[Open-data-census] Open Data Census Methodology - how to compare federations with centralized Gov. systems?

Wed Jun 12 15:25:19 BST 2013

Greetings all;

First, the initiative should be lauded for its efforts.  It is a great
initiative.

However, I do think that if we are working with data, and we wish to do
evidence based decision making based on the results of the census, census
methodology and data comparison issues need to be made obvious to ensure
accuracy and comperability.

For example, the Canadian census results are problematic, not because the
information submitted is inaccurate, but because it is grossly incomplete.
 Canada is a federation, with explicit divisions of power (
http://www.pco-bcp.gc.ca/aia/index.asp?lang=eng&page=federal&sub=legis&doc=legis-eng.htm)
whereby
jurisdictional responsibilities between the Federal government, provinces
and territories and cities are constitutionally set.  Jurisdictional
divisions therefore dictate divided responsibilities which translates in
different things being administered  resulting in many kinds of
administrative data residing in multiple governments under different rules.

Natural resources for instance are managed by both the federal and
provincial and territorial governments, and the data associated with
pollution will differ depending on the resources in question and where they
are located.  Data reside in 14 jurisdictions & innumerable departments and
access to these data and how they are collected will be determined by each
of these.  A citizen would have to contact each jurisdictional or
department separately to access these.  This is not an insignificant task.

In addition, cities are governed by provinces and territories and there is
no real incentive for cross jurisdictional collaboration across provinces
and territories, although it does happen.  Canada is a big place and what
happens on the west coast is not the same as the east, the prairies, or
Quebec, Nunavut or Ottawa. Cities do not have jurisdiction over some of the
items listed in your census, and even though, they are the innovators in
Canada on the Open Data front, they would score low on the census as they
just do not produce the data sought after in the census.

Also, even though Canada has a federal Open Data Program, there are other
assaults on knowledge, the Census of Canada was cancelled, Library and
Archives has been decimated, there is muzzling and control of government
scientists, access to information requests are slow and contentious,
scientific monitoring stations are being closed, think tanks lost their
funding, surveys on vulnerable peoples cancelled and data releases
carefully controlled, databases like the gun registry being destroyed.  And
this is just the tip of the iceberg.  The Federal gov. may have open data,
but if knowledge and data producing institutions are being decimated, does
it really matter?

Of the 13 provinces and territories, only 3 have open data programs, the
data produced by these are the juicy ones however, this is health,
population health, education, social welfare, resources, roads, etc.  These
are the institutions that deliver services to Canadians, and these are the
ones we want to have data from, but these are the late bloomers in terms of
data sharing, transparency, disclosure, access to information and so on.
 It is nice to have federal data, but we really want admin data from
provinces and territories.  Also, business registries are federal, and
there are also provincial and territorial ones.  The open data census does
not capture this nuance.  The results from Canada would be much lower if
this nuance was taken into consideration.

There are thousands of cities and municipalities in Canada, currently we
have 36 with open data pilots or catalogs
(http://datalibre.ca/links-resources/#Open
Data Cities).  The data they contain are cute, it is generally  parks,
recreation, monuments, and so on.  There are a few with restaurant health
inspections and so on. This is the jurisdiction that manages city
infrastructure, where we live, and they do not have demographic data and so
on.  I would love to know more about brown fields, dump sites, waste
management, boil water advisories and smog alert days, but these are not in
the open data city portals.

Thus, when the question of public transportation comes up under the Open
Data Census, it is problematic as the national/federal government does not
have jurisdiction over transit unless a system crosses provinces and
territories, but that is not transit per say, it is airlines, the national
rail lines or Greyhound buses.  Transit is administered at the level of the
municipality, and that picture differs in each city.  There is no way to
give an overall mark, and then of course, there is the issues that
Montreal, Toronto, Ottawa, Calgary, Edmonton and Vancouver, are big cities
(for Canada), they have larger transit systems, smaller cities would not
have the resources necessarily to release their transit data if in fact
they even had gps on their buses and so on.  So how to compare these?

In other words, the picture is complex, and we may get good marks for one
jurisdiction, but that mask the reality of the others.  Which brings us to
the open government partnership, I know not a Census topic, but it only
allows for national government representation, in Canada, this is very
problematic for the reasons discussed above.

My question then, is how will the Open Data Census capture the differing
governance issues between the nation states compared?  How will a user of
the results be advised of the comparative caveats?  Can countries really be
compared? Is there recognition that this first census may in fact be the
first to understand the differences and that this one will find out the
methodological issues and the next one will aim to address these?

Sincerely
Tracey

-- 
Tracey P. Lauriault
Post Doctoral Fellow
Geomatics and Cartographic Research Centre
https://gcrc.carleton.ca/confluence/display/GCRCWEB/Lauriault
http://datalibre.ca/
613-234-2805
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/open-data-census/attachments/20130612/7aa4a593/attachment.htm>