² If the ‘new’ history is new at all, it is because it is less élite-oriented, it attempts to implement social science techniques – especially quantification and model building – and, at its best, it emphasizes the comparative nature of urban and social research.

³ Poverty and Progress: Social Mobility in a Nineteenth Century City. (Cambridge, Mass., 1964)Google Scholar.

⁴ See Thernstrom, Stephen, The Other Bostonians: Poverty and Progress in the American Metropolis 1880–1970 (Cambridge, Mass., 1973), especially pp. 222–3CrossRefGoogle Scholar, for a concise presentation of current studies, and Thernstrom, and Knights, Peter R., ‘Men in Motion: Some Data and Speculations about Urban Population Mobility in Nineteenth-Century America’, Journal of Interdisciplinary History 1 (Autumn 1970), 7–35CrossRefGoogle Scholar, for an overview of the question.

⁵ In his The Plain People of Boston, 1830–1860: A Study in City Growth (New York, 1971)Google Scholar, Knights performed the Herculean task of tracing a small number of individuals throughout the state of Massachusetts by searching in large numbers of local records. See especially his exhaustive approach in Chapter Six.

⁶ Although it should be noted that people's memories often are surprisingly misleading!

⁷ See, for example, Knights, Peter R., ‘City Directories as Aids to Ante-Bellum Urban Studies: A Research Note’, Historical Methods Newsletter 4 (09 1969), 1–10CrossRefGoogle Scholar; and Edward Pessen, ‘The Occupation of the Ante-Bellum Rich: A Misleading Clue to the Sources and Extent of Their Wealth’, Ibid., 5 (March 1972), 49–52.

⁸ On the general issue of the applicability of models in historical research, see Bogue, Allan G. (ed.), Emerging Theoretical Models in Social and Political History (Beverly Hills, 1973)Google Scholar (originally published as a special issue of the American Behavioral Scientist 16 [05–06 1973]Google Scholar). For good discussions of models and migration research, see Mangolan, J. J., with the assistance of Cornelia Morgan, Human Migration: A Guide to Migration Literature in English 1955–1962 (Lexington, Ky., 1968), pp. 1–20Google Scholar; Bogue, Donald J., ‘Internal Migration’, in Hauser, Philip M. and Duncan, Otis Dudley, The Study of Population: An Inventory and Appraisal (Chicago, 1959), pp. 486–509Google Scholar; and Lee, Everett S., ‘A Theory of Migration’, Demography 3 (1966) 47–57CrossRefGoogle Scholar. For presentation and testing of several hypotheses see, for example, Anderson, Michael, ‘Urban Migration in Nineteenth Century Lancashire. Some Insights Into Two Competing Hypotheses’, Annales de Demographie Historique 1971 (Paris, 1972), pp. 13–26Google Scholar; Gallaway, Lowell E., Vedder, Richard K., and Shukla, Vishwa, ‘The Distribution of the Immigrant Population in the United States: An Economic Analysis’, Explorations in Economic History 11 (Spring 1974), 213–26CrossRefGoogle Scholar; Jeffrey G. Williamson, ‘Migration to the New World: Long Term Influences and Impact’, Ibid., 11 (Summer 1974), 357–90; and various issues of the excellent International Migration Review. The Ellemers model is found in Ellemers, J. E., ‘The Determinants of Emigration. An Analysis of Dutch Studies in Migration’, Sociologia Neerlandica 2 (1964)Google Scholar and its application is in Akerman, Sune, Cassel, Per Gunnar and Johannson, Egil, ‘Background variables of population mobility: an attempt at automatic inter-action detector analysis. A preliminary research report’, The Scandinavian Economic History Review 22 (1974), 32–60CrossRefGoogle Scholar.

⁹ For a fuller description of the Soundex and its uses, see Stephenson, Charles, ‘Tracing Those Who Left: Mobility Studies and the Soundex Indexes to the U.S. Census’, Journal of Urban History 1 (11, 1974), 73–84CrossRefGoogle Scholar.

¹⁰ Briefly, the initial letter of the surname is not coded; the letters a, e, i, o, u, y, w, and h are ignored. The Soundex coding guide is: (1) b, p, f, v; (2) c, s, k, g, j, q, x, z; (3) d, t; (4) 1; (5) m, n; (6) r. Each Soundex Code is a letter and a three digit number. When two ‘key’ letters or equivalents are adjacent, they are coded as one. Lloyd becomes L300, Hildebrand is H431, Engebrethson is E521, Kuhne is K500, and Stephenson S315. The Soundex does not deal consistently with certain prefixes. McDonnell, for example, might be found under M235 or D540. The most concise guide to the Soundex is the ‘Guide to the Soundex System’, Federal Population Censuses, 1790–1890: A Catalog of Microfilm Copies of the Schedules (Washington, D.C., 1972)Google Scholar.

¹¹ The restrictions apply primarily to insure the privacy of those enumerated. While the 1900 Census can be used only in the National Archives, this includes the thirteen Regional Archives Branches throughout the country.

¹² The 1880 Soundex includes 2,367 reels, the 1900 is composed of 7,844.

¹³ Some gaps have been discovered in the 1900 Soundex, and the National Archives is in the process of correcting them.

¹⁴ It is interesting to note that, while Federal Censuses since 1900 remain closed, a large proportion of the population can be found in state censuses up to about 1935. For a guide to state censuses see Dubester, Henry J., State Censuses: An Annotated Bibliography of Censuses of Population Taken After the year 1790 by States and Territories of the United States (Washington, 1948)Google Scholar. New York, for example, took excellent censuses through 1925.

¹⁵ In 1885, for example, Florida, Colorado, Nebraska and the New Mexico territory were enumerated. Later censuses were taken – in El Paso, Texas, for example – but the enumeration books either are closed or no longer exist.

¹⁶ Double enumerations were also taken for New York City and Philadelphia in 1870.

¹⁷ Mrs. Elizabeth Trimmer, of the Chicago Regional Archives Branch was kind enough to find this information for me.

¹⁸ We are using the 1905 census only for New York County (Manhattan) at this time.

¹⁹ For other studies of the South, see Hopkins, Richard J., ‘Occupational and Geographic Mobility in Atlanta, 1870–1896’, Journal of Southern History 24 (05 1968), 200–13CrossRefGoogle Scholar; and ‘Status, Mobility, and the Dimensions of Change in a Southern City: Atlanta, 1870–1910’ in Jackson, Kenneth T. and Schultz, Stanley K. (eds.), Cities In American History (New York, 1972), pp. 216–31Google Scholar; and Worthman, Paul B., ‘Working Class Mobility in Birmingham, Alabama, 1880–1914’, in Hareven, Tamara K., Anonymous Americans: Explorations in Nineteenth-Century Social History (Englewood Cliffs, 1971), pp. 172–213Google Scholar. Each of these fine studies depended upon Federal Censuses, city directories, and other local records, and were thus confined to the immediate metropolitan areas. The lack of state census data for the South has been a major impediment to producing more wide-ranging studies.

²⁰ For more on this type of tracing, see Stephenson, ‘Tracing Those Who Left’.

²¹ St. Charles Biographical Directory and Census Report 1885 … (Chicago, n.d.).

²² The influence of fraternal orders in American working-class life is a crucial but surprisingly neglected topic. An important dissertation currently under way by Brian Greenberg, ‘Worker and Community: The Social Structure of a Nineteenth Century American City – Albany, New York, 1850–1886’, and research by David Montgomery at the University of Pittsburgh should go a long way toward stimulating interest and further research in this area.

²³ See Gallaway, Vedder, and Shukla, ‘Distribution of the Immigrant Population’.

²⁴ Because of the sheer bulk of the Soundex, confining research to only a few initial letters could make a more efficient use. Obviously social variables would be little affected by alpha selection – it has never been supposed, for example, that people whose last names begin with B, or T, or whatever, are richer or have more children than those whose last names begin with N or J. However, an ethnic bias is possible, and such a bias could affect other variables. We have tried to avoid obviously ethnically-oriented letters – Ms, as an example. Another problem which might cause difficulty is the likelihood of sampling an inordinate proportion of people from the same family. We are aiming at a sample size of approximately 1000 in most cases. While we believe that larger samples will be required for more far-ranging results, in each case the numbers we are drawing will be adequate to test the feasibility of our method.

²⁵ They are organized in Soundex order by initial letter of first name. Thus Allan A250 will precede Ben A250 or James A250.

²⁶ See the census compendia for rough information on migratory streams. As a matter of efficiency, and in so far as we are able, we will group together all those for whom we are searching in a particular state. Thus when we search New York in 1900 we will look not only for those who were in that state in 1905, but also in Florida in 1885, or Wisconsin in 1905.

²⁷ McGraw-Hill (New York) has just issued the long-awaited revised and updated manual for the most recent version of SPSS.