By Deirdre McCloskey
Real national income per head in Britain rose by a factor of about 16 from the 18th century to the present. Other cases, such as that of the U.S. or Korea, have been even more startling, historically speaking. Like the realization in astronomy during the 1920s that most of the “nebulae” detected by telescopes are in fact other galaxies unspeakably far from ours, the Great Fact of economic growth, discovered by historians and economists in the 1950s and elaborated since then, changes everything. And 16, if one follows William Nordhaus’ persuasive arguments about quality improvements in (say) lighting, is a very low lower bound: the true factor is roughly 100. As Maxine Berg has argued, changing quality of products was as important as changes in process. But the gain is not to be measured by pot-of-pleasure “happiness studies.” These are questionable on technical grounds, but especially on the grounds that they do not measure human fulfillment. They ignore the humanities, pretending to scientific precision. It makes more sense to stay with things we economists can actually measure, such as the rise of human scope indicated by the factor of 16 or Nordhaus’ factor of 100, or by what Sen and Nussbaum call “capabilities.” Of course, what we really care about are the scope or capabilities of the poor. These have enormously expanded under “capitalism”—though a better word is simply “innovation,” arising from bourgeois dignity and liberty. It is the Bourgeois Deal: let me alertly seek profit, and I will make you rich.”
In her paper Deirdre McCloskey makes a call for economist to discuss “things we economists can actually measure”. It was never so appropriate, I think. In this second posting I have taken a different approach to select the paper for discussion. As noted in a private exchange with Deirdre, as editor of NEP-HIS and the blog, I filter contributions twice. Instead, she argued, lets look for a more “democratic” way of selecting papers for discussion.
After giving this some consideration (including a brief discussion with other people running NEP), I came up with the idea of going back and looking at past issues of NEP-HIS reports with the aim of selecting the paper that had the greatest number of downloads. How long back I had to go was not initially clear but, due to a hardware issue at NEP, the latest report from which download statistics were available was that of 2009-10-24. This sort of settle the issue or at least enabled me to start.
It so happened that in that report it was one of Deirdre’s papers that had the greatest number of downloads (90 downloads as of 2009-11-20, with the runner ups having 75 and 74 respectively). That is the paper quoted above.
Before proceeding to discuss Deirdre’s paper, however, let us consider some issues/differences between the selection of papers in the first two blog posts. On the one hand, in the first blog, the I choose a paper at the same time that I was editing the report the paper was being distributed. This is the “double filter” described above. So what? Well this action might bias downloads as the selection could influence people, who would not otherwise consider that paper, to actually download it.
On the other hand, on the second blog, the I chose a “recent” report and from it, the “most popular” paper as measured by number of downloads. As mentioned, the selection of the report and cut off date to measure the number of downloads were both arbitrary (but I am sure some of my friends at NEP/RePEc will have a view as to how to “optimise” these). Anyway, such selection could equally result in greater number of downloads for that paper, for exactly the same reason as discussed above. However, such action will also add to and thus increase the difference between papers with few downloads (presumably of little interest) and the “most popular”. I am under the impression that search engines tend to favour in their ranking pages (in this case papers) with higher number of previous “hits”. Probably you now get a feel of where I am heading, in the context of Deirdre’s paper above, selecting the “most popular” will increase the difference between “download rich” and “download poor” authors. Is that “fair”?
But it also strikes me that choosing “the most popular” download risks guiding this forum towards discussing “how many angels sit on the head of a pin”, that is, reinforcing “the main stream”. Always selecting papers for discussion solely based on the highest number of downloads will certainly foreclose the possibility that the NEP/Blog editor chooses a contribution which highlights an interesting and perhaps unusual area or a topic that is running through several other papers posted through NEP-HIS and which one in particular seems to summarise well. Perhaps that is not the case and NEP-HIS subscribers will identify such “jewels”.
I am sure there are other ways to make the process of selection for papers to be discussed in the blog much more democratic and transparent. For instance, a voting system which is independent of downloads and in which subscribers directly express their preferences. That sort of application will not even need an editor in the blog. However, NEP at the moment lacks such system as well as the resources in its infrastructure to do it (but as an “open platform” would be more than happy to support anyone who wishes to consider making a contribution!).
There is, of course, the possibility that people actually write to me as NEP-HIS Editor and suggest papers (from other authors) to be included in the discussion. I honestly don’t expect that to happen regularly but happy to entertain and action any and all such suggestions. That would still be the result of “editor’s choice”, though.
So perhaps a balance could be struck and every other week the editor selects a paper while the rest of the time, the “most downloaded” is chosen for discussion. What do you think? Input here would be must appreciated.
Now lets talk about the paper, here I am acting solely as facilitator as you have “spoken”, please lets have your comments on the content as well….