Matt Yglesias wrote a piece that interpreted research by economist Brigham Fransden that suggested that unionizing private sector workplaces is not good in the end for those workers because newly unionized workplaces close down more often and see older and higher paid workers move on to be replaced by lower-paid, less experienced workers.
Limiting the pay of the highest-earning members of a particular company in the context of a marketplace where most companies aren’t unionized naturally has the effect of inspiring many of those higher-earning workers to seek new jobs elsewhere. This plausibly also explains the finding that recently unionized companies are more likely to go out of business. If you lose your star performers to the competition, you put your business at risk.
I’m not really convinced of this, but the question is worthy of further research. The presentation of Fransden’s paper says much about the problem with Vox-style (also, 538-style) reporting in that it takes a single article and presents it as the God honest True Data that gives us the one single take we all need to know. Is there something to Frandsen’s conclusions? It is possible. I do not know. Is a single study enough that Yglesias should be presenting it as the truth on private sector unions? No, absolutely not. Is the question worth more studying, perhaps by someone actually exploring real workers in a field outside of economics, as well as by more economists? Yes, absolutely. Then maybe we can come to some conclusions on the matter.
It’s also worth noting that Fransden’s paper is self-published. It’s not listed under his publications on his department website and there’s no evidence from the linked PDF of publication. So in other words, the conclusions that led Yglesias to write a well-read piece and present it as truth have not gone through the peer review process. This doesn’t per se mean the research is not valid of course. I don’t really know about economics, but a self-published piece in history, even something as sophisticated as a statistical-heavy working paper, would have about as much relevance as a blog post.
Yglesias and his cohort treat data as something sanctified. But what about human bias? What’s are Fransden’s politics? How are they affecting his data? It’s a question we have to ask. And then of course, there’s Yglesias’ own political leanings on labor, which have been on display for years. While broadly sympathetic for reducing income inequality and the like, as far as I’ve ever seen, Yglesias has never publicly supported a labor struggle. Meanwhile, he has dismissed the deaths of 1100 Bangladeshi workers as a reasonable price to pay for the nation’s economic development, he slammed the Chicago Teachers Union struggle (see Farley’s dissection of this here) and teachers unions over and over, supporting the worst kind of Rheeist interpretations of issues in public schools. It also goes without saying that he opposed the Huffington Post boycott. He also decided the best response to the firing of high ranking tech executive and misogynist Pax Dickinson was to troll the labor internet. Naturally then, his response to the BART workers strike was to say that paying public transportation workers more was a bad idea.
So given this long-established history of talking about unions and strikes in a negative way, we also have to ask ourselves if the Yglesias writing on this topic tells us more about his own predilections than the relative success of the recently unionized workplace. Like how my writings about labor should be considered within the context of my support for organized labor, Yglesias’ writing on labor need to be considered within the context of his history of not supporting union efforts. Instead, it’s published as politically neutral. Just the facts, Ma’am.