Writing in the Financial Post, Macdonald-Laurier Senior Fellow Philip Cross says that Canadians should not expect the federal government to produce all the data they need.

Philip Cross

He argues that we need to take a page from the United States, where those who need data are willing to pay for it from the private sector.

“We should not reflexively assume that only government can provide good statistics efficiently, without exploring private sector alternatives”, Cross writes.

By Philip Cross, Dec. 2, 2014

The discussion of data never dies down, hardly surprising for a society that suffers from quantophrenia — an obsessive reliance on statistics, reflecting our collective mania about the appearance of evidence-based decision making.

Two things are striking about the latest round of hand-wringing about Canada’s so-called data deficit. The first is the stark difference between how Canadians and Americans view the role of governments in providing data. The second is the lack of understanding of the inherent limitations of data, especially for transactions taking place outside the marketplace.

In Canada, the call for more data invariably means more data from government agencies, whether Statistics Canada for labour market statistics or the CMHC for housing data. Both are currently meeting with stakeholders to discuss how the volume of statistics can be expanded. These government agencies are asked to be “information utilities,” with the connotation that they monopolize the supply of data.

Contrast this with the attitude in the U.S., where government is regarded as only one link in the supply chain for data. South of the border, government is seen as just the starting point for data, as the head of The Weather Company said in a recent interview with Bloomberg. While their weather analysis begins with satellite data collected by the government, which shows the “top” level of the weather, it is then supplemented and refined with data collected from airplanes which fly through the “middle” level of the weather. Air balloons supply data on the weather near ground level.

The same attitude of Americans is evident for economic statistics. Despite government statistics on U.S. business being available a month faster than in Canada (the trade-off, of course, is that they are less accurate), other sources have sprung up. The monthly jobs data from the Bureau of Labor Statistics are preceded by the ADP payroll report, which comes out a couple of days quicker with comparable accuracy. Metrostudy boasts that its data on housing starts is better than the government’s because its researchers drive 200,000 miles per quarter to inspect every new home subdivision, a larger sample than the Census Bureau’s. The Billion Prices Project conducted by two professors at MIT uses real-time data from the Internet to produce daily estimates of the Consumer Price Index which are more than sufficient for most people including those in financial markets, but without the $200 million cost to the U.S. government to calculate the CPI (you can see the results at their PriceStats website).

The greater supply of data in the U.S. partly reflects higher demand from financial markets for information about the world’s most important economy. Frankly, Canada isn’t important enough to justify the same statistical infrastructure. But it is also partly a difference in culture. Canadians passively expect government to produce all the data, rather than seeing an opportunity for an enterprising firm or individuals to fill the gap with faster, better or cheaper data sources. An encouraging example of the latter is a recent report from Macdonald Realty Ltd that one-third of homes sold in Vancouver went to people with ties to mainland China. In releasing this data, a company spokesperson said, “People always say there are no stats. Well, here are the stats.” This is the attitude needed for more private sector involvement.

However, even with additional private sector sources, there will never be enough data. For example, one of the crying needs in the housing sector is information on the role of non-traditional sources of demand, with Asian buyers in the Macdonald Realty report as one main example. However, a report this week from the Canadian Association of Accredited Mortgage Professionals found that housing demand is also being fuelled by parental gifts to their children. Not that the result is surprising. I recently sold a house to someone whose parents co-signed the offer and the purchase. A good chunk of the proceeds will help a niece buy a house. The redistribution of wealth function of families puts the federal government’s efforts to shame.

Transfers within the family are an example of the limitations of understanding the world through data alone. Statistics on income and wealth, for example, ignore the impact of intra-family transfers, one reason the distribution of consumption and living standards (the outcome we should be interested in) is much more equal than the distribution of incomes. Most families collectively support elderly parents, help children and relatives get through school and then buy their first home, start a new business, or just meet the inevitable financial emergencies all are prone to. However, information on intra-family transfers is not well-understood, since many people understandably will not reveal to outsiders (whether a government agency or the private sector) the full extent of their most intimate financial dealings inside the family.

We should not reflexively assume that only government can provide good statistics efficiently, without exploring private sector alternatives. Remember too that the cry for more data can never be satisfied, because, in the words of Barron’s Gene Epstein, “statistics are useful for telling a story about history, but they are not useful for gaining a full picture of reality.” Relying on statistics and their illusion of precision (mathematics is precise, statistics are not) leads to quantophrenia, a condition worth avoiding.

Philip Cross is a Research Fellow with the C.D. Howe Institute and the former Chief Economic Analyst at Statistics Canada

MLI would not exist without the support of its donors. Please consider making a small contribution today.