So ... the large discrepancy between weak ADP and the huge upside in #NFP for November caused some consternation last Friday. For example, @JulianMI2 pointed out that
But I also noticed this comment - Thanksgiving timing effect "could be at least 100k" - attributed to @EconguyRosie via @carlquintanilla
Could timing of Thanksgiving have such a large impact on payroll numbers? Haven't looked into hiring rates just before and after Thanksgiving, but 2019 was the first year since 2013 and 2014 where Thanksgiving occurred during a last full workweek ending the month.
Less noticed on the same day was that Canada missed employment expectations by 80K, which is huge for Canada and represents one of the largest downside surprises in the last 10 year (the number is based on a household survey and different from BLS methodology).
The dots show the delta between NFP and (2x normalized) Canada employment change in Nov 2013, 2014 and 2019 when Thanksgiving fell in the last full workweek ending the month. The deltas are among of the highest since 2012. In 2013 and 2014, NFP did weaken the following months.
I didn't include prior years. Thanksgiving calendar was the same in 2008, but hey that was 2008, though even then there was a small tick up. And like 2019, 2013 and 2014 were decent years for the economy. So was there pull-in of hiring this November beyond returning GM workers?
Maybe this helps to explain the much stronger NFP? However, the sample size is only 3 and only data on hiring around Thanksgiving would build or undermine the case. That will take some work unless @EconguyRosie can lend a hand with his analysis of "could be at least 100k"?
