Census And Public Policy Research Paper

1. History Definition

The archaic definition of a census is ‘poll tax,’ indicating its early association with the state’s function of revenue extraction. Although the term census dates to ancient Rome, where it referenced the registration of citizens and their property, the idea of a census predates Roman times. For example, Numbers, the fourth book of the Hebrew Bible, starts with the Lord instructing Moses: ‘Take ye the sum of all the congregation of the children of Israel, after their families, with the number of their names, every male … from twenty years old and upward.’ That is, take a census of those ‘able to go forth to war.’

It is probable that taxation or military conscription or both gave rise to other census operations indicated by archeological records, such as the Nepohualco or ‘counting place’ built by the Chichimecas a millennium ago. This site, near today’s Mexico City, consists of stones deposited by each inhabitant of the region, an early census databank. Other archeological records from various ancient civilizations indicate similar preoccupation by the state with carrying out systematic statistical studies. In these ancient records it is difficult to detect all of the state purposes that might have been served by a census. Certainly there was interest in economic data, as is suggested by statistical records of harvest and grain storage. Ancient Chinese records, dating from the Han dynasty (206 BC to AD 220), indicate that detailed population counts were made after each harvest. Subsequent systematic census records appear throughout Chinese history, and many scholars concur that China should be credited with the first comprehensive census (Domschke and Goyer 1986).

There were attempts to establish a periodic census in early modern times, such as the ambitious seventeenth century proposal for an annual census directed to France’s Louis XIV:

Would it not be a great satisfaction to the king to know at a designated moment every year the number of his subjects, in total and by region, with all the resources, wealth & poverty of each place?… [Would it not be] a useful and necessary pleasure for him to be able, in his own office, to review in an hour’s time the present and past condition of a great realm of which he is the head, and be able himself to know with certitude in what consists his grandeur, his wealth, and his strengths? (cited in Scott 1998, p. 11).

In the modern period, however, it was the United States that initiated the first periodic census. The US Constitution (1787) requires that an enumeration of the population be taken every 10 years. This decennial census, first taken in 1790, served two state-building functions. One was periodically to reallocate seats in the House of Representatives, the legislative branch to which positions are allocated proportionate to size of population in each of the member states. Because the population might grow at different rates from one state to the next, provision was made to take a census every decade and then to calculate how many legislative seats would go to each state. The second nationbuilding task addressed the challenge of geographic expansion. The country was expected to expand into new territories, and these areas were to join the nation on equal footing with the original states. The decennial census measured population growth and its geographic dispersion and thereby served as the mechanism regulating how territories were added to the state.

Though the US was early in establishing a periodic census, most European nations shortly followed in instituting some form of regular census to assist in nation-building and policy-making—the United Kingdom and France in 1801; The Netherlands, shortly after independence, in 1830; Sweden in 1855; Spain in 1857; Italy in 1861; the German Empire in 1871; and the Russian Empire in 1897. In nearly every instance, the postcolonial nations of Latin America, Asia, and Africa established a periodic census as an early step in nation building. By the end of the twentieth century only a handful of countries do not have a census.

2. Census And Public Policy

A wide array of public policy purposes are served by a regular official count of the population and its characteristics. Before turning to these purposes, it is necessary to distinguish between a census and a sample survey. The vast majority of state-generated statistical information comes from sample surveys, which elicit information from a part of the population in order to estimate characteristics of the entire population. A census customarily describes the effort to count everyone in the population. There is an important statistical interaction between sample surveys and the census, for the latter provides the frame necessary to draw samples for survey purposes. It also provides the basic denominator that serves as a control to determine whether a sample is demographically and geographically representative of the country. That is, does the sample have the correct proportion of age, race, ethnic, gender, and other demographic categories that have been measured in the population census, and does it reflect how the population is distributed geographically, again as measured in the census. The close relationship of a census and surveys indicates that many policy functions that rely on sample data are, in fact, drawing from a prior census.

An important distinction between the census and the survey is that the former allows estimates of population or housing characteristics to be made at comparatively small levels of geography. Policy purposes that require fine geographic detail will directly draw on census information. Policy purposes that are targeted at higher levels of aggregation tend to rely on sample survey data and thus have an indirect relation to the census.

A census supports policy-making in four broad arenas: election systems, taxation, public funding of social services, and social control.

2.1 Election Systems

When the basis of election to a legislature is size of the population being represented, census information is used to draw the boundaries of election units. As indicated above, this was the primary purpose for which a census is required by the US Constitution. It is now the practice in dozens of democratic countries, and not just for national legislatures. Election districts for city governments and other local and regional governments are frequently drawn to ensure approximate equal population size for each geographic area that constitutes an election unit.

Size is not the only population characteristic that can be taken into account in drawing election districts. Demographic traits such as ethnicity, language, and race have been used. This occurs when the state wants to ensure that the legislative body is ‘demographically representative’ of the population at large, for instance, that racial minorities have the opportunity to elect someone from their race to the legislature. Under these circumstances, census data are commonly used to monitor compliance with election laws. In the United States, for instance, census information on ancestry, gender, race, ethnicity, language spoken at home, place of birth, and year of entry into the country are critical for monitoring compliance with voting and civil rights laws.

When political party affiliations vary across different ethnic, language, or racial groups—as they often do—the use of the census to promote ethnic, linguistic, and racial balance or access will invite partisan disagreements. Census counts become the target of such disagreements, as in the events that led to the Nigerian civil war. Even census methodology can come under attack, as has happened in the United States in connection with its 1990 and 2000 census.

Basing political representation on population counts also invites jurisdictions to inflate their numbers. This incentive confronts census taking with serious methodological challenges. Unless there is a corresponding disincentive—such as basing tax rates on population size—a census can be vulnerable to strong political pressure by those with a direct stake in high counts.

2.2 Taxation

The most direct relation of a census to taxation policy occurs when taxes are collected from a jurisdiction proportionate to its population size. Thus a central government might calculate the tax obligation of its administrative units depending on the size and wealth of the unit, and use a census to make the determination.

A more indirect use of census (and survey) data occurs when the state measures the economic life of the nation, using that information to help design a taxation system—whether to tax manufacturing activity, natural resources, individual income, property, at what rates and in what combinations. For example, population projections based on census data are used to design social security tax rates and to estimate probable social security payment levels well into the future.

2.3 Public Funding And Planning

Social welfare policies are generally targeted to specific population groups, such as families in substandard housing, children in poverty, the physically disabled, minorities historically disadvantaged or mistreated, victims of particular diseases. Assessing the size and geographic distribution of groups targeted for welfare payments or subsidies is the task of a nation’s statistical system. Frequently public funding formulae are based on census counts of particular population groups (Nathan 1999). In the United States, for instance, in 1999 approximately $US185 billion in medical, education, housing, transportation, unemployment, and other payments depend in part on census data. Similar funding formulae using census information occur in many countries with advanced social welfare systems.

When public funds are allocated to target groups based on their size, they have a strong interest in being fully counted. But these are also population groups that are difficult to reach in standard census operations, and even if reached, will sometimes resist being counted because of concerns that census information can be used against them. This situation has given rise to advocacy organizations that work with and put pressure on census bureaus to assure as full a count as possible of their particular constituencies. Where this has occurred, census taking is no longer the domain only of professional statisticians and their government minister, but becomes of much more generalized political interest.

In addition to targeted welfare spending, national governments frequently use census information for program planning across many policy domains such as public transportation, health and education services, job training, police and fire protection, sanitation and sewage disposal, and land use.

The census also plays a role in national security policy. As indicated above, since biblical times the state has relied on census information to assess the numerical strength of its fighting age population, and also at times to design and implement actual conscription. Of course, a reliable count of the fighting strength of one nation’s population will be used by another nation to calculate whether that nation poses a military threat. The first US census in 1790 was declared by the nation’s leaders to have missed a sizeable portion of the population, and the Secretary of State took care to announce that fact to European powers that might otherwise have concluded that the young nation was too weak to protect its independence. Of course not just population censuses but also economic censuses help the state assess its military capacity and that of other nations.

The relative strengths and weaknesses of contending groups within a nation are also revealed by census information, and have often played a role in strategic choices that precede internal warfare. The suppression of census data in several African countries reflects a decision by the ethnic group in power to disguise the comparative numerical strength of ethnic groups challenging its rule. In the period preceding the Civil War in the United States the more rapid growth of ‘free-states’ compared to ‘slave-states,’ as revealed in the 1850 census, was a factor leading to the succession of the South from the national union.

2.4 Social Control

More generally the state uses the census as an instrument of social control, as it was routinely used during the colonial era. Great Britain, as early as the seventeenth century, introduced a census in many of its colonies in order to control more effectively its distant subjects and to assure that resource extraction proceeded smoothly. Decisions about which territories needed policing, required additional labor, could benefit from infrastructure investments, and so forth were facilitated if a reliable census could be taken. Great Britain and other colonial powers were early practitioners of what has now become commonplace. Social control policies that police borders, population movement, and even reproduction and fertility are fashioned in part on the basis of census information.

The boundary between routine social control and abusive exercise of power by the state is not easily drawn, but certainly history offers many examples of the census as an instrument of the latter. Family planning policy and fertility regulation in China and India, for instance, is often cited in this regard.

Population displacement is a further example. Census data have been used to place restrictions on freedom of movement or to force population groups to live in places they would not have chosen. An important instance of the latter is the role of US census data in the nineteenth century relocation of native American Indian tribes from their homelands to reservations. US census data were also used to facilitate the internment of Japanese-Americans at the beginning of World War Two, because they were said to pose a security risk shoul Japan invade America’s west coast.

Census data have also been used in genocidal policies, the most tragic example being the Holocaust. Population registers or special censuses identifying Jews greatly assisted in the location of Jews and their transport to concentration and extermination camps. The 1939 German census, for example, was used to identify specific individuals. In The Netherlands small area tabulation of a census of religions taken in 1930 was used in dot-maps of Amsterdam to indicate the density of the Jewish population. Poland, France, and Norway offer additional evidence linking census data to Holocaust operations (Seltzer 1998).

These historical instances of gross misuse of census information have had one beneficial outcome. There is now much more professional and political sensitivity to issues of data confidentiality, the independence of statistical agencies, and related ethical considerations (Seltzer 1994).

3. Census Data And Democracy

Census information connects to politics, and especially to democratic politics, in ways other than its use by the government to design and implement public policy. Statistics and especially statistical series also advance key features of democratic politics. One purpose is helping the electorate to assess government performance and hold it accountable. Another is in providing a platform for social reform. A third is in the assertion of group interests (Prewitt 1999).

3.1 Democratic Accountability

The theory of electoral accountability holds that competitive elections offer to the electorate alternative portrayals of how well the government in power has performed or will perform against future challenges. Aspirants for political office present themselves in terms of their past accomplishments and their promises of future accomplishments in managing the economy, protecting the nation’s security, and generally enhancing national well-being. Voters then elect, re-elect, or evict accordingly.

This simple model presumes that the voters have some way of assessing government performance. Statistical trends and social indicators play a major role. They indicate whether the economy is growing or stagnating, whether education or health or housing is improving, whether crime rates are down, whether the environment is being protected, and so forth. Arguments about who can take credit for improvements, or should be blamed for failure, are the common currency of competitive elections and are often advanced by citing statistical trends.

Consider, for example, the sharp political debate over immigration policy in many wealthier countries as they face heavy demands for entry from Central Asia, Northern Africa, and Latin America. These debates are framed in terms of the ‘proper’ balance between native population groups and newer immigrant groups, with census and related data providing the subtext. Governments are then judged in terms of how well they protect racial purity with anti-immigration policies; or, conversely, display humanitarian compassion or perhaps, as in the United States at present, replenish the labor pool with more open immigration policy. In either instance, as in so many areas of government programming, long-term statistical trends are invoked when assessing policy.

3.2 Social Reform

Democratic theorists have long been interested in how issues get on the political agenda, especially issues that matter most to the unorganized or disenfranchised interests of society. For instance, how do issues such as child labor or male–female wage differentials become politically salient? One route to the political agenda is though social reform action that takes as its starting point statistical profiles. Since the early days of industrialization, reform groups have used statistics to document the social conditions that they were committed to eradicating (Cohen 1982). Examples are poverty laws, prison reform, child abuse, racial discrimination in education, housing, and employment. Reform activists mobilize political participation and inform public debate by transforming previously unnoticed social conditions into highly visible social injustices. In this way resource-poor groups compensate for their lack of political organization or low levels of political participation. They use state provided statistical information as a political resource (Lacey and Furner 1993).

3.3 Group Interests

Census information can be the basis on which to establish group identity. When, for instance, the census differentiates the population into its component racial, ethnic, or tribal groups, it promotes these categories in public consciousness and in public policy. To be measured is to be politically and socially visible, and this visibility can become the basis for group representation. Public policies such as affirmative action targeted to India’s lower castes or racial minorities in many countries are formulated on the principle of statistical proportionality. That is, there is pressure to ensure that racial or ethnic groups historically discriminated against are represented in high status occupations or the legislative bodies proportionate to their numbers in the general population.

This process, once started, gathers momentum. More and more groups insist on being fully counted in a census. The leaders of these groups then assert political claims on the basis of the counts. A census which is judged to undercount particular groups, such as the homeless or the linguistically isolated, is attacked by advocates for the undercounted population groups. The census and even census methodology can become a political battlefield.

The increasingly central role of statistical proportionality in democratic politics and public policy is welcomed by some but resisted by others. Proponents argue that decades or even centuries of injustice can be reversed through social policies that correct for a history of discrimination. Because this discrimination was and even remains group-based, it requires group-based measures to overcome it. Others express concern that attaching policy benefits and penalties to persons on the base of their race or ethnicity is to abandon the liberal principle of individual rights in a democracy.

The census and what it measures is of great political interest to both sides of this debate. What is unmeasured cannot easily become the target of public policy, and certainly cannot be used in policies that employ statistical proportionality to achieve social purposes.

4. The Census As A Subject Of Public Policy

The census, being an instrument of nation building, policy-making, public planning, resource allocation, and democratic governance, will itself be the focus of policy. The overriding policy consideration for a government considering its census is how to strike the proper balance between accountability and autonomy. On the one hand, the census is instituted and funded primarily because it serves government purposes. If it serves no useful purpose it will be underfunded and eventually discarded. On the other hand, the census shares with other scientific enterprises the requirements of professional independence. When a government does not respect the independence of census operations, it places at risk the credibility of the counts and can easily be accused of manipulating census numbers in order to protect itself or enhance its prestige.

This tension between policy accountability and scientific autonomy plays itself out in multiple ways within the policies that govern a census. At one extreme, census results and other official statistics are only very selectively released. For example, China during its Cultural Revolution (1966–76) tended to suppress all statistics or to make them available only through highly controlled government reports. In the former USSR, the 1937 Population Census was never made public, in part to keep from public attention the magnitude of the 1932–3 famine. In less extreme instances, particular statistical series are presented in a manner designed less to inform than to persuade the public. In democracies, for example, the release of inflation rates or poverty measures can be timed to benefit the incumbent party during an election period. Practices such as suppression of data or political manipulation of its presentation and release will of course undermine public confidence in census counts and related statistics.

At the other end of the policy continuum is the assumption that the census is a public good. From this perspective, national statistics serve broad social purposes that reach well beyond the immediate needs of the government in power. This perspective, of course, is compatible with granting considerable independence and autonomy to a census bureau staffed less with political appointees than with professional statisticians.

In most countries, census design and operations exist uneasily in a policy space that simultaneously includes close government oversight, on the one hand, and acceptance of the importance of scientific autonomy, on the other. The weight assigned to one or the other will vary depending on the ideology of the government in power. When the conservative government of Margaret Thatcher came to power in the United Kingdom, an explicit decision was taken to narrow the functions of the statistical service and make it serve specific, immediate policy needs. In the United States, a number of questions previously asked were eliminated from the 2000 census forms because they were not directly related to a federal program. This narrowing of the function of the US census dismayed demographers and sociologists who pointed out that the census was becoming less analytically useful, that is, was less of a public good.

The mission a census is expected to serve—narrow, immediate, and even secret government functions at one extreme or broad public good functions at the other—is the first-order policy issue that confronts a government as it establishes and funds the periodic census. There follow, then, a host of subsidiary policy issues. How are government statisticians expected to balance their duties as civil servants with their obligations to their professional peers? What criteria should be used to select the director of a census operation? How much autonomy is granted a census agency in its ongoing effort to improve methods and concepts, especially in redefining politically sensitive measures in areas such as race and ethnicity or income and poverty? What rules govern the timing and nature of data release? What is the appropriate tradeoff between data confidentiality and respondent privacy, on the one hand, and the society’s demand for detailed information, quickly provided, on the other? Should a census agency only provide raw data, or should it have its own analytic capacity and thereby provide reports and interpretations? How can appropriate government oversight of a census operation guard against waste of public money or agency incompetency without itself being accused of partisan intrusion in the census?

All of these and other issues are the proper subject of political debate in the articulation of policies that govern the funding, organization, and implementation of a census. The issues are similar in nature to those that arise in the government sponsorship of science or the arts, each being an activity that requires a high degree of professional autonomy but each also being an arena of great political sensitivity.

With respect to a census, or statistical activities more broadly, thoughtful observers have articulated useful policy principles. In essence these principles recommend that the statistical products should be designed to serve the needs of multiple political interests and not just those of the government currently in power, and should be sensitive to future policy issues rather than focused only on the present. Success in institutionalizing this principle will maintain both the relevance and the credibility of a census (Seltzer 1994).

5. The Census As A Subject Of Social Science

The social sciences make heavy use of census data across a wide variety of topics. The quantitative study of society is severely hampered in the absence of census information. It is perhaps surprising, therefore, that studies of the census itself are so few. That is, though the census is everywhere the subject of public policy, it is seldom the subject of public policy analysis.

The exceptions to this general proposition are important, and should not be overlooked. There are historical treatments and case studies of particular censuses that are often rich in analytic detail (Anderson 1988, Anderson and Feinberg 1999, Arowolo and Daramola 1982, Medvedev 1989). There are studies of particular policy issues that emerge in the census context, such as analysis of confidentiality or of the census as an instrument of social control or the role of the census in revenue sharing schemes.

What is generally lacking, however, is social science theory about the emergence, role, function, and consequence of the census. It is largely assumed that the modern nation-state needs a census, else it would not have emerged in nineteenth century Europe and North America. This assumption, however, has not been subjected to investigation in political sociology. Theoretically rich treatments of nation building and citizenship (Bendix 1964) or of the state’s effort to improve society (Scott 1998) provide a platform for studies of the political causes and consequences of the census, but with few exceptions (Alonso and Starr 1987) the work is yet to be undertaken.

The political economy literature is similarly deficient in its theoretical treatment of the census as an object of inquiry. It is easy to find assertions that the modern economy requires fine-grained demographic data for making decisions about labor market dynamics, location of retail outlets, types of products that will meet future consumer needs, and so forth. From this latter assertion, it is an easy step to claim that census data are a public good and to invoke microeconomic theory to explain the government’s decision to fund a census. Using the United States as a case study, one sustained examination of this hypothesis reports that public good theory offers an insufficient explanation of the rise of the modern census (Kelman 1985).

Neither political sociology nor political economy has yet turned its analytic tools on the census as a feature of the modern state and modern economy. This will likely be corrected. If metaphors about how the ‘knowledge society’ will replace industrial society are taken seriously, so also is the assertion that information is the emerging infrastructure for the knowledge revolution just as railroads or canal systems were the earliest infrastructure for the industrial revolution. An accurate and credible census is basic to establishing an information infrastructure and thereby requires systematic attention by the social sciences.

Bibliography:

1. Alonso W, Starr P (eds.) 1987 The Politics of Numbers. Russell Sage Foundation, New York
2. Anderson M J 1988 The American Census: A Social History. Yale University Press, New Haven, CT
3. Anderson M J, Feinberg S E 1999 Who Counts. Russell Sage Foundation, New York
4. Arowolo O, Daramola O (eds.) 1982 Philosophy of Population Census in Nigeria. National Population Commission, Lagos, Nigeria
5. Bendix R 1964 Nation-Building and Citizenship. John Wiley & Sons, New York
6. Cohen P C 1982 A Calculating People: The Spread of Numeracy in Early America. University of Chicago Press, Chicago
7. Domschke E, Goyer D S 1986 The Handbook of National Population Censuses: Africa and Asia. Greenwood Press, Westport, CT
8. Economic Commission for Europe 1992 Handbook of Official Statistics in ECE Member Countries. United Nations, Geneva
9. Kelman S 1985 Why should government gather statistics, anyway? Journal of Official Statistics 1(4): 361–79
10. Lacey M J, Furner M O 1993 The State and Social Investigation in Britain and the United States. Cambridge University Press, Cambridge, UK
11. Medvedev R 1989 Let History Judge: The Origins and Consequences of Stalinism. Columbia University Press, New York
12. Nathan R 1999 The politic of printouts: The use of official numbers to allocate Federal grants-in-aid. In: Alonso W, Starr P (eds.) Who Counts. Russell Sage Foundation, New York
13. Prewitt K 1999 Public statistics and democratic politics. In: Alonso W, Starr P (eds.) Who Counts. Russell Sage Foundation, New York
14. Scott J C 1998 Seeing Like A State: How Certain Schemes to Improve the Human Condition Have Failed. Yale University Press, New Haven, CT
15. Seltzer W 1994 Politics and Statistics: Independence, Dependence, or Interaction? United Nations Department for Economic and Social Information and Policy Analysis, Working Paper Series No. 6. New York
16. Seltzer W 1998 Population statistics, the Holocaust, and the Nuremberg Trials. Population and Development Review 24(3): 511–22