote in her submission to the Committee:

It is widely agreed that the NHS cannot produce data of with scope, detail, or reliability of the mandatory long form census for three reasons: the voluntary feature will reduce the overall response rate—predictions are that it will only reach 50% to 60%, compared with the 94% response rate for the census; it will skew the rate of response from different sectors of society in unpredictable ways; and, without a simultaneous compulsory census data to calibrate differences in response rates, it will be impossible to accurately weight the NHS responses to increase their accuracy.[21]

Doug Norris of the Marketing Research and Intelligence Association told the Committee:

In particular, the concern with the loss of the mandatory census is the loss of our ability to track change over time and it's often the trends that are important rather than a level of something which is of interest.[22]

Specific impacts related to women

Central to the Committee’s study is the possible impact of the changes to the Census, including its shift from mandatory to voluntary surveys. A witness representing professors in Quebec told the Committee that of 75 university-level research projects using Census data, many were focussed on women and their status.[23]

Françoise Naudillon told the Committee:

Women in particular would have everything to lose if the mandatory long-form census, as conducted until 2006, was cancelled. Cancelling the form opens a giant gap in the Canadian heritage of statistics, and the coherence, reliability and comparability of data will be lost. It is actually due to the production and systematic analysis of gender-based data that it was possible to implement programs to fight gender stereotypes and inequality between men and women.[24]

She concluded: “Cancelling the mandatory long-form census will condemn women to see, hear and say nothing about themselves...”[25]

The Committee heard about three possible impacts: the undercounting of vulnerable women and girls, the lack of data to conduct adequate gender-based analysis (GBA) of programs and policies, and the lack of data to evaluate programs and policies and to make strategic investments with respect to advancing the status of women.

With respect to counting women and girls, as noted above, witnesses including representatives of the Canadian Council on Social Development[26] and the Canadian Women’s Foundation[27] suggested that people in vulnerable groups—low-income, single-parent, Aboriginal, and immigrants, for examples—would be less likely to respond to a voluntary survey and this could result in skewed data on the total numbers and the circumstances of women in these groups. The Canadian Women’s Foundation told the Committee: “[A] move to a voluntary survey will mean that Canada's most economically disadvantaged women and girls will no longer be properly counted.”[28]

A witness from Status of Women Canada explained the role that data play in GBA:

Having a strong set of data that is gender disaggregated is [...] fundamental to fulfilling the government-wide commitment to performing and entrenching the practice of gender-based analysis, otherwise known as GBA. GBA is the process of examining a policy, program, or initiative for its impact on women and men in all their diversity. It provides a snapshot in time that captures the realities of women and men affected by a particular issue.[29]

Status of Women Canada also told the Committee that Statistics Canada advises them on the data sources that “best illuminate the circumstances of women in a variety of spheres.”[30]

However, Kathleen Lahey and Céline Duval expressed the view that without mandatory Census data, such analysis by governments and women’s organizations would be impaired.[31] Ms. Duval told the Committee: “Depriving organizations of reliable data that allow them to support their arguments in all sectors is the same as muzzling or destroying their work of assessing Canadians' needs.”[32]

Similarly, some witnesses testified that the mandatory long-form Census data were essential to identifying inequalities and to assessing policies and programs to reduce such inequalities. These concerns were summarized by a witness from the Canadian Federation of University Women, who said “The mandatory long-form census is a critical tool to monitor the status of women and to formulate policy to advance women's equality.”[33]

Martha MacDonald, an economics professor at St. Mary’s University told the Committee: “Of course, advocacy on women’s equality relies on the data from the census. Without it, groups will have difficulty making their points and women’s ongoing inequality will become invisible.”[34]

Finally, the Committee heard from the Canadian Marketing and Intelligence Association that detailed Census data were used by employers, large and small, to assess their hiring, recruitment and retention of women to ensure they reflected the labour force in their communities.[35] Mr. Norris, representing the Association, also spoke from his experience as a demographer with Environics Analytics: “In that capacity, I work with many large businesses, small businesses, nonprofit organizations, and governments at all levels in using the census information and other information to help them make their business decisions.”[36]

Status of Women Canada officials indicated that it was too early to tell what the impact of the change might be.

Privacy issues

The Minister responsible for the Census, the Honourable Tony Clement, cited privacy concerns of Canadians as the reason for the cancellation of the mandatory long-form Census. The Committee also received a written submission citing privacy concerns:

The real concern is that compelling someone to release personal information, regardless of what is done with it, is unavoidably an invasion of privacy. May I draw your attention to the fact that what is invaded here is not the data? It is the person. My affairs are private when I, and only I, decide whom to share them with. They are no longer private when the state, using powers of compulsion, breaks in upon them.[37]

While privacy concerns were not explicitly within the scope of the Committee’s study, as defined in the motion above, some witnesses responded to Members’ questions about privacy. Responses cited the close care taken by Statistics Canada with Census data;[38] the importance of the data as a counterpoint to any privacy concerns;[39] and the mandate of Government to remove those questions it deems to be invasive of privacy.[40]

Elimination of Census questions on unpaid activities

As noted above, the question on unpaid activities, which was included in the mandatory long-form Census in 1996, 2001 and 2006, focused on unpaid housework, yard work or home maintenance; looking after children without pay; and providing assistance or care to one or more seniors without pay. Beverley Smith, a witness before the Committee, said:

When I heard of the voluntary survey, I was dismayed that the unpaid labour question would no longer be compulsory. Imagine my surprise to learn that the question wouldn't be asked at all.... To take off the unpaid work sends the message that women's unpaid work does not matter.[41]

Adding the questions

The question on unpaid activities was added to the Census four years after the time-use cycle of the GSS was introduced. The GSS covers a broader range of activities with responses taking the form of a 24-hour diary. The Census question asks for an estimate of time devoted to these more specific activities in the previous week. A witness before the Committee who was employed by Statistics Canada at the time and who was involved in evaluating the Census questions, said:

In inserting those items [in question 33] into the census, the idea was that communities concerned with services pertaining to child care or pertaining to elder care could turn to Statistics Canada for assistance in connection with background data.[42]

Statistics Canada consultations on the 2011 Census

In questions about the removal of the question from the 2011 Census or National Household Survey, committee members heard conflicting accounts of the consultations leading up to its removal. The confusion may be based in the production of two different reports by Statistics Canada about the 2011 Census. The first was the discussion document that was to be the basis of consultations, and the second was the report on the findings of the consultations.

The discussion document released in July 2007[43] identified several questions for review including family structure, ethnocultural characteristics, and Aboriginal questions. There was no reference to consideration being given to changing or removing questions related to household activities, which included the questions on unpaid activities in this document. A year later in the report on results of this consultation, Statistics Canada reported that “...during most in-person consultations, participants were asked about the use and importance of unpaid work data.”[44]

That same report indicated that of 72 comments on this question, 30% had “stressed the need for unpaid work data collected on the census”, saying that “the data are used to analyse gender equity, understand economic divisions, measure the volume of volunteer work and develop policies.”[45] Another 30% favoured the removal of the question from the Census, because “the household activities categories are too broad and don’t provide sufficient context, the question would need to be improved to make the results meaningful—especially as it relates to the time references, and, there are alternative data sources.”[46]

Of the comments received about unpaid activities, 51 came from federal departments and agencies and other levels of government, with only four coming from non-government organizations and six from advisory committees.[47] Statistics Canada noted that the responses to these questions are used to evaluate and monitor “federal legislation, policies and programs including: Immigrant Settlement and Adaptation Program, National Advisory Council on Aging, Women’s Program.”[48] This section of the report concluded with “careful consideration is being given to excluding household activities from the 2011 Census questionnaire.”[49]

Before the Committee, Statistics Canada officials reported that a “series of highly focused consultations” followed the release of this second report in July 2008.[50] When questioned about these consultations, Statistics Canada officials offered the following description:

Certainly, we talked to the governments of our provinces and territories, as Statistics Canada does have territorial and provincial focal points. As well, we had a meeting with federal department managers from a number of departments with interests in the census information. We also did a follow-up and a specific call with provincial and territorial status of women offices, and that was organized by Status of Women Canada. Then also, as part of all of our consultations on the census, we talked to our advisory committees. We have a National Statistics Council, and particularly on this one there was an advisory committee on social conditions, which has advised us on gender work over time…. What we heard back is that there was very, very little use of the census unpaid work questions but that there was a great deal of use of the general social survey questions, which are very extensive.[51]

The documentation requested by the Committee confirmed that a follow-up meeting on unpaid activities and place of work in June 2008 identified the above-noted organizations as “partners and stakeholders” to be consulted by senior management in the summer of 2008. The consultations were “to determine the extent unpaid work data are being used and their policy relevance.”[52]

Three witnesses pointed out that the data are being used. Ms. MacDonald told the Committee, “In terms of the unpaid work questions, in my own experience there certainly is academic research on it. It’s also extremely useful for teaching and for general public education, that sort of lower level of research.”[53] Ms. Lahey told the Committee:

I don’t have any problem believing that when the Statistics Canada gender experts went looking for evidence of use of these data, they may have had a hard time finding a great deal of it. Status of Women Canada did fund a huge amount of research on this. That research has all been taken off the web page, hidden in government archives, not available on the Internet, and is not available for sale.[54]

Ms. Naudillon described the importance of the data in discussions at the United Nations Commission on the Status of Women: “...in 2008, the main issue was decent work for women and recognizing their work at home, including the care given to the elderly.”[55]

Comparisons with General Social Survey

As noted above, the Statistics Canada consultation report in 2008 indicated that several respondents referred to “alternative data sources”. Witnesses before the Committee indicated that the General Social Survey (GSS) provided detailed data on time use every five years, starting in 1992.

Two witnesses who have published articles using the data from both surveys said that the Census questions provided greater specificity, with respect to “gendered unpaid work” and that  elder care was omitted from the GSS,[56] and that greater geographic detail in the Census allows for examination of “small communities and rural-urban differences.”[57]

These two witnesses, and others,[58] described the Census and the time-use cycle data from the GSS as “complementary.”

Mr. Fellegi, who described himself to the Committee as the father of the GSS, offered the following description of how the surveys work together:

The GSS was really as much a teaser as it was an attempt to answer every question. At the same time, it has answered a lot of questions, and it has resulted in extremely interesting and useful analysis, but a survey is typically an inter-censual indicator; most of the time the census provides the detailed picture once every five years. They are complementary in their roles, and that complementary aspect works very well between the GSS and the long form.[59]

Compliance with international obligations

The motion passed by the Committee for this study mandated the Committee to “take into account the United Nations Beijing Declaration and Platform for Action to which Canada committed.” Witnesses were asked to comment on whether relying on other voluntary surveys, rather than the Census, for this information, would meet the recommendations in the Beijing Platform for Action. There are several parts of the Platform related to unremunerated activities, a sample of which is provided below.

The Platform calls on national and international statistical organizations to:

Devise suitable statistical means to recognize and make visible the full extent of the work of women and all their contributions to the national economy, including their contribution in the unremunerated and domestic sectors, and examine the relationship of women’s unremunerated work to the incidence of and their vulnerability to poverty.[60]

Governments were urged to:

Seek to develop a more comprehensive knowledge of work and employment through, inter alia, efforts to measure and better understand the type, extent and distribution of unremunerated work, particularly work in caring for dependants and unremunerated work done for family farms or businesses, and encourage the sharing and dissemination of information on studies and experience in this field, including the development of methods for assessing its value in quantitative terms, for possible reflection in accounts that may be produced separately from, but consistent with, core national accounts.[61]

Some witnesses before the Committee expressed the view that the elimination of the Census (or National Household Survey) question on unpaid activities would diminish the capacity of governments and other organizations to “monitor and to value the contribution of unpaid voluntary and care work.”[62] In particular, the witnesses from Association féminine d'éducation et d'action social expressed their concerns about being able to measure the extent to which the economy relies on unpaid activities,[63] echoed by a witness from the Canadian Council on Social Development, noting that unpaid activities are not otherwise captured in Canada’s formal economic accounts.[64]

When asked whether the GSS data are sufficient to meet the requirements for providing both rural and urban analyses of unpaid activities, an official from Statistics Canada responded:

Rural and urban information we can, to some extent, produce from the GSS. It just depends on the amount of data required or the detail of the analysis. We still produce estimates on rural and urban—maybe not for every province, but we still do that from the GSS.[65]

She also responded to a question on whether the smaller sample size would have an impact on the results, saying “Yes, it is a smaller sample, and we will not get the same level of information that we would with a larger sample.”[66]

Recommendations

Based on the testimony and submissions received by the Committee, the Committee recommends:

• the reinstatement of the mandatory long-form Census for 2011;
• that questions on unpaid activities be included in the 2011 long-form census; and
• that Statistics Canada conduct a comprehensive public consultation on the possibility of expanding the unpaid activities questions for the 2016 Census.