When people who don’t understand the basics of time series analysis try to talk about time series analysis

From 2019pay special attention to the statistical analysis of CoRev (at the end):

Reader JBH writes:

“Employment has grown remarkably well under this president.”

I laughed and laughed and laughed when I read this. Why? Let’s take a look:

Figure 1: Non-farm employment (dark blue) and stochastic trend (red). Stochastic trend estimated using data for 2010-2016 and regression of the difference of the first logarithm by a constant. Source: March 2019 BLS employment report and author’s calculations.

We are doing just as well as we did in 2010-2016, even after massive tax cuts and the end of trillions in spending caps.

But kudos to JBH for living in a fantasy world… it’s definitely nicer than my world. Thanks Drumpf!

Update, 04/12/2019 4:15 PM PT: Reader CoRev Requests:

Can you provide source data?

Here data set used, downloaded directly from FRED.

Update, 04/16/2019 9:00 AM PT: A CoRev reader provided his control file for my trend analysis. No comment here:


About three years after CoRev introduced his “anomaly analysis”, I still don’t fully understand what he does. It looks like a deviation from the mean. If anyone can tell me why this confirms his view of the world (i.e. my choice of using a stochastic trend and time sampling bias versus seeing Trump’s job boom up to this point), please tell me. I’m (still) dying to understand.

In the meantime, here is a chart showing problems with linear (deterministic) trends for non-farm payrolls (reproduced from “Why Friends Don’t Let Friends Apply Deterministic Time Trends to Off-Farm Employment”):

Figure 1: Nonfarm payrolls, 000, sa (black) and linear deterministic trends estimated over 20-year subsamples. Recession dates as determined by the NBER are in grey. Source: BLS, May employment report, NBER and author’s calculations.