The Population Research Center Library at the University of Texas began collecting population censuses in the early 1960s initially with funds from the National Science Foundation. The first five year core grant from the National Institutes of Health in 1971 enabled the Center to hire a full-time librarian to organize the collection. Since then the collection has grown to include 30,000 volumes either in print, microfilm or microfiche and recently in CD-ROM form. We estimate that about 85% of all known population censuses can be found in the Center's library. Only the Library of Congress and the NYPL have a similar universal coverage policy.
So how do we get a hold of these census volumes then?
In the past the US dollar was more widely accepted as a desirable currency so that we could pay most of the invoices in dollars. However, recently many nations wish to be paid in their own currency. This poses a problem when the currency cannot be handled by the banks in Austin or the University of Texas, Accounting Dept.. I recently tried to order some censuses from Tunisia but was unable to pay for them in the dinars they wanted. I ended up having to borrow the volumes from another library and copying them. For the countries that list their prices in their own currencies but will generally accept dollars, I have to go hunt for a copy of the Monday edition of the Wall Street Journal to find the exchange rate for such exotic moneys as the Mauritian Ouguiya or the Slovenian Tonar. Sometimes, a country will offer a free copy of its census but has no mechanism in place or is prohibited by law from accepting payment for postage.
Another major obstacle in treatment and use of the census report is language. While most are written in English or the language of the country and some English translation, several are only published in the language of the country. The majority of these remaining languages are French, Spanish, Portuguese and German which are within my reach. However, e.g., I recently received a number of volumes from the newly formed republic of Macedonia which were exclusively written in Serbo-Croatian. I am still working on trying to translate the title pages.
Until recently our holdings were still included in a card catalog. Under pressure from our grant givers, National Institute of Health, we needed to look for ways to replace the card catalog. The unique nature of census materials and our staff shortage made a more conventional on-line catalog an unlikely choice.
In order to get some sort of on line catalog devised as quickly as possible and with minimal tedious data entry, I decided to electronically scan the catalog cards for each country into a word-processing file and from there edit out any mistakes.
The country entries were then organized by continent, geographical regions based on United Nations classification, and by country. Each country file was converted into an html file and loaded unto the World Wide Web as part of the Population Research Center's library home page. This whole process took about 1 and ½ years and was completed last December. This has taken care of listing the citations for the census volumes. However, contrary to monographs, censuses are almost never searched by author or title. What is important is contents and year. So we are still faced with finding a convenient way to search for contents of the volumes. For instance, if you wanted to do a comparative study on internal migration in Latin American countries, you would need to scan all tables of contents of all censuses you are interested in for all years you need to see if what you are looking for is included. One of the things I am hoping to look into soon is perhaps scanning in the tables of contents and somehow linking the topics in some sort of database.
What will you not find in a census: vital statistics such as number of deaths, births and marriages. Income is only included in only 40% of national censuses, religion is not very common either. Finally, in the censuses we collect you will not be able to trace your forefathers since we are mainly dealing with statistical tables not containing personal names.
As was mentioned earlier, much of the information in the censuses still in print version can only be mastered by personally scanning the table of contents for each volume for each year for the specific country. This means the researcher or staff member must spend considerable time searching through each publication in order to acquire a full understanding of the material. I would like to explore the possibilities of facilitating access to the census data still in print version. I welcome anyone who has any suggestions.
Finally, future censuses, at least for the developed countries, will most likely be exclusively in electronic format. Some of us are familiar with the entire 1990 US census made available by a commercial vendor on one CD-ROM. I also recently purchased the 1989 Soviet Union on one CD-ROM. It remains to be seen, however, if most of the developing countries can afford the high price of this technology or if libraries can afford the high price of electronic media. Population data are also beginning to appear on the Internet. Several organizations such as the Census Bureau and the Population Research Center have established links to those countries that have a web page for their statistical agencies. Most of those web sites include the latest population estimates for their country.
In short, if most countries decide to publish their census data on the Web for free, the expenses of purchasing some data on CD-ROM may be offset by the free data on the Internet. It is, however, still too early to make budget judgements based on this.