They show that even within a specific (job-title, company, state) tuple, higher unemployment led to higher skills requirements, in a probably-causal way.
That doesn't mean the skills gap is a lie, or any other nefarious bad behavior on the company's part. The study doesn't go into that.
I'd say they are worse than Fox News. Fox News gets a lot wrong, yes, but also everyone knows their entire prime time lineup is opinion...Vox makes no such distinction. If you're pretending to be straight news but you're opinion...that's a big problem to me and worse than misleading commentary that you know is commentary.
I've seen Matt literally copy something straight from MMFA (a liberal news watcher) and Think Progress (a liberal think-tank) and treat it as news...not commentary.
I suppose it’s a bit funny to me that Vox articles often make it to the front page of HN, but FoxNews just don’t. But yea, Vox is worse.
On Twitter he is basically considered a joke on both sides of the aisle.
He seems to be considered highly reputable by many on both sides of the aisle as well. Certainly more reputable than an hacker news commenter making misleading insinuations as to why somebody might want to delete tweets, and overconfident, baseless proclamations of who is and is not considered a joke by the world at large.
Can you explain the causal bit?
The sneaky detail that stands out to me is that this is the 'share' or '%' of jobs with high skills went down, but not necessarily the absolute number of high skills jobs.
So during a recession, there are not a lot of low-end jobs available. As things warm up, more Ops. needed, so more jobs.
But the 'long term view' R&D jobs may have always been there in absolute terms. They might have only changed in relative terms because of the number of regular jobs going up and down.
It implies more causality between 'regular jobs' and 'the economy' but maybe less so between 'high end jobs' and 'the economy'.
Ergo, the 'skills' gap still exists - companies still want high end folks and can't find them. Which makes sense as well because I believe one could crudely say there is a 'skills gap' - anyone hiring high-end people can feel this.
So, you'd probably be better off skimming the paper for the details, but the gist is that they estimate the effect of unemployment on skills requirements in a few ways, and the results all come out about the same. They use state-to-state variation to take out some of the bigger macroeconomic factors. They use the return of a huge number of troops when we pull out of Afghanistan for another change that is mostly uncorrelated with economic conditions and do some other clever stats to make that assumption more solid. (I briefly skimmed it hours ago when I made the parent comment, so I might be a bit off now, but that's the gist of what stood out to me).
Showing causality is generally about finding apples-to-apples comparisons and finding sources of randomness causally upstream of what you want to measure the effect of, to get a kind of mini-experiment. They do this. It's not proof, but it's some decent evidence.
------------
On your question about percent vs absolute:
Your idea is totally plausible and doesn't contradict the study, as far as my skimming goes. There's no answer to to "why" just a decent argument if I suddenly made unemployment higher, more of the job postings you see would have higher skill requirements. My pet theory on "why" is that filtering out the noise gets really hard when tons of people start applying, so they up the bar.
Companies will always want the highest end folks, as it makes them more competitive and productive. The question is are they willing to pay competitively for that talent. More often than not, the answer is no.
I feel that the 'skills gap' was just a convenient excuse during the recession. Companies and hiring managers will always want the best bang for their buck, so when unemployment is high, they aimed higher but still complained they weren't getting what they wanted. And now that unemployment is low, they are just blaming their hiring woes on that.
This is just an anecdote, but during the depths of the recession, the company I worked for was hiring folks with PhDs for low level admin jobs. And at the same time we were 'struggling' to fill some of our technical roles because we couldn't find folks with the 'right skills'. The funny thing was, I had got hired into that technical role with absolutely no experience, fresh out of college, right before the recession started.
[edit]: To be fair, I think it was partly done to help filter out resumes, as each job posting would get a TON of responses. And there was another issue (if I re-call correctly) where the rate of people changing jobs also dropped pretty dramatically (probably for job security reasons), so the job market for mid-level folks was pretty illiquid at the time. So combine those two things, and that could lead to the perception that there was a skills gap.