The stink about the census


in Canada, Geek stuff, Politics

One of the biggest challenges in statistics is collecting a representative sample: finding a subset of the population that will do a good job of approximating the whole group. When a dataset contains a lot of sampling bias and is not reflective of the general population, it is essentially worthless as a guide. That cannot be fixed by using a larger sample side, nor can it be dealt with via fancy mathematics.

The classic example of sampling bias is the ‘Dewey Defeats Truman’ headline, from The Chicago Tribune in 1948. The newspaper got their prediction wrong because they sampled people with telephones, at a time when telephones were comparatively rare. Most of the people who had them were rich, and rich people were more supportive of Dewey. As a consequence, telephone polling provided bad information about the likely voting behaviour of the whole population.

This clearly relates to the decision of Canada’s federal government to make the 2011 long-form census optional. With a mandatory census, you more closely approximate an unbiased sample (it isn’t perfect, because some people will refuse to fill in even a mandatory form). With a voluntary census, you are always vulnerable to the possibility that the sort of people who will make the effort to complete it will differ from those who will not. In such a situation, the data in the census could be a poor reflection of the situation in the population as a whole.

That is why it is foolish for the Fraser Institute to advocate the use of voluntary polling or market research, in place of a census. The quality of data from such sources can never be as good, because sampling bias will always make it suspect.

Zoom also has a post on scrapping the mandatory long census.

{ 34 comments… read them below or add one }

Milan July 19, 2010 at 6:28 pm

I must say, I have never seen Canadians so concerned about data quality before…

. July 19, 2010 at 6:32 pm

UPDATED – CensusWatch: Suddenly, everyone wants to talk statistics with Tony Clement …

* July 19, 2010 2:52 PM |
* By Kady O’Malley

Earlier today, an ad hoc coalition of bankers, economists, academics, pollsters and other data enthusiasts (including not one but two former clerks of the privy council) released the following request for a meeting with the minister to discuss their census-related concerns:

Read the full text of the letter here

Milan July 19, 2010 at 6:37 pm

The letter linked above makes indirect reference to sample bias:

It is of course difficult to judge in advance the sort of overall response rate that would be realized for a voluntary survey. Tests of a move from mandatory to voluntary in the U.S. yielded unsatisfactory results and the process was dropped. Past experience indicates that the responses will not be representative of the total population. And in particular the responses from key communities of concern, such as the very poor, Aboriginal communities, recent immigrants and some ethno-racial communities, will likely be quite low. Effective health services would be compromised. Consequently, the impact of these changes will be disproportionately borne by those who are already most vulnerable. (Emphasis mine)

It is also signed by some heavy-hitters, include two former Clerks of the Privy Council.

. July 20, 2010 at 8:34 am

Clement accused of misrepresenting census impact

Statscan insiders say Industry Minister’s comments playing down effects of voluntary survey enraged staff

John Ibbitson and Tavia Grant and John Lorinc

Ottawa and Toronto — From Tuesday’s Globe and Mail Published on Tuesday, Jul. 20, 2010 3:00AM EDT Last updated on Tuesday, Jul. 20, 2010 3:02AM EDT

Growing unrest within Statistics Canada and growing anger from groups of every political persuasion have left Stephen Harper’s government facing a revolt over its plans to end the mandatory long-form version of the census.

Officials within Statscan believe Industry Minister Tony Clement is misrepresenting the advice he has been receiving from the internationally respected agency charged with gathering data on Canadians.

Mr. Clement has said Statscan officials reassured him the agency can manage the 2011 census effectively without forcing some people to fill out the longer version of the form.

That’s not what Mr. Clement has been told, according to a source close to the story who asked not to be identified, and Statscan officials expect chief statistician Munir Sheikh to come to the agency’s defence by saying so.

The minister’s claims have angered and demoralized a staff already under pressure from budget cutbacks, which have led, for example, to less thorough regional analysis of statistical trends.

This assault on the integrity of the general census is, for many of the 6,000 workers within the agency, particularly galling.

“The census is the flagship of the department,” said one person inside Statscan. “If you want to demoralize staff, there’s no bigger target.”

. July 20, 2010 at 8:38 am

“Mr. Drummond, who recently stepped down as chief economist of the TD Bank, said the council unanimously believed that abandoning the mandatory long-form census would skew the 2011 results, causing a statistical break with previous surveys that would it make impossible to read and project trends accurately. “

. July 20, 2010 at 9:48 am

Statistics aficionados, start your engines: Liveblogging the census debate at Industry … or not.

* July 20, 2010 7:30 AM |
* By Kady O’Malley

You know, if anyone had told me back when the House adjourned in June that just a few weeks later, I would be on the edge of my seat over the prospect of liveblogging committee hearings on the 2011 census, I’m pretty sure I wouldn’t have believed them, but here we are. It’s a funny old world.

I should probably warn any similarly obsessed readers that at the moment, it’s not even clear whether they’ll even get around to the census at all at today’s meeting. which was originally scheduled to deal with the closure of the Shell refinery in Montreal. Still, if there’s even the chance that a battle over the balance between privacy rights and mandatory data collection will break out, I’m not going to miss a second of it, so check back at 8am for full coverage. Oh, and no, I’m not sure if any representatives of the pro-census Jedi lobby, knights or otherwise, will be in attendance, but one can always hope, right?

. July 20, 2010 at 11:23 am

A creepily-catchy YouTube sing-along census defence

* July 20, 2010 10:25 AM
* By Andrew Davidson

I honestly don’t know what to make of this, but it certainly shows some people can unleash their pent-up creative energy on literally anything.

It’s almost like there was a competition calling on people to squeeze as much fun out of the most boring song subject matter (the threat of the mandatory long-form census’s demise) and longtime Toronto NDP organizer and gay-rights activist John Campey — accompanied by his merry “Data Hounds” — won. Good on them for that achievement, at the very least.

If listened to over and over (as I have done for the past 20 minutes) the tune becomes creepily catchy. It sounds like something a California cult would come up with after spending a decade analyzing the first verse of The Beatles’ “She Loves You” and Mamas & the Papas’ studio outtakes to capture their signature acoustic recruitment brainwave trigger-hook. It’s no Pet Sounds, but it’s got something.

. July 20, 2010 at 1:43 pm

Counting people
Leviathan’s spyglass
The traditional census is dying, and a good thing too

Jul 15th 2010

GOD is, according to the Bible, in two minds about censuses. The Book of Numbers is so named because of God’s command to Moses that he should count the Israelites in preparation for war. Years later when King David does the same thing, the Lord wastes no time in smiting him for his trouble.

Perhaps God’s ambivalence springs from uncertainty about whose side He is on. Historically, rulers liked censuses, because they enable them to conscript and tax their people. Citizens disliked them for the same reasons. But, as governments became less malevolent, an exercise designed to extract value from the populace became one whose purpose was to improve the quality of administration.

Now this centuries-old tradition is slowly coming to an end. If statisticians in Britain get their way, for instance, the census planned for next year could be the country’s last. Instead, they are considering gathering information from the vast, centralised databases held by government, such as tax records, benefit databases, electoral lists and school rolls, as well as periodic polling of a sample of the population. It is a global trend, pioneered, inevitably, in Scandinavia. Denmark has been keeping track of its citizens without a traditional census for decades; Sweden, Norway, Finland and Slovenia, among others, have similar systems. Germany will adopt the approach for its next count, also due in 2011.

There are two reasons for the change. The first is that computerisation allows statisticians to interrogate databases in a way that was not possible when information was stored on cards in filing cabinets. The second is that counting people the traditional way is getting harder, and less useful. Rising labour mobility and the accelerating pace of societal change mean that information goes stale more quickly than ever. Since its last census in 2001, for instance, Britain has seen hundreds of thousands of immigrants arrive from new eastern European members of the EU. Local governments complain that out-of-date information ignores these newcomers, leaving schools overcrowded, budgets stretched and houses scarce. At the same time, filling in the forms has become more onerous: what started as a short questionnaire about who lived where has turned into an inquisition about everything from toilet and car ownership to race and religion. As a result, compliance rates are falling. The decline of deference raises worries about reliability: last time, when asked about their religious affiliation, 0.7% of Britons replied that they were Jedi Knights.

R.K. July 20, 2010 at 5:44 pm

I must say, I have never seen Canadians so concerned about data quality before…

It’s less anger about the census itself, I think, and more an opportunity for the opponents of the Harper government to hammer them on something that almost everyone thinks they have gotten wrong.

. July 21, 2010 at 3:48 pm

“OTTAWA — The head of Statistics Canada cancelled a Wednesday afternoon town hall meeting with employees through a terse email memo, igniting rumours he may resign amid controversy over changes to the census.

“In light of today’s media coverage, I am cancelling the scheduled town hall meeting,” Munir Sheikh, the chief statistician, wrote in an email to employees just before 1 p.m. “I am reflecting on my position and that of the agency and will get back to you soon.”

Mr. Sheikh, who has been in his post for two years, has remained silent throughout the public uproar over changes the Conservative government made to Canada’s census. The agency itself has been largely muzzled, no longer permitted to grant interviews and responding to inquiries only with brief emailed statements.

Former chief statistician Ivan Fellegi, a legend who spent 50 years at the agency and 23 in the top job, has said he would have resigned if such a change had been forced on him when he was in charge.

“Every ounce of my professional knowledge, which is certainly the same professional knowledge that is very widespread at Statistics Canada, tells me that this is statistically absolutely wrong,” he said in an interview with Postmedia News.

At the end of June, it became public that the government had scrapped the mandatory long-form census and replaced it with a voluntary survey — a move that met with heavy criticism from community and religious groups, municipalities, think-tanks and academics. Critics say a voluntary survey will produce a skewed national demographic portrait and fundamentally undermine the integrity of the statistics on which countless programs, funding models and other surveys rely.”

. July 22, 2010 at 8:50 am

Canada by the numbers

Industry Minister Tony Clement claims that Statistics Canada has told him that moving from the mandatory long-form census to a voluntary survey would not create problems (Refusing To Reverse Their Census Decision, Tories Find A New Way To Go Populist – July 16). Ivan Fellegi, who was the chief statistician for 23 years, has explained why that is not true and suggested that the people inside Statistics Canada must actually be very unhappy with this decision. Where is the current chief statistician in all of this?

Decades ago, we established that the Bank of Canada needs to operate at arm’s-length from political interference. The same should be true of the national statistical agency. If statistical collection changes with the ideological whims of the government, the very basis of government decision-making, transparency and trust is shattered.

We need a chief statistician who is willing to stand up for Statistics Canada as an independent institution. Where is Munir Sheikh?

David Green, economics department, University of British Columbia, Vancouver

Milan July 22, 2010 at 10:00 am

This resignation reveals the tension that always exists within the federal government, between ministers who have legitimacy based on popular support and civil servants who have expertise and an obligation to be impartial.

When senior civil servants feel sufficiently strongly about something to openly defy the government, it is pretty clear that there is a major tension between expert advice and the government’s political instincts.

. July 22, 2010 at 10:11 am

Long or short, Tories must retreat on the census

John Ibbitson

Ottawa — From Thursday’s Globe and Mail Published on Thursday, Jul. 22, 2010 12:00AM EDT Last updated on Thursday, Jul. 22, 2010 9:33AM EDT

The Harper government’s refusal to listen to reason on the long-form census caused the chief statistician to quit Wednesday night – an extraordinary move – and left Statistics Canada in open revolt. Having gotten into this mess, the Conservatives must now retreat, or put the very future of the census itself at risk.

Chief Statistician Munir Sheikh’s letter of resignation, from a career public servant who has served in the Privy Council Office, in Finance and as deputy minister of Labour, is exceptional.

Clearly, Mr. Sheikh was not prepared to let the government order up a flawed census, while offering Statscan’s own official silence as proof that everything is fine.

City and provincial governments had already warned that replacing the mandatory long-form version of the census with a voluntary survey would effectively gut the census.

Every kind of industry, labour, academic and charitable organization had sent petitions, letters and press releases saying the same thing, backed by a truckloads of statisticians and economists.

Now, in his letter or resignation, the nation’s chief statistician offered the final verdict: On “the question of whether a voluntary survey can become a substitute for a mandatory census.

“It cannot.”

. July 22, 2010 at 10:26 am

Under the circumstances, I have tendered my resignation to the Prime Minister.

I want to thank him for giving me the opportunity of serving him as the Chief Statistician of Canada, heading an agency that is a symbol of pride for our country.

To you, the men and women of Statistics Canada – thank you for giving me your full support and your dedication in serving Canadians. Without your contribution, day in and day out, in producing data of the highest quality, Canada would not have this institution that is our pride.

I also want to thank Canadians. We do remember, every single day, that it is because of you providing us with your information, we can function as a statistical agency. I am attaching an earlier message that I sent to Canadians in this regard.

In closing, I wish the best to my successor. I promise not to comment on how he/she should do the job. I do sincerely hope that my successor’s professionalism will help run this great organization while defending its reputation.

Munir A. Sheikh”

. July 22, 2010 at 11:16 am

“Ian McKinnon, chair of the National Statistics Council, issued a statement saying Mr. Sheikh’s obligation to keep secret his advice to cabinet ministers “left him unable to defend his professional competence or respond to statements that tended to cast doubts on the professional competence of Statistics Canada.”

Mr. McKinnon was referring to the controversy about statements made by Mr. Clement in the last week that left the impression Statscan management was satisfied with the voluntary-form alternative for the next census.

“With Dr. Sheikh’s resignation, Statistics Canada, and indeed the nation’s statistical system, has lost the committed services of a man of integrity and honour,” Mr. McKinnon said.”

. July 22, 2010 at 3:34 pm

Media advisory: 2011 Census

July 16, 2010

OTTAWA — Statistics Canada is not in a position to answer questions on the advice it gave the Minister in relation to recent statements the Minister has made.

Milan July 22, 2010 at 3:36 pm

As noted by Kady O’Malley, this generic text has taken the place of the statement quoted above on the StatsCan website.

. July 22, 2010 at 3:39 pm

The long form long list

For those of you scoring at home, you can add the town of Smith Falls to the list of those opposed, alongside provincial governments in Manitoba, Ontario, Quebec and Prince Edward Island, representatives from the United Way, Canadian Labour Congress, Toronto Board of Trade, Canadian Nurses Association and Canadian Public Health Association, city officials in Edmonton, Calgary and Red Deer, Ottawa city council, former clerk of the Privy Council Alex Himelfarb, the chief economist of the Greater Halifax Partnership, the French Language Services Commissioner of Ontario, the executive director of the Société franco-manitobaine, the Canadian Medical Association Journal, the Canadian Jewish Congress and the Evangelical Fellowship of Canada, the Marketing Research and Intelligence Association, the Quebec Community Groups Network, the president of the CD Howe Institute, the Canadian Council on Social Development, the Association of Municipalities of Ontario, the director of Toronto Public Health, Mr. Census, the Statistical Society of Canada, the Federation of Canadian Municipalities, the Inuit Tapiriit Kanatami, the Canadian Marketing Association, the Canadian Federation of Francophone and Acadian Communities, the Executive Council of the Canadian Economics Association, the director of the Prentice Institute at the University of Lethbridge, the senior economist at the Canadian Centre for Policy Alternatives, the Canadian Institute of Planners, the Canadian Association for Business Economics, the co-chairman of the Canada Census Committee,, the Canadian Association of University Teachers and the former head of Statistics Canada. Not to mention, the man who was, until last night, Canada’s chief statistician.

. July 22, 2010 at 6:02 pm

Clement won’t back down on census

Industry Minister Tony Clement has dismissed growing calls for him to reverse his decision to scrap the mandatory long-form census, saying he and Prime Minister Stephen Harper are on the same page on the issue.

“There’s not a micron of difference of opinion between myself and the prime minister on this,” Clement told the CBC’s Rosemary Barton in an interview on Power & Politics with Evan Solomon.

During the interview from London, Clement said the government has taken a “compromise position” between privacy concerns and ensuring usable data from the next census in May 2011.

Clement’s comments came a day after Munir Sheikh, the head of Statistics Canada, the national statistical agency, resigned in protest over the move to scrap the mandatory survey.

. July 23, 2010 at 4:54 pm

An independent mind has always characterized Sheikh

Tavia Grant

From Friday’s Globe and Mail Published on Friday, Jul. 23, 2010 12:25AM EDT Last updated on Friday, Jul. 23, 2010 11:22AM EDT

Munir Sheikh navigated his career as a public servant through eight prime ministers over the course of nearly four decades. A string of successively more senior positions led him to oversee a $100-billion tax-reduction policy and help craft the 2005 budget. One former colleague described him as the best economist in the federal government.

In the end though, it was a questionnaire that toppled the 62-year-old.

It’s a stunning turn for a man who dedicated his working life to public service. Dr. Sheikh’s Wednesday night resignation as Statistics Canada’s chief statistician over the census is all the more remarkable because of its rarity. In a world where loyalty is king, bureaucrats of his standing do not tend to quit over differences of opinion.

He did. In doing so, he displayed qualities that have emerged through his 38-year career: stubbornness and independence of mind.

Indeed, debate was an intrinsic part of the daily fare at the Sheikh household in Pakistan, where the 10 siblings would tussle over economics, society, politics and culture.

That intellectual wrestling left the family’s second-youngest son, Munir, well prepared for the rigours of public service. It also honed his skills as an independent thinker, willing, to a point of stubbornness, to defend his arguments.

“He is a very strong individual and he holds his ground. If he believes he’s right, he’s right,” says his brother, Shamim Sheikh, a University of Toronto civil engineering professor. “Debate within the family environment is part of our fabric.”

Independence is at the core of extraordinary developments in Ottawa, which saw Dr. Sheikh resign from one of the world’s most respected statistical agencies. In doing so, he stood up for what economists, professors, city planners and historians have said for weeks: One cannot substitute the mandatory long-form census for a voluntary one.

. July 26, 2010 at 8:51 am

Harper’s census push months in the making

Michael Valpy

From Monday’s Globe and Mail Published on Monday, Jul. 26, 2010 3:00AM EDT Last updated on Monday, Jul. 26, 2010 8:00AM EDT

Prime Minister Stephen Harper decided at the end of December to scrap the mandatory long-form census despite being told by Statistics Canada officials that important data would likely be lost or impaired as a result.

He considered going further by making the whole census voluntary, people familiar with what transpired have revealed. On the long census form, he overrode objections from his own officials in the Privy Council Office and senior finance department staff, although Finance Minister Jim Flaherty said on the weekend that he thinks census data can be collected voluntarily without being compromised.

The government announced at the end of June that the long form would be voluntary in the 2011 census.

Opposition MPs will get their first chance to question Industry Minister Tony Clement, the minister responsible for Statscan, and former chief statistician Munir Sheikh, who has quit in protest against the change, when they appear before a parliamentary panel on Tuesday.

Mr. Harper’s decision has baffled political analysts familiar with his thinking – people including political scientist Tom Flanagan, who played a key role in Mr. Harper becoming prime minister – but not University of Calgary economist Frank Atkins, his graduate thesis supervisor.

. July 27, 2010 at 4:25 pm

Posted by Parker on 23 July 2010 at 14:41

Harper spokespeople argue that sending the voluntary census long form to a larger number of people will compensate for any loss of data quality due to the newly voluntary nature of the form.

. July 29, 2010 at 2:08 pm

John Ivison: Conservatives have a knack for picking small hills to die on

It’s just as well for Tony Clement that he helped save a woman from drowning last weekend — it’s likely to be the last good news story he features in for quite some time.

The Industry Minister’s testimony at a specially convened committee meeting on Parliament Hill on Tuesday will have won few converts to the government’s argument that scrapping the mandatory long-form census is a smart and necessary move.

The world’s most boring political scandal rumbled on, as MPs were recalled to hear from a long list of witnesses, who almost unanimously condemned the government’s plan.

Walking past the tourists and the marching bands on the lawn of Parliament Hill to cover the committee felt like a descent into the ninth circle of Hell. Inside a stifling, packed committee room, Mr. Clement made the case that the government should not threaten people who fail to fill in the census with jail time.

On this, he has a point. It is also conceivable that some people may find census questions about home repairs and spousal support to be “coercive and invasive,” as the Industry Minister suggested.

But the government’s solution — to make the mandatory form voluntary — has received such blanket condemnation — even from those who are normally staunch Conservative allies, such as seniors organization CARP — that the government is looking not so much out of touch, as out to lunch.

. August 9, 2010 at 3:02 pm

Why Stephen Harper thinks he’s smarter than the experts
By John Geddes | August 9th, 2010 | 10:53 am

For the government relying on academic research is bad politics

Painful experience lay behind Harper’s conscious move away from the influence of academic research. His former chief of staff, Ian Brodie, talked candidly about the transition at Montreal’s McGill University last year, in a panel discussion on the role of evidence in policy-making. Brodie recounted how Harper had run in the 2004 election on a tax-cuts platform carefully constructed along lines favoured by tenured economists. “We promised a comprehensive system of moving brackets around, cutting bracket rates, multi-year this, multi-year that, a corporate income tax cut as well,” he said. “A program so well thought out that even the people who wrote it can’t remember the details now.”

The Conservatives lost that election. The setback, Brodie explained, led Harper and his advisers to radically rethink their approach. By the 2006 campaign, Harper was pitching a simple idea, cutting the Goods and Services Tax, which was almost unanimously opposed by mainstream economists. But if experts would have overwhelmingly preferred reducing the tax burden on income and investment, voters liked the sound of Harper’s uncomplicated pledge to slash the widely resented consumption tax. That GST promise helped them win, and Harper’s team learned to treat conventional wisdom among specialists with a certain disdain.

On another key Tory policy theme—law and order—Brodie touted conflict with academics as good politics. Most university criminologists say there’s no evidence to back up the Tories’ heavy emphasis on imposing longer prison terms. They point to studies showing that more jail time doesn’t reduce crime. At the McGill panel, though, Brodie said voters tend to side with Conservatives when they argue with “sociologists, criminologists, defence lawyers and Liberals” about prison terms. “Politically, it helped us tremendously,” he said, “to be attacked by this coalition of university types.”

So not only do Harper’s advisers suspect that following expert advice leads to unsaleable policies, they also think battling the experts can boost their popular standing. In the census controversy they seem willing, almost eager, to take on virtually the entire Canadian research establishment. Among the many groups arguing for keeping the mandatory long-form census, which Harper is turning into a less reliable voluntary survey, are the Canadian Economics Association’s executive, the C.D. Howe Institute’s president, the Canadian Medical Association Journal, and the Canadian Institute of Planners.

. August 10, 2010 at 5:56 pm

The long shadow of the long census cancellation: A politician’s nightmare
By Duncan Cameron | August 10, 2010

Who would have expected Stephen Harper to be so foolhardy as to ignore the outcry? As of Monday, 272 national organizations and prominent individuals are listed as opposing the cancellation of the mandatory long census. The prime minister should have had enough confidence in himself and his government to admit a mistake, and move on. By standing by his beliefs instead, he has created a politician’s nightmare, a controversy that allows his opponents to grow support… by doing nothing.

The Conservative option — make the long-form questionnaire voluntary — has been described by a Laval University economist as an “appallingly stupid idea” because “there is simply no way any useful data will emerge from this exercise.” As well, the cost of the move is more expensive than the current census.

Why does the Conservative decision to spend more money on the voluntary survey fail the social science test? Statistical analysis requires that information be gathered and compared over time. Analysts are looking for trends, effects that can be traced to causes. Changing the census from mandatory to voluntary requires charting trends based on two different methods for collecting data. With incompatible data sources, the results are unreliable, and charting trends accurately becomes impossible.

The Harper Conservatives have made no secret of what they believe. There are those who think that government can act on behalf of the people to better their existence, and those (Conservatives) who think government best respects individuals when it acts least. Harper clearly has decided his government only acted in accordance with his minimalist role for government views when it decided to abandon the long census.

. August 11, 2010 at 2:20 pm

Federal Court to fast-track census challenge

OTTAWA — Federal Court Justice Roza Aronovitch has agreed to fast-track legal proceedings against the Harper government to reverse its decision to scrap the mandatory long-form census.

A federation representing Francophone and Acadian Canadians launched the challenge this summer arguing that the government’s decision contravenes the Official Languages Act and its obligations to provide services for minority French-speaking communities across the country.

The data from long-form census questions on language used in the workplace and at home help determine which regions and offices require services in both official languages.

. August 26, 2010 at 1:47 pm

Globe editorial
Canada goes off the statistical standard

The leading science journal Nature is right to criticize Canada’s abandonment of the mandatory long-form census

From Thursday’s Globe and Mail Published on Wednesday, Aug. 25, 2010 8:00PM EDT Last updated on Thursday, Aug. 26, 2010 8:56AM EDT

The mandatory long-form census is the “gold standard” in government information collection, say two scientists in an opinion piece published this week in Nature, an influential British scientific journal. It follows that Canada, which is replacing the mandatory part of the census with a voluntary one, is choosing some lesser standard. Doesn’t Canada need the gold standard? Shouldn’t it want it?

Apparently not. The country that set out to Own the Podium at the Vancouver Olympics (and succeeded in winning the most golds) is content to have inferior data collection. The Conservative government has announced that the mandatory long-form sent to 20 per cent of Canadians and with detailed questions about work, home life, religion and ethnicity is being replaced for the 2011 census with a voluntary questionnaire sent to 30 per cent. It’s as if the country suddenly lost its ambition and aimed for a “personal best.”

No one thinks the substitute is the gold standard. The head of Statistics Canada, Munir Sheikh, resigned to correct a misimpression fostered by Industry Minister Tony Clement that he believed it was. Wayne Smith, now the acting chief statistician, calls the voluntary survey “usable and useful,” but says it “will, of course, never be comparable to census data.”

. September 9, 2010 at 4:04 pm

“A June, 2010, internal study obtained by The Globe and Mail under the access-to-information law offers an inside look at how new census-taking rules could skew data in a range of areas from housing to demographics.

Statistics experts warn its findings demonstrate how minorities and groups such as renters could be measurably underrepresented or miscounted in the coming 2011 census.

The Statscan study, Potential Impact of Voluntary Survey on Selected Variables, was prepared with full knowledge of Ottawa’s census change plans. It attempts to re-create what the 2006 long-form census questionnaire would have yielded had respondents not been compelled to answer it. These simulated results were contrasted with real 2006 census data.

Statscan researchers found the voluntary approach produced less accurate results – a problem that was especially significant in small population groups, according to outside statistics advisers who reviewed the report for The Globe.

For instance, the real 2006 census long-form found that renting households as a percentage of the population in Canada had dropped by 3.08 percentage points from the 2001 census.

But when the Statscan study simulated the results of a voluntary 2006 long-form – which reflect the lower response rates expected in optional surveys – it got a markedly different answer. Calculations instead indicated that rented dwellings in Canada as a share of the population declined by 8.07 percentage points from 2001.

The difference – nearly five percentage points – suggests a voluntary survey in 2006 would have massively undercounted renting households.

. October 7, 2010 at 11:22 am

Court rejects francophone challenge to Canada census

Canada’s Federal Court has rejected a French-speaking group’s attempts to stop the government scrapping the mandatory long-form census.

Prime Minister Stephen Harper’s Conservative government has recently come under fire for making the next federal long-form census voluntary.

The Federation of Francophone and Acadian Communities argued the move violated the Official Languages Act.

But a judge said the act did not state how information should be collected.

. February 16, 2011 at 10:24 pm

Other countries do things different ways, Wayne Smith, Stats-Can’s newly appointed chief statistician told the Globe and Mail on Saturday. For example, some northern European countries don’t have a census. They collect data from national registries. “The government wants to step back and say, OK, ‘Let’s look at those other models: what is possible in Canada?’ ”

This is funny stuff. If you have a black sense of humour.

See, what makes registry-based data collection work is — wait for it — mandatory reporting. And not just the modest level of mandatory reporting required by the census. No, with registry-based data systems, every person is given a government ID number, required by law to register a raft of personal data, and further required to constantly notify the government of any change in address, job, family relations, car ownership, etc. And all this information is linked and centralized.

In short, anyone who thinks the mandatory census is unacceptably intrusive should find this positively Communist.

But Stephen Harper has to replace the television he smashed with something. And it can’t be the TV he smashed, after all. That would make him look stupid.

Ladies and gentlemen, it’s my job to be blunt. So I’ll just say it: Stephen Harper is incompetent.

I know that’s not his reputation. Lots of people accuse him of being ruthless, or an ideologue, but he’s usually credited with being a basically competent manager.

He doesn’t deserve that credit. His government is badly run and incoherent. Promising fiscal conservatism, Stephen Harper spent money like crazy, expanded the federal government, cut taxes, and turned a surplus into a structural deficit (yes, it’s structural, as even the International Monetary Fund agrees). He has no real plan for getting the budget back into balance.

Most of Harper’s key tax policies were horribly designed (economists opposed cutting the GST and want to see the tax code simplified, not larded up with micro-credits). On climate change, he favours command-and-control regulation, not market-based policy (contrary to most expert opinion). His government doesn’t even pretend to have evidence that its mandatory minimum sentences reduce crime. (Heaps of evidence says they don’t.) His handling of foreign investment issues has been so arbitrary and political that UBS Investment Research recently warned foreign investors “may begin to perceive Canada as not ‘open for business.'”

Harper said setting a deadline to end the mission in Afghanistan would endanger the troops. Then he set a deadline. Then he agreed to a new mission without even consulting the military.

. July 7, 2011 at 11:33 pm

Statscan settles for incomplete long-form surveys in 2011 census

Jennifer Ditchburn
Ottawa – The Canadian Press
Last updated Wednesday, Jul. 06, 2011 6:19PM EDT

Census workers are settling for incomplete long questionnaires in the final push of the summer collection period, raising concerns the data will be even more compromised than originally feared.

The new, voluntary National Household Survey was the controversial replacement for the long-form census, eliminated last summer by the Conservative government.

The Tories said it wasn’t right to threaten Canadians with jail time or fines for not answering the detailed questions on everything from religion to education levels.

. September 21, 2011 at 12:34 am

Sheikh’s version: Ex-chief statistician picks apart cancellation of long census

By: Heather Scoffield, The Canadian Press

09/20/2011 5:26 PM

OTTAWA – The federal government cancelled the long-form census with little heed to the consequences of its decision, according to a new first-hand account of the drama that unfolded a year ago.

An essay by former chief statistician Munir Sheikh says the census decision has shaken Statistics Canada’s neutrality and independence, and put at risk the government’s own work in many areas.

In the essay, Sheikh warns statisticians working at the federal agency to “guard against political intervention” until better solutions are found.

Sheikh also raises concerns over poor data on aboriginal populations, especially housing on reserves, and about the government making key decisions on pension reform without having reliable information on wealth in Canadian households.

But it’s the cancellation of the long-form census that compounds the weakness in other data, he writes.

. December 10, 2011 at 10:46 pm

Public service, government need “moral contract” to stop partisan exploitation of bureaucracy

OTTAWA — Canada needs to set ground rules for a new “moral contract” between ministers, public servants and Parliament because the existing rules are too weak to stop the partisan exploitation of the bureaucracy, says a former senior bureaucrat who helped write some of those rules.

Ralph Heintzman, a research professor at the University of Ottawa, is calling for a new charter for the public service to set boundaries for a bureaucracy operating under a powerful prime minister’s office that’s obsessed with communications control — what’s known in academic circles as new political governance.

He argues the Conservatives’ centralized communications command is riding “roughshod” over the federal communications policy, breaches ethics guidelines and risks turning the independent and non-partisan public service into a propaganda arm of the government. He said the deputy ministers, including the country’s top bureaucrat, who aren’t stopping it are also violating the codes.

. September 19, 2015 at 5:14 pm

The Canadian Conservative government’s war on science, statistics and evidence have been a great boon to its ability to create policy that helps its friends and destroys the country, but the deep and arbitrary cuts to science and statistics have eroded Canada’s ability to know what is happening in the country to a terrifying extent.

Even as other countries are moving to “open data” as a matter of course in government policy, Canada has embraced a deepening, suspicious secrecy. Data that used to be open is gone — but ministers like the outgoing Tony Clement praise the openness of government with a straight face.

Reading Anne Kingston’s story in Macleans was like being on the receiving side of a boxer’s speed-bag practice, one punch after another, one bad policy after another. I pasted in a very small sample below, but if there’s one article you read in full this weekend — whether or not you’re in Canada — this is the one. If you want to understand how a digital transition in government can be a stealth attack on the very idea of truth and knowledge, read this.

Canada has a new underground of scientists and statisticians and wonks who’ve founded a movement called LOCKSS — “Lots of copies, keep stuff safe” — who make their own archives of disappeared data, from the libraries of one-of-a-kind docs that have been literally incinerated or sent to dumpsters to the websites that vanish without notice. There’s an election this October — perhaps we can call on them then to restore the country’s lost memory.

Leave a Comment

Previous post:

Next post: