From many perspectives genealogy research seems
better than ever, with the future bright and shining. More of us are seeking our
ancestors, aided by a torrent of new books telling us how, what and where to research.
We look around and see all kinds of files recreated on CD-ROMs, many with indexes.
Searching PERSI on CD-ROMs will be a joy after wrestling with two dozen thick volumes.
The Internet has a multitude of sites devoted to genealogy, growing every day, with
the promise of rewarding on-line research. LDS members continue to microfilm new
records throughout the world, and the Family History Library continues to add to
its magnificent collection. Dozens of software programs, written exclusively for
genealogists, await our call to be compiled on cheaper, ever faster computers.
So prospects for 21st century genealogy look great, don't they? Well,
maybe. Maybe not.
In 1996 the National Genealogical Society and the Federation of Genealogical
Societies joined together in establishing a Records Preservation and Access Committee.
Still in its formative stage, the Committee will be composed of a liaison person
from each state. Chair Jack Brissee has expressed some of the concerns with which
the Committee will deal:
Historical records and our access to them are of vital importance to
genealogists and historians. Few of us would get very far in our research without
them.
Concerns about preservation and access are therefore always with us
and, indeed, increasing. The growth in sheer volume of paper records, the changing
nature of how records are created and stored, concern over misuse of records, and
the growing realization of the importance of non-governmental records, all of which
are complicated by resource limitations, generate significant challenges related
to preservation and access.1
As a contribution to the Committee's efforts, future issues of the New
Mexico Genealogist will describe and analyze some of the problems facing 21st century
genealogists. The problems include the apparent trend toward the closing of certain
public records, the seemingly diminishing number of private record sources, the
impact of electronic record-keeping on governmental records at all levels, the preservation
of and access to Social Security records, and the strategic role of the National
Archives at the federal level.2 The series begins
in this issue with a look at the federal censuses 1930-2000 from a genealogical
perspective.3
The 1930 census, available to us in the year 2002, will look familiar
because the basic question categories have changed in only minor respects from 1920.
The absence of plans for developing a complete Soundex is the most alarming problem
connected with the 1930 census. Perhaps we could take a lesson from the 1930s, when
the 1880, 1900, 1920 and some 1930 Soundex files were created by the Works Progress
Administration under the direction of the Bureau of the Census. Could not the Bureau
today initiate a bidding competition by which a private company would generate an
electronic data base of the 1930 records? Terms of the contract might include a
provision for employing and training persons leaving the welfare rolls.
Although little noted at the time, the 1940 census was significant because
of the use for the first time of statistical techniques such as probability sampling.
The Bureau of the Census found that utilizing such techniques provided several advantages:
Sampling in the 1940 census allowed the addition of a number of questions
for just 5 percent of the persons enumerated without unduly increasing the overall
burden on respondents and on data processing, and also made it possible to publish
preliminary returns 8 months ahead of the complete tabulations. The Bureau was able
to increase the number of detailed tables published and, also by sampling, to review
the quality of the data processing with more efficiency. 4
For genealogists, however, probability sampling translates into data
devastation. In 1940 only five per cent of the populace was asked questions about
the place of birth of father and mother; mother tongue (or native language); whether
a veteran or wife, widow, or child of a veteran; whether social security deductions
were made; usual occupation; whether a woman had been married more than once and
number of children born.5
The 1950 census continued probability sampling, with one in five answering
questions about place of residence a year ago; country where father and mother born
(one country assumed for both); highest school grade attended; income of head of
household as well as relatives living in the house; and service in the Armed Forces.6 In 1960 one in four was asked questions about birthplace;
birthplace of father and mother; residence; and highest school grade attended. Detailed
questions about work, occupation and income were answered by 25 per cent of the
population.7
The 1970 sampling questions were comparable in most respects to the
1960 census. Differences included additional questions about language spoken in
person's home when a child (Spanish is the first check mark), completion of vocational
training, and duration of a disability.8 In 1980
sample questions were asked of 50 percent of the populace in places with fewer than
2,500 inhabitants and of 17 percent of the populace elsewhere.9
In 1990 households were sampled on a 50 percent basis for places with fewer than
1,500 persons and on a 1-in-6 or 1-in-8 basis elsewhere.10
The sample questions for both 1980 and 1990 generally reflected those asked in previous
years.
As genealogists we may conclude that our chances are becoming increasingly
dismal of finding detailed personal information in the censuses from 1940 through
1990. The only questions asked on a 100-percent basis in the six censuses are name
and address, household relationships, sex, age, color or race, and marital status.
Color and race vary from decade to decade in the kinds of questions asked. In 1980
and 1990 the questions include Spanish/Hispanic origin.
But when the census for the year 2000 becomes available in 2072, we,
or rather, our descendants, will experience the biggest change of all. The Census
Bureau plans to enumerate only the first 90 percent by actual count, then estimate
the remaining ten percent. The Bureau justifies this decision because of the enormous
additional expense of enumerating the last ten percent and because of its confidence
in the sampling techniques.
In Living Within Limits, Garrett Hardin states that "the
cost of the 1990 census was over $2 billion, or slightly more than $8 dollars per
person. A scientific sampling method could yield equally reliable results at a cost
of perhaps $50 million."12 In other words,
the scientific sampling method has the potential to eliminate census enumeration
entirely at a vastly lower cost.
In fact, the Netherlands government abandoned traditional census taking
in 1971. They rely on birth, death, emigration and immigration numbers for population
statistics while using sampling for all other questions. Hardin concludes that "the
population of the Netherlands is probably known more accurately without a census
than is the population of the United States with one."13
Whether the United States will adopt procedures similar to the Netherlands we cannot
know, but it seems certain that the census beyond 2000 will continue to decline
as a major genealogical source. The United States census began in 1790 because of
political considerations in apportioning the House of Representatives. Through the
years the census has developed into a vital tool for planning and research purposes
by business, government and academia.
This may be a good time for we genealogists to remind ourselves that
the records we refer to as genealogical records were not created with us in mind.
For example, the creation of vital records is sanctioned by government for legal
purposes as well as in the past by churches for religious purposes.
Does this mean that genealogists can only hope for whatever crumbs are
swept from the banquet table by the major consumers? On the contrary, genealogists
must actively monitor the creation and preservation of, as well as access to, all
records with a genealogical content, and take appropriate measures to ensure record
viability. That is the mission of the Records Preservation and Access Committee,
but it should also become the mission of all genealogists.
The problems involving 21st century records are complex, and the New
Mexico Genealogist plans to keep you informed. In the next issue we will explore
the question of whether there is a trend toward closing public records.
1 Jack Brissee letter of 14 December
1996 to the President of New Mexico Genealogical Society.
2 Although archivists have maintained
a lively discussion in their journals about the future of records, there has been
little involvement by the genealogical community. This article has been stimulated
and informed, however, by Robert Charles Anderson's comprehensive presentation at
the National Genealogical Society's 1996 Conference in the States. Robert Charles
Anderson, On The Impossibility of Doing Research in the Year 2046
(Hobart, Indiana: Repeat Performances, 1996), sound cassette F-123.
3 For an excellent background
article on the census, see William Dollarhide, "Census Records: Look Again!,"
Genealogy Bulletin, 15 (September-October 1996), pp. 1, 6-8, 10-18.
4 U. S. Bureau of the Census,
200 Years of U. S. Census Taking: Population and Housing Questions,
1790-1990 (Washington: U. S. Government Printing Office, 1989), p. 5
5 Ibid., pp. 64-67.
6 Ibid., pp. 74-75.
7 Ibid., pp. 78-79.
8 Ibid., pp. 83-88.
9 Ibid., pp. 90-97.
10 Ibid., pp. 99-105.
11 U. S. Bureau of the Census,
The Plan for Census 2000 (rev. draft of April 5, 1996), p. II-6. In
an Addendum of September 12, 1996, the Bureau backed away from the original plan
to interview a 1-in-10 sample of the last ten per cent and stated that a new sampling
plan would be announced later.
12 Garrett Hardin, Living
Within Limits: Ecology, Economics, and Population Taboos (New York: Oxford
University Press, 1993), p. 34. As cited by Robert Charles Anderson, end note 2.
13 Ibid.
