Census Indexes and Finding Aids
Census Indexes and Finding Aids The census is a clear reflection of population growth in the United States. The millions of names and figures added to the census totals over the years have made indexing, particularly of the most populous states, a formidable and expensive task. Recent developments in technology have facilitated indexing and publication, and now a significant and ever-growing number of later statewide and even nationwide census schedules have been indexed. While mistakes and omissions exist in census indexes, it is generally agreed that even an imperfect index can be an invaluable timesaver and is certainly better than no index at all. Improved technology and better editing are making most new compilations more inclusive and more accurate. Indexes on CD-ROM and Online Over the past several years, online census indexes have increased in both number and scope. This increase has been driven, largely, by a number of commercial or institutional indexing projects. Ancestry.com, for example, offers a searchable, every-name index of every available U.S. federal census with links to images of the actual census returns. Genealogy.com offers an every-name index to several U.S. federal censuses, including 1790–1820, 1860, 1870, and 1890–1910 on its subscriber site. Another site, Census4All.com, has an every-name index to the 1910 U.S. Federal Census with Vermont, Rhode Island, and New Hampshire presently available. The Census4All indexes can be searched for free, and a list of other members of the household or copies of the actual census schedules desired can be ordered for a fee. Unlike the 1880 Soundex, which is only a partial index, the LDS Church’s FamilySearch 1880 United States Census and National Index is an every-name index, with entries including name, relation to head of household, sex, marital status, race, age, birth place, occupation, and father and mother’s birth place. The index is searchable online at FamilySearch and as part of the census collection at Ancestry.com. 1880 Census CD-ROMs are also available for sale at the FamilySearch site <www.familysearch.org>. In addition to its U.S. federal census records, Ancestry.com also offers a large number of state and local census indexes. HeritageQuest’s Family Quest Archives CD-ROM collection, <www.heritagequest.com>, contains complete head of household indexes to U.S. Federal Censuses for the 1790, 1800, 1810, and 1870 censuses. Partial indexes are available for 1850, 1900, and 1910. Fortunately for researchers, libraries with good genealogy collections usually make it a priority to acquire these popular and important indexes, on CD or by subscription on the Internet, as soon as they become available. In addition to the commercial online census index projects, volunteers from around the country have responded to the need for finding aids by producing, in various formats, indexes to many of the federal censuses. One such volunteer effort, The USGenWeb Census Project can be accessed online at <www.rootsweb.com/~census/states.htm>. Other sites provide links to census data online. These include the following: Census Links.com <www.censuslinks.com> Census-Online.com <www.census-online.com> Cyndi’s List—Census <www.cyndislist.com/census.htm> Census Index Limitations A common mistake beginners make is to consult an index, find a name, extract the index information, and go no further. Many seem unaware that census indexes are simply finding aids. While there is a certain element of excitement in discovering an ancestor’s name in an index, there is greater satisfaction in store for those who view the fuller picture provided in the actual census schedules. A well-prepared index includes a preface explaining the index parameters (for example, whether it is an every-name index or if only heads of household are included) and identifying specific problems encountered in the process of compiling the index. The wise researcher will read every preface carefully. In most published census indexes, only the heads of households are listed. If an individual was a child when the census was taken, and if the name of his or her father, mother, or other head of the household in which he or she lived is not known, a long and tedious search may be in store. It may be necessary to look at different census schedules for every entry for a given surname in an index before the correct household is found. Regardless of the care taken by the creator of an index to make it accurate, no index is perfect. Omissions, misinterpreted names, and misspellings creep into virtually all census indexes. Some indexes are not useful for tracing individuals because information was culled from microfilm that was nearly impossible to read, and sometimes the microfilmed version itself lacks certain information. Examples of the latter are the published federal census indexes from 1790 to 1840. Like the censuses themselves, the indexes are of limited use in finding individuals because only heads of household are listed. Likewise, most post-1840 census indexes include only heads of household and “strays.” Frequently, names are actually included in an index but cannot be found because they are misspelled to the extent that they are unrecognizable. Some surnames have been incorrectly alphabetized when indexers could not decipher even the first letter of a surname. In some handwriting styles, the letter L resembles an S; thus, the handwritten surname Lee might become See in an index. Handwriting styles have caused indexing problems when certain similar-appearing letters have been confused, including T and F; J, G, and Y; I and J; K and R; O and Q; P and R; and U and W. Page Numbering Problems Pages of census schedules were originally numbered by the census taker; when the schedules were later arranged and bound, they were often renumbered with a hand stamp. It is common for some volumes to have two or more series of page numbers. A stamped number, when it is present, is usually the page reference used in printed census indexes. It is very important to determine which page number the census indexer was using. Sometimes in an index, for example, the indexer was inconsistent with the page number that he or she used, making it difficult to find names. History and Quality of Census Indexes Computer technology has revolutionized the process of indexing census schedules. Computer-generated census indexes are becoming increasingly available in book, microfiche, CD-ROM, and online forms. Despite the advanced technology, however, no index is error-free. Misinterpretations of handwritten census manuscripts and transcription mistakes continue to thwart research, particularly when the first letter of a name is entered into an index incorrectly. While a number of individuals and genealogical societies have used computers to create census indexes, most such indexes have been created by commercial firms. The oldest of these firms is Accelerated Indexing Systems. Accelerated Indexing produced indexes for every extant state and territory census through 1860, and some for later years, as well as a number of special censuses and census substitutes. These indexes are available online and in CD-ROM format (see page 185). The schedules for some states and areas have been indexed more than once by different organizations and commercial publishers. But though the year and the locality indexed may be the same, formats and contents can differ dramatically. Names may have been interpreted differently; some publications may include names missed by others; and some may include much more than county, township, and page and microfilm numbers after the names of heads of households. It is wise to check every index when more than one is available for a given time and place. Misspellings have occurred on several levels. The census enumerator may have misunderstood the name and written it incorrectly. (See table 5-1). Even if the enumerator got it right, the indexer may have misread the enumerator’s handwriting or had other difficulties reading the old and fading microfilm. (See table 5-2). Many indexes, up to and including the 1920 census, cover individual counties only. They can prove especially useful when a name or names cannot be found in a statewide compilation. Because local indexes are frequently compiled by genealogical societies and indexers who tend to be familiar with local name spellings and geographical distinctions, their reliability is sometimes greater than the larger indexes. Statewide censuses are sometimes interfiled with other sources in single personal name indexes available in state archives. The addresses of state archives and state historical societies are given in the appendixes section. Indexes from 1790 to 1840 The federal government led the way in publishing census indexes when, in the early 1900s, it published indexed volumes of the extant 1790 census schedules for each state. The individual state volumes have since been privately reprinted and are widely available in libraries with genealogy collections. Some indexes for the years 1790 to 1820 also include the tallies listed for each family. These tallies would be listed as such: 1790 a. free white males age sixteen and older b. free white males under age sixteen c. free white females d. all other free persons e. slaves 1800–10 a. free white males to age ten (under age ten) b. free white males to age sixteen (of ten and under sixteen) c. free white males to age twenty-six (of sixteen and under twenty-six) d. free white males to age forty-five (of twenty-six and under forty-five) e. free white males over age forty-five f. free white females to age ten (under age ten) g. free white females to age sixteen (of ten and under sixteen) h. free white females to age twenty-six (of sixteen and under twenty-six) i. free white females to age forty-five (of twenty-six and under forty-five) j. free white females over age forty-five k. other free persons (except Indians not taxed) l. slaves 1820 a. free white males to age ten (under age ten) b. free white males to age sixteen (of ten and under sixteen) c. free white males between ages sixteen and eighteen d. free white males to age twenty-six (of sixteen and under twenty-six) e. free white males to age forty-five (of twenty-six and under forty-five) f. free white males over age forty-five g. free white females to age ten (under age ten) h. free white females to age sixteen (of ten and under sixteen) i. free white females to age twenty-six (of sixteen and under twenty-six) j. free white females to age forty-five (of twenty-six and under forty-five) k. free white females over age forty-five l. slaves m. free colored persons Indexes from 1850 to 1870 Many statewide indexes for censuses after 1850 include only heads of household and the names of persons in households whose names were different than that of the household head. Obviously, then, a large percentage of the actual population of a state is excluded from such an index. This is especially a problem with common names, and when a child’s or woman’s name is known but that of the head of household is not. 1880 Soundex Until recent years, the fastest method for finding names in the 1880 census for most states was to use the Soundex, a partial index that includes only households with children ten years old and under in residence. Compiled by the Work Projects Administration (WPA), the Soundex index was designed to identify those who would be eligible for Social Security. (An explanation of the Soundex coding system follows this section.) It is important to remember when using the 1880 Soundex that, while a large portion of the population is not indexed because many families had no children ten years old or under, all individuals and families were supposed to have been included in the original census schedules. Some of the original Soundex index cards survive and have been distributed among various state and local agencies; others have apparently been destroyed. Some of the 1880 cards were lost or misfiled before or when they were microfilmed. Use the Soundex to determine surname distribution throughout the state. This can be an important clue if you don’t know which county to search for a family. You can identify family naming patterns (because each person in the family is listed on the Soundex card, with relationships stated), find orphaned children living with persons of other surnames, and identify grandparents living under the same roof. They are listed in the census schedule, even though they may not be indexed separately. As mentioned, the LDS Church offers an every-name index to the 1880 census in electronic format (see page 185). 1890 Index A card file to the names on the surviving 1890 schedules is available on two rolls of microfilm titled Index to the Eleventh Census of the United States, 1890 (National Archives microfilm M496). The index is also available in printed form. Ken Nelson, comp., 1890 Census Index Register (Salt Lake City: Genealogical Society of Utah, 1984), is an index to the 6,160 names in the surviving fragments of this census. Available on microfilm (1,421,673, item 11), it can also be found in the reference area of the Family History Library (Family History Library book Ref 973 X2n 1890). Also see Helen Smothers Swenson, Index to 1890 Census of the United States (cited earlier), and “Veterans Schedules, 1840–1890,” on page 197. 1900 Index The Soundex index to the 1900 census is regarded as one of the most inclusive and accurate of the federally-created indexes. It serves as an efficient key to locating households and individuals in the most genealogically informative census ever taken. Unlike the 1880 census, the 1900 census identifies all heads of household and every adult whose name is different from that of the head of household. 1910 Indexes The most notable problem with the 1910 census has traditionally been the lack of indexes for most states. Miracode (a slightly modified version of Soundex) and Soundex indexes exist for only twenty-one states: Alabama, Arkansas, California, Florida, Georgia, Illinois, Kansas, Kentucky, Louisiana, Michigan, Mississippi, Missouri, North Carolina, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Texas, Virginia, and West Virginia. Since Soundex/Miracode indexes are not available for the remaining states, researchers must rely on city directories, county landowners’ atlases, enumeration districts, or specially created finding tools (such as the special index to streets and enumeration districts for certain cities), or conduct tedious, page-by-page searches of the census schedules. (For a detailed description, see “Census Indexes and Finding Aids” on page 185.) Soundex and Miracode indexes were created by the Bureau of the Census for the twenty-one states that lacked a centralized vital statistics bureau at the time the indexes were created. The Miracode system uses the same phonetic code and abbreviations as the Soundex system, but Miracode cards list the visitation numbers assigned by the enumerators, while Soundex cards show the page and line numbers on the appropriate census schedules. With the exception of Louisiana, which uses both, the following states have been indexed using either the Soundex or Miracode systems: Alabama, Arkansas, California, Florida, Georgia, Illinois, Kansas, Kentucky, Michigan, Mississippi, Missouri, North Carolina, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Texas, Virginia, and West Virginia. Both indexing systems give the surname, first name, state and county of residence, city (if applicable), race, age, and place of birth, as well as the volume number and enumeration district number of the census schedule from which the information was obtained. Some large cities are indexed separately in the 1910 census. Be sure to see separate Soundex listings in the National Archives microfilm catalog for some metropolitan areas in Alabama, Georgia, Louisiana, Pennsylvania, and Tennessee. The 1910 Census City Street Finding Aid The Federation of Genealogical Societies (FGS) promoted and coordinated the funding to put to microfilm an important finding aid for the 1910 census. Known as the Cross-Index to Selected City Streets and Enumeration Districts, 1910 Census, it was produced in 1984 by the Bureau of the Census to facilitate its work of searching the original schedules for age and other personal data in response to inquiries from individuals and government agencies. This index to city streets and census enumeration districts for thirty-nine cities in the 1910 federal population census is widely available on fifty sheets of microfiche. The index enables users of the population schedules to translate specific street addresses into the appropriate enumeration district number and corresponding volume number of the microfilmed schedules. The city schedules were selected for indexing by the Bureau of the Census based on the frequency of requests for information. The indexes were originally in bound volumes, but they were unbound for microfilming. With the exception of several of the larger cities, the index for each city occupied a single volume. The original arrangement of the indexes has been preserved, with the exception that the boroughs of Manhattan, the Bronx, Richmond (Staten Island), and Brooklyn have been placed under the heading “New York City.” There is no index for the borough of Queens. Entries in the index give for each city a list of city streets and house numbers and show the appropriate enumeration district. The records are arranged alphabetically by name of city and thereunder by street. Named streets, arranged alphabetically, are listed first, followed by numerical streets. Immediately preceding the index portion of each volume is a table listing the enumeration districts covered in the volume, with a cross-reference to the corresponding volume of the original population schedules. The thirty-nine cities included in the 1910 index follow: • Akron, Ohio • Atlanta, Georgia • Baltimore, Maryland • Canton, Ohio • Charlotte, North Carolina • Chicago, Illinois • Cleveland, Ohio • Dayton, Ohio • Denver, Colorado • Detroit, Michigan • District of Columbia • Elizabeth, New Jersey • Erie, Pennsylvania • Fort Wayne, Indiana • Gary, Indiana • Grand Rapids, Michigan • Indianapolis, Indiana • Kansas City, Kansas • Long Beach, California • Los Angeles and Los Angeles County • Newark, New Jersey • New York City (excluding Queens) • Oklahoma City, Oklahoma • Omaha, Nebraska • Patterson, New Jersey • Peoria, Illinois • Philadelphia, Pennsylvania • Phoenix, Arizona • Reading, Pennsylvania • Richmond, Virginia • San Antonio, Texas • San Diego, California • San Francisco, California • Seattle, Washington • South Bend, Indiana • Tampa, Florida • Tulsa, Oklahoma • Wichita, Kansas • Youngstown, Ohio The 1910 street index can dramatically reduce the problems and time expenditure involved in searching large cities for which there are thousands of pages of census entries. 1920 Census Soundex The 2,074 rolls of microfilm for the 1920 census are Soundex indexed on 8,590 rolls of microfilm. The Soundex includes all of the states as well as the then territories of Alaska and Hawaii. The Canal Zone, Puerto Rico, Guam, American Samoa, the Virgin Islands, and military, naval, and various institutions are also indexed. 1930 Census Soundex The 1930 U.S. Federal Census was released by the National Archives and Records Administration on 1 April 2002. Soundex indexes are available for the states of Alabama, Arkansas, Florida, Georgia, Kentucky (only Bell, Floyd, Harlan, Kenton, Muhlenberg, Perry, and Pike counties), Louisiana, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, and West Virginia (only Fayette, Harrison, Kanawha, Logan, McDowell, Mercer, and Raleigh counties). To aid in locating entries for other areas, geographic descriptions of census enumeration districts are reproduced in NARA microfilm publication T1224. In addition, NARA has purchased a large number of city directories from a commercial vendor for use in its facilities. More details can be found on NARA’s website at <http://1930census.archives.gov>. Beyond the Index Experienced researchers know that there is much more to a census search than merely checking an index—whether that index is a book, a microfilmed version of the Soundex, or a computerized database. Unfortunately, too many beginners give up the search if the name sought does not appear in the index; if it does appear, they often seem content with the minimal information found in the index. Those who do not take the time to get the full picture provided by careful study of the actual census schedules usually miss important information and clues to further research. The study should include not only the subject of the search but the general area in which that person lived. To focus on only one name or one family in a given census is to see only a partial picture—somewhat like reading one chapter of a fascinating book.