The Spatial Mobility of Ordinary People

The Making of ... [link to the preprint text]

This is an easy one to write. I have always liked making lists, and have quite a methodical mind. The bare bones of The Right to Manage, for example -- the 5 x 8" notecards containing bibliographic details and abbreviated notes on every book and article that I read, and that went into making it, between 1972 and 1981, sit in a shoebox on my office shelves as I write this, and could still serve their original purpose, providing me with a key to my boxes and boxes of notes on A4 paper. Then, when I began the work that led eventually to Bloodless Victories, I worked out that I needed to get some basic data about the hundreds of firms that made up Philadelphia's metal-working industries organized into a usable form, and that using notecards for this purpose was likely to be unsatisfactory. By the early 1980s an alternative was becoming available -- computerization! (on a visit to Madison, Wisconsin in 1982, I met a young social science academic who had been given a terminal in his home, to see if he could find any use for it) -- and by the middle of the decade I began to acquire some basic skills myself, using a program called Report Generator on Durham's "mainframe" (what a dated term!) in order to sort and classify the thousands of items in a bibliography of American labor history that I had been commissioned to produce for the British Library. In 1986 I went to the Woodrow Wilson Center for Scholars at the Smithsonian Institution in Washington, DC, where one of my temporary colleagues, Jon Sumida, had a first-generation Mackintosh computer on his desk, with which he could do what seemed to be extraordinary things, and when I got home I decided to get something of my own that I could afford, and learn to use by myself. My original PC -- an Amstrad 8256, soon upgraded to 8512 -- had very limited capabilities, but it did enable me to buy a flat-file database program written to run under CP/M and designed by a freelance software developer, John Campbell. Masterfile was the beginning of a sort of middle-aged love-affair with databases, and then spreadsheets, and the ability they give any historian to organize, manipulate, and interrogate thousands or tens of thousands of bits of data.

At the same time, while I was researching the history of labor relations in the U.S. metal trades around the turn of the 19th-20th centuries, I also came across a dataset that seemed, to me, to have enormous potential. I was reading the Iron (later International) Molders' Journal, and as well as looking at the interesting bits at the front of each monthly issue (articles, speeches), I began to think about the routine monthly reports from local secretaries at the back. I had gone to them for snippets of information about strikes and other developments among the Philadelphia Molders' locals, but soon realized that their detailed statements about the membership -- who had joined, resigned, been expelled, died, arrived from another town, or left looking for work? -- offered a goldmine to any social historian prepared to put in the hundreds of hours required in order to transcribe and organize it. One could, in principle, reconstruct the careers of tens of thousands of otherwise almost anonymous American skilled workmen, from the 1860s onwards, and understand the working of craft labor markets and the unions that regulated them much better as a result.

I did nothing with this idea while I was concentrating on Bloodless Victories, but after I had sent the manuscript off to Cambridge University Press in 1995 I took advantage of the fact that I could buy most of the essential raw material for this enquiry into the skilled molding workforce (a long run of the Molders' Journal) on microfilm. So I did just that, and began to figure out how time-consuming it would be to build a database tracing molders' careers. I started at the beginning, because I knew that the job would become a lot bigger but also easier if I could move from the Civil War era into the late 1860s, when the union began to give each of its thousands of members a unique identifying number, listed thereafter alongside his name until the turn of the century -- at which time the reporting of names ended, because the union realized it was simply making a free gift of valuable information to its organized employer foes.

I managed to do 1864 and the first third of 1866 before I ran out of the will to live. I had by then identified 4,920 individual unionized molders, 40 of whom (all local leaders and activists) I could trace in 1859, 64 in 1860, 54 in 1861, 24 in 1863 (after the national union's effective collapse at the start of the Civil War), then 3,404 in 1864, and 1,795 in the first third of 1866. (I forget what happened to 1865's data -- as I only have summaries of local reports from January of that year in my database, the answer is probably "I didn't have the heart to do it after I had flogged through 1864, which is why I skipped to 1866.") I concluded fairly rapidly that this was the sort of research task that needed collaborators, research assistants, lots of money, and also a clear purpose -- I was accumulating data in an "if I build it [the database], they will come [ideas]" frame of mind, and soon decided that this was not very clever. I also worked out better (easier, more fun) things to do with my time -- the foundry industry research that took me to the Hagley in 1995 through 1998, and of course the rewriting of Bloodless Victories to meet Cambridge University Press's requirements.

However, in 2001 an opportunity arose to go to the always very agreeable meeting of European American Historians at Middelburg, a beautiful early-modern city in Zeeland, one of the nicest bits of the Netherlands and conveniently close to Bruges. I had been there to give a paper before, and knew that they were very hospitable and there was a good prospect that if I wrote another halfway decent paper, it would find its way into their conference volume too. Any academic is always looking for a publication opportunity, especially if it comes with an expenses-paid foreign trip attached. The conference theme for 2001 was "Mobility in American History," and the obvious thing for me to do was to resurrect my Molders' database, read other work (classic local case studies from the heroically naive age of quantitative social history) on mobility, and sandwich the literature review and my own little case study together. The result is not a great piece of work, but it's at least moderately interesting, I think, and communicates some of the data's potential.

I have translated my old databases into a more user-friendly spreadsheet format, so that the results are available to anybody who may stumble over them. One of these days, somebody is going to realize that the Iron Molders' Journal really is the goldmine that I thought it was, especially now that an increasing number of its volumes is being digitized so that one no longer has to sweat over a microfilm reader. But the only other person I know of who has worked on it (Tom Klug of Marygrove College in Detroit, whose 1993 dissertation at Wayne State on "The Roots of the Open Shop: Employers, Trade Unions, and Craft Labor Markets in Detroit, 1859-1907," Chapters 5-11, is the only thorough research on stove-making that I know of, apart from mine) does not seem to have broken into print with the results yet.

Extras:

Molders' Union Locals, 1859-1870 -- this simply summarizes information about locals contained in Grossman's biography of William Sylvis.
Reports on Individual Molders' Location and Affiliation, 1859-1870 [very incomplete!] -- this would be the big one, if I had had the heart to complete it... As it is, it just contains information about small numbers of activists except for 1864 and the first quarter of 1866 (both complete). The major problem with compiling it was the number of variant forms of members' names, meaning that tracking individuals' moves from local to local was often difficult and relied on a combination of detective work (e.g. a person is reported to have arrived in Local B with a card from Local A, but Local A has no record of having granted a card to a man of that name; but there is one with a similar name who is recorded as having left with a card, and who has not shown up anywhere else) and guessing (given that some variant names were so unlike one another that one could not be very confident that they really were referring to the same man).
Summary of Iron Molders' Journal Reports, 1864 -- probably the most useful of my old spreadsheets, compiled from all of the information about members reported in the monthly reports, and a brief version (mostly just particularly interesting bits) of the local officers' comments on local conditions and events.
Distribution of Iron Molders' Union membership, 1864 -- a map derived from the above data, showing cities in descending order of average reported monthly membership. The IMU was still mostly a stove molders' union -- the four biggest stove manufacturing centers (Albany, Cincinnati, Philadelphia, and Troy) made up 40 percent of reported national membership. The map also shows where the companies represented at the 1866 Albany Convention of N. American foundry operators came from. Not coincidentally, there is quite a close fit between reported IMU strength and employers' interest in forming an anti-union alliance, though attendance also seems to have been affected by ease of travel to Albany. When I have done the same thing with 1865 and 1866's monthly reports as I have for 1864's, the fit will probably be even closer.

[t.b.a.]