Economic Intelligence

Watching the Numberphile video The Clever Way to Count Tanks caused me to look up and read An Empirical Approach to Economic Intelligence in World War II by Richard Ruggles and Henry Brodie

This is a good paper. The title sounds like the whole paper is dry, but it isn’t. It’s surprisingly interesting to find out what industrial products were of interest (airplane tires and military tires, V-1 and V-2, tanks) and what information got used to estimate production. The authors explain what got done (but not how it got done) by the Economic Warfare Division of the US Embassy in London during WW2.

This paper does emphasize how they decoded serial number formats, and the detective work to figure out what companies did the manufacturing, which may be as important to directing bombing raids as the rate of production.

As always, the weird little details about the astonishingly massive WW2 war effect are maybe as interesting as the topic of the paper. For instance, the US State Department had an “economic war division”, that employed statisticians or applied mathematicians. A footnote on page numbered 74 says that the “EWD personnel consisted of analysts loaned by the Office of Strategic Studies, the Foreign Economic Administration, and the State department”.

Relationship with ULTRA

In 1947, the fact that the Enigma machine cipher had been cracked was absolutely not something that was declassified or talked about. That was an important secret until 1974. The authors of the paper, and the Economic War Division may or may not have been aware of ULTRA, although some EWD analysts were on loan from OSS, and may have had clearance for ULTRA.

I have some questions. How much of the work of this paper was a “parallel construction” of some German war materiel production to cover the fact that ULTRA gave precise figures? How much of this work was double-checking some ULTRA details, how much was done because it was worthwhile to know about tank and V-2 production, but those items weren’t mentioned in Enigma-encoded radio transmissions? Harold C. Deutsch doesn’t mention anything like the gritty details of what ULTRA provided, giving only general categories, and emphasizing order of battle, which appears to be a term of military art, and absolutely does not include discussion of tire production, but could conceivably include discussion of tank production.

Typography

This paper looks like it was set in Times New Roman. It’s in a single, justified, column, with footnotes. The footnotes are set in smaller type. There are italic sub-section heading lines, and small-cap section headings. Here’s an example, from page numbered 87:

The tables all have the lines separating headers from data. This table appears in the middle of the page, Footnote 1 attaches to the table, there’s a large paragraph between the footnote and the end of the page.

I believe this was typeset using a Linotype machine. Whoever did the typesetting was skilled, and did a wonderful job. You’d be hard pressed to author something this well done using a WYSIWYG word processor.

Multiple versions on the web

Usually, Numberphile is pretty good at pointing you to their source materials, but this time they didn’t provide a link. At about 13:26 in the Numberphile video, there’s a graphic showing the paper’s title and authors.

All of the freely-available PDF versions of this paper I can find on the web are clearly scans (of better or worse quality) of the a paper copy of the article. The Journal of the American Statistical Association is jealously guarded by Taylor & Francis Online, who want $64 to purchase the PDF.

Apparently, I typed the title of the paper into Google search. The first thing Google showed me was a PDF from Universidad de Buenos Aires This PDF contains perhaps the cleanest scan available. It does not have a cover sheet.

The CIA PDF is clearly scanned from a bound paper copy. The cover sheet has some extra text rubber stamped on it:

This Document Is
A Source Reference In
A Historical Paper
DO NOT DESTROY

The cover sheet says Reprinted from the JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, March, 1947, Vol 42, pp. 72-91.

Some of the pages, mostly even numbered, have what looks like ink blobs. Someone pored over the CIA’s paper copy of this article. The even numbered papers have other scanning artifacts that look like vague images of the edges of the paper pages being scanned. The odd numbered pages feature slightly off square images, and curved lines of text, indicating that the bound copy wasn’t pressed flat to the bed of the scanner. Depending on where a page in a bound volume lies, the left or the right hand copy will be a little more off square than the other. This appears to be an independent scan from a different paper copy.

There’s a Sci-Hub PDF, too. It has two cover sheets indicating that it originated from Taylor & Francis Online. Each page has a text-processor-paste-in running vertically on the left hand margin, “Downloaded by [Michigan State University]” at 11:23 10 January 2015" This PDF doesn’t show the tell-tale marks of scanning from a bound paper copy.

Official Citation

Ruggles, R., & Brodie, H. (1947). An Empirical Approach to Economic Intelligence in World War II. Journal of the American Statistical Association, 42(237), 72–91. https://doi.org/10.1080/01621459.1947.10501915