Prologue: Selected Articles
Spring 2002, Vol. 34, No. 1
The WPA Census Soundexing Projects
By Claire Prechtel-Kluskens
"Your letter received this morning brings me great joy."— Geneva F. Hoult, Jan. 26, 1938, a floor supervisor during 1900 census indexing project, upon being offered a similar position during the 1920 census indexing project.
"Someone has simply been taken up with a new toy. . . ."— Thomas A. Devor, Chief Clerk, 1900 census indexing project, reacting on Aug. 25, 1936, to the Census Bureau's decision to adopt the Soundex system.
Genealogists are familiar with the Soundex indexing system as a useful tool for finding people in the 1880 - 1920 censuses and in twentieth-century immigration records. When the 1930 census opens on April 1, 2002, however, researchers will be disappointed that there is a Soundex for only twelve states. The existence of these indexes is an accident of history. The indexing projects were done to put unemployed people to work during the Great Depression of the 1930s and because the Bureau of the Census needed indexes for its own needs. The projects ended as manpower was diverted to increased military and industrial needs for World War II.
This article draws upon several files from Record Group 29, Records of the Bureau of the Census, to provide a brief history of the administration of the 1880, 1900, and 1920 indexing projects from 1935 to 1941, and a glimpse into the plans for the 1930 index. 1
During the Great Depression, New Deal agencies sought to put people back to work. The Works Progress Administration (established in 1935 and renamed Work Projects Administration in 1939) assumed a dominant role in work relief activities aimed at fighting the depression by providing federal jobs for the unemployed. WPA projects encompassed a wide spectrum of tasks whose usefulness ranged from fleeting to long-lasting. WPA project that have been a tremendous help to genealogists are the indexes to census schedules and immigration records.
The 1900 census was the first WPA census indexing project. On January 26, 1935, William L. Austin, director of the census, outlined seven projects that might be undertaken with "[a]llotments from the $4,000,000 appropriation now before Congress for public relief and public works." 2 Number seven was the "Card Index of Population by Families in 1900." Austin believed the index would prove necessary if Congress passed legislation for a federal system of old age annuities and unemployment insurance. For many eligible for old-age benefits, "population census records provide the only acceptable evidence of place and date of birth, since dates of birth in most cases, and for some years to come, will antedate the establishment of birth registration." 3 An index would speed up the search process.
Reference demands had already grown so rapidly that the bureau was "three months behind in current searches, and . . . getting progressively farther behind." 4 The bureau submitted its application on June 15, 1935, and it was approved by mid-July.
The bureau then sought office space in St. Louis and found it in the "Old Custom House and Court House" at Eighth and Olive Streets. The bureau opened an office on Monday, September 15, 1935, in "one of the smaller court rooms in the building" and then began indexing operations in mid-October. 5
Nearly a score of permanent bureau employees formed the core supervisory staff. Gerald Ryan was administrative officer and project manager, and Thomas A. Devor was chief clerk. There were eighteen supervisory personnel, of which thirteen later worked on the 1920 project. 6 Some 2,865 persons were employed. 7
Upon completion of the 1900 index on April 23, 1937, 8 its nearly 34,000,000 cards in 1,200 crated cases were shipped to Washington, D.C., in eight standard railway storage cars. 9 In Washington, the bureau microfilmed the cards with WPA funds. 10 In September 1937, sixteen camera operators each filmed an average of 1,639 cards per hour! 11 Microfilming was completed before June 30, 1938. 12
Choosing the Soundex System
About May 1936, the bureau began considering how the index would be arranged before microfilming. 13 On June 17, 1936, Dr. Timothy F. Murphy, the bureau's chief statistician, noted that "the problem of determining the best system of indexing the cards . . . is a most difficult one, due to the size of the index, the large number of foreign names and the variation in spelling by the enumerators of many common names." Murphy recommended that the Census Bureau consider offering the Globe-Wernicke Company, Remington Rand Company, and the Shaw-Walker Company an opportunity to demonstrate their indexing systems. Globe-Wernicke Company's system was a straight alphabetical index; Remington Rand Company had several systems including the "Automatic" and the Russell Soundex methods; and the Shaw-Walker Company had "a different system," which Murphy did not describe. 14On June 18, 1936, Austin selected Dr. Vergil D. Reed (Assistant Director), Oliver C. Short (Executive Assistant to the Director), and Dr. Timothy F. Murphy (Chief Statistician), to recommend an arrangement system. 15 After studying the three companies' demonstrations, these men recommended the Remington Rand Company's Soundex system. When Devor learned that the bureau had chosen the Soundex method, he wrote to the bureau's chief clerk, Arthur J. Hirsch:
Why in the name of Heaven this contract was ever awarded to the Soundex people by the Bureau authorities is more than I can understand. . . . All this foolishness about putting every one of these thirty million cards through an elaborate numerical coding system is pure bunk. It certainly can add nothing whatever to the efficiency of the system. . . . I think it is positively absurd that intelligent people . . . consider a system that requires constant training and which may be made obsolete at any time by the whim of the inventor. . . . It looks to me as if someone has simply been taken up with a new toy. . . . Get after Austin and tell him to cut it out. 16
In response, Hirsch reported that Austin said "he was not going to change his order and that you [Devor] and I [Hirsch] were too old to understand anything new." 17
After the 1900 indexing project was completed, the bureau did not plan any additional indexing projects. However, demand for the bureau's age verification service continued to climb. During fiscal year 1939, the bureau received 163,132 requests for this information, compared to 4,166 in fiscal year 1929. 18
Col. Brehon Somervell, administrator of the WPA in New York City, wrote to the census director on September 27, 1937, that the New York WPA "would be happy to consider an application for an indexing project." 19 After Somervell pitched his proposal to the secretary of commerce, the secretary sounded out President Roosevelt on Nov. 22, 1937, and "the President . . . specifically indicated . . . this project was worthwhile and should be approved." 20 The project proposal estimated the cost at $2,250,000 and would take 3,535,662 hours of work. 21
Gerald Ryan was placed in charge of the 1920 project with Thomas A. Devor again as chief clerk. 22 The administrative supervisory structure was described as follows:
We were asked to furnish for the work population records of the 1920 Census and also the same small supervisory force we had used on the 1900 card indexing project at St. Louis. We agreed to give to the general supervision of the project eighteen employees, thirteen of whom come from our civil service rolls. . . . Only one of the eighteen we are sending to New York is to be retained on our pay rolls, and this one is Mr. Gerald Ryan who is in charge of the New York project and who is responsible for the safe keeping of our records. 23
The bureau began preliminary preparatory work, such as securing a building at 63 Vesey Street and getting it ready for workers, in mid-February 1938. 24 Actual work started on April 1, 1938, on the seven states for which the bureau received the most age inquiries: Illinois, New York, Ohio, Pennsylvania, Texas, Massachusetts, and Michigan. 25 By July 30, 1938, information had been transcribed from census volumes onto nearly 12.4 million index cards. 26 On July 1, 1939, there were about 2,465,269 cards to transcribe, 14,015,0143 to verify, and 25,972,320 to Soundex. 27 An experienced clerk could Soundex fifty cards an hour. 28
On July 28, 1938, Col. Richard C. Patterson, Jr., assistant secretary of commerce, made an inspection tour of the 1920 indexing project. 29 Director Austin wrote that Patterson "knows . . . that you [Ryan] and Devor are doing a better job in New York than you did in St. Louis, and both the projects are considered the model jobs for the WPA in the United States." 30
Ryan was administrative officer until April 1939, followed by Walter B. Carter, who headed the project until he was succeeded by Henry F. Stewart in March 1940. 31
On June 30, 1938, 3,029 persons worked on the project, 32 dropping to 2,864 by November 15, 1938, 33 to 2,200 by June 30, 1939, 34 and to 1,888 on June 30, 1940. 35 The 1920 indexing project was completed about June 15, 1940. 36 The WPA began shipping indexes for completed states to Washington by truck after July 15, 1939. 37
As the completion of the 1920 project drew near, the bureau considered indexing the 1880 census. At the same time, because the New York WPA census projects staff had become experienced in various census research activities, other projects were proposed to piggyback the 1880 project for administrative and funding purposes. 38
Director Austin appointed a committee chaired by Assistant Director Vergil D. Reed to study proposed WPA census projects. On December 1, 1939, Reed's committee made a lukewarm recommendation that an 1880 census index be undertaken, noting that "the results of this indexing will only serve our purposes for a relatively few years" and "will not contribute much to attaining the long-range objectives of the Bureau." 39 However, the bureau's project application made the expectant results sound more valuable:
This index will be used . . . in the handling of the large volume of requests for age information and verification from various public agencies. This card index will have a definite value both to the bureau . . . and to the general public. It will greatly facilitate the handling of requests for age certification, and will reduce the costs of the individual search. 40
Reed's committee made several specific recommendations, including requiring all employees be required to pass "an examination in penmanship and in reading handwritten material." The last recommendation noted that the original agreement between the bureau and the WPA administrator for New York, dated February 6, 1938, had required that "all clerks must qualify" for the work, but in the intervening time, Mr. Carter had been compelled to accept anyone sent to him. 41
The President approved the "partial index of the 1880 census" project in April 1940. 42 It began May 23, 1940, and was completed June 30, 1941, using 2,068,032 hours of work, at a cost of $1,129,470.65. 43 On June 30, 1941, 1,260 persons were employed on all the Census Bureau's WPA projects in New York City. 44
Age verification applications continued to grow dramatically due to Social Security, national defense needs, passport applications, and other reasons from 97,000 during the 1937 fiscal year to 216,000 during the 1940 fiscal year. From January to April 1941 alone, the Census Bureau received more than 120,000 requests. 45 Dr. Timothy F. Murphy, chief statistician for the bureau, noted that the cost of furnishing the information from the 1930 census was $2.50 per application. To make a search in the 1920 index cost about $ .54 and, he estimated, if the 1930 census were Soundexed, an application could be answered for much less. 46
The bureau's April 1941 application to index the 1930 census estimated that it would cost $4,525,277, and take 6,773,140 hours of work. 47 Despite the application, indexing may not have been undertaken during the WPA era. The New York WPA did age verification searches until September 1, 1942, when its personnel were reassigned to the bureau's remaining research projects. 48
The declining interest in WPA projects was reflected in the rapid turnover in the bureau's supervisors there from 1940 to 1943: Henry F. Stewart, 49 Elliott M. Murphy, 50 and finally Walter B. Carter, who remained with the project until its completion. 51 All New York City WPA census projects were terminated Friday, March 12, 1943. 52
The WPA census indexing projects remain a vital source of access to the 1880 through 1920 censuses.
1. All persons mentioned in footnotes were Bureau of the Census employees unless otherwise specified in the footnote or main text. All records mentioned are from Record Group 29, Records of the Bureau of the Census, unless otherwise specified. Series 142 is "General Records Maintained by William Lane Austin, 1933 - 41," who was census director during the time these projects were undertaken. Series 149 is "Miscellaneous Records, 1900 - 53," kept by the Office of the Chief Clerk, later the Administrative Service Division. Series 192 is "Annual Reports, 1899 - 1949." UD series 9 is "Work Projects Administration Projects Proposed to the Census Bureau, 1936 - 39." UD means "undescribed in any inventory."
2. Memorandum, Director to Secretary of Commerce, Jan. 26, 1935, folder 25-6, file 353, series 149. This file contains a draft memorandum, marked "not sent"; the author has not located the final memorandum, which is referred to in memorandum, Stuart A. Rice to Mr. Kerlin, Dept. of Commerce, Apr. 6, 1935, in the same folder.
3. Memorandum, Rice to Kerlin, Dept. of Commerce, Apr. 6, 1935, folder 25-6, file 353, series 149. See also W. L. Austin to Dallas W. Dort, Admin. Asst., WPA, Washington, DC, June 26, 1937, folder 25-12, file 353, series 149.
4. Memorandum, Rice to Kerlin, Dept. of Commerce, Apr. 6, 1935, folder 25-6, file 353, series 149.
5. Austin to Hon. John J. Cochran, House of Representatives, Aug. 31, 1935, Folder 25-13, File 353; Series 149. J. F. Peoples, Director of Procurement, Treasury Dept., to the Secretary of Commerce, July 22, 1935, folder 25-13, file 353, series 149. The project site was originally to have been Atlanta, GA, but not enough qualified persons were available. Rice to Corrington Gill, Asst. Administrator, Federal Emergency Relief Admin., Aug. 22, 1935, folder 25-11, file 353, series 149. There was insufficient office space in Washington, DC.
6. The thirteen were Gerald Ryan, Thomas A. Devor, Abraham Levin, Walter B. Carter, George W. Harmon, Harry E. Watkins, Grace E. Parker, Barbara Z. Bower, Ada M. Hastings, Camilla J. Patterson, Constance B. Schultz, Virginia Welch, and Laura E. Womeldurf. See Memorandum to the Director from Oliver C. Short, Jan. 24, 1938, file 57-B, series 142. Other supervisors in St. Louis included Mr. Lillard; Edgar N. Harbit, and Dorothy I. Beall. George W. Harmon, who was an assistant to administrative officer in New York, was in the chief clerk's office in St. Louis. Memorandum to the Director from Oliver C. Short, Jan. 24, 1938, file 57-B, series 142. Additional supervisors in St. Louis were Grace E. Rankin (who is probably Mrs. Grace E. Parker on the list of 1920 project supervisors), Dorothy Jarret, and Bessie Case. Letter, Chief Clerk to Thomas A. Devor, Apr. 18, 1936, folder 25-11, file 353, series 149. Geneva F. Hoult worked on the 1900 project but was not a supervisor; she became a supervisor for the 1920 project.
7. Annual Report of the Bureau of the Census, year ending June 30, 1936, p. 33 [reprinted from the Annual Report of the Secretary of Commerce . . . 1936], series 192.
8. Annual Report of the Bureau of the Census, year ending June 30, 1937, p. 27 [reprinted from the Annual Report of the Secretary of Commerce . . . 1937], series 192.
9. Thomas A. Devor to Arthur J. Hirsch, Nov. 24, 1936; folder 53-5, file 1100, series 149. Devor to Hirsch, Jan. 5, 1939, folder 26-13, file 362, series 149.
10. Rice, acting chairman, Coordinating Committee of the Central Statistical Board and Works Progress Admin., to Austin, July 13, 1937; see also Franklin D. Roosevelt to Secretary of the Treasury, July 8, 1937, both in folder 25-12, file 353, series 149. See also Austin to Dallas W. Dort, Admin. Asst., WPA, Washington, DC, June 26, 1937, folder 25-12, file 353, series 149.
11. Untitled attachment, dated Oct. 8, 1937, to memorandum, T. F. Murphy to the Chief Clerk, Oct. 16, 1937, folder 25-12, file 353, series 149. The filmers were surnamed Bonta, Bridget, Dawson, Deacon, Ellswoth [sic], Green, Lucas, Marks, Morgan, Norris, Pinto, Rudisill, Schley, Swartz, Trinter, and Wade.
12. Annual Report of the Bureau of the Census, year ending June 30, 1938, p. 33 [reprinted from the Annual Report of the Secretary of Commerce . . . 1938], series 192.
13. Devor to Hirsch, Aug. 25, 1936, folder 25-11, file 353, series 149. Devor writes: "We have been delayed here since early in May . . . because the Bureau in Washington has been fooling around trying to find something that was better than a simple, straight alphabetical index."
14. Memorandum, Chief Statistician to the Director, June 17, 1936, folder 25-11, file 353, series 149.
15. Memorandum, Director to Dr. Reed, Mr. Short, and Dr. Murphy, June 18, 1936, folder 25-11, file 353, series 149.
16. Devor to Hirsch, Aug. 25, 1936, folder 25-11, file 353, series 149.
17. Hirsch to Devor, Aug. 28, 1936, folder 25-11, file 353, series 149.
18. Annual Report of the Bureau of the Census, year ending June 30, 1939, p. 40 [reprinted from the Annual Report of the Secretary of Commerce. . . 1939], series 192.
19. Brehon Somervell to Austin, Sept. 27, 1937, File "Works Progress Administration— New York City," UD series 9 (NN-364-101, MLR 109886).
20. Oliver C. Short to Brehon Somervell, Oct. 6, 1937, File "Works Progress Administration— New York City," UD series 9 (NN-364-101, MLR 109886); memorandum to Austin from Kerlin, Dept. of Commerce, Nov. 23, 1937, file "Works Progress Administration-New York City," UD series 9 (NN-364-101; MLR 109886).
21. Index of 1920 Census WPA Project Proposal, O.P. No. 324869, undated carbon copy, folder 26-13, file 362, series 149.
22. Memorandum, Austin to Kerlin, Dept. of Commerce, Jan. 26, 1935, file 57-B, series 142. See also Devor to Hirsch, Jan. 26, 1939, folder 53-6, file 1100, series 149.
23. Director to W. M. Steuart [former director], Feb. 4, 1938; File 57-B, Series 142. For details on names of supervisors and their salaries see Memorandum to the Director from Oliver C. Short, Jan. 24, 1938, file 57-B, series 142. See also Austin to Somervell, Feb. 8, 1938, file "Works Progress Administration— New York City," UD series 9 (NN-364-101; MLR 109886).
24. Director to Steuart [former director], Feb. 4, 1938, file 57-B, series 142.
25. Gerald Ryan to Austin, Dec. 1, 1938, file "Works Progress Administration— New York City," UD series 9 (NN-364-101, MLR 109886); memorandum to the Director from Oliver C. Short, Feb. 7, 1938, file "Works Progress Administration— New York City," UD series 9 (NN-364-101; MLR 109886).
26. "Table Showing Total Number of Cards Transcribed Monthly," folder 26-13, file 362, series 149.
27. W.B.C. [Walter B. Carter] apparently to Devor, June 22, 1939; folder 26-15, file 362, series 149.
29. Ryan to Austin, July 20, 1938, file 57-B, series 142.
30. Director to Ryan, July 13, 1938, file 57-B, series 142.
31. Memorandum, Oliver C. Short to the Director, Mar. 29, 1939, and Short to Somervell, July 28, 1939, both in file 57-B, series 142; also memorandum, Ryan to the Director, Mar. 5, 1940, and Director to Col. Brehon Somervell, Apr. 4, 1940, both in folder 26-14, file 362, series 149.
32. Annual Report of the Bureau of the Census, year ending June 30, 1938, pp. 33, 37 [reprinted from the Annual Report of the Secretary of Commerce . . . 1938], series 192.
33. Director to Ryan, Nov. 25, 1938, file 57-B, series 142.
34. Annual Report of the Bureau of the Census, year ending June 30, 1939, p. 40 [reprinted from the Annual Report of the Secretary of Commerce . . . 1939], series 192. Given as 2,292 on p. 45.
35. Annual Report of the Bureau of the Census, year ending June 30, 1940, p. 59 [reprinted from the Annual Report of the Secretary of Commerce . . . 1940], series 192.
36. Cover Sheet, "File 362-WPA Projects conducted in New York City," n.d., folder 26-13, file 362, series 149.
37. Chief Clerk to Devor, June 6, 1939, folder 26-13, file 362, series 149.
38. Memorandum, Austin to Dr. Reed, Nov. 22, 1939. See also memorandum, "Committee" to the Director, Dec. 1, 1939. The piggybacked studies were Dr. Dedrick's fertility study and Dr. Truesdale's reliability studies. The reliability study would compile the age, race, nativity, and other basic characteristics of a small sample population as reported on the 1900 through 1930 or 1940 censuses, inclusive. Onondaga County, NY, was chosen for this purpose. See also Somervell to Austin, Apr. 10, 1940; all of these records in folder 26-14, file 362, series 149.
39. Memorandum, "Committee" to the Director, Dec. 1, 1939, folder 26-14, file 362, series 149. Committee members were Asst. Dir. Vergil D. Reed, chair, and Edward J. Gardner, Timothy F. Murphy, Calvert L. Dedrick, Arthur J. Hirsch, and Gerald Ryan.
40. Paper headed "Local WPA Projects Conducted in New York City," June 30, 1941, folder 26-14, file 362, series 149.
41. Memorandum, "Committee" to the Director, Dec. 1, 1939, folder 26-14, file 362, series 149.
42. Somervell to Austin, Apr. 10, 1940, folder 26-14, file 362, series 149, with copy in file 57-B; series 142.
43. Paper headed "Local WPA Projects Conducted in New York City," June 30, 1941, folder 26-14; file 362; series 149.
44. Annual Report of the Secretary of Commerce for the Fiscal Year Ended June 30, 1941, p. 58, series 192.
45. T. F. Murphy to Irving Huie, Administrator, WPA, New York, NY, May 8, 1941, File 362A, series 149.
46. Ibid. For similar data and reasons, see "Index of 1930 U.S. Census" WPA Project Proposal, New York City, State Application 15006 [undated carbon copy], attachment regarding item 28, file 362A, series 149.
47. "Index of 1930 U.S. Census" WPA Project Proposal, New York City, State Application 15006, undated carbon copy, file 362A, series 149. The author found no records describing any indexing work done as a result of the application.
48. J. C. Capt, Director, to Irving Huie, Administrator, WPA for the City of New York, Aug. 31, 1942, file 26-15, series 149. The next day, the 1930 census volumes returned to the Census Bureau building, Suitland, MD, by truck.
49. "Index of 1930 U.S. Census" WPA Project Proposal, New York City, State Application 15006 [undated carbon copy], attachment regarding item 26, file 362A, series 149.
50. Capt to Huie, July 23, 1942; see also Capt to Elliott M. Murphy, July 20, 1942, both in folder 26-15, file 362, series 149. Murphy resigned July 24, 1942, to join the Remington-Rand Corporation.
51. Memorandum, F. R. Pitman to Mrs. Almon, Mar. 13, 1943, folder 26-15, file 362, series 149.
52. W. B. Carter to Pitman, Mar. 10, 1943, folder 26-15, file 362, series 149.
Claire Prechtel-Kluskens is a microfilm projects archivist at the National Archives and Records Administration specializing in records of high genealogical value.
|Articles published in Prologue do not necessarily represent the views of NARA or of any other agency of the United States Government.|