Historical Statistics of the United States is the premier source of quantitative evidence on American economic, social, political, demographic, and institutional history. Introduced in 1949 as a time-series supplement to the Statistical Abstract of the United States, it has inspired similar efforts in countries around the world. At the time of this writing, when the current edition is a quarter-century out of date, Historical Statistics of the United States is still a basic reference generating hundreds of citations annually in academic, professional, and journalistic publications. This essay describes the origins of this influential work and how it evolved over time.

Historical Statistics of the United States was initially developed by the U.S. Census Bureau as an historical supplement to its Statistical Abstract, a convenient one-volume compendium of economic, social, political, and demographic data, reported for the United States as a whole and by state, compiled from data collected by the highly decentralized federal statistical system.¹  Statistical Abstract has been published annually (with only a few exceptions) since 1878 by various statistical units of the federal government.²

In 1943, it became the responsibility of the Statistical Reports Section in the Office of the Assistant Director for Statistical Standards, Bureau of the Census. Shortly thereafter, most of these historical statistics, generally uneven in coverage and quality, were dropped from the Statistical Abstract to make room for expanded coverage of contemporary subjects.

The Census Bureau developed the idea of supplements to the Statistical Abstract in the 1940s to satisfy a growing appetite for quantitative data. Statistics were becoming increasingly important as a form of evidence, especially in the social sciences, and new statistical methods were being developed to analyze quantitative data. Many public policy concerns of the time focused on the distribution of income and assets, stimulating interest in statistics that could facilitate comparisons between subgroups of the population. Other public policy concerns focused on the rate of economic growth and the international competitiveness of the United States. Economists, impressed by the wartime mobilization experience and hoping to avoid another Great Depression, were studying the determinants of economic growth and stability. The newly published theories of John Maynard Keynes generated an interest in national income growth as an aggregate phenomenon.

The first edition of Historical Statistics of the United States was published as a Statistical Abstract supplement in 1949, following on the heels of the Cities Supplement (1944) and the County Data Book (1947).³  The Cities Supplement and the County Data Book reported up-to-date figures for small geographic areas, designed to reveal phenomena obscured by aggregation to the state and national levels. Historical Statistics of the United States was a companion volume designed to provide consistent time-series data for the national aggregates covering as long a period as possible.

Historical Statistics of the United States was originally envisioned as a serial that would be published at one-decade intervals, with mid-decade updates that would include adjusted data and new figures for recent years. The first edition of Historical Statistics of the United States, Volume I of the series, appeared in 1949, and the Continuation to 1952 was published five years later. Volume II was published in 1960 as the second edition, Historical Statistics of the United States: Colonial Times to 1957, and was followed by the Continuation to 1962. Volume III, subtitled Colonial Times to 1970, was published in 1975 as the Bicentennial Edition (U.S. Bureau of the Census 1949b, 1954, 1960, 1965, and 1975). This third edition marked the end of Census Bureau publication of historical compendia. The present Millennial Edition has been prepared by the academic community (see the Preface).

Back To Top

The original Historical Statistics of the United States project at the Bureau of the Census was headed by Morris B. Ullman, at that time Chief of the Statistical Reports Section in the office of Morris H. Hansen, Statistical Assistant to the Director of the Bureau of the Census.⁴  Historical Statistics of the United States was only one of many new statistical reporting projects developed at the Bureau of the Census by Ullman and Hansen, a team referred to as "The Two Morrises.”⁵ The Office of the Assistant Director at that time was a place of innovation and intellectual stimulation, where professionals on the Bureau of the Census staff interacted closely with academics and other nongovernmental researchers at the forefront of the rapidly advancing science of statistics. The Statistical Reports Section was designing and implementing new, user-friendly ways to disseminate data to a wider audience. The Statistical Methods Section was developing sampling theory and applying it to the survey activities of the Bureau of the Census, playing a major role in establishing the modern fields of statistics and sampling theory.⁶  Hansen frequently testified before Congress about the Census Bureau's important new breakthroughs in data-processing technology, especially the potential savings of time and money promised by the development of computers (Scott 1968; Duncan and Shelton 1978).⁷

In this heady environment, one of Ullman's first initiatives was to place a return postcard in all copies of the 1942 Statistical Abstract of the United States requesting feedback on its use and soliciting suggestions for improvement. This informal survey indicated that most copies went to libraries and to governmental agencies where they provide quick access to statistics for the writers of term papers, speeches, and reports. Suggestions for improvement included requests for more and longer historical series as well as more breakdowns of the aggregate into data for smaller areas (Ullman 1949). Size limitations made it difficult to include either of these in the Statistical Abstract of the United States itself, but the idea of developing supplemental publications for this purpose was appealing. The first "trial balloon,” no doubt benefiting from the new data-processing equipment that made it easier and less costly to produce small-area statistics, was the Cities Supplement published in 1945. Its success led to further work on the County Data Book that appeared in 1947. A time-series project was next in line, but its implementation was more ambitious and took longer to come into focus.

Like the Statistical Abstract of the United States itself, Historical Statistics of the United States was a collection of data from secondary sources in an era when every table had to be manually typeset. If it made relatively little direct use of new technology, it would benefit greatly from the statistical expertise of the Census Bureau staff.⁸  This staff included not only professional statisticians but also specialists in the collection, design, and editing of statistical tables, experts who worked with a similarly specialized unit at the Government Printing Office on the production of statistical reports. These in-house skills were a major asset for the production of Historical Statistics of the United States at the Bureau of the Census.

Historical Statistics of the United States also required the expertise of nongovernment statisticians scattered in universities, research institutes, and foundations throughout the country. These individuals were collected into a variety of professional associations that, in their turn, formed umbrella organizations for purposes that crossed disciplinary boundaries. The Social Science Research Council (SSRC) was such an organization, formed in 1924 by seven professional associations in the social sciences.⁹  The stated goals of the SSRC were to support basic research in the social sciences, to improve the infrastructure for social research, and to include the "human or social dimension” in the nation's scientific projects (Prewitt 2000).

When J. Frederick Dewhurst decided in 1945 to promote compilation of a source book of historical statistics, he turned to the American Economics Association (AEA), the American Statistical Association (ASA), and the SSRC for sponsorship. Dewhurst was the senior economist at the Twentieth Century Fund (TCF), a private foundation in New York focusing since the early 1930s on in-house projects to investigate "flaws in the economic system, and the positing of solutions to them” (Century Foundation 2000).¹⁰  In the mid-1940s, Dewhurst was Research Director for America's Needs and Resources, a major TCF undertaking that analyzed historical trends in production, income, and consumption to assess the U.S. economy's economic potential (Dewhurst 1947). Drawing heavily on new time-series data being constructed elsewhere, most notably at the National Bureau of Economic Research (NBER) in New York, the TCF project exemplified a growing interest in research that used historical data and applied quantitative methods for policy analysis.

In a memorandum dated April 12, 1945, Dewhurst proposed a new project "involving the completion and publication of a comprehensive source book of economic statistics” to be accomplished cooperatively by the AEA and the ASA. At its April 1945 meeting, the Executive Committee of the AEA named three members to a committee "to explore the merits and feasibility of such a project,” the Board of Directors of the ASA followed suit in May, and the Economic History Association was invited to participate as well.¹¹  Chaired by Dewhurst, it was called the "Joint Committee for a Source Book of Economic Statistics.”  The proposal was discussed by the SSRC Problems and Policy Committee, where "one member” suggested the possibility that the Census Bureau might be able to handle the preparation of such a book (Copeland 1946). The member was not identified in the minutes, but it may be noted that Phillip Hauser, one of the ASA members of the Joint Committee, was at that time Deputy Director of the Bureau of the Census.

The Dewhurst initiative struck a responsive chord at the Census Bureau in the Statistical Reports Section, where the historical supplement to the Statistical Abstract of the United States had been accepted in principle but not yet developed (Ullman 1949). By August 1945, letters had been exchanged between SSRC and the Bureau of the Census, the Secretary of Commerce had approved the project, and the Census Bureau had agreed to include the project in its budget request for the 1946–1947 fiscal year (Bell 1946, pp. 880–1; Copeland 1946). In November 1945, Dewhurst's Joint Committee met at the SSRC offices in New York City, and Hauser agreed that the Census Bureau would conduct "exploratory and planning work.”  Ullman's proposal for a time-series supplement to the Statistical Abstract of the United States thus became the basis of discussions between the Census Bureau and the committee members (U.S. Bureau of the Census 1946a). At its meeting in April 1946, the Joint Committee accepted the Census Bureau's proposal, declared the project feasible, and dissolved itself after only one year of existence.

The Bureau of the Census agreed to handle the publication, the working title of which was "Source Book on Economic Statistics.”  From the beginning, however, the Census Bureau recognized the ambitious scale of the project and the fact that its staffing requirements differed from those of their usual statistical reports. As observed in its report to the Joint Committee: "Because of the nature of the work involved, the amount of clerical work necessary for the preparation of the various chapters is greatly subordinated to the amount of professional effort needed…. It is, therefore, suggested that … the Joint Committee will recommend, and assist the Bureau in selecting and obtaining cooperation from, a sufficiently large number of consultants each of whom will be asked to prepare the material on a particular subject” (U.S. Bureau of the Census 1946a).

To implement this recommendation, the SSRC appointed a Committee on the Source Book of Historical Statistics (Advisory to the Bureau of the Census), again with Dewhurst as chairman. The members of this Advisory Committee were actively involved in the preparation of Historical Statistics of the United States, and their participation – as a committee and as individuals – is liberally acknowledged in the Introduction to the 1949 edition.

Even though it was common for research carried out by the federal government to draw on the expertise of advisory board members and to use nongovernment researchers as consultants, by all accounts the degree of collaboration between the Bureau of the Census and SSRC on Historical Statistics of the United States was unusually close and extensive (U.S. Bureau of the Census 1949b; Goldfield 1998; William Lerner, personal interview, Suitland, Maryland, August 5, 1998). The SSRC Committee on Research in Economic History, whose chairman (Arthur H. Cole of Harvard University) was a member of Dewhurst's committee, even provided a grant to support a full-time Executive Secretary. This position was filled by A. Benjamin Handler, who had been working with Dewhurst at TCF as Assistant Research Director for America's Needs and Resources. Handler would end up spending much of his time in Washington and worked so closely with Census on Historical Statistics of the United States that he is said to have functioned as a de facto member of the Statistical Reports Section.¹²

Back To Top

Initial guidelines for the Historical Statistics of the United States project were clear that the goal was to produce a statistical source book, asserting explicitly that "text material will describe the statistical significance of the data but will not involve analysis of their economic or social significance” (U.S. Bureau of the Census 1946c). But the compilation of historical data itself is often an example of economic or social research. Many of the nongovernment economists, demographers, and sociologists recruited by the SSRC had acquired their expertise in historical statistics through just such analyses, and this kind of research was often the framework for assessing the importance of including a particular series.¹³  Some controversy over the appropriate boundary between data reporting and substantive analysis was inevitable.

The pragmatic solution to this problem was to include detailed notes on the source of each series and to add a set of "substantive” appendices for special topics. Source notes for each series and section would present "the concepts used, the principal points of comparability, some discussion of the methods used in preparing the series, other information available which is not shown in the supplement, and such cross-reference information as is needed” (U.S. Bureau of the Census 1946b).

The original estimate was that on average a page of tables would require nearly a page of text. Specialists would need to look elsewhere for more details, but Historical Statistics of the United States could serve as a reference to provide direction for this search. This function of the Bureau of the Census as a clearinghouse for statistical expertise was deemed so important that the notes for each series and section in Historical Statistics of the United States would include as a matter of policy the names of individual contributors both inside and outside of government.¹⁴

Topical coverage for Historical Statistics of the United States was to take as its point of departure the Statistical Abstract of the United States, to which it was a supplement. Coverage in the Statistical Abstract was itself based on a prior consensus as to which statistics were both interesting and accurate, and the presumption was that these same criteria would make it useful to carry them back historically.¹⁵  This was not expected to be limiting, however, for if a "new” series were deemed worthy of being added to the Supplement, current data on the subject would be considered for inclusion in the Statistical Abstract (U.S. Bureau of the Census 1946a). The data from one should feed directly into the tables of the other, and series-by-series cross references were published both as an appendix to the Statistical Abstract of the United States and as source notes to the historical tables. The major exception would be "lapsed” historical series for which data were no longer collected, as in the size of the slave population, or for which the data had undergone important conceptual changes in construction, as in the early figures for housing and for manufacturing (U.S. Bureau of the Census 1946d; 1949b, Appendix II).

Coverage in Historical Statistics of the United States was generally to be confined to data for the continental United States as a whole, primarily for reasons of space but also because Historical Statistics of the United States was intended as a time-series supplement to the Statistical Abstract of the United States itself and not to its small-area supplements.¹⁶  Anticipated exceptions included cases where subnational statistics would be essential to interpreting the aggregate (as in merchant marine statistics, reported separately by coast and inland waterways), where an aggregate would not be meaningful (as in data on rainfall or internal migration), or where a subnational series could be interpreted as an indicator of the aggregate (as in prices on the New York Stock Exchange or the production of anthracite in Pennsylvania). Exceptions were also made for cases of important series that were available only for a limited geographical area, as for the early years in which figures were limited to the Atlantic seaboard (U.S. Bureau of the Census 1946d; 1949b, Appendix II).

Coverage in Historical Statistics of the United States was limited to series available on an annual or census-period basis that would extend back for at least twenty years. Some historical series were available on a monthly or quarterly basis, but it was decided that data presented at quarterly, monthly, or weekly intervals would be included only if they were deemed "of paramount importance.”  A substantive appendix on the turning points of business cycles would include "a few illustrative series of basic significance presented on a monthly or quarterly basis,” but readers were cautioned that "this should not be counted upon to round out a subject or time-period presentation in the main part of the volume” (U.S. Bureau of the Census 1946d).

The order in which time series were arranged in a Census Bureau report generally depended on its purpose, arranged backward in time (beginning with the latest date) to give perspective for current data or forward in time (beginning with the earliest date) to show "historical progression or development” (U.S. Bureau of the Census 1949a, paragraph 1122).¹⁷  Historical Statistics of the United States could be viewed either way, for while historians would find the forward ordering most natural, general readers writing a speech or report on current policy would typically focus on recent decades to provide a context for data in the Statistical Abstract of the United States (William Lerner, personal interview, August 5, 1998). The Census Bureau advisory committees were comprised mainly of historians and tended to favor forward ordering (Evans 1952). Because series varied widely in their earliest dates, however, forward ordering would have resulted in far more "white space” on the printed page and would have increased the publication's size and cost.¹⁸  For these and related technical reasons, Historical Statistics of the United States used backward ordering (U.S. Bureau of the Census 1949a).

Back To Top

By December 1946, the Bureau of the Census had prepared a plan for publishing 3,000 data series and was soliciting comments from outside experts that the SSRC had enlisted (U.S. Bureau of the Census 1946e). The experts, encouraged by the potential of the project, made suggestions for the inclusion of a number of additional series, some of which would have required major research projects in order to be developed. In the meantime, the Census Bureau funds for printing were set to expire in June 1947.¹⁹  Ullman solved both problems by recasting the 1949 edition as an extended prospectus, the "first draft” of a multistage project, enabling subsequent editions to benefit from the feedback of users. Preparation of the 1949 and 1960 editions should thus be viewed as two phases of a single project, the first relying primarily on the federal statistical system and the second including relatively more contributions from experts in academia and research institutes.²⁰

In the interest of speed, nearly all series in the 1949 edition were obtained within the federal government from individuals who already had a working relationship with the Statistical Reports Section.²¹  Contributors were asked to document sources, define terms, and provide explanatory notes for all changes, revisions, and adjustments necessary to compile the time-series figures. The "selection, assembling, posting, correction, preparing of tables for the printer, and the writing of text” were concentrated between October 1946 and June 1947, when the volume went to the printer, and any series not in hand by that date was simply dropped (Ullman 1949). This meant that most of the included series would have a counterpart in the Statistical Abstract of the United States, but it also meant exclusion for most "lapsed” series for which data were no longer collected. Other topics were missing because data were not readily available and there was insufficient time for finding and evaluating appropriate material.²²  Of the three essays originally proposed as substantive appendices, only one was prepared in time to be included.²³

Both the original prospectus for the 1949 edition and the final publication included 3,000 series arranged into fourteen major chapters. Many of the proposed series (and their corresponding chapters) had to be dropped, but the federal statisticians who responded in time provided more series than expected, and some of the planned chapters were split into two or more.²⁴  Often documenting their historical statistics for the first time, these statisticians uncovered a variety of unanticipated problems of fact and interpretation. The Census Bureau team (including Handler) focused on editing text for completeness and clarity. They fact-checked all notes and references, and by the time the final copy went to press, every source cited had been followed up by a member of the editorial staff in the Statistical Reports Section. The unexpectedly large number of changes considerably lengthened the elapsed time between galley and page proofs, although they reveal a process of feedback that improved the quality of federal statistics beyond the Census Bureau itself (Ullman 1949). An appendix in the 1949 Statistical Abstract of the United States updated the historical series for the years 1946–1948 and cross-referenced each series with its counterpart in Historical Statistics of the United States.

Despite the obvious limitations of its scaled-down first effort, the 1949 edition of Historical Statistics of the United States was well received. It satisfied an immediate need while at the same time stimulating suggestions for improvement and expansion. By September 1950 (about one year after publication), an advisory committee chaired by G. Heberton Evans Jr. had been appointed by the Economic History Association at the request of the Census Bureau. Its report, issued two years later, contained general and specific recommendations for revisions, most of which would be implemented in the 1960 edition (Evans 1952). Many of the shortcomings identified in the Evans report were the direct consequence of the decision to scale down the production of the 1949 edition, and the recommended revisions brought the 1960 edition in line with the original plan.²⁵  Thus, while the 1960 edition would have a total of 8,000 series arranged in twenty-four chapters (a substantial increase over the 3,000 series and fourteen chapters in the 1949 edition), its Introduction characterizes it as "intended to achieve the purpose foreshadowed in the original volume” (U.S. Bureau of the Census 1960, p. ix).

The 1960 edition of Historical Statistics of the United States was also prepared in the Statistical Reports Section in cooperation with the SSRC. Funded by a grant from the Ford Foundation, SSRC engaged more than 125 specialists as consultants on various chapters. These individuals are cited by name in the Introduction. Herman P. Miller directed the project at the Bureau of the Census, served as Executive Secretary of the SSRC Committee on Historical Statistics of the United States chaired by Evans, and participated in the selection of consultants. The final product would go through a cycle of five printings, and some 40,000 copies would be sold before the supply was exhausted.

At mid-decade, the Census Bureau published its Continuation to 1962 and Revisions (U.S. Bureau of the Census 1965).²⁶  By then, however, policy interests had shifted and neither the Census Bureau nor the SSRC was prepared to devote major funding to the preparation of a new edition of Historical Statistics of the United States. The Bicentennial Edition, published in 1975, was eventually undertaken by the Statistical Reports Division as an updating exercise. Drawing on previous experience and on voluntary contributions from a network of experts already largely in place, the major chapter headings remained the same as in the 1960 edition. The 1975 publication encompassed some 12,500 time series, an increase of more than 50 percent over the 1960 edition. Some of these were excluded from earlier editions because they had yet to meet the twenty-year-minimum criterion for a historical series and others because they had not yet been "discovered or properly developed.”²⁷

Back To Top

Historical Statistics of the United States was first developed in the mid-1940s by the Statistical Reports Section, Office of the Assistant Director for Statistical Standards, Bureau of the Census. The project fulfilled multiple goals of that unit. Produced as a supplement to the Statistical Abstract of the United States, it greatly expanded the Abstract's coverage of time-series data. As part of the Census Bureau's mission to improve statistical reporting, it raised standards for historical statistics and correspondingly stimulated the development of statistical expertise. As a joint project between the Bureau of the Census and the SSRC, it created and enhanced a network of professional experts on historical statistics.

When Frederick Dewhurst first proposed developing a source book for economic statistics, modern research in quantitative history was in its infancy. Expertise in the collection of historical statistics was the domain of a few widely scattered individuals in the statistical units of various departments and agencies of the federal government as well as in academia. Today quantitative historical research occupies a central place in the social sciences, in no small part because of the signal role played by Historical Statistics of the United States.

Back To Top

The spirit of two important teachers and mentors informs this essay: the late Morris B. Ullman, formerly of the U.S. Bureau of the Census, and the late Jonathan R. T. Hughes of Northwestern University. Carmella Chiswick acknowledges the helpful support of David Pemberton in the Census History Unit, Bureau of the Census, the suggestions of William Lerner and Edwin Goldfield, and the comments by Margo Anderson and Barry R. Chiswick.

Back To Top

Anderson, Margo J. 1988. The American Census: A Social History. Yale University Press.

Bell, James Washington. 1946. "Proceedings of the 58th Annual Meeting of the AEA: Report of the Secretary.” American Economic Review 36 (2): 870–86.

Century Foundation. 2000. "The Fund's First 75 Years (1919–1994).” Internet site of the Century Foundation.

Cole, Arthur H. 1950. "Review of Historical Statistics of the United States, 1789–1945: A Supplement to the Statistical Abstract of the United States.” Review of Economics and Statistics 32 (3): 273–4.

Copeland, Morris A. 1946. "Proceedings of the 58th Annual Meeting of the AEA: Progress Report on Source Book of Economic Statistics.” American Economic Review 36 (2): 934–6.

Deming, W. Edwards. 1950. Some Theory of Sampling. Wiley.

Dewhurst, J. Frederick. 1947. America's Needs and Resources: A Twentieth Century Fund Survey Which Includes Estimates for 1950 and 1960. Twentieth Century Fund.

Dubester, Henry J. 1950. Catalog of United States Census Publications, 1790–1945. U.S. Government Printing Office.

Duncan, Joseph W., and William C. Shelton. 1978. Revolution in United States Government Statistics, 1926–1976. Office of Federal Statistical Policy and Standards, U.S. Department of Commerce. U.S. Government Printing Office.

Eckler, Ross. 1972. The Bureau of the Census. Praeger.

Evans, G. Heberton, Jr. 1952. "On the Revision of Historical Statistics of the United States, 1789–1945.”  Report of the Committee of the Economic History Association (Advisory to the Bureau of the Census), April 15, 1952.

Goldfield, Edwin G. 1998. Personal interview. Suitland, Maryland, August 6, 1998.

Hammack, David C. 1999. "The Politics of Numbers: Historical Statistics of the United States and Public Policy, 1945–2000.”  Presented to the 1999 Annual Meetings of the Social Science History Association.

Hansen, Morris H., William N. Hurwitz, and William G. Madow. 1953. Sample Survey Methods and Theory. 2 volumes. Wiley.

Prewitt, Kenneth. 2000. "SSRC: A Brief History of the Council.” Internet site of the Social Science Research Council.

Scott, Ann Herbert. 1968. Census, U.S.A.: Fact Finding for the American People, 1790–1970. Seabury Press.

The Statistical History of the United States from Colonial Times to the Present. 1966. With an Introduction by Ben J. Wattenberg. Horizon Press.

The Statistical History of the United States from Colonial Times to the Present. 1976. With an Introduction and User's Guide by Ben J. Wattenberg. Basic Books.

Truesdell, Leon E. 1965. The Development of Punch Card Tabulation in the Bureau of the Census: 1890–1940. U.S. Government Printing Office.

Ullman, Morris B. 1949. "Statistical Abstract Program, with Special Reference to Historical Statistics of the United States.”  Presented to the American Statistical Association, Philadelphia Chapter, February 18.

U.S. Bureau of the Census. 1946a. "Report to Joint Committee on Status of Source Book Project.” April 1. Mimeographed, 3 pages plus exhibits.

U.S. Bureau of the Census. 1946b. "Source Book of Historical Statistics (A Supplement to the Statistical Abstract of the United States).” September 9. Mimeographed, 3 pages.

U.S. Bureau of the Census. 1946c. "Source Book of Historical Statistics: General Description of Project.” December 10. Mimeographed, 4 pages.

U.S. Bureau of the Census. 1946d. "Source Book of Historical Statistics: Basic Premises for Data Selection.” December 10. Mimeographed, 6 pages. Reprinted with minor changes as "Appendix II. Statement of Basic Premises” in U.S. Bureau of the Census, Historical Statistics of the United States, 1789–1945. U.S. Government Printing Office, 1949.

U.S. Bureau of the Census. 1946e. "Source Book of Historical Statistics: Detailed Outline.” December 10. Mimeographed, 21 pages.

U.S. Bureau of the Census. 1949a. Bureau of the Census Manual of Tabular Presentation: An Outline of Theory and Practice in the Presentation of Statistical Data in Tables for Publication. By Bruce L. Jenkinson. U.S. Government Printing Office.

U.S. Bureau of the Census. 1949b. Historical Statistics of the United States, 1789–1945. U.S. Government Printing Office.

U.S. Bureau of the Census. 1954. Continuation to 1952 of Historical Statistics of the United States, 1789–1945 with Revisions of Selected Series. U.S. Government Printing Office.

U.S. Bureau of the Census. 1960. Historical Statistics of the United States: Colonial Times to 1957. U.S. Government Printing Office.

U.S. Bureau of the Census. 1965. Historical Statistics of the United States, Colonial Times to 1957: Continuation to 1962 and Revisions. U.S. Government Printing Office.

U.S. Bureau of the Census. 1975. Historical Statistics of the United States: Colonial Times to 1970. U.S. Government Printing Office.

Back To Top