Canadian research study hits snag over lousy population data – The Globe and Mail

2015/05/08 1 Comment

Good piece on some of the limitations of the National Household Survey, particularly at the Census Tract level.

Nice comment on the irony between more corporate big data, with all the related privacy implications, and less effective government big data (although linking the NHS to CRA income information will improve quality and is appropriate use of big data):

The project, produced with the financial backing of the Maytree Foundation, used 2010 to 2013 data from the Canadian Community Health Survey, a voluntary annual survey with a sample size of 65,000. The project didn’t use the last voluntary national household survey due to difficulties in comparability and in assessing data quality for smaller communities.

The researchers want to see the mandatory long-form census not only reinstated, but expanded to include more questions on wealth and health. They recommend more sharing of information between federal government departments and more tools, such as online searchable databases, to make data more accessible and useable.

The irony is that the lack of data on the public side comes as Big Data is giving private firms unprecedented access to rich details about customers’ lives.

“The trend in the private sector, for companies like Google and Uber, they’re making brilliant decisions, strategically, because they’re collecting more and more information, mining it and refining their services and products,” said Mr. Johal.

“And what we’re seeing in Canada is governments stepping away actively from getting that information – so that makes it really hard for us to make those smart decisions and invest in our future.”

Canadian research study hits snag over lousy population data – The Globe and Mail.

Filed under Government Tagged with , ,

Something of a compromise on the census?

2015/04/20 Leave a comment

This is a significant improvement, given that income reporting through tax returns should be more accurate than self-reporting. Of course, the broader issues related to the NHS versus Census remain:

But StatsCan has made one notable change to the process.

. . . in order to reduce the time required and to make it easier for Canadians to respond to the National Household Survey, income questions will be replaced with more precise tax and benefit data that have been available to Statistics Canada since 1985. As this will be done for all Canadians, income information for 2016 will be the most accurate in the history of the census.

What does this mean?

Within the 2011 national household survey were various questions about income. But respondents were also given the option to skip those, if they agreed to let Statistics Canada access information already provided in their tax filings.

… In addition to saving respondents time, StatsCan says it will save the agency money. In the case of 2016, income data will be linked to census responses.

Liberal MP Ted Hsu had proposed a private member’s bill to, among other things, reinstate the long-form census. That bill was defeated in February, but, at the time, he told me about a possible compromise: adding some number of questions to the short-form census to provide a better statistical basis. UBC economist Kevin Milligan dubbed this the “medium-form census” and Milligan says he’s somewhat pleased with the change. “I think it is a step toward a medium-form census,” he says. “If I couldn’t have a long-form census, and I was asked for one change to the short form, this is what I would have asked for.”

Hsu would still rather have the long-form census—something the Liberals are committed to reinstating if they form government—and notes that the voluntary nature of the NHS will still create problems. Hsu also thinks it would be useful to link tax and benefit data to the NHS so that income could be correlated with dwelling, education and labour market information. Milligan would also rather have income data linked to the NHS, but says, “Having it for short form gets you pretty close to the tools you need to make some decent weights that make all the other surveys (like NHS) more useful.”

And though Hsu doesn’t think the use of tax and benefit information is a big deal, he does think there needs to be a conversation about possible privacy implications if the government moves further to use administrative data that it already possesses. Back in 2010, when debate arose over the government’s decision to eliminate the long-form census, the Scandinavian model of data collection and use raised as a possible alternative, but the databases maintained by those countries might raise questions for Canadians about the handling of personal information.

Something of a compromise on the census? – Macleans.ca.

Filed under Government Tagged with ,

Did losing the long-form census weaken Canada’s jobs data?

2015/02/18 Leave a comment

toronto-census-nhs-756x1024More on the Census and National Household Survey, and Kevin Milligan’s analysis of how the shift has degraded the quality of the Labour Force Survey:

I say ‘presumably’ because the most recent complete methodology document I can find is from 2008. In that 2008 methodology document, it is very clear that the Census plays a very important role in the methodology of the Labour Force Survey. In fact, a quick ‘control F’ search of the document reveals the word ‘census’ to appear 127 times. Looking through these instances, you can see the Long-Form Census was used in a variety of ways. Sometimes, it was to cross-check an assumption or a decision they made in designing the Labour Force Survey. In other places, it is clear they used the Long-Form Census explicitly to pick which households get surveyed. To give one precise example, page 19 of the methodology document explicitly references income taken from the 2001 Census—and income is not available on the Short-Form Census.

In short, I think Mr. Smith’s ‘myth’ is a miss.

I have complete confidence in Mr. Smith’s claim that the 2015 Labour Force Survey methodology now uses only the Short-Form Census for selecting and weighting households. However, it is equally clear that until now the Labour Force Survey relied on the Long-Form Census. This change raises several important questions: Why make a change? Why was the National Household Survey discarded? Is the Labour Force Survey improved by ignoring the National Household Survey? Why and how?

I have a pretty good guess why Statistics Canada discarded its own expensive National Household Survey for designing the Labour Force Survey. The reason: the response rate from the National Household Survey is low, and varies strongly within regions. As one example, above is a map made by Dwight Follick that compares the response rate at the ‘dissemination area’ level for Toronto from the 2006 Long-Form Census and the 2011 National Household Survey. Similar maps are available from Mr. Follick for Montreal, Ottawa, London, Calgary, Edmonton, and Vancouver. The pictures aren’t pretty—the response rate fell dramatically, rendering the results of the National Household Survey much less reliable.

Mr. Smith might be correct that the Labour Force Survey methodology changes aren’t large enough to make much difference. But we will never actually know, since we can’t compare the new results to the previous methodology because the 2011 Long-Form Census doesn’t exist so we can’t check.

I will continue to use and trust the Labour Force Survey, myself. However, the switch to the National Household Survey has degraded some of the tools that Statistics Canada has used until now to make the Labour Force Survey—and other Statistics Canada products—reliable.

Did losing the long-form census weaken Canada’s jobs data? – Macleans.ca.

Filed under Government Tagged with , , ,

Cities to weigh loss of long-form census for community planning

2015/02/05 Leave a comment

Yet another group weighing in on the ongoing implications and costs associated with replacing the Census with the National Household Survey:

Across the country, cities are feeling the impact of the census changes, said Brad Woodside, president of the Federation of Canadian Municipalities and mayor of Fredericton.

“We’ve heard from our members that the change to the new National Household Survey is impacting their ability to effectively plan and monitor the changing needs of their communities,” he said in an e-mailed statement to The Globe. “We support all efforts to increase the reliability of the data from the census.”

Local governments rely on this information to understand the changing needs of communities, and make a range of decisions, “from where to establish new bus routes, build affordable housing and provide programs for new Canadians, he said. “We continue to call on Statistics Canada to work with municipalities to provide communities of all sizes the most reliable information from the available data.”

Mr. Tory said he will raise the issue with the mayors of the country’s largest cities when they meet in Toronto later this week. The topic is not on the agenda for the gathering, which begins Wednesday evening, but he said he can bring it up in informal discussions.

“I believe you really should try to have the best possible evidence in front of you when you are making important decisions,” he said Tuesday. “I can ask if this is a problem they are facing.”

Cities to weigh loss of long-form census for community planning – The Globe and Mail.

Globe editorial makes the same point but equally unlikely to have much effect:

There is now incontrovertible evidence the Conservative government’s 2010 decision to scrap the mandatory census questionnaire, which quantified everything from family income to ethnicity to regional demographics, was an unalloyed catastrophe.

Opposition has come from think-tanks of every political persuasion, business leaders, charities, public administrators and basically anyone with a PhD. Thanks to a deliberately sabotaged census, we know less about Canada in 2011 than we did about Canada in 2006. Who thinks that’s a good idea?

What’s more, conducting a halfwitted census turned out to be more expensive. The 2011 voluntary household survey increased errors, reduced accuracy, chopped the response rate by 30 per cent – and cost an extra $22-million. Congratulations: The Harper government figured out how to spend more for less.

The decision to kneecap the census was transparently ideological, a rash exercise in partisan narrow-casting, and was quickly exposed as such.

Dozens of experts predicted the damage that would be wrought. It’s time for the Conservative government to finally acknowledge how right they were.

The next opportunity for the House to revisit the Census Act will come next month via another private member’s bill – this one tabled by Conservative backbencher Joe Preston.

It would remove two aspects that are problematic to some Conservatives: jail for refusal to complete the form, and automatic public disclosure after 92 years.

There is still resistance in Mr. Preston’s party to bringing back the mandatory long form. We hope that removing the central justifications for killing it represents an evolving mindset.

Some mistakes are easy to reverse. It may be too late to restore a proper census in 2016, but a return in 2021 should be inevitable.

The census: Little knowledge is a dangerous thing – The Globe and Mail

Filed under Government Tagged with , ,

Long form census: Duelling backbencher bills revive House debate

2015/02/04 Leave a comment

Continued triumph of ideology over reason, the Government’s refusal to reinstate the mandatory census:

The response rate for the 2006 long-form census was 93.5 per cent, compared with 68.6 per cent for the voluntary National Household Survey that replaced it in 2011.

Statistics Canada withheld information on thousands of smaller Canadian communities because the information was unreliable.

The census tract of Elgin, in Preston’s southwestern Ontario riding, had a non-response rate of 26.1 per cent for the National Household Survey.

Critics say the problems with the data are compounded by the fact that the survey results cannot be compared with the results from the mandatory censuses going back many decades.

Hsu said filling out the census forms is a civic duty, just as Canadians have a duty to pay their income taxes.

“The fight over this bill is a fight over the soul of this country,” Hsu told MPs last week.

“It is a fight over whether Canadians should collect information about ourselves so that we may have solid evidence with which to govern ourselves wisely.”

Long form census: Duelling backbencher bills revive House debate – Politics – CBC News.

Filed under Government Tagged with ,

Damage from cancelled census as bad as feared, researchers say

2015/01/30 Leave a comment

The impact of an ideologically motivated decision, impacting both social research (the intended target, as suggested by Paul Wells in his book The Longer I’m Prime Minister) as well as the business community and municipalities who use census data for planning purposes (e.g., store and school locations):

“It has certainly impacted my own work on what has been happening to middle-class earnings in Canada,” says Charles Beach, professor emeritus of economics at Queen’s University.

More broadly, it has “inhibited research into inequality and identifying winners and losers in economic growth, research into understanding the national problems of the have-nots in the economy, and research into how best to provision local government services.”

In the private sector, the Canadian Chamber of Commerce, whose network represents 200,000 businesses across the country, is publicly calling on the federal government to restore the mandatory long-form census.

Some researchers – such as those working on a sweeping long-term study on income polarization in Canadian cities – are choosing to abandon using the NHS altogether. They may be settling for less-detailed tax-filer data, while others, such as some public health units, are still using outdated 2006 census data.

In Canada’s largest city, “it has definitely had an impact in the way we plan for services” for people such as seniors, single parents, youth and immigrants, says Harvey Low, manager of social research at the City of Toronto. “We are less sure ” about the characteristics of people served in communities.

Damage from cancelled census as bad as feared, researchers say – The Globe and Mail.

Filed under Government Tagged with , ,

Ending mandatory long-form census has hurt Canada – Globe Editorial

2014/11/08 Leave a comment

The Globe on the ideologically driven decision to cancel the Census and the private member bill to restore it:

The warnings were prophetic. The compulsory long-form census in 2006 had a 93.5 per cent response rate. The voluntary one in 2011 had a 68.6 per cent response rate, even though more surveys were sent to more homes. When the 2011 data were released, they came with prominent warnings about contamination due to “higher non-response error.” Information gathered about more than one quarter of all Canadian communities wasn’t released because too few people in those places filled out the voluntary form. Aboriginal communities were particularly underrepresented.

Think-tanks, economists, scientists and academics in Canada and around the world have dismissed the 2011 data as fatally flawed. It can’t be compared in a meaningful way with the 2006 data, because they were gathered using different methodologies. Vital research projects on issues like income, unemployment and poverty that require long-term data have been compromised. And Statistics Canada can’t provide an accurate picture of how Canadians are faring, relative to 2006, since the 2008 economic crash.

Statisticians are statisticians so we don’t have to be. If they say they need accurate, regular, comparable census, then that’s what they should get from the government. Mr. Hsu’s bill may be doomed, but it will go down fighting to reverse a decision that has harmed the country in tangible ways.

Ending mandatory long-form census has hurt Canada – The Globe and Mail.

Filed under Government Tagged with , , ,

Policy making suffering in Canada without the long-form census

2014/11/07 Leave a comment

Unfortunately, the Government will not take advice to reinstate the Census:

Perhaps the biggest casualty of the switch to the new survey is the ability to analyze trends over time – among the most critical components of any research tool. The household survey and the long-form census are so different that we are no longer able to compare different periods in a statistically rigorous way.

We see nothing wrong in requiring Canadians by law to complete a survey as important as this one. Even in the U.S., where trust in government is not exactly high, the American Community Survey is mandatory. The authorities have reasoned – and few citizens have objected – that a mandatory response is the only way to ensure adequate data quality.

If the government in Ottawa can be persuaded to bring back the mandatory long-form survey for the next census, due in 2016, we will have a gap of 10 years since the last such exercise.

That may not be ideal, but it would be acceptable. The full census was conducted at 10-year intervals prior to the introduction of the current five-year cycle in 1986. Indeed, a 10-year break would be less disruptive than continuing with the new household survey, which leaves us with a complete break in historical data.

Prime Minister Stephen Harper said recently that “you can’t manage what you can’t measure.” His comments were made in the context of child and maternal health, but they apply equally to other areas of policy making, and drive better performance in business planning and economic analysis.

My experience in working with the NHS and in talking to those more experienced than me, confirms these weaknesses of the NHS compared to the Census.

A short-sighted and ideological move that weakens evidence-based policy for governments and needed information for business decisions.

Policy making suffering in Canada without the long-form census – The Globe and Mail.

Filed under Government Tagged with , ,

StatsCan considering ‘virtual census’ to replace head count | Ottawa Citizen

2014/09/02 Leave a comment

Makes sense, in terms of efficiency and potentially accuracy (i.e., income data from tax forms may be more reliable than self-reporting).

Ironically, given that it is essentially a “big data” approach, it could be more intrusive than the former mandatory census that the Government abolished citing its intrusiveness (and not mentioning less than a handful of complaints).

Information sharing and privacy will be the harder policy issues.

But good that this is being considered:

A virtual census relies heavily on administrative data: giant caches of information collected by government in the regular course of business.

Statistics Canada already uses 500 databases drawn from federal, provincial and municipal governments, and the private sector. Among many other things, those data files provide information on individual incomes taxes, corporate taxes, payroll deductions, employment insurance, building permits, births and deaths, even telephone bills.

A handful of European countries — Finland, Holland, Sweden, Denmark and Germany — have scrapped their survey-based censuses in favour of population counts that rely heavily on an amalgam of administrative data. Sometimes, that data is combined with information from sample surveys.

“If you look around the world, particularly in Europe, you’ll see many countries have moved away from going out with a survey the way we do,” Smith told the Citizen.

“They’ve said there are data files available that we can use to effectively count the population: we don’t need to spend hundreds of millions of dollars and bother every household in the country to count the population.”

StatsCan considering ‘virtual census’ to replace head count | Ottawa Citizen.

Filed under Government Tagged with , ,

Non-profits benefiting from data access – Post-Census

2014/05/16 Leave a comment

Good piece in the Globe on the use of data at the neighbourhood level and how the private sector has partially filled the gap caused by the cancellation of the Census and its replacement by the weaker National Household Survey (which does not offer the same level of granularity – the Census went down to the 250 household level):

“We’re not saying we didn’t need the mandatory census or that these data would be as good as if Statistics Canada had done a mandatory long-form census, but businesses absolutely rely on income and ethnicity data for small areas and Statscan didn’t release them,” said Jan Kestle, president of Environics Analytics.

“It’s easier to do when you’re only five years out from a [mandatory long-form] census. In five years time, we’re going to either need more mandatory questions or we’re going to need better access to good quality administrative data.”

Most people use the company’s data in conjunction with a mapping tool and segmentation analysis, which sorts the population into lifestyle categories such as “Middleburg Managers” and “Young Digerati,” to better understand their habits and tastes. A library, for example, found that despite having a large population of senior citizens, programs advertised to “seniors” were a bust. Having looked more closely at their income and lifestyle data, they targeted the same group as “mature adults” and had much more success.

“Often, the real power is in the melding of the data. They know things about their users, but not their neighbourhood, then they marry them,” said Doug Norris, chief demographer at Environics Analytics.

Non-profits benefiting from data access – The Globe and Mail.

Filed under Government Tagged with ,

← Older posts

Newer posts →