[CivicAccess-discuss] Free US Census data leads to analysis

[Tracey P. Lauriault]

Census Data belong to Statistics Canada and not to the public, yet is it a
public institution and the public pays for those data via taxation.  Those
same data are also sold to other levels of government and federal
departments.  Meaning the public pays many times for the same non rivalrous
good.  Census data are available to the public at a high cost and under very
restrictive user licenses.

Census data, to be mapped require a digital file that is called a framework
dataset.  Framework data, are geo-enabling files, they allow you to map
census data, elections data, ecological zone data.  Electoral boundaries,
census tract, city boundaries (cd - csd files), postal code files, etc. are
also only available for sale from our public institutions and these are
extremeley expensive (thousands of dollars).

For example i made a request for 3 data variables from statistics canada on
some poverty variables in order to understand the risk of homelessness.  The
price for those data was 60 000$.  The grant for the research was 50 000$,
so we did the work but not the work we wanted to do.

Polling data are the property of the private companies who conduct the
polls.  They are the result of a small number surveys and normally.
However, Elections Canada provides access to the results on a somewhat
restrictive way, I guess exit polling data.

What CNN and others were doing in the US, was, comparing census data (which
are free in the US) with vote results, which gives you a demographic
breakdown of each county.  These are estimates, but you can imagine if you
have a county that is 95% white, black or hispanic, or 95% between the ages
of 18-25 (university/college counties) and the result of the vote is 90%
republican or democrat, then you have a pretty good predictor of which
demographic voted for what.

This information is also useful for developing campaingn strategies,
imagine, in the Case of Apathy is Boring, seeing the results of the youth
vote, the youth voter turn-out and the distribution of youth in Canada by
neighbourhhod.  You would know where best to concentrate your resources.
This is the kind of knowledge we as citizens do not have free and open
access to.  We can get it, but it would be at a huge cost.

Where to focus dialogue? I speculate one organization at a time, and the
gold standard at the moment is StatCan, and that would mean changing public
policy on access to data and cost recovery.  This would mean I think
garnering public will from our elected officials and that requires a really
good strategy.

We can get exit poll / electoral results from elections Canada, here -
http://enr.elections.ca/home1.aspx but these are a little limited as you
will see as it is hard to zoom in and also to compare the whole country.

Cheers
t

On Thu, Nov 13, 2008 at 11:27 AM, Ilona Dougherty
<ilona at apathyisboring.com>wrote:

>  Just a quick question from someone who doesn't know very much about this
> stuff....but is learning. :)
> Is it an issue of the demographic electoral data not be collected in canada
> (ie no exit polls)? Or of the data not being accessible?
>
> Is this something that it would be useful to have a dialogue with Elections
> Canada about or is it a gov't wide thing?
>
> I am always trying to understand what the possible action steps are.....one
> thought I had would be that an independent NGO might be a good way to do
> exit polling....
> I need to understand better how it is done in the states....
>
> Ilona
>
> On 12-Nov-08, at 7:04 PM, Tracey P. Lauriault wrote:
>
> Judyth;
>
> I think we agree on most points.
>
> See the blog post on Datalibre.ca re-what the media provided during the
> elections here (
> http://datalibre.ca/2008/10/16/elections-2008-mashup-uncertainty-and-voice/)
> and what the US had here (
> http://datalibre.ca/2008/11/06/data-maps-and-the-geographic-imagination-us-election-visuals/
> ).
>
> The debate is not so much about what we get in the media, it is about what
> our public institutions are doing to communicate to us with our public data,
> or more importantly what they are not communicating and how they prevent us
> from doing so.  In a democratic country, I am not always comfortable to see
> important results on a variety of issues only through the lens of the media
> or industry or the government for that matter.  There are times when
> citizens need to look at the information and data closely, come up with
> their conclusions and develop conversations around what they discover.  At
> the moment we are missing out on a bunch of conversations.  Read this post
> on the high cost of studying poverty in Canada here - (
> http://datalibre.ca/2007/11/30/paying-for-data-to-study-poverty/).
>
> I responded inline to your message below!
>
> On Wed, Nov 12, 2008 at 3:17 PM, Judyth Mermelstein <
> judyth.mermelstein at mail.mcgill.ca> wrote:
>
>> "Tracey P. Lauriault" <tlauriau at gmail.com> wrote:
>> >[...]
>> >Has anyone seen in the private media the breakdown of the Canadian vote
>> by
>> >electoral district and a discussion of who lives there?
>> >
>> >Have you noticed the silences in blogs about organizing and stategizing
>> >around certain demographic groups, the pattern of the vote and etc.?
>> >
>> >Have you read in Canada about the anglophone vote, Haitian vote, Chinese
>> >vote, aboriginal vote, eastern European vote, youth vote, poor vote, rich
>> >vote, senior's vote, womyn's vote, men's vote, etc.?
>> >
>> >Well we cannot even imagine having that conversation because our public
>> data
>> >are locked up in the deep lurkium dbases of Statistics Canada, while our
>> >public electoral maps are locked up, funny, also in the same dbase, and
>> the
>> >key to all of these are the Cost Recovery Price!
>>
>> I'm enough of a dinosaur to look for my election information in print as
>> well as online so I did see the breakdown of votes by electoral district in
>> the Montreal Gazette, as well as various articles before and after dealing
>> with specific cultural communities. The CBC provided reasonable television
>> coverage of political parties targetting specific ethnic or other
>> demographic groups.
>
>
> Yup, it was ok.
>
>>
>>
>> Much as I agree with the idea that ideally this information should be
>> available online as soon as it is gathered, and presented creatively in ways
>> that make sense to the public, StatsCan has very little to do with this
>> particular problem. They don't do exit polls broken down by ethnicity.
>
>
> Canadians provide these data in the census, census data can be compared
> with electoral results.  This is what they did during the US elections.
> Census data are however extremely expensive and under sever user
> restrictions in Canada and are Free in the US.  The frameworks data (maps)
> associated with the census are also extremely costly.
>
>
>>
>> The strategizing and organizing by political parties is considered the
>> property of those parties, and it's not the kind of information politicians
>> want shared with the public. One of the obvious jobs of *journalists* during
>> an election campaign is to watch what the parties are actually doing in
>> various ridings and to report on things like tailoring messages differently
>> when speaking to different demographic groups. Personally, I'd be happier if
>> more journalists (including journalist-bloggers) were out in the field
>> tracking these things rather than riding the leaders' campaign trails and
>> reporting what the parties want them to.
>
>
> Yes, it is true that it is the job of journalists, they however, did not do
> that job because I speculate, they did not have the census data to do so.But
> it is also the job of citizens, watchdog organizations, etc. to also do this
> job.  Journalists are not always objective and newspapers are partisan and
> issues are just missed. For instance how did Canada get turn Tory? Who is
> voting Tory? Seniors? People in rural Areas? Evangelicals? Etc.
>
>
>>
>> Meanwhile, Canada is not yet at the degree of connectedness where all
>> voters will self-report with their ethnic data, etc. on the Web. Some will,
>> of course, in their own blogs or by commenting on blogs or forums, but most
>> won't -- and wouldn't necessarily tell pollsters anything either.
>
>
> I would never ask anyone to do this online.  This information is derived
> from the census and you do matching analysis and proportional analysis by
> riding or polling station.
>
>>
>>
>> Meanwhile, too, politicians aren't usually too happy about citizens
>> scrutinizing electoral districting: it becomes too obvious that riding
>> boundaries are gerrymandered. For example, in Quebec, if a voter in the
>> Gaspé counts as one vote, a voter in an urban riding like mine (Jeanne Le
>> Ber) only counts as 2/3. Politicians like to do this with the poorer urban
>> ridings where people are more likely to vote against them for their past
>> records and agitate for change than vote for them by tradition or because
>> somebody's promised to build yet another road to nowhere.
>
>
> Precisely why we need those datasets so that we can report these dubious
> boundaries.  They are first and foremost our public data as we paid for
> them.
>
>
>>
>>
>> >How are we to imagine our country in all its diversity if we are not even
>> >allowed to make pictures that tell us part of our story?  These last two
>> >elections - back to back - made obvious the real cost to locked up data -
>> a
>> >collective national ignorance!
>>
>> Well, the old-fashioned way to "imagine our country in all its diversity"
>> was the CBC (radio as well as TV) which made a point of telling all its
>> listeners about what was happening in every region, and the relative
>> cheapness and ease of travel by rail or road within most of Canada. The
>> former has been pretty much gutted -- regional production shut down, fewer
>> reporters outside Toronto, decisions to market to different demographics in
>> imitation of commercial broadcasters, etc. -- and the reach of rail travel
>> practically eliminated without reflecting on the consequences of fossil fuel
>> price-increases when everyone is dependent on personal cars to get around.
>
>
> So you agree with the point then, that citizens need other options, one is
> a solid public media, the other is access to those data.
>
>>
>>
>> >I have been reading blogs about Proposition 8 (same sex marriage) vote in
>> >the US, the demographic breakdown of those results, and was drooling over
>> >the visuals i watched on election night.  Seems to me, the data, the
>> >infrastructure and the pictures can help us have a conversation about
>> >sameness and differences.  Canada however remains silent!
>>
>> Maybe it's just me but I'm not convinced pretty graphics make up for the
>> lack of understanding of real events. No matter how many times I see those
>> coloured maps and pie charts, I can't forget that polarizing states into
>> "red" and "blue" visually gives the totally false impression that the
>> populations of those states are fairly homogeneous -- which is rarely the
>> case -- and the thinner slices of "pie" represent people who don't really
>> matter.
>
>
> Precisely, the data we got on the Canadian elections did the same thing,
> see the blog posts above. It is the detailed data that tells you the story,
> the CNN maps dove into the actual counties which showed a much more diverse
> story.  The information in the US is free, so it is possible to match census
> data at a micro scale to county data at a district scale, which is precisely
> what people were doing and that is what they were talking about.
>
>>
>>
>> To me, it's worth looking at a riding like Jeanne Le Ber and asking *why*
>> it voted strongly for the Bloc Québecois in the past two elections, when it
>> was a "safe" riding for the Liberals before that. Hint: it has little to do
>> with the "sponsorship scandal" issue which is all anyone talked about in the
>> media, and a great deal to do with truly lousy representation of the
>> constituency in recent years, though in the Neverendum Referendum era the
>> voters felt obliged to support the Liberals to keep the country together and
>> the local economy from collapsing further.
>
>
> Yup
>
>>
>>
>> That's not something pollsters are normally paid to look at (in fact, you
>> could argue they're usually hired to elicit particular answers to
>> politically-expedient questions) and therefore you don't have the data with
>> which to make a meaningful graph. The same is no doubt true of at least a
>> hundred other ridings around the country. The parties themselves try to
>> gather and analyze the information but keep it under wraps as much as they
>> can because it's needed for the next round of strategic manoeuvres.
>
>
> Again, you reaffirm the points I made earlier.  We are not getting unbiased
> reporting from the media nor posllsters!
>
>>
>>
>> The government in power is not an impartial body dedicated to full public
>> disclosure: it's a political party with a strong interest in keeping the
>> public ignorant of how elections are really won. Elections Canada rules can
>> only go so far and mainly look at financing. not cultural and political
>> matters. Statistics Canada can report on demographics without voting
>> information; Elections Canada can report voting patterns but has no access
>> to demographic breakdowns; polling companies survey only what politicians or
>> media types are interested in knowing, and then having published or
>> suppressed depending on the circumstances. The question then is who can
>> collect the data we would like to have graphed, and how meaningful would it
>> be really to know that 36% of Indonesian Canadians voted NDP or whatever?
>
>
> Imagine if, Elections Canada did not have to pay for StatCan data which
> they have to do now! Further, Elections Canada cannot give out the FSA file
> because StatCan sells it! Imagine if we all had access to these data? Would
> we use them more? Would we inform new conversations? Would we have evidence
> based policy formulation? Could we keep government - both bureacracy and
> government accountable.  The data mean something per community.  For
> Instance do the poor vote differently and if so why? The blog post I looked
> at re-Proposition 8 was really reflexive about double discrimination, an
> important converstation to have.
>
>>
>>
>> Seems to me the big news about California's Proposition 8 vote was the
>> $200 million the Mormons spent to help get it passed and I got that from an
>> e-mailed report, not a blog or graphic. My next question is, when will
>> anyone be willing to spend that kind of money on public-data-collection
>> instead of lobbying and advertising?
>> Just wondering ...
>
>
> The data are already collected, we just do not have access to them!
>
>>
>>
>> Anyway, I'm with Bob Gibson <bob at gibsonlibraryconnections.ca>:
>> >Because it must affect the efficacy of advocacy groups across the
>> spectrum,
>> >if there were a way to consolidate their wish lists, abolition of CC
>> might
>> >make its way into the government mindset just because it is so unanimous.
>> >Like closing Guantanamo. Is there such a thing: an umbrella/consolidation
>> of
>> >Canadian advocacy groups' wishlists? If there ain't, there should be!
>>
>> Judyth
>>
>> --
>> N.B.  I am off-line much of the time these days: please
>> phone if you need me urgently, and bear with me if less
>> urgent matters take a while to resolve.
>> ##################################
>> Judyth Mermelstein   "cogito ergo lego ergo cogito..."
>> Montreal, QC         <judyth.mermelstein at mail.mcgill.ca>
>> Canada H4G 1J4       <lapomme at postaccess.com>
>> ##################################
>> "A word to the wise is sufficient. For others, use more."
>> "Un mot suffit aux sages; pour les autres, il en faut plus."
>> _______________________________________________
>> CivicAccess-discuss mailing list
>> CivicAccess-discuss at civicaccess.ca
>> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>>
>
>
>
> --
> Tracey P. Lauriault
> 613-234-2805
> https://gcrc.carleton.ca/confluence/display/GCRCWEB/Lauriault
> _______________________________________________
> CivicAccess-discuss mailing list
> CivicAccess-discuss at civicaccess.ca
> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>
>
>
> _______________________________________________
> CivicAccess-discuss mailing list
> CivicAccess-discuss at civicaccess.ca
> http://lists.pwd.ca/mailman/listinfo/civicaccess-discuss
>



-- 
Tracey P. Lauriault
613-234-2805
https://gcrc.carleton.ca/confluence/display/GCRCWEB/Lauriault
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.pwd.ca/pipermail/civicaccess-discuss/attachments/20081113/b1bdfe72/attachment.html>