What is a Census?
A government survey that is conducted to obtain information about the population and economy of a country is termed a census. The word comes from the Latin censere, meaning 'to tax', and was originally applied to the lists of adult males and their property in ancient Rome. Registrations of the people in certain parts of the country-or of certain groups of people, such as landowners-were also made in ancient Babylon, Egypt and Greece, most probably for taxation purposes.
The first attempt to undertake a census in the modern sense-that is, a complete count of a country's population-was in China in AD 1370. The evidence for this early census is contained in documents that give instructions for undertaking the population count and contain an outline of the penal ties for anyone refusing to com ply with the census. A completed census schedule also exists, but no records of national results have survived.
In Europe, as early as the fifteenth century, local population counts were carried out, often to estimate food requirements of a city in time of siege. In the seventeenth century, colonial powers began to undertake a series of censuses of their foreign territories.
In 1624 Britain undertook the first of a series of population and resource surveys in Virginia and in 1666 France made a comprehensive census of its Canadian provinces. The first attempts to survey the entire population of any European country was probably in England and Wales in 1695, although it is not certain that the total population of both countries was actually surveyed. Attempts to calculate the precise population and economic conditions of a country in early censuses were usually related to political and economic events. The 1695 English and Welsh census was undertaken to estimate the funds that could be raised for the war against France and in Iceland in 1703 a national census was undertaken to establish the extent of pauperism in the country.
The practice of undertaking censuses at regular intervals was initiated in the United States, where the first federal census was carried out in 1790. The US Bureau of the Census has continued to conduct a census of population every 10 years since then. During the nineteenth century, most European countries began to undertake censuses, Russia being the last to do so in 1897. By the end of the nineteenth century, approximately one half of the world's population was covered by a census, but many were incomplete or inaccurate.
Efforts were made to improve the reliability of census statistics with the use of new techniques of data collection and analysis. The comparability of census statistics among countries was improved after 1853, when the first international statistical conference was held. Such conferences have established standard methods of data classification. The reliability of census data in many developing countries has been improved since 1945 by assistance from the United Nations and other international agencies who have provided advice and training for census personnel and manuals on the processes of census taking.
At present, at least 90 per cent of all nations undertake some form of census, usually a population count.
Some countries carry out regular censuses to obtain information not only about population numbers but also about its health, education, housing, leisure, income and residence movements. The statistics obtained are published and used by government agencies to predict future needs in housing, education, health and welfare, and also by manufacturing and retailing bodies to plan production and marketing policies. Many countries also conduct separate censuses on housing, agriculture, manufacturing, commerce and mining. At 5-yearly intervals, the United States conducts a census of governments, which obtains information on all local governments in the country.
Careful planning of a census is required to ensure that all relevant data are obtained and that questions will elicit the appropriate data and not be misunderstood. It is also important for questions to be carefully phrased to avoid alienating respondents, who may then deliberately give inaccurate answers. No census can ever be completely accurate, because some people will refuse to answer questions or seek to conceal information, and others will misunderstand questions or forget relevant information.
Statisticians take such factors into consideration when compiling census statistics. A census is usually one of two types.
It may be a self-enumerated census, in which census forms are distributed to respondents, who answer them and return them to the census bureau, or an enumerator census, in which interviewers go from door to door and fill in forms with the respondents' answers.