Uses of Census Long-form data – Question Justification

Graphic By http://www.socialsignal.com/ As part of my PhD dissertation research I have been investigating the Census of Canada. I have a dbase of all the questions since 1871 being edited at the moment and needless to say copious notes. Here is a small extraction that is of relevance to the debates about the census at the moment.

The list is not exhaustive and not fully edited, but does provide insight as to why those questions are asked and why those who know the census are outraged.  It also remains uncertain how the changes will be able to address the legislative requirements.

  1. Aboriginal Identity: Employment Equity Act; Indian Act; Multiculturalism Program; Aboriginal Human Resources Development Strategy, used by aboriginal governments, communities, and organizations, characteristics of on and off reserve populations
  2. Activity limitations: Data are used to develop and manage transportation, housing, communications, employment equity and other programs.
  3. Certificates, diplomas and degrees, field of specialization: Employment Equity Act, Immigration Act, Canada Student Loans Program,
  4. Citizenship: Citizenship Act, Canadian Multiculturalism Act, Immigration Act, voting and electoral planning.
  5. Common-law status: first asked in 1991, track changes in family structure and family relationships, and prevalence of cohabitation. De facto marital status is also assessed. Common-law data from 1986 and 1981 were derived from relationship to person questions. changes in family structure and family relationships, prevalence of cohabitation, first time in 1991, formerly it was derived, opposite sect included, Same acts as marital.
  6. Difficulties with daily activities: Started in 1991. Used for the Participation and Activity Limitation Survey (PALS). Canadian Charter of Rights and Freedoms, Employment Equity Act, Canada Health and Social Transfer.
  7. Ethnic origin: Canadian Charter of Rights and Freedoms, Multicultural Program, diversity measure in Canada, and characteristics of ethnic groups, also required for the Multiculturalism Act and Canadian Charter of Rights and Freedoms.
  8. Full-time or part-time work: EI, Old Age Security Act, Canada Pension Plan.
  9. Household activities: included for the 1st time in 1996, to measure contributions made from unpaid work, to “give a picture of both the market and non-market component of Canadian Society” (Statistics Canada, 1997:69).
  10. Housing: assess current state of hosing stock, evaluate future needs, national Housing Act and Canada Mortgage and Housing Corporation Act, mortgage loan insurance programs, new homeowners insurance programs, land management programs, housing assistance programs, etc. this was asked of 20% sample instead of a 80% sample.
  11. Income: social assistance, EI, Old Age Security Act, indicator of the economic well being of Canadians, only source of small area information on income, draw comparision between groups, sources of income, and to analyze income composition,
  12. Indian Band/First nation membership and Treaty/Registered Indian: Employment equity act, Indian Act, Aboriginal Business Canada Program, and Aboriginal Human Resources Development Strategy.
  13. Knowledge of English and French: Official Languages Act; Citizenship Act, Official Languages Support Program.
  14. Knowledge of non-official languages: Canadian Charter of Rights and Freedoms, Citizenship Acts, Multiculturalism Program.
  15. Labour Market: can be used for EI, social assistance, education and training, Incorporation status, place of work, mode of transport to work, language at work
  16. Language of work: insight into the vitality of official languagues among official and non official minority language groups, use of language on the job, linguistic integration assessment, integration of allophones immigrants into the labour force
  17. Language: to administer Canadian Charter of Rights and Freedoms. Also need of language training, assess language skills, home language was added in 1971, to study linguistic assimilation in Canada and to evaluate language programs to help linguistic groups maintain their heritage and to assess which official languages are learned.
  18. Languages learned at home: Canadian Charter of Rights and Freedoms, Citizenship Acts, Multiculturalism Program.
  19. Languages spoken at home: Canadian Charter of Rights and Freedoms, Citizenship Acts, Multiculturalism Program.
  20. Marital Status: for producing family data, population estimates, prior to 1991 common-law couples responded as married, analysis between legal marriage and co-habitation. Canada Child Tax Benefit, Old Age Security Program, Canada Pension Plan and Same sex couples
  21. Mobility: Canada elections act, Federal-Provincial Fiscal Arrangement Act, Official Languages Support Program. Measure pop growth, migration in intercensal years, benchmark data to adjust intercensal estimates, migratory statistics, pop growth, mechanism used by the labour market to smooth out income and employment disparities, population estimates needed for the Federal-Provincial Fiscal Arrangements and Federal Post-Secondary Education and Health Contributions Act. I year mobilities was added here,
  22. Mode of transportation to work: first time, meet the needs of users such as transportation planners and engineers, boards and market analysis. To plan urban growth and transportation networks, environmental impact and energy consumption with transportation.
  23. Number or rooms / bedrooms: helps to evaluate overcrowding, dwelling size, housing condition and quality of life,
  24. Period of construction: life cycle of buildings, housing renovations, emergency repair programs, RRAP programs,
  25. Place of Birth of Parents: Citizenship Act, Immigration Act, Canadian Charter of Rights and Freedom (but this is a new question so…), second and third generation Canadians.
  26. Place of birth: Citizenship Act, Canadian Multiculturalism Act, Immigration Act, adaptation to Canadian culture.
  27. Place of work: used to assess commuting, and its impact on the lives of Canadians, to identify requirements for transportation, public service locations (e.g. schools), help urban transportation planners, traffic patter analysis,
  28. Population Groups – Employment Equity Act, Official Languages Act, Canadian Multiculturalism Act,
  29. Relationship to person 1: Canada Child Tax Benefit, Old Age Security Program, Canada Pension Plan.
  30. Religions: Cultural Integration Program, Cultural Enrichment Program, Multiculturalism Program
  31. Schooling: illiteracy, remedial programming, continuing education market, refreshing workers skills, (no legislation or program requirements with this one)
  32. Shelter cost: National Housing Act, Canada Pension Plan, Canada Health and Social Transfer.
  33. Unpaid Work: Immigrant Settlement and Adaptation Program, Women’s Program and National Advisory Council on Aging
  34. Visible Minority – employment equity, status thereof,
  35. Year of Immigration: Programs – Language instruction for New Comers to Canada; Independent, Sponsored Immigration and Refugees; Immigrant Settlement and Adaptation Program,
  36. Yearly payment on shelter costs: estimate costs, developing and evaluating housing, welfare and public service programs

20 comments

good article.. only hope our federal government does ‘see the light’ and reverse their decision

Monika Awizen

Monika Awizen’s avatar

Very good article.The problem is that Harper think he IS THE LIGHT.

The list is not exhaustive and not fully edited, but does provide insight as to why those questions are asked and why those who know the census are outraged.  It also remains uncertain how the changes will be able to address the legislative requirements.
+1

I don’t see how these questions are even really a matter of privacy. Most of these things could be found on a Facebook profile, and that corporation is using our information for targeted ads rather than useful government services. Thank you for putting these up, they were very helpful.

Jared Langdon

Jared Langdon’s avatar

I agree with the government on this one. It’s fine for them to ask for the information on the long form, but it isn’t okay for them to demand it at gunpoint. Filling out the form should be like voting. You may consider it your civic duty, but you should be free to decide whether or not to take part. The short form on the other hand (the one that just counts people) is necessary for democracy, since we need to make sure the ridings are properly allocated. That’s what the census is for – nothing more.

“I have a dbase of all the questions since 1871 being edited at the moment and needless to say copious notes”

Tracey,

Will you be releasing this dataset as Open Data, some time after you complete your PhD?

-g

Many countries including Australia have made voting compulsory which is somegthing I would support. The long form must be mandatory as the information gathered has so much impact on all of us and is used to make so many important decisions.

It’s a matter of personal privacy. If I were to get a long-form census I would refuse to answer any questions I didn’t want to, and for that I deserve deserve in jail?

They don’t deserve to know my ethnic origin, that’s just racist, and only used to fuel racist government “affirmative” action hiring programs.

They don’t deserve to know my “household activities”, that’s none of their business how I spend my time.

They don’t deserve to know my Indian Band/First nation membership and Treaty/Registered Indian status – those questions are only justified so as to propagate a racist apartheid reserve system in Canada.

Likewise my religion. Separation of church and state, huh? Religion should be the private domain of the private citizen.

Sometimes principles (like privacy) are more important. There are lots of people who fudge the long-form census (The perfect example of which was the chain email encouraging bilingual francophones to fake an inability to speak English to get more franco services in non-franco areas). So by being honest and forthright by refusing to fill it out, rather than randomly ticking off bullshit responses as so many others do, somehow I deserve JAIL?!?!

There are some things on the long form that are indeed useful, but others are simply offensive and invasions of personal privacy. Thanks to the government for seeing our perspective on this.

I will make my data public once assembled and published. I have already made many data finds public on this blog to help provide some facts on the issue.

I agree with Marcus, if everyone is in such a snoot and thinks its important to have the long form mandatory then just answer the questions, there ya go, problem solved. For those that believe some of the questions are nobody’s business, you want them jailed, or fined? get real.

First of all, the long form is not being abandoned, merely made voluntary. The information will still be available; you just won’t be threatened if you don’t want to complete the form.

Secondly, I don’t buy the claim that its gleanings are used very much. Today the mayor of a city of 20,000+ said he could not remember the last time they had used that information. I suspect this is simply a political football and citizens are swayed by their favourite politician’s opinion. Start thinking, folks, and you will find the matter deserves to be ignored.

Making the long form “merely voluntary” is precisely the problem. If it’s not mandatory, it looses its value as the reference point for ALL other surveys – whether they’re done by public or private entities. Without the census, we will no longer be able to address sample biases in surveys. Ironically, we’ll be spending more money for this useless data than we would have done for a proper census!

The fact that elected politicians are not aware that they (or rather, their administration) use census data everyday simply means that they don’t do the analysis themselves – the data sources are totally transparent to them, and that’s ok. What’s not ok is their disregard for the evidence basis required for them to make informed decisions.

As someone said last week, the Conservatives are bringing us into the era of decision-based evidence making…

Ah! It is used alot! Just persuse the list of serious data users and producers on census watch. Also, Government users are huge! I like it when the gov is informed don’t you?

For some historical context, until the others make their databses available, here’s:

- an historical overview, which pts out if you think the long form now is bad, consider the fact that the one everyone had to do in 1871 had 211 q’s, and all the various ones you might have got stuck doing in 1901 had 561 altogether:

http://www12.statcan.ca/census-recensement/2006/ref/about-apropos/hist-eng.cfm

- here’s a table of when the various q’s were introduced (and note, the infamous no. of bathrooms was only ever asked _once_, in 1981) :

http://www12.statcan.ca/census-recensement/2006/ref/dict/app-ann001-eng.cfm

- and here’s an account of when & where the 10 or 20% sampling on some longer forms started happening (in 1941; not that Fellegi was wrong about his introducing the shorter forms in ’71; before that, it’s just everyone had to do long forms _and_ some had to do even longer forms):

http://www12.statcan.gc.ca/census-recensement/2006/ref/rp-guides/rp/sw-ep/sw-ep_p03-eng.cfm

- and here’s where you can see images of the main population q’re for all the censuses from 1971 back to confederation:

http://www.uottawa.ca/academic/arts/cdn/ccri/CCRI/1971.html

(Just look for the “Population Census Schedule” link at the bottom left of each year & click on the year in the middle right for the others)

The benefit of mandatory long forms is that they can be distributed randomly. This allows statisticians to generalize from a small randomized population to a larger population.

Allowing people to self-select whether they participate destroys this, making the whole exercise a waste of paper.

The benefit of the mandatory, is an even sample everywhere, which allows for cross jurisdictional comparison and also longitudinal analysis. We start at ground zero with this one and loose comparing local areas with the past.