Busted: The Sad Data Manipulation of Prof Mark Perry

@TBPInvictus here:

It was exactly one year ago – here – that I penned my first piece on Seattle’s then-new (just going into effect) minimum wage. That piece focused on interesting (and wholly unsubstantiated) claims that the law – at that time it had been in effect about a week – was already a failure and had caused restaurant closures and job losses. Many minimum wage critics latched on to a March 4, 2015 Seattle Magazine article titled Why Are So Many Seattle Restaurants Closing Lately? – this despite the fact that the article included the following somewhat confusing and misleading line (my emphasis):

Though none of our local departing/transitioning restaurateurs who announced their plans last month have mentioned this as an issue*, another major factor affecting restaurant futures in our city is the impending minimum wage hike to $15 per hour. [Ed note: Curious that a supposed “major factor” would not be mentioned by even one departing restaurateur.]

That was one year ago, and what a long, strange trip it’s been. I’ve penned some 20+ articles on that topic here since then, most of them refuting and rebutting deliberately misleading misinformation straight from its nexus – Professor Mark Perry (for whom I once had some respect before he sold his soul to AEI). Sadly, it has always been the case that Perry’s false narratives have made their way through the right-wing echo chamber – up to and including Murdoch-owned entities Fox News and the NY Post – notwithstanding the fact that they rarely stood up to even a modicum of scrutiny. Professor Perry’s bona fides include a Ph.D. and M.A. in economics from George Mason University, and an M.B.A. in finance from the University of Minnesota. (Editor’s Note: GMU is a huge recipient of Koch Brothers dollars, including the aptly acronymed Anthony Scalia School of Law)

And so, one year on, Perry has now suffered the most grave of indignities – he was recently busted for inappropriately mixing and matching two totally disparate datasets to further his making disingenuous anti-minimum wage arguments:

“Thus, our assessment is that Perry used two different sources of data in the same chart to compare the surrounding metropolitan region and Seattle. His ultimate calculation of the surrounding MSA excluding Seattle’s employment involved subtracting the number of total nonfarm jobs in the MSA (using CES) by Seattle employment (using LAUS). […]

It appears that the conclusion drawn from figure 1 was in large part a matter of using two different forms of data. More broadly we suggest caution in regard to declaring anything about the minimum wage hikes until more and better data become available and more sophisticated analyses are conducted. [Ed note: I’ve long maintained that patience and analysis would be paramount in drawing conclusions about Seattle’s minimum wage hike. And, in fact, a multi-year study is underway to do just that. In the meantime, we’ve seen no evidence of ill effects in the past year.] “

Bad Ideology makes for a bitch of a research assistant.

I’d note that Perry’s abuse of LAUS data – in and of itself – was the subject of this piece by Michael Hiltzik at the LA Times.

Upon being busted, Perry removed the chart(s) in question, and in its place put this unapologetic note:

Update: A chart that originally appeared below incorrectly used Seattle area employment data from two different data sources. An updated chart will be prepared and posted shortly.

The chart “incorrectly used Seattle area employment data from two different data sources.” No freakin’ idea how that happened!

As a heavy user of many government databases, I can attest to the fact that one doesn’t draw data “from two different data sources” by accident. Far, far from it. As good as most government databases are (and they are good), it can be a bit of a laborious undertaking to drill down to exactly what one is looking for.

Its not an accident that the two different datasets Perry’s been abusing have been mixed – they need to be accessed separately and then combined; mined from two very different locations at BLS.gov.

The takeaway here is that Professor Perry has been a dishonest broker of information on Seattle’s minimum wage hike from the very start. This was only the most recent, and most egregious, affront to honest, objective and reasoned analysis of data.

To the NY Post, Fox News, Stuart Varney, Dori Monson, Jazz Shaw, Tim Worstall, IBD, AEI, and myriad other outlets, know – if you didn’t already – that you’ve hitched your wagons to a fraud.

This will change nothing to many of these folks, because neither facts nor fraud matter to ideological zealots. The record will now reflect beyond a shadow of a doubt that data abuse and manipulation are simply business as usual to advance that agenda and pursue an ideology.

Cognitive dissonance is also a bitch.

And a note to Arthur Brooks and AEI: If you have so little concern for your reputation you allow such shoddy, dishonest work to regularly be published under your imprimatur, then reputationally, you get what you deserve. I was previously not in the camp of our host, who has made the claim that all research from AEI, Cato, Manhattan Institutute, Hoover, etc,. is completely without merit, but I am moving in that direction (BR: I said so much of it is that it’s not worth the time to separate the bullshit from the mere junk).  Either you are unaware of this, or you condone it, or you simply do not care, none of which reflects well upon you.

At this point, I suspect you have thrown in the towel of objectivity, and are now partisan hacks, doing the work of your monied overlords.

The greatest tragedy in all this, by a long shot, was that I didn’t make the catch, having long ago written off Professor Perry as unworthy of serious consideration.  Damn you, Reid Wilson and H. Luke Shaefer.

Additional reading: Differences Between Data Series

~~~

UPDATE: April 9, 2016 4:30pm

This is looking like a regular habit for Perry, as several readers have directed me to this from 2011:

Why Perry’s Chart Was Dishonest, Lazy, or Thoughtless

I wonder what other related errors I may have overlooked?


123