Switzerland Socio-Economic Variables, 1870-1930
History of the Data
Francine Vanderwalle compiled socio-economic data for Switzerland for the census years 1870, 1888, 1910 and 1930. The data are contained in separate data files; some variables are available at only one point in time, while others exist for all years. The main source for the data was Statistique de la Suisse. The data are aggregated by canton and district, with a label to identify the canton and district. These internal labels, unfortunately were different for each year. The total number of districts was 181 in 1870. In 1888, one new district was added in Canton 15 (App Ar), bringing the total to 182. In 1910, 5 new districts were added in Canton 11 (Solothurn), bringing the total to 187. In 1930, four districts were dropped: one in Canton 12 (Basel); one in Canton 17 (St. Gall); and two in Canton 25 (Geneve), reducing the total to 183. Variables may not be comparable in these districts in the affected years.
In 1983, Susan Watkins and Beverly Harris converted the data into ascii format, put it on tape, and created codebooks for the data, as well as a list of common variables and a key to match the labels in the data to the cantons and districts. They also contacted Francine Vanderwalle for clarification of the meanings of some variables. All of the documentation was added to the tape along with the data. Beverly Harris passed along a folder of notes to Barbara Vaughan when she left OPR.
In 1994, Barbara Vaughan moved the data from tape to the OPR Unix archive. She noticed a major problem with the key to identify the cantons and districts: Two of the years (1870 and 1930) seemed to have district labels that were not unique. After examining the data, she determined that in 1930, the internal labels in the data were unique, but did not match the key in the documentation. Fortunately, all four data files had a series of variables that were supposed to be identical (the years that Ig and infant mortality reached certain levels). Using the two years that did have accurately described and unique labels, she attached these variables to the key and then matched the other two years to these variables in order to determine the canton and district. To avoid the need to use the printed key at all, she added a unique 4-digit label (taken from the key listing), as well as the name of the canton and district, to the end of the data records.
Some editing of the codebooks was also done at this time; for instance, the altitude variables were previously labelled as "feet" although Francine Vanderwalle's notes indicated that they were metres. Also, there were some obvious errors in the list of canton and district names, which were corrected from Francine's notes.
Contents of the Data
The four indices of fertility (If, Ig, Ih, and Im) are available for all four years, as well as infant mortality and the proportion of women single at age 20-24 and age 45-49. Proportions belonging to major language groups and religions are available for all years. Variables measuring the education level of draftees are available through 1910, as are the proportions living at various altitudes, which was found to be correlated to fertility trends. Most years have comparable variables showing the breakdown of the labor force by sector, and female labor force participation. The 1910 data set has variables obtained from a special 1905 "census of enterprises".
All of the variables (except those added by Barbara Vaughan) are in 10-column fields with three decimals. The terms "proportion" and "per cent" are used rather loosely in the codebooks.
- The variables that should be identical for all four years (the years that infant mortality rate and marital fertility fall to certain levels) are in fact sometimes different for some districts in some years. In most cases, this appears to be because of a redrawing of districts within a canton. This should perhaps make researchers wary of comparing other variables in the affected cantons and years.
- The district labelled "Toggnbrg, Alt" in St. Gall (Canton 17) in 1930 may in fact be "Tablat". According to the documentation, "Toggnbrg, Alt" was dropped in 1930. However, because of other errors in the documentation for 1930, the districts were matched based on the variables that were the same for all four years. There was no match for Tablat, although there was a match for "Toggnbrg, Alt".
- Labor force participation statistics for 1870 and 1888 are based on statistics of the proportion employed in "agriculture", "industry" and "administration, commerce and transportation"; in 1910 and 1930, these are the proportions employed in primary, secondary and tertiary industries.
- According to Francine Vanderwalle's notes, all birth and mortality statistics include stillbirths.
- Draftees are all 18-year-old men.
Availability of Data and Codebooks
All of the data and codebooks, as well as these notes and the list of common variables, are available online. You can click here for a file list. The codebooks can be viewed online:
- Van der Walle, Francine. "Education and the Demographic Transition in Switzerland." Population and Development Review, vol. 6, no. 3. 1980.
- Van der Walle, Francine. "Migration and Fertility in Ticino". In Population Studies, vol. 29, 1975. pp. 447-462.