Contemplate the next three examples of seasonally adjusted vs. not seasonally adjusted knowledge.

Determine 1: Nonfarm payroll employment, 000’s, seasonally adjusted (blue), not seasonally adjusted (tan). NBER outlined peak-to-trough recession dates shaded grey. Supply: BLS by way of FRED, NBER.

The deep trough related to the 2020 recession may cause distortions within the means of estimating seasonal components. It’s not as if the statistical companies are unaware of the challenges of maximum occasions. Right here’s the BLS take from 2022 on employment, and for CPI/PPI.

Right here’s the image for GDP.

Determine 2: GDP in bn Ch.2012$, annual charges, seasonally adjusted (blue), not seasonally adjusted (tan). NBER outlined peak-to-trough recession dates shaded grey. Supply: BEA by way of FRED, NBER.

And right here is items exports.

Determine 3: Items exports, hundreds of thousands$/month, seasonally adjusted (blue), not seasonally adjusted (tan). NBER outlined peak-to-trough recession dates shaded grey. Supply: Census by way of FRED, NBER.

Right here’s a Census presentation on adjustment post-pandemic.

What to do within the wake of massive occurrences just like the pandemic? Jonathan Wright writes on “echo” impact in seasonals, put up Lehman (BPEA 2016), and put up Covid (NY Fed). Abeln and Jacobs (JBCR 2022) talk about Covid and typical filters.

Earlier dialogue of seasonal adjustment within the Covid period, right here.

What about anomalous climate (though what anomalous climate is in these occasions is getting more durable and more durable to outline)? SF Fed has an adjustment, counting on county stage knowledge. Blue is the usual collection on change in nonfarm payroll employment.

Determine 4: Month on month p.c change in nonfarm payroll employment, seasonally adjusted (blue), adjusted with regional heterogeneity (tan) and with out regional heterogeneity (inexperienced). Supply: SF Fed, accessed 3/1.

It’s exhausting to see the significance of this climate adjustment utilizing this span of knowledge. right here’s a element

Determine 5: Month on month p.c change in nonfarm payroll employment, seasonally adjusted (blue), adjusted with regional heterogeneity (tan) and with out regional heterogeneity (inexperienced). Supply: SF Fed, accessed 3/1.

Basing the adjustment to 2015, the SF Fed calculates January change in NFP at 389 to 393 thousand, versus official 517 thousand.

Boldin and Wright (BPEA 2015) current a number of approaches of adjusting for climate.

If you wish to complain in regards to the seasonally adjusted knowledge, by all means, accomplish that. However you need to establish what side of the seasonal adjustment is bohering you. And you need to accomplish that after studying what the statistical company in query has accomplished. On the whole, it’s normally potential to do a greater job, however at the price of utilizing some type of process that appears advert hoc from the skin. In different phrases, there’s typically a tradeoff between higher procedures and transparency of strategies, in turbulent occasions.