I think. I haven't read the post. I just read a post on the post by Drum and flipped out.
Matt Yglesias notes a tension in lefty thought today: the stuff we all support (better healthcare, more teachers, childcare, new infrastructure, etc.) is in the non-manufacturing sector, and yet we all cheer when President Obama calls for increased focus on manufacturing. So which do we want? More people working in manufacturing or more people working in service and construction industries? It's hard to have both, after all.
I support high employment in manufactuging. The reason is that I believe that people are paid more if they work in manufacturing than if they work in other sectors. These labor market rents are not considered by employers when deciding how many people to hire. The aim isn't just a sustainable trade balance but also "good jobs for good wages." The evidence supporting the hypothesis that high wage sectors provide better jobs (not just compensating differentials) to the same workers is overwhelming.
update: Is it still true ? Angus wonders. FRED failed me with an index of hourly compensation in manufacturing only from 1987 on. For what it's worth it doesn't show a trend of relative decline compared to overall non-farm, but 1987 was mostly post Reagan and, in particular, post a period of grossly over valued dollar and huge trade deficits.
Yes economists argued for decades that this can't be true and that our models show that people are paid (counting non pecuniary amenities) based on their ability. But believing in those models is like believing in Phlogiston. They don't fit the facts. If there are labor market rents, then the logic of propmoting manufacturing employment is clear. People get something for nothing if they switch from employment in services to employment in manufacturing -- well the data show they lose big if they move the other way. This happens when the shift can't possibly be a sign that the manufacturing employer learned that they weren't as able as they seemed, because it happens when whole plants are closed.
Two decades ago, there was an interesting academic debate on the topic. I claim it was settled. The people who were wrong (as usual) just changed the subject and ignored the data (as usual). Google scholar Larry Katz. In any case, you have to admit that we might want higher employment in one sector because of labor market rents. The argument that the market knows best requires the assumption that the labor market clears. That is that there is no unemployment. I promise you that this is clear.
" It's hard to have both, after all." Huh ?!?!? Have you noticed that the unemployment rate isn't exactly zero right now ? More generally that the employment/population ratio varies a lot and we might want it to be high. I will consider this to be just a slip (mostly). But one point is that, if the labor market doesn't clear, then we don't have any reason to think that we prefer employment to unemployment but don't prefer employment in one sector to employment in another. There are models in which one can prove that the government shouldn't favor one sector. They are models in which there is 0 unemployment. Relying on them is not a good idea.
On more narrow topics.
I don't think you can understand the widespread leftist and centrist support for infrastructure spending without considering the unemployment rate. Some might think that we should spend more on average over the next decades, but others think we should pull spending forward so that we spend more now and less in the future -- that the point of spending now is that we will have non falling down bridges then and won't have to spend then. Supporters of more infrastructure spending now and in the future can agree with supporters of more infrastructure now and less in the future about what to do now. But it just isn't true that Obama says he supports high government spending now and in the future.
Also the supporters of high spending on education and infrastructure don't argue that it is an substitute for manufacturing -- that our goal is to be able to recite Shakespeare we drive our foreign made cars down nice new highways. Rather it is that infrastructure and education are needed for high manufacturing productivity so that the effect will be that employment now in infrastructure and education will cause higher manufacturing employment in the future.
Finally (whew) there are sectors other than manufacturing, infrastucture, education and health care. For example FIRE. It is very easy to argue that we would be better off if fewere people worked in financial services (Yglesias argues this frequently). One might suspect that they are working away separating fools and their money, that is they would be out of a job and their clients would be better off if those clients decided to buy and hold the market portfolio instead of trying to beat the market and losing. This is definitely your view and Yglesias's view. But somehow you forget it when discussing sectors other than manufacturing. This was not a slip. This was contrarian BS.
Or how about real estate. And now that I mention it construction of housing. One might think we would be better off with smaller houses and more manufacturing, infrastructure education and health care. One might note that the current policy isn't neutral but involves huge subsidies for the mortgage interest deduction. One might be Matthew Yglesias proving that Matthew Yglesias wrote that post as a contrarian dweeb.
Or how about insurance ? Some people think we could get just as much health insurance with much lower employment and cost if we had single payer or a public option or Medicare buy in. The evidence for this view is overwhelming. One of those people is named Matthew Yglesias. But he forgot about the bloated private health insurance industry when aiming to channel his inner Kinsley and write some pointless contrarian BS.
He knows better. You know that your devastating critique is just one of many valid devastating critiques. His post was a provocation. That is, he is going over not to the dark side but to they many shades of grey side.