Foreign Affairs

Cold War International History Conference: Paper by Margaret O'Neill Adams

Session VI: Non-traditional Resources and Research Opportunities

Historical Electronic Records Related to the Cold War Preserved in the National Archives and Records Administration

by Margaret O'Neill Adams
Reference Program Manager, Center for Electronic Records
Electronic and Special Media Records Services Division
National Archives and Records Administration

Copyright 1998 by Margaret O'Neill Adams.


Data files constitute most of the electronic records currently accessioned into the National Archives. They come from the use of data processing technologies, tools that have supported the running of government and business throughout the 20th century. The West has dominated the data or information processing industry from the beginning of the century, through the epoch of the Cold War, to the present. Note 1 It follows then that data processing applications supported a wide range of U.S. federal government programs related to the Cold War and that the electronic records that remain from computer-based data processing are valuable primary source material for this epoch. Further, electronic records series include technical documentation for their respective data files, adding to the corpus of primary sources.

Information is recorded in data processing records as coded alphabetic or numeric characters, or some combination of such values, sometimes interspersed with literal textual fields. Most frequently, this form of electronic records is not suited to traditional "reading" by a researcher, not only because the records consist of digital bytes, but also because they do not represent narrative material. However, technical documentation describes the structure of the records, identifies the values of any coded data elements, and includes other information necessary for using the electronic data. Sometimes there is additional contextual or supplementary information. Most technical documentation for electronic records from the Cold War era is eye-readable textual material, usually on paper or microfiche, that may offer evidence or perspective not readily accessible in any other records.

The National Archives and Records Administration (NARA) has had a program for electronic records (formerly called machine-readable records) since 1969. At present, the Center for Electronic Records is the unit in NARA with custodial responsibilities for this records type. The Center's program does not include retrospective conversion of eye-readable paper documents into a digital format. NARA has a digitization project underway, known as the Electronic Access Project (EAP), but the Center has no programmatic relationship to the EAP.

Most of the electronic records in the Center's holdings have been created since the 1960s, when the U.S. federal government adopted mainframe computer technology for data processing. However, the Center's holdings also include data files and their technical documentation from the earliest pre-mainframe days of the Cold War. Some of these are records that agencies originally created on punched cards. In most such cases, the creating agencies migrated the punched card records they needed to an electronic format when they adopted newer technologies than the card tabulation equipment used in the first 60 years of applied data processing. By the time agencies transferred these records to NARA, they were in digital files and came to be called electronic records. Since those early days of computer-based data processing, the universe of electronic records has expanded to include many different kinds of electronic records, but data files were the first kind.

A more recent category of electronic records, familiar to anyone who uses a personal computer, is the kind we create with office automation utilities or tools like word-processing or e-mail or other messaging systems. They generally resemble traditional archival records, i.e., documents, in ways that electronic data records do not. Most electronic records created with office automation store free form text and are human-readable and comprehensible when displayed or printed. This explains, at least in part, why contemporary historians, archivists, journalists, lawyers, and judges are preoccupied by the challenges of preserving for posterity the essential evidence created with office automation. While this attention is warranted, it would be unfortunate if it obscured the research value of electronic data records.

To date, electronic records from office automation applications have not been transferred to NARA in substantial volume because their creators are still using them or because they are too new. While we anticipate receiving a variety of the newer kinds of electronic records series in the near future, we have very few such series at present, and they are not the focus of NARA's current electronic records custodial activities. Thus this paper does not discuss them.

While data files are the earliest form of electronic records, they continue to be produced and used by agencies of the U.S. federal government -- and by governments, businesses, institutions, and organizations everywhere. The computerization of information in the 20th century makes retention of, access to, retrieval from, or analysis of data possible in ways that simply were not practical when data were recorded on clay tablets, in manuscript formats, or in modern types of paper records. Data recorded on all those media tended to be voluminous and expendable. Punched cards added to the volume, and few people appreciated their documentary value for secondary analysis. So most federal data records series from the punched card era that agencies did not migrate to digital formats were not appraised as having long term value. Electronic data files are the descendants of punched card files, albeit in a format that makes the issue of volume relatively moot. Given their lineage, however, it has taken a while for many people to recognize the long-term value of this post-World War II form of documentary material. Fortunately, some understood. As a result NARA now preserves a significant body of electronic records from the early application of computer-based data processing in governmental programs, including hundreds of files related to the Cold War.

Electronic Records from the Cold War

As mentioned above, most of NARA's electronic records date no earlier than the 1960s, when the U.S. federal government adopted mainframe computer technology as a basic tool supporting many of its programs. Coincidentally, and research may confirm causally, this technological adaptation came in the midst of the Cold War. Note 2

A significant portion of the over 100,000 electronic records files in NARA's custody (well over 400 gigabytes) are quantitative data. The federal government relies upon computer-based data processing to facilitate large-scale records collection, management, and analysis. Historical research concerning many late-20th century federal policies and programs necessarily requires analyzing these data. This does not mean, however, that the research value of electronic data files is limited solely to quantitative analyses. Researchers who do not anticipate embarking on data analysis may nonetheless wish to review technical documentation for relevant electronic records series. In the course of doing so they may discover that technical documentation can be revelatory in unexpected ways. Note 3

The undated technical documentation for the Army's Korean conflict casualty electronic records file illustrates discoveries relevant for Cold War research. It lists the values for each of the coded data elements in the records of 109,975 deceased, wounded, missing, or captured U.S. Army casualties of the Korean war. The values for the coded "type of casualty" field, for example, suggest an expectation that some Army personnel might die as a result of being gassed in action or as a result of radiation received in action, or that some might be seriously wounded in those ways. Through analysis of the data, we know that none of the records contain these codes. Yet their presence in the technical documentation suggests that in the minds of the Army personnel who designed this record-keeping system, the possible use of lethal gas or radiation was sufficiently plausible that they assigned individual codes to these potential casualty types. The system's planners also seem to have anticipated widespread warfare throughout the Far East. For the codes for "place of casualty" included, in addition to South Korea Sector and North Korea Sector, the Indo-China Sector, Tibet Sector, Mongolia Sector, Honan Sector (sic), Manchuria Sector, North Japan Sector, South Japan Sector, South China Sector, and Formosa Sector. In 99.9 percent of the records, only the codes for North and South Korea Sectors appear, and most of the variations in the remaining 92 records are codes that are not in the documentation. Thus they may represent mispunched data. The codes for "race" of the casualties reveal a blurring of race, ethnicity, and nationality in the U.S. Army of the 1950s, and may reflect a similar cultural phenomenon in the general population. They may also suggest those whom the Army was targeting for recruitment. The "race" codes were White, Negro, Chinese, Japanese, Hawaiian, American Indian, Filipino, Philippine National, Puerto Rican (other than Negro), Puerto Rican (Negro) and "others." Note 4

Technical documentation from opinion polls provides a different form of evidence and can tell its own story, independent of the responses or their analysis. NARA has a significant body of electronic data files and their documentation from polls gathered throughout the world in Cold War-related research by the Department of Defense and by the United States Information Agency (USIA). NARA's textual (paper) records also preserve case files and/or reports of the official government analyses of poll data, often products of the very same polls for which NARA preserves the original individual responses in an electronic format.

Some of the earliest opinion polling undertaken on behalf of the U.S. government predates the Cold War. Among the documentary materials of this activity are records of the Army Research Branch from a project that encompassed well over 200 surveys of the opinions of U.S. military personnel during and at the end of World War II. They are collectively known as "The American Soldier in World War II." Note 5 Some of the questions posed in these surveys are prescient in using terminology that a few years later would be considered reflective of Cold War thinking. One series of 15 polls with electronic records in NARA's holdings, taken between June 1944 and August 1945, queried soldiers concerning a variety of issues related to the post-war world. In the technical documentation for a survey of 2981 enlisted men in the European Theater interviewed in August 1945, just after the news of victory over Japan, we find a simple question like: "How do you think we will get along with Russia after the war?" Immediately following, the question was repeated, substituting Britain for Russia. A bit further into each interview, the soldiers were asked whether they agreed or disagreed with statements like, "Russia is more interested in dominating or controlling the world than she is in building a truly democratic world. Britain is more interested..."

In this case the technical documentation includes a distribution of the frequency of each category of response to each question. From it we learn, for example, that 32 percent of the soldiers responded to the first question above by choosing the statement "we will disagree about some things [with the British] but manage to get along," while 24 percent indicated the same expectations for future relations with Russia. More positively, 41 percent of the soldiers disagreed with the second statement, that Russia was more interested in dominating the world than in building a truly democratic world, while 45 percent disagreed with this statement as applied to the British. Note 6 Researchers interested in analyzing these responses by controlling for individual characteristics that might have influenced the soldiers' responses, or in comparing the relationship of these responses to others, can acquire copies of the complete data files and analyze the records, for as long as they wish, using whatever computing hardware and software they have available for their work.

An even more substantial collection of federal poll data are from the United States Information Agency (USIA), Record Group 306. From the beginning of its creation in the mid-1950s to the present, the USIA has relied upon opinion surveys in countries throughout the world to evaluate its informational programs and publications, as well as to gauge international public opinion on a wide-ranging set of foreign policy issues. Frequently the issues surveys seek comparative measures of how respondents perceived the relative positions of the United States and the Soviet Union. A recent article by two NARA archivists provides an overview of the rich body of research records that USIA has transferred to NARA in the form of textual reports and memoranda based upon the poll data.Note 7 NARA's electronic records collection includes over 1400 electronic files with the original data on which at least some of those analytical reports are based. The electronic records that USIA has transferred to NARA span the Cold War era from 1955 through 1989, but are more comprehensive for surveys taken from the 1970s forward than for the earlier decades.Note 8

Electronic records preserved at NARA also reflect the uses of data processing technology for basic record-keeping on the administration of foreign military sales, assistance, and direct defense expenditures and receipts throughout the world during the Cold War, and on the "prime" contracts awarded by the Department of Defense during this period. All are in Record Group 330, Records of the Office of the Secretary of Defense. They include Foreign Military Sales, Item Detail and Case Report Systems files, FYs 1950-1986. These data record sales of military equipment and services by the U.S. to allied and friendly foreign countries under the terms of the Foreign Military Assistance Sales Act, as amended. Data for FYs 1950-1963 are combined, but after that the records of the Master Item Detail file appear to be on an item sold basis. There are over 300,000 records in the file; among the data elements are identification of the country and the values of the items sold.

A related series comes from the Military Assistance Program. A single file, known as the Master Program/Delivery Detail file for FYs 1950-1986, records military assistance by the U.S. to friendly foreign countries, in accordance with the Foreign Assistance Act of 1961, as amended. Approximately 160,000 records identify the country to which the U.S. gave loans and grants for specific military equipment and/or training to senior military officers in the U.S., and detail about the loan or grant. As with the foreign military sales data, military assistance data for FYs 1950-1963 are cumulative.

The third series are the data known as the International [military] Balance of Payments (IBOP) for FYs 1960-1985. The Department of Defense compiled more than 160,000 IBOP records in order to report to the Bureau of Economic Analysis, Department of Commerce, and to the Department of the Treasury on the outlays for goods and services purchased abroad under all U.S. defense programs by the Military Departments and the Defense Agencies. There are six categories of expenditures. One of them captures information on U.S. payments for the commonly funded NATO infrastructure program. Each record identifies the U.S. military service involved, the recipient country, and the goods or services and their dollar value.

Military Prime Contracts Files cover contracts valued at $10,000 or more awarded by the Department of Defense FYs 1966-1975, and the Defense Contract Awards Data System records similar information beginning in FY 1976 for contracts valued at $25,000 or more. Each of the above series uses some data elements in common with one or more of the other series, or has a code serving as a cross reference. Therefore, taken together or individually, all of these electronic records series, initially created to serve an administrative function, represent valuable primary documentation for any researcher seeking to analyze the economics, as well as policy aspects of U.S. military sales, assistance, direct expenditures and contracts during the Cold War.

NARA's electronic records also include an example of a large-scale database system that served to abstract and catalog major collections of textual material in the U.S. Agency for International Development (AID). AID's Development Information System dates from 1976, but includes records on documents and projects from FY 1950 through mid-1989. Its databases form an institutional memory for the development assistance projects and programs AID has supported worldwide.

Electronic records at NARA also document aspects of the hot wars in which the U.S. military participated during the Cold War era. From the Korean war, NARA has two series of electronic records on casualties, one of which we discussed above, and two others on repatriated POWs.Note 9 Further, NARA preserves a substantial collection of electronic records on casualties from throughout the Southeast Asian combat area and a mass of detailed electronic records on air, land, and sea operations during the Vietnam conflict. Most of these records are well-described elsewhere. Note 10

Electronic records at NARA document another engagement of the U.S. military during the Cold War, the 1983 military action on the island of Grenada. The combined forces of the U.S. military and of the Organization of Eastern Caribbean States found roughly 35,000 pounds of what they considered to be documentary evidence of the activities of the Grenadian government and New Jewel party leaders. It was brought to the U.S., where the Defense Intelligence Agency microfiched it and created an electronic index to the microfiche. While the U.S. returned the "captured documents" to Grenada, NARA preserves the electronic index and the microfiche in R.G. 307, Records of the Defense Intelligence Agency.

All of the electronic records related to the Cold War discussed above came from a remarkable number of programs within the Department of Defense and from two then-independent foreign policy-related agencies, USIA and AID. However, the impact of the Cold War on the U.S. government and society was more pervasive than the above selection of electronic records suggests. It affected U.S. federal record-keeping in subtle and surprising variety, well beyond the series discussed so far. Thus we turn to describing a few of the electronic records series at NARA that offer evidence of some of the non-defense, non-foreign policy facets of the Cold War.

Since the end of World War I, the U.S. government has kept rosters on scientific and technical personnel to use in the event of a national emergency. Shortly after the National Science Foundation (NSF) was established in 1950, at the beginning of the Cold War, NSF became responsible for producing the registers and expanded their scope. They approached this task systematically, relying on data processing and collaborating with professional scientific associations nationwide. NARA now preserves the electronic National Register of Scientific and Technical Personnel for 1954, 1958 and biennially through 1970; and the National Engineers Register for 1964, 1967, and 1969, in R.G. 307, Records of the National Science Foundation. They offer insight into scientific manpower planning during some of the Cold War.

The records in the 1954 register file include detailed information on each scientist's selective service classification, military or reserve status, foreign language proficiency, foreign geographic area knowledge, dates when this knowledge was gained, and how it was gained. Curiously, by 1970, as if the information was no longer relevant, despite the war then raging in Southeast Asia, the data elements exclude any selective service or military status information but do capture information about the scientist's foreign language proficiency. By 1970, the overall focus was each person's areas of scientific specialization. Note 11

A recent transfer of electronic records from the Federal Emergency Management Agency (FEMA) provides yet another perspective on the pervasive nature of Cold War planning. The files of the National Shelter Surveys include records on approximately 1.8 million buildings, located in all fifty states, collected from 1964-1992. The authority for this massive data collection was the Federal Civil Defense Act of 1950, as amended. Some of the geographic codes in the records use an Office of Civil Defense coding scheme dating from the 1960s. The shelter surveys identified and described buildings that were suitable for population protection against the effects of a nuclear attack and for reception and care from the effects of a natural or technological disaster. Note 12

Some of the most telling human dramas of the Cold War relate to the waves of immigration of political refugees to the U.S. through this period. Electronic records now preserved at NARA from the Office of Refugee Resettlement (ORR), Department of Health and Human Services (DHHS), provide comprehensive data of the refugee population, 1975-1990. This record-keeping enabled DHHS to fulfill its requirement under the Immigration and Nationality Act, as amended by the Refugee Act of 1980, to report to the Congress annually on the Refugee Resettlement Program. The ORR Master File, 1975-1990, in R.G. 292 (Records of the Administration for Children and Families) uses an immigration status code for each of the 1.2 million persons with a record in the file to indicate his or her qualification for the refugee resettlement program. The codes identify those who meet the definition of refugee or asylee as found in the Refugee Act. The same field separately identifies Amerasians, qualifying immigrants, parolees (humanitarian or public interest) who arrive as part of a "refugee flow" but who do not qualify for refugee status and so do not qualify for ORR benefits, qualifying nonimmigrants, entrants (a term applied "to certain Cuban and Haitian nationals"), permanent resident aliens, naturalized U.S. citizens, and deceased persons. Each record captures information on country of origin, educational background, English-language facility, and occupational experiences of the refugees, suggesting the focus of the resettlement program. Note 13

As we have discussed, government and the general society have been using computers and predecessor tabulating machines to facilitate recordkeeping throughout the 20th century, especially in the past four decades. Thus fewer and fewer mid- to late 20th century historical subjects can be comprehensively researched without at some point involving analysis of relevant electronic data records. This essay is selective rather than comprehensive in suggesting the scope of federal electronic data records on Cold War topics that NARA preserves. Researchers who utilize any of the electronic primary source material we have described have an opportunity to discover for themselves the conclusions to be drawn from analysis of the records. In the course of their discovery, they may also gain new perspectives on the evolving relationship of technology and the historical record.

End Notes

  1. James W. Cortada, Before the Computer: IBM, Burroughs, and Remington Rand and the Industry They Created, 1865-1956, Princeton, N.J.: Princeton University Press, 1993. See especially the section on "Technological Innovation and Transfer" in the Conclusion, pp. 281-286. [Back]

  2. Cortada, pp. 222-246. [Back]

  3. All of the electronic records files discussed in this paper that had some security classification have been declassified. Access to all of NARA's federal electronic records files, as to all of NARA's federal records is subject to the terms of the Freedom of Information Act (FOIA). [Back]

  4. Documentation for the Korean Conflict Casualty File, 1950-53, Records of the Office of the [Army] Adjutant General, R.G. 407, The National Archives at College Park (MD). [Back]

  5. Benjamin L. DeWhitt and Heidi Ziemer, compilers, Records Relating to Personal Participation in World War II "The American Soldier" Surveys, Reference Information Paper 78, Washington, D.C.: National Archives and Records Administration, revised 1997. [Back]

  6. Documentation for The American Soldier in World War II (AMS), AMS-235: Attitudes Toward Post-Hostilities Problems, Records of the Office of the Secretary of Defense, R.G. 330, The National Archives at College Park (MD). [Back]

  7. Kenneth W. Heger and David A. Langbart, "An Untapped Resource: Research Records of the United States Information Agency," SHAFR NEWSLETTER, Vol 29, No. 2, June 1998. [Back]

  8. See a description of some of the electronic poll data from USIA preserved by NARA in, Tim Wehrkamp, compiler, A Brief Introduction to National Archives Records Relating to the Cold War, Reference Information Paper 107, Washington, D.C.: National Archives and Records Administration, 1998, pp. 10-11. [Back]

  9. Tim Wehrkamp, compiler, American Prisoners of War and Missing-in-Action Personnel from the Korean War and During the Cold War Era, Reference Information Paper Number 102, Washington, D.C.: National Archives and Records Administration, 1997, pp. 82-86. [Back]

  10. Donald Fisher Harrison, "Machine-Readable Sources for the Study of the War in Vietnam," Databases in the Humanities and Social Sciences - 4: Proceedings of the International Conference on Databases in the Humanities and Social Sciences (July 1987), pp. 179-191; Donald Fisher Harrison, "Computers, Electronic Data, and the Vietnam War," Archivaria, No. 26 (Summer 1988), pp. 18-32; Margaret O'Neill Adams, "Vietnam Records in the National Archives: Electronic Records," Prologue, Vol. 23, No. 1 (Spring 1991), pp. 76-84; Charles E. Schamel, compiler, Records Relating to American Prisoners of War and Missing in Action from the Vietnam War, Reference Information Paper 90, Washington, D.C.: National Archives and Records Administration, 1996, pp. 26-32. [Back]

  11. Documentation for The National Register of Scientific and Technical Personnel: 1954, 1958, 1962, 1964, 1966, 1968, 1970, Records of the National Science Foundation, R.G. 307, The National Archives at College Park (MD). [Back]

  12. FEMA Manual 9620.1/October 1990, National Shelter Survey Instructions, pp. 1-1 to 1-4 from Documentation for National Facility Survey/Reception and Care (NFS/RAC) Survey Files, Records of the Federal Emergency Management Agency, R.G. 311, The National Archives at College Park (MD). [Back]

  13. Documentation for Office of Refugee Resettlement (ORR) Master File, 1975-1990, Records of the Administration for Children and Families, R.G. 292, The National Archives at College Park (MD). [Back]