THE LONG FORM CENSUS, DISASTER PLANNING AND OTHER USES YOU DIDN'T KNOW YOU'D BE MISSING Ottawa Statistical Society 8 th Annual Seminar Wendy Watkins Carleton University Library Data Centre October 25, 2010
THE LONG FORM CENSUS, DISASTER PLANNING AND OTHER USES YOU DIDN'T KNOW YOU'D BE
MISSING
Ottawa Statistical Society 8th Annual Seminar
Wendy Watkins
Carleton University Library Data Centre
October 25, 2010
INTRODUCTION
Purpose of this presentation is to introduce you to things you wouldn’t normally associate with a census
To do this we will: have a brief look at products from the 2006
census check out the levels of geography available with
the 2006 census look at the continuum of access from free to
restricted examine a couple of examples of things we will
NOT be able to do with the National Household Survey (NHS)
CENSUS DATA PRODUCTS
GEOGRAPHY
GEOGRAPHY
_______________________________________________
OR, MORE SIMPLY ….
2006 census geography Canada Province Large cities Municipalities Neighbourhoods Areas smaller than neighbourhoods Blocks
2011 National Household Survey Likely to lose everything below the municipal
level because of data quality issues
CONTINUUM FROM DIY TO DISTC Completely Open
Statistics Canada Census 2006 website Open on location
Depository Services Program Level of detail is slightly better here
Open with membership Data Liberation Initiative
Academics only; desktop analyses
Restricted (requires intervention by Statistics Canada) Research data centres
Largely academics; some government All output vetted for disclosure by Statistics Canada
Custom tabulations Anyone with money
Remote job submission Experimental; real thing coming soon
WHAT WE WON’T BE ABLE TO DO AS WELL
Many research questions require several types of data Aggregate data at fine levels of geography
Not likely to survive Census public use microdata (PUMF)
No PUMF for 2011 Too few variables
NHS PUMF No idea what it may or may not be able to contain Depends on data quality Could produce a distorted picture of Canadians as
richer, whiter and better educated than we really are Not likely useful for policy-making
3 EXAMPLES OF THINGS YOU DIDN’T KNOW YOU’D BE MISSING
Public health H1N1 vaccination plans
Study of refugee adaptation Very detailed study 5 years after Project 4000
Disaster Planning Sophisticated modelling using detailed Census
data to increase the ratio between rescue and recovery
PUBLIC HEALTH – H1N1 PLANNING
PUBLIC HEALTH – H1N1
Officials used 2006 Census data to locate clinics in neighbourhoods with: high density low income high proportion of new Canadians
Also used language data to determine whether languages other than English and French were required
Can certainly get density information from 2A form
Cannot get the same level of refinement for planning with the NHS
TONY CLEMENT, JULY 21, 2010 There has been considerable commentary about the federal
government's decision to replace the 2011 mandatory long-form census with the voluntary National Household Survey. The government made this decision because we do not believe Canadians should be forced, under threat of fines, jail, or both, to disclose extensive private and personal information.
We believe it is not appropriate to compel citizens to divulge how many bedrooms they have in their houses or what time they leave for work in the morning.
PROJECT 4000, REFUGEE ADAPTATION STUDY
PROJECT 4000, REFUGEE ADAPTATION STUDY
PROJECT 4000, REFUGEE ADAPTATION STUDY
Complex study of Vietnamese refugees Used interviews, survey questionnaires and
Census data Census variables used:
Crowding Number of bedrooms Number of people in the household
Employment Education Official language acquisition Language spoken at home Language spoken at work Etc., etc.
DISASTER PLANNING
DISASTER PLANNING
Remember Clement’s intrusive variables: Number of bedrooms Time people leave for work
Complex study undertaken by an Environmental Studies graduate student and a researcher in the US
Developed a model to be used to try to increase the ratio of the rescued (alive) to the recovered (dead)
Used a wide selection of long form data
DISASTER PLANNING (CON’T)
Housing stock variables Age of dwelling State of repair
Time people left for work Distances travelled Crowding
Number of bedrooms Number of people in the household
Developed scenarios for first responders based on which structures were likely to collapse when certain areas were likely to be occupied or
empty (time of day, day of week, etc.) Based response decisions on various scenarios
WHERE DOES THIS LEAVE US?
Census decision may actually end up costing more than data quality, but perhaps for the government, the following is more important:
Canadians will not be able to hold the government accountable for its policies since its election without comparable data.
QUESTION
Could this be what’s really behind the decision to kill the long form census?
MORE INFORMATION
http://datalibre.ca/census-watch/ http://www.facebook.com/home.php?#!/
group.php?gid=134479453241171 http://
www.evaluationcanada.ca/affichage/ccsd_20101029_e.pdf
http://www.theglobeandmail.com/report-on-business/economy/economy-lab/the-economists/economists-unite-you-have-nothing-to-lose-but-data/article1829559/
http://www.cbc.ca/canada/story/2010/12/12/ns-long-form-census-being-challenged-in-court-again.html
http://www.evaluationcanada.ca/affichage/ccsd_20101029_e.pdf