Found in the wild:
The idea is that the pandemic distorted things so much that it’s best to ignore it. If that wasn’t acceptable, that era could be presented in a separate chart.
As also seen in XKCD:
Looking at that graph of labour market tightness makes me very suspicious. This is the sort of graph editing that I don’t think is valid.
There is a clear agenda to say that the effect of Covid magically vanished at the end of the first quarter of 2021. This is visually provided by what looks like a single month of flat demand that is at the same value as before the start of Covid. So clearly all the effects of Covid are over, and we can get on with treating the market as if Covid never happened. So what if the very next month starts a one year run that sees the number of jobs versus unemployed nearly double? Covid was over wasn’t it? That isn’t defensible.
If one elides the numbers I would want a much better story than - special circumstances, you don’t need to know.
There is a clear distortion due to Covid here, and thus I would want to know what happened in the elided time to understand what happened after.
The graph as presented looks more like overt cherry picking of data with very thin justification for the exact period the data was cherry picked over. It isn’t as if Covid lasted exactly a year.
Best name for a concept I’ve learned in years! I love that it sounds made-up, but isn’t.
This is a point: “The Pandemic” isn’t a sharp, well-defined span of time. Some things got back to “normal” quickly, others took longer, yet others transitioned to a “new normal”, and some transitioned to a “pandemic state” and then to a “new normal”. And some of the “new normals” might yet transition back (or might not: Who knows?). How do you know which times to blank out, for any given data set?
For picking scales, time ranges , data to eliminate or presentation method the person making the graphic ( won’t get into graph vs plot vs chart vs anything else etc) really needs to know why they are choosing to present the data in a graphic and what use is to be made of the data .
Are they using it to illustrate a trend or a conclusion already reached. Are they using it to study the data to discover a trend or conclusion or make a quantitative calculation or highlight a correlation with other factors , is it being used to monitor the current situation and identify degrading or emergency situations.
Once the need for graphic is known you can then go about selecting graphic type, scales , choosing what data to eliminate or time ranges to view.
The same data can be viewed in different ways and scales for different needs. A graphic in the absence of context or use case is mostly useless.