Name: Addresses (Adresser) Period(s): 1953, 1960, 1963, 1967, 1970, 1971, 1975, 1979, 1980, 1981, 1982, 1983, 1984 (SMS) Code book: III

Information

Rich information about the cohort members’ addresses is available for analysis. These were collected on several occasions over time. Variables that measure the addresses are available for the years: 1953, 1960, 1963, 1967, 1970, 1971, 1975, 1978, 1979, 1980, 1981, 1982, 1983 and 1984. These variables are taken from a number of different files; Recide, Adress, Adress2, Adresser 1958-1975, Adresser2 1953-1975 and Kommunerna. The variables in the Recide-file were originally from the files Fob60 (1960), Fob70 (1970) and Fob80 (1980). The remaining years are from the files Adress and Flytt3.

The variables from these files have been combined to form a new file named “addresses datasets united” to give a file that contains all the relevant variables. In some cases there were duplicates which have now been erased; in other cases variables have been constructed derived from original variables. Moreover, some variables measuring the same phenomenon are available in more than one version (ie. the version in the different data files do not completely correspond to each other). In such cases, the version from the Recide-file is accounted for in this code book. This is due to earlier revisions for which we have no documentation. Further on, for the same reason, the Recide-file has precedence when variables measuring same phenomenon from different files have been merged to enable us to account for as many respondents as possible. In this codebook, the variables are shown by year and their place in the new united data file (addresses datasets united), with a reference underneath to their original data file. In the new data file the variables have been given a new ending to enable identification of their original data file location.
The cohort member’s address is measured by a ten digit code scheme that accounts for county, municipality, parish and district of residence. This code scheme was given to the addresses when the coding work was done in 1954, with the two first numbers specifying county and the next two municipality. Addresses outside the metropolitan area have a four digit code scheme, beginning with number three and followed by a three digit code scheme for country, which is coded in the same columns as county and municipality. For birth addresses outside the metropolitan area, the code number was assigned with the aid of Statistic Sweden’s established classification for counties and municipalities from 1954. Further on, diplomatic passports were given county code 37 and municipality code 99. The real district code for the metropolitan cohort member is the last six digits, of which the first three give the parish and the last three the FOB-district of residence (Martinnen, 1976). For example, lan53 refers to digits one and two; kom53 refers to digits three and four; for53 refers to digits five, six and seven, while fob53 refers to digits eight, nine and ten in the code scheme. These variables, except for the one measuring count (lan), can only be understood in conjunction with the other variables. For this reason, variables measuring municipality (kom), parish (for) and district (fob) of residence are given by showing valid n instead of frequency tables. New variables were derived to account for municipality (lankom) but are not shown in tables; valid number of observations are also shown for these variables.
Some guidance to how the addresses should be interpreted;
1. If the cohort member moved abroad, he is counted as living in Sweden during the year of emigration. County and community numbers are collected from the year before the emigration. In cases where the person returned he is counted as living in Sweden that year.

2. If the cohort member died, he is counted as living in Sweden. County- and community numbers are assumed to be the same in the year of death as the year before. 3. Cohort members with diplomatic passports are assumed to have their address in the metropolitan area.

4. Cohort members with no address are assumed to live in Sweden (except county).
5. There is no information about those living in the county but outside the metropolitan area for 1967 and 1971. Those whose information is missing have been given number 3, i. e. living in Sweden, other county of residence than Stockholm (not, however, if they died or emigrated that year).

For further information in detail about the coding of addresses for the years 1953, 1963, 1967, 1971 and 1975, see Martinnen (1976).

When compiling this codebook, Statistics Sweden’s established classification of counties and municipalities (January 1952) and the report on statistical co-ordination (SCB, 1986:5) that accounts for Swedish municipalities between 1952 and 1986 were used. The report gives the local government district divisions and the numerical codes and changes between the years.

The codes for counties were 01-25 until 1968, when Stockholm city and Stockholm county where combined into one county, after which codes 01 and 03-25 are used instead. The last two codes are used for numerating the respective municipalities within the county.

To maintain personal integrity and avoid backwards identification, frequencies of thevariables measuring addresses are only shown when categories contain 50 or more observations.

References

Martinnen, E. (1976), Rapport ang addresskodningsarbetet inom project Metropolit [Report on the coding of the Metropolitan project, 20.5.1976. Eero Martinnen]. Not published.

Statistiska centralbyrån [SCB] (1986), Sveriges kommuner åren 1952-1986. Förändringar i kommunindelning och kommunkoder [Changes in the local government district division and the numerical codes]. Reports on Statistical Co-ordination 1986:5. Stockholm: Statistics Sweden.