The following is a transcript of comments from Wayne Smith, chief statistician at Statistics Canada. He spoke to The Globe and Mail in a lengthy sit-down interview in Ottawa last week. Comments have been condensed and edited.
On emerging priorities at the agency:
- Consumer price index. “We're making a number of investments to improve the quality of the index. For example, we are expanding the sample, measuring some things more frequently than we used to, and we're introducing the new weights based on changing consumer patterns more rapidly than we have in the past. It's all in the effort to make sure that all of these decisions that are being made based on the consumer price index are being based on very solid statistical information.”
- Financial security. “We have good measures of income distributions, wealth is an issue in Canada, the measurement of wealth. We have a vehicle, the survey of financial security that we run periodically. The last time was early 2005, but on a reduced sample. So we've made a commitment to conduct a new round of that survey in 2012, and release it in 2013. There's no permanent provision in our budget for it...we'd like to get it on a five to six-year cycle.”
- Retirement. “We'll get some [information]from the financial security survey, but we're very interested in what's going on with regard to defined benefit and defined contributions and what impact that has on the incomes of the retired, and decisions to retire. We are paying attention to that dimension. Among demographic groups – it's people transitioning from work into retirement. I think we're going to see some very new phenomena there, the old notion where you're working one day and you're retired the next is gone. We'll see people with continuing attachment to the workforce, for all kinds of reasons. We need to be able to monitor that.
- The environment. “It's an area where we think we're not providing adequate information to the Canadian population and we'd like to do better. The problem at the moment is to develop a conceptual framework...what are the measures that we require, to what extent is the information currently available and how can we put it together in ways that are helpful for Canadians to understand the relationship between the environment and the economy – and how the environment is changing over time, for better or for worse. That's not there now, and it's an area where we're giving some priority too...the idea is that we are going to continue to invest our resources going forward into improving environmental statistics. We're interested in absolute measures of air quality or water quality or the extent of pollution, and we're interested in the connections between environment and health. We've created an advisory committee on environmental statistics, so we're going out and talking very broadly to all the stakeholders [including federal and provincial governments as well as academics]to try to get some sense of how we should prioritize this.”
On possible budget cuts at the agency:
"Given the government's fiscal situation, I have no expectation that Statistics Canada gets a pass...My expectation is that we will be asked to make some reductions in our program."
On how and where it would reduce costs:
"We will make sure that what we're doing is of high quality. We would rather reduce the scope of the program in order to maintain the quality of what we're producing, than degrade the quality in order to maintain the scope of the program.
"We're not going to bring the quality down to a level that it would compromise the uses and the decisions that are being made on the basis of this data."
On lower priorities at the agency (or what types of surveys could be cut):
"We tend to assign a lower priority to things where the information is of poor quality, the information has a very narrow user base, or the data isn't sufficiently frequent to really respond to the needs of users, or information is available from some other source... those are the things, if we're under budget pressure or we have a very high priority somewhere else..we're going to look at, as things that we could possibly discontinue."
On tracking both social and economic trends:
"We need to maintain balance...it's as important going forward to understand the health of Canadians, their levels of education, their income distributions, their physical security, we need to be able to measure and report on those things, just as well as on the economy itself.
"The one area where we think we're out of balance is the environment. We would say there are three pillars that we need to be concerned about, the environment, society and the economy. We want to do a good job in all three domains, and the one where we don't think we're doing as well as we should is the environment."
On whether it would introduce an index of well-being, similar to that of France or the U.K.:
"We're providing the base data. Indexes of well-being involve enormous value judgements. You can ask people directly about their sense of well-being, and we have done that and continue to. But in terms of constructing an index of well-being, there are enormous value judgements necessary to do that. We see our job as ensuring that a broad base of measures and indicators that would go into such a measure exist in Canada. And the environmental side is the piece where we need to do better."
On the quality of information from new 2011 national household survey:
"There's still an enormous amount of work to do to understand how good it is. We're in the process of coding and processing...later this year we'll get a sense...of whether we are seeing any problems in the data.
"I can guarantee you there will be some problems in the data -- there always are, even in 2006 we saw some issues in the data. Are we going to see something worse than in 2006? We don't know that. All I can tell you is we're going to be very rigorous in assessing the quality of the data and we're going publishing or make public all of the information we know about the quality of the data."
On the ability to compare 2011 census data with 2006 census:
"Whenever you change anything, there's a risk to comparability, but it's not a given that there's going to be significant problems with comparability.
"The danger in 2011 is that because of the significantly lower response rate that we're going to see response bias creep in...some groups that would have responded in a mandatory context have not responded in a voluntary survey, and that's what we're going to look at and assess.
"To say it's a given that it must have happened is wrong. It's statistically not defendable to say it must be true that this data, the estimates, will be biased and flawed and therefore not comparable to 2006.
"You could in fact make comparisons to 2006, potentially.
On recommendations for the 2016 census:
"We've looked at what had been done in other countries and we're confident we've seen all the models.
"We are now in a position where we can go to the government and say 'here are the various models, here's what's achievable for 2016, here's what's achievable in the longer term, here's what Statistics Canada would recommend.' We're very close to that point. We'll be doing that in the next few months.
"I can't discuss what we'll recommend."