Feat. the one stupid party trick you need to amaze the economists in your life…
substack.com
Interesting post by Oren Cass (former Romney advisor) on the subject of U.S. manufacturing.
1. i am inherently suspicious of anyone who claims to have some revolutionary or unique insight based on one piece of publicly available data. are we expected to believe that nobody in the last 20 years has considered the data he discusses? isn't it more likely that he's made an error of some sort? or, if not an error, a misleading analysis.
2. he makes much of the so-called decline in american manufacturing output over the past two decades. productivity growth is down, and has recently turned negative; industrial output over the past 15 years has been in decline. his conclusion is that americans have outsourced so much manufacturing that we have forgotten how to do it. thats a paraphrase but he makes that point distinctly.
lets dig in a little. suppose there's an auto factory that produces 1M cars a year. the cars sell for 100k each. there are 1000 workers who make the cars. that's a 100m a year in revenue, and the productivity is 100m a year / 40000 labor hours (assuming 40 hour weeks). now a company opens a non-union factory in the adjacent right-to-work state. the company attracts 500 of those workers and starts to sell cars at 80k each. the first company has no choice but to match the 80k price. now the two plants together produce the same number of autos and employ the same number of workers. except now they are splitting all those labor hours across 80m in revenue, not 100. productivity has dropped by 20%!!!!
in reality, what happened is that the price of cars dropped. and because productivity is measured in terms of output, a 20% decline in prices meant that the output declined. the same amount of cars were produced, at the same price. nobody forgot how to make cars. all that happened was that prices adjusted.
disclaimer: i am no econometrician, and its possible that there are more sophisticated techniques for measuring productivity -- but as far as i know, this is basically the gist of all productivity measures. anyway, it is possible i am missing something but at the moment i don't think so.
3. anyway, lets change the hypothetical slightly. it's not a new state; its a new country. maybe china, or somewhere else. whatever the case, businesses in this new country make goods more cheaply than american companies, and thus the prices are forced down. note that tariffs wont solve this problem for any american industries that export. so what we would expect to see in the data is a drop in productivity and output, even though the same amount of stuff is being produced by the same number of people. note that my example above required some of the workers to move from one plant to another. that was just stylistic to make the model as mathematically and intuitively simple as possible. the same story is true if labor is not portable.
if this is true, then offshoring isn't really the problem. the expansion of manufacturing in other countries will have this effect by its very existence. the exact thing that we want economies to do -- make stuff cheaply, so there is more of it -- is causing some of those economies to look as though they are moving backward.
which is to say, cass is focusing on something that is a pure artefact of data. it isn't a sound basis for economic policy. again, i could be wrong about the productivity measure, but i haven't found anything on the internet to contradict my previous understanding so i am going for now with what i know.
4. i honestly have no idea what point he thought he was making in the worker augmentation module piece. that might be an anticipatory response to an argument i don't see. otherwise, i think its pretty silly.
jacobin.com
www.nytimes.com
youtube.com