Fields sounds basically right about the some weaker contributions of technological development for the war, and I think this should be obvious too. The technology we use in battle is not obviously applicable to economic growth. I'm curious to see what Fields actually wrote, though, because this seems to have been glossed over too easily. Clearly there's no automatic correspondence between war tech and peace tech, but any improvements in transportation and communication for the war effort were likely to have been adapted with a lag. I'm concerned Fields and/or Henderson is not giving that point the credit it deserves. Of course you're not going to see war time technological development pay dividends in the domestic economy. But do you see it adapted, integrated, and paying dividends five or ten years later? That seems more likely. The other thing that the war did for technological development is lay the foundations for post-war science. David Kaiser talks about how the war (and the subsequent Cold War) dramatically changed the way physics was taught, and of course war-time agencies evolved into the post-war NSF and NASA, which have been very important to post-war technological development. So I guess I'd just say that yes, clearly the war-time work was dedicated to war and clearly that doesn't have peace-time analogs, but it seems very important to think and talk about this in terms of lagged effects. I'm not ready to dismiss WWII as a major contributor to U.S. technological development.
The major downside of Henderson's review is that he repeats this argument of his about post-war predictions and Keynesianism and tries to shoe-horn that into a claim that the war didn't end the depression. I think maybe I'll write something up about these post-war predictions over the Christmas break.
Arnold Kling has an essay on the impact of technological change on "middle-class" jobs. In the beginning he contrasts it with both Keynesianism and the "business uncertainty", which I don't think is entirely necessary for him to do. I agree strongly with Kling about the role that technological change has to play in trends in unemployment and the restructuring of the job market, but I still maintain a Keynesian understanding of the job market problems we've been struggling with since 2008. We don't want the Schumpterian processes to go away even though they may cause problems for workers, because in the long-term they are a good thing. But even if in the short-term those Schumpetrian processes did go away I still think we'd be dealing with a substantial Keynesian depression right now. Anyway - the essay is good. Sometimes with such short-term problems right in our face we forget the bigger trends shaping our era.
Finally, I was at the Association for Public Policy Analysis and Management (APPAM) conference yesterday presenting my preliminary results on the Georgia job creation tax credit. But after that I went to an interesting session on public financing of research. The first paper wasn't particularly interesting - just some regressions showing that grant funding improves research output. The obvious problem is the endogeneity of grant funding - not that I doubt it improves research output, but it's also true that productive researchers get the grants. That paper was very much one of those "I have data, I know how to run a regression, so I will run a regression" papers, without much thought to what they were doing. The second paper was more descriptive but more interesting - the guy did a survey of stem cell researchers and looked at the relationship between state revenue sharing requirements for stem cell research and the entreprenuerialism of the researchers. As the discussant noted, it was one of those papers where a null finding was very important. Essentially whether they were in a state that had revenue sharing requirements or not, and whether they knew about those requirements or not the researchers were all about equally entrepreneurial. His interpretation was that at this stage in the research, profiting off the research wasn't the most important thing on their minds, and for those who did have more entrepreneurial ambitions, a state revenue sharing requirement was very low on the list of things they were worried about. They just weren't that responsive to the policy, which suggests at least at this point its not a policy we should lose much sleep over.
I had more problems with the third paper - and I voiced them in the Q&A and the discussant subsequently agreed with me. The guy was trying to show that federal R&D spending crowded out state R&D spending. These two variables are also closely related, of course - and unlike the first paper he was aware of that and tried to solve it. He used a state's representation on House and Senate appropriation committees as an instrument for federal spending on R&D that was independent of state spending on R&D. My concern was that the lion's share of federal R&D spending going to public universities (that's what he was looking at) is competitive and is not assigned to states by Congressional appropriations committees. Maybe the little that they do have control over is sufficient for identifying the model, but it seems unlikely to me. He found crowding out of 13-18%. I would suggest that is likely researchers who did not get sufficient state funding and thus went after federal grants more determinantly. The other thing that was tough in accepting his work was that "state R&D spending" is pretty vague. He's looking at direct grant monies from states, when presumably state general higher education funding (which he doesn't look at) gets directed to R&D at the university level.