Well, I mentioned him a couple weeks ago. Now, John Cochrane says that Moretti and Chang-Tai Hsieh have a new working paper. They claim that the wage differentials between cities are productivity differentials. Instead, they might be cost-of-living differentials. Or, to be more precise, cost-of-living-among-affluent-people differentials.
I think you’re right to question how much of it is productivity. Some might be compositional, but much of it is a transfer of economic rents to real estate owners who collect income above unencumbered costs. In either case, expanding housing would increase real national production.
Just about all closed access city households reduce real housing consumption to minimize these rent transfers. High income households can reduce it enough to pull nominal housing expenses down to regular levels, but low income households can’t. So the cost of living isn’t so much from lifestyle differences as it is from an access fee administered through rent.
http://idiosyncraticwhisk.blogspot.com/2016/05/housing-part-143-closed-access-and.html
Too much “GDP factory” talk about this – “if these cities had rules like other cities GDP would be higher by …” -> except that at least SF and I suppose to some degree NY has various geographic and structural issues as well.
Worse, if the same people are doing the same work for more money, but in a higher cost of living city (Janitor in SF vs Janitor in Bakersfield) – GDP will “look bigger” because more dollars flowed around in circles. But actual wealth will not have changed much, or at all.
Finally, I will point out that Seattle metro (where I live) is NOT facing rising housing costs caused by “GDP factory” or a *general* influx of highly paid people. It’s Amazon, Microsoft, Google (probably in about that order) and the 2nd and 3rd tier technical firms they hire. Put another way, it’s a *pattern* of trade which brings a high volume of dollars by selling a regionally exportable good (software, web services.) Janitors, and for that matter Lawyers and Doctors, are not particularly more productive here than elsewhere. But they’re competing for houses with the Amazon/Microsoft/Google gang.
(Boeing employs a great number of people, but mostly at middle class wages. Safeco is a big employer too, but seems relatively static in size.)
Measure GDP with better cost adjustments than a national average deflator and you will get more accurate numbers.