Credible projections curb the inflation cost of looking through
Credible inflation and output projections make weaker responses to supply-driven inflation more effective in a Bank of Canada experiment. Published October 16, 2026, the research finds that detailed policy-rule information alone offers little benefit.
Projections narrow the policy trade-off
Without projections, the weaker Look-Through rule prolongs inflation and raises the risk of unanchored expectations relative to the stronger Lean-Against rule.
Inflation returns to steady state in roughly ten periods versus five; policy rates peak near 400 versus 800 basis points.
Output gaps fall to about -300 versus -600 basis points.
Long-term inflation expectations peak near 300–400 versus 250 basis points.
Adding quantitative rule information to qualitative guidance produces little improvement.
Providing six-quarter inflation and output projections instead anchors expectations and stabilizes outcomes.
With projections, inflation differences between the rules are not statistically distinguishable, while Look-Through retains greater output stability.
A controlled test of persistence
Olena Kostyshyna, Luba Petersen and Jing Yang compare two policy rules and three communication protocols using seven-person undergraduate groups from Simon Fraser University and McMaster University.
Each economy runs for 50 periods.
In period 19, a large cost-push shock arrives and shock persistence doubles from 0.3 to 0.6. Look-Through then cuts the inflation-response coefficient from 2.2 to 1.2; Lean-Against retains 2.2. Projections use the rational-expectations solution.
The authors' views are independent of the Bank's Governing Council.
The advantage is conditional
Six Lean-Against sessions without projections entered deflationary spirals and were excluded from the main estimates.
Its reported inflation advantage therefore applies only outside those episodes.
The authors advise looking through supply inflation only with credible projections that people can use.