- Randall Bartlett, Deputy Chief Economist • LJ Valencia, Economist
Economic Viewpoint
Diagnosing the Data Quality Crisis
June 25, 2026
- The unexpected decline in Q1 2026 real GDP in Canada has led to a lot of soul searching among economists. How did so many forecasters get it so wrong?
- The primary reason is that forecasters rely on monthly real GDP by industry to forecast quarterly GDP. But monthly GDP has become much less reliable as an economic indicator since the end of the pandemic. It is now much more heavily revised than it was prior to 2020. And at the individual industry level, large sectors like wholesale trade and manufacturing see some of the most substantial revisions.
- But the concerns don’t end there. Series that had been useful for forecasting monthly industry output aren’t any longer. The most egregious example is the retail trade component of monthly real GDP, which has become completely uncorrelated with monthly retail sales volumes since the end of the pandemic.
- The causes of this problem are wide-ranging. Increased challenges in data collection, greater swings in underlying population data, fewer resources available to collect and compile the information and a resulting reliance on imputed data are all to blame. Some of these challenges can be overcome with more funding. But others, like concern related to trust in public institutions, may be a more intractable problem.
- All parties using official data, whether in the public or private sector, should collectively make the case for high-quality information to support evidence-based decision making. Without good data, we’re all lost.