A population register, broadly interpreted, is a list of persons who belong to a predefined group, containing information identifying the members in a unique way. In that broad sense, a list of members of a club, union, or society could be called a population register if it contains some characteristics like name, address, and date of birth which (in combination) provide unique identifications.
In a more formal sense, the term population register denotes a list (register) of persons who are citizens or residents of (or in some other sense "belong to") a country or a sub-national region. That list typically includes each person's name, current address, and date of birth as an external identifier, and a personal identification number as an internal identifier. This identifier should be unique.
Brief History and Status
Population registers in the more formal sense of the term have been known for several centuries. The parish registers found in many European countries are an example. Local civil registers were established with the development of social security and similar programs at the end of the nineteenth century, when municipal and other authorities needed lists of persons under their responsibility. The demand for national registers intensified in northwest Europe after World War II, with the development of the welfare state. For some countries, that meant a fully centralized register; in others local registers were maintained but under central leadership and coordination (to ensure that every person was included in only one local register).
Since the 1960s, registers have normally been kept in an electronic format, at local and/or national level. This format makes for greater usefulness but is not a prerequisite for a well-functioning population register.
As of the 1990s, most European countries had some sort of population register, local or central or both, and assigned some kind of personal identification numbers. Ireland and Romania are exceptions. Aarno Laihonen's 1998 article provides an overview of the situation. Many less developed countries either already have some kind of population register or are planning to develop one. In Eastern Europe during the Communist period, population registers were used for control as well as administrative purposes, and the successor regimes for the most part have not maintained them. The U.S. lacks any sort of national population register.
The use of registers and identification numbers for statistical and analytical purposes, and the laws regulating this use, vary greatly by country. The Nordic states, the Netherlands, and Belgium have fully centralized registers available for analytical purposes. Israel and Slovenia have similar systems; Austria has one under development.
Content and Linking
The number and nature of variables included in a population register are determined by the political and administrative intention for use. The normal identifiers are name, date of birth, and sex; marital status is recorded; information on children and family may be included, and in some countries, ethnic group. Address or locality information makes it possible to combine regional registers or integrate them at the national level. Some registers on principle include as little information as possible, to avoid items that change frequently. Sex, age, marital status, and a numerical identifier are probably the minimum for a basic register. Other registers that may be linked to it would include the variables needed for specific purposes, like address, births, or co-residence. A register with such variables will have to be updated often.
The key determinant of the ease of use of a register is whether or not it assigns a personal identification number (PIN-code) to each name. Such a number is typically made up of digits indicating date of birth and a sequence number, a code for sex, and one or more control digits. The encoded information should not, of course, include items than may change.
Use of PIN-codes makes it easy to link information from different registers, so the content of any particular register is less important. The variables in the whole cluster of registers can be used simultaneously. For example, in the 2001 Norwegian census about 30 registers were combined.
Where two or more registers of the same population can be linked, the system of registers permits refined analysis of population processes in combination with a broad range of covariates (Aukrust and Nordbotten (1973), Statistics Denmark (1995)). Indeed, linked systems are like social and demographic surveys, but without the surveys' response problems and sampling errors. Ideally, a register keeps a record of all previous variable values, along with the date of change. This is important for historical studies and for longitudinal micro-level analyses with time-varying covariates.
Population registers belong to a broader family of administrative registers. These may be numerous, often including employment and unemployment registers, income registers, social security registers, registers of criminality, education registers, etc. Because they are established and run for administrative purposes, their content may not always be well-suited for scientific purposes. For instance, analysts of internal migration may prefer a different definition of place of residence than the one used by tax authorities. However, both the responsible authorities and the researchers share a common interest in appropriate definitions and timely updating.
The main advantage of an administrative register is quality. A register run for purely statistical or scientific purposes, can seldom keep high enough quality. Those registered will not have serious interests in informing the register about every change of status. An administrative register is often linked to, and uses administrative routines, which also can be used for controlling the register content. It will often be in the interest of both the registered persons and the users to keep the register correct. Neither register users nor the registered persons can accept that important decisions are based on registers with wrong information. Frequent use increases the possibility to detect mistakes.
A population register adds persons at birth or at entry and deletes them at death or departure. Coverage problems mainly concern migration. There may be a strong personal interest in not having a correct register status. In Scandinavia, however, there are probably more persons (but still below one percent of the total population) who have left the country without notification than are persons living there illegally. Remaining clandestinely in highly regulated countries like the Nordic states without being registered is very difficult.
Internal migration causes particular problems for local registers, due to delayed or completely missing migration notifications. Moreover, for researchers, the rules for registration of place of residence may be ill-suited to demographic analysis. The more actively the register cooperates with other institutions (mail and telephone systems, electricity, schools, health care, etc.), the better its quality.
Statistics based on population registers share some of the same problems as those based on sample surveys. Persons who want to hide themselves or their characteristics will probably not be better covered by sample surveys than by registers. The two sources can, however, be combined. Registers can be used as a sampling frame for a survey, or certain pieces of information can be taken from the register instead of bothering the respondent. In addition, having a register linked to a survey allows survey non-response to be properly analyzed; whereas in a register-based system non-response is virtually non-existent. If variables needed for some purpose are not included in the register, or included with unacceptable definitions, sample surveys or other direct methods of data collection may be necessary.
Data Protection and Privacy
Linking of registers provides information of potentially great value to government, but also great potential for misuse. The same is true for social scientist users of these data. In both cases, confidentiality and protection of privacy have to be properly taken care of. Virtually every country with register systems that allow linking has a Data Surveillance Authority or equivalent agency with the aim of strict enforcement of privacy requirements. These usually specify that information should never be available for identifiable persons; individual data are to be available only for making statistical tables or estimates. In Europe, this work is regulated by the European Union (EU 1995).
Aukrust, Odd, and Svein Nordbotten. 1973. "Files of Individual Data and their Potentials for Social Research." Review of Income and Wealth 19(2):189–201.
European Union (1995): Directive 95/46/EC of the European Parliament and of the Council of 24 October 1995 on the protection of individuals with regard to the processing of personal data and on free movement of such data. Brussels, Belgium.
Laihonen, Aarno, Ib Thomsen, Elisabetta Vassenden, and Britt Laberg. 1998. Final Report from the Development Project in the EEA: Reducing Costs of Censuses through use of Administrative Registers. Documents 98/1. Oslo: Statistics Norway.
Statistics Denmark. 1995. Statistics on Persons in Denmark. ECSC-EC-EAEC. Brussels-Luxembourg.