[ SkipToMainMenu ]

Bill C-36: An An Act to Amend the Statistics Act - Third Reading

Honourable senators, I rise today to speak at third reading as critic of Bill C-36.

The stated intent of Bill C-36 is to strengthen the independence of Statistics Canada and, in particular, the independence of Canada’s Chief Statistician. It also seeks to improve the way in which Statistics Canada collects data. These are laudable goals, and I do support the intent of the bill.

The extent to which the bill achieves all of its stated goals, however, is a matter of debate.

The role of Statistics Canada is to produce statistics that help Canadians better understand our country. It has been said many times during this debate that statistics are a public good, and I wholeheartedly agree with this. Indeed, many of us rely on the data produced by Statistics Canada in our work as senators. The vital data collected by Statistics Canada is also utilized by policy-makers, academics and economists, enterprise and industry, and local communities and interest groups.

Simply put, the need for accurate, reliable data, produced using transparent means, cannot be overstated.

Bill C-36 seeks to enhance the independence of Canada’s Chief Statistician by enshrining into law his or her decision-making authority over operations and statistical matters. It attempts to provide this independence through the following provisions.

First, it will be more difficult for the government to remove the Chief Statistician during their appointment. Second, the term of the Chief Statistician is limited to five years with the option to renew one time, a change from the current practice of no term limit. Finally, if the minister chooses to implement a method, procedure, operation or statistical program that the Chief Statistician does not agree with, a directive must be issued and tabled in both houses of Parliament, allowing for public debate.

Unfortunately, during our hearings, many witnesses demonstrated that there remains a gap between the legislation’s goals and its proposed solutions.

The primary concerns relate to how the Chief Statistician is appointed. The process has not changed from the current practice of a Governor-in-Council appointment. Dr. Ivan Fellegi, who served as Chief Statistician from 1985 to 2008, describes the process as follows:

It was like any other deputy minister. Somebody was appointed and God only knows how.

This, in the words of Wayne Smith, the Chief Statistician of Statistics Canada from 2010 to 2016, makes the appointment process contained in Bill C-36 egregiously flawed. He said:

Where it fails, and critically so, is in the process of selection of the Chief Statistician. Bill C-36 sets down no provisions, no requirements for the selection process itself. This is somewhat surprising from this government given that, in opposition, it tabled private member’s bills that featured a clearly prescribed process for the selection of the Chief Statistician as a key element in instituting his or her professional independence. . . .

The government is saying “trust us,” but it could have said as much for any provisions of Bill C-36. The selection process is the most fundamental provision. What point is there in protecting the professional independence of the Chief Statistician if the Chief Statistician can be selected based on his or her willingness to do the bidding of the government?

Dr. Fellegi added:

I strongly urge you, in a case of vacancy, to consider requiring the establishment of a search committee of eminent and appropriately knowledgeable people for the purpose of searching for and putting forward to the Prime Minister a short list of qualified persons.

John Pullinger, the United Kingdom’s National Statistician, contrasted Canada’s approach, which is contained in Bill C-36, with the multiple-step process that is used in his own country.

He explained:

. . . I am appointed by the Queen, which puts it above politics, but I’m appointed according to the terms and conditions set down by my board, which is separated, again, from any kind of political interference. . . .

We have public appointments commissioners, an independent group that deals with public appointments in the U.K. I went through that process. There were seven different stages of it with different professional and administrative elements to it. The final panel that appointed me had a mix of people. There was the chair of my board. There was the chair of our audit and risk committee. There was the head of the civil service, the head of the treasury department and the civil service commissioner. They were the five people who were making that recommendation.

Although Mr. Pullinger was not required to appear before a parliamentary committee before his position was ratified, the chair of the United Kingdom’s Statistics Authorities appointment is ratified by Parliament. If the intent of Bill C-36 is to make the appointment process more independent, perhaps the government should have looked to follow the U.K. model more closely.

Finally, Mel Cappe, Clerk of the Privy Council from 1999 to 2002, observed that the renewability provision in Bill C-36 for the Chief Statistician’s term is in conflict with the notion of independence. He said, “Renewability makes the person, arguably, more subject to the government of the day.”

For these reasons, the committee appended an observation to Bill C-36. It reads as follows:

During the committee’s hearings on Bill C-36, An Act to amend the Statistics Act, many witnesses expressed concern regarding the appointment process of the Chief Statistician.

These concerns included the term of the Chief Statistician being renewable, the lack of Parliamentary approval and the absence of a search committee.

Therefore, the committee urges the government to consider using tools including Executive Search Committees or Parliamentary approval to ensure the Chief Statistician is a non-partisan appointment who is independent from the government.

I sincerely hope that the government takes this observation seriously and ensures that the appointment process of the Chief Statistician is in keeping with the tenets of independence and non-partisanship.

However, even if the government follows this observation, we should not be under the false illusion that Statistics Canada will now be insulated from government intervention and completely error-proof by extension. This is a misguided concept. Statistics Canada remains an instrument of the government by design and purpose.

For this reason, Philip Cross, Senior Fellow at the Macdonald-Laurier Institute, noted that the fixation of the government on independence may be misguided. Mr. Cross explained:

. . . based on 36 years of working at StatsCan, I would argue there has been excessive emphasis on the virtues of independence and not enough on the importance of accountability.

For example, he highlighted an error in the 2016 census that led to an unusually high increase in the number of anglophones in Quebec communities.

Mr. Cross observed that increased independence would not have prevented this error and that the reputation of Statistics Canada is damaged when unreliable data is produced, saying:

If ever the public or users lose confidence in the accuracy of StatsCan data, it would take years to regain.

Another example of the fallibility of Statistics Canada also presented itself in the 2016 census when the population of the Canadian Jewish community was reported to have shrunk by 56 per cent between 2011 and 2016. This blatantly false finding indicates a failure by Statistics Canada to obtain accurate data through its own choices of methodology.

These two examples demonstrate errors that are not related to the independence of the institution. A wall between the political arm of government and the methodological branch of Statistics Canada does nothing to address these problems and in fact may contribute to them.

Rather than increased independence, these errors suggest the need for increased oversight at Statistics Canada.

Unfortunately, this government has made a major amendment to the Statistics Act that eliminates the National Statistics Council, a body that provides advice on the full range of Statistics Canada activities, and replaces it with the statistics advisory council.

Unlike the current council which has a membership of 40 individuals that can easily have representatives from each province and territory, the new council only has a membership of 10. This guarantees at least three provinces or territories will have no representation on the council, although as we heard in testimony, there is no intention to have regional representation on this body at all. As the minister himself told us at committee, “There could be four people from P.E.I. That’s the beauty of this process.”

It is regrettable that the government chose not to create a statistics advisory council with representation from all provinces and territories.

I am concerned that this change in Statistics Canada’s oversight body will cause even more methodological and process issues moving forward, resulting in data being more error prone. If any region is left off the council, we can only imagine the errors in the 2021 census that might occur, similar to those that occurred in 2016.

In closing, honourable senators, while Bill C-36 purports to make wholesale changes to Statistics Canada in the name of independence and transparency, once you dig into the details of the bill, it appears to change little about the way that Statistics Canada functions.

The government continues to have the unilateral ability to appoint whoever it wants to the position of Chief Statistician; the government continues to write the questions for the national census; the statistics advisory council will not represent all of our provinces and territories; and there is a lack of clarity regarding whether Canadians will be able to retroactively consent to their data being released after 92 years.

During Senator Cordy’s third reading speech she indicated that the Minister of Innovation, Science and Economic Development is considering further amendments to the Statistics Act as early as this coming year. Given that the amendments contained in Bill C-36 are not urgent, the minister’s approach to bring forward multiple pieces of legislation to amend the Statistics Act over a short period of time appears to be haphazard.

I believe Parliament would have been better served with a single piece of good legislation, rather than a situation where the minister will come back and clean up some of the mistakes and omissions contained in Bill C-36.

I encourage all senators to read the committee testimony and consider whether this legislation achieves the goals it sets out to achieve. For the reasons outlined today, I am confident you will determine that Bill C-36 comes up short.

Back to: Speeches and Statements