Australia, Brazil and Singapore still pass the bounds test and stay inside their published intervals when any control but the pandemic-trough one is dropped. Without that one, Canada's off-ladder estimate roughly halves.
Conditions and method
Status: exploratory.
Each market’s model outside the United States is refitted with one event control dropped at a time; the seasonal terms are always kept. The chart shows the long-run GDP elasticity and the bounds statistic under each drop.
The United States row is not refitted here: the range of its other specifications is on the ladder chart.
India and the United Kingdom fail the bounds test in this refit, so the chart treats their long-run elasticities as not meaningful: the rows for them show how the number moves, not a figure to quote. Canada passes the bounds test but is off the ranked ladder because its two traffic series disagree, so its rows are a diagnostic only.