Labor force by trade
Trade and employment statistics have been an integral part of the Swiss census since 1860, and data relating to trade affiliation, i. e. the social structure of employees and their dependents, exists since the 1888 census. It is, however, impossible to compile homogenous time series at the line of business level covering the years 1860 to 1980. (The results of the 1990 census were not yet available at the time of completion of this chapter). In truth, each individual census significantly varies from its predecessor, which greatly increases the difficulty of comparisons between censuses. In addition, the quality of the data from earlier censuses leaves much to be desired. For this reason, the 1860 census should only be interpreted with utmost caution, and substantial reservations, at least, are in order for the 1870 census. This is already evident in the opening commentary to the 1880 employment statistics, which, among other observations, complains that the 1860 and 1870 census takers had been burdened with too many ancillary tasks – in 1860 they were ordered to perform a count of rifles and in 1870 one of businesses. The 1880 census, too, had serious omissions and a plethora of ambiguous descriptions “on the majority of the forms of this vast counting venture,” so that in April 1881 the Swiss Department of the Interior decided, “that return be arranged by the canton governments of all census forms containing such ambiguities and omissions to the appropriate census taker for the purpose of completion, with a detailed notice of the omitted in each case. In addition, each canton was provided a list of returned forms, with each form having a list of the omissions to its completion attached to it; there were detailed records of all returned materials, for unless the three census volumes remain unwritten, all proper census data had to be recovered and accounted for. This did indeed happen, though by far not all ambiguities and omissions had been resolved, partly because the census takers were unable to ascertain the required information from certain people.” (Statistical delivery [“Statistische Lieferung” / “Livraison statistique”] No. 59: 1880 Census, Vol. 3, p. III–XIV).
These complaints did not fail to have an impact: There was a substantial quality increase in general census results between 1880 and 1888 which applied to the employment statistics as well. However, the counts from 1888–1960 still fall short in several areas. This problem is addressed in the introduction to the third volume of the 1888 census, which covers the topic “population by trade” as follows: “Experts have long known the presentation of the population’ s employment condition as the most complex part of a census, and its results as the least complete. Those familiar and well known problems again surfaced in the current census and, despite all efforts, could not be completely overcome.
These problems mainly consist of the fact that the pertaining terminology describing occupation and occupational activities in general and often that of specific occupations, does not describe them precisely and with enough differentiation from similar areas; that many of those terms have border areas, often of substantial size, whose belonging to the term in question remains uncertain. This inevitably leads to ambiguity and uncertainty, starting with the recording of occupations on the census forms and their statistical compilation, which may impact the comparability of results or even put them in question altogether – and this applies not only to statistics from different countries, but even to compilations from a single country if the date originates from different times and different tallies.”
Nonetheless, the Office felt compelled to energetically counter the criticism of certain municipal governments on the “characteristics and usefulness of the figures arrived at”: “The experience gained by the Statistical Office during its comprehensive preparations did not indicate that occupation data on the forms would be complete and error-free by any means, but, in general, neither can this data be considered unsatisfactory; in general, it was better and better suited for detailed analysis than any data from any prior census. The goals of this census, and what had been expected from it, has been fulfilled for the most part. The primary task for the future will be to achieve the same in a simpler, more reliable manner, and thus one less cumbersome for all concerned.” (Statistical Delivery No. 97: 1888 Census, Vol. 3, p. 3*–11*).
Even today such a differentiated and well measured contemplation is surely quite supportable. But the resolution to design future tallies not only more simply but also more reliably, proclaimed in the closing sentence, ignored the dynamics of the structural changes which trades became subjected to in the late 19th and early 20th century. For the 1900 census, the Federal Statistical Office dropped the system of economic classification that the 1888 census was based on and replaced it with a more complex version. This happened again for the 1910 and 1920 census.
Viewed from the perspective of the late 20th century, the counts of 1870 and 1880 and those of 1920 to 1960 appear the most comparable among themselves. The most important changes in the counting methodologies occurred between 1880 and 1888, 1888 and 1900, 1900 and 1910, 1910 and 1920, 1960 and 1970, and 1970 and 1980. We cannot describe here in detail where two successive counts differentiate in major ways. (Note, however, the respective remarks attached to tables). Despite the frequent changes made to categories by the authorities, we feel it is still quite acceptable to present tables covering different trades over longer periods of time.
For certain occupations, primarily in the textile industry and its sub-parts, we can present newer estimates which reach as far back as to the first half of the 19th century, if not farther. At times those estimates disagree substantially. With respect to this it must first be taken into account that Gruner’s data only covers one part of the employed population, namely manufacturing workers. Second, one may safely assume the estimates of Dudzik and Siegenthaler to be generally better supported than those of Kneschaurek. Third, it must be considered that, for his own estimates, Ritzmann-Blickenstorfer did not use new sources, but merely compared the estimates of Gruner, Dudzik, Siegenthaler, and Kneschaurek against each other and with the results of the 1870 and 1880 census. (The latter estimates were performed in order to weight the trade salary indexes for the years 1821–1890).
The first official trade statistics divided by sectors and occupations for all of Switzerland covers the era of 1870–1960. The values for 1870 and 1880 are based on a retrospective table which we took from the 1900 census (Statistical Delivery No. 151). A comparison with data from the 1888 census enabled us to apply the newer category system also on the years 1870 and 1880. We found a second retrospective summary in a table volume of the 1960 census; from that statistic originate the values for 1888 up to and including 1960. We quote from the text accompanying that table: “In general, time series have been taken as far into the past as it was possible to put together comparable numbers. The compilation of comparable partial sets is complicated because the schema of trade classification [...] changed from census to census as older trades or occupations disappeared and their functions became themselves trades or merged into others. In retrospective summaries, such trades and occupations are generally assigned to those groups which they most belong to by their nature. Only a small number cannot be assigned conclusively, and they do not diminish the validity of the numbers.” (“Statistical Source Books of Switzerland” [Statistische Quellenwerke der Schweiz” / “Statistiques de la Suisse”], vol. 385).
In this spirit, we feel that the restatement of categories which we applied to the 1870 and 1880 occupation statistic categories does not represent a problem.
It was now tempting to also demonstrate the development in the cantons through a combination of the two mentioned retrospective tables. However, this was not advisable as those retrospective summaries, at the canton level, covered only the three main sectors plus a few of the major branches. In the industrial sector, especially, there is no differentiation into trades; we exclusively have data on the population employed in mining, “industry and trade,” construction, and energy at our disposal. The years 1860–1960 and 1888–1960, however, did not only bring the well known occupation migration from the first into the second and then third sector in most cantons, but there was a most significant structural change within the second sector. In order to demonstrate this change, it is necessary to know the number of people employed at different times in the main branches of the industrial sector, primarily the textile industry and its subbranches and in the metal and machine industry. For this reason we engaged in the very time-consuming attempt to reconstruct, based on the 1860 to 1960 census numbers for the employment structure in each individual canton, the numerical progression of employment in the major branches of the first, second, and third sector. Since the 1970 and 1980 census combine fulland part-time employees while the tallies in earlier counts only cover full-time occupation we chose not to extend the branch statistics for individual cantons beyond the era of 1860–1960. For all of Switzerland, however, we have retrospective tables at our disposal which originate from the census year 1980.
We acknowledge that there is room for refinement in four directions even in these detailed presentations: First, it would be desirable to have the proportion of male and female employees for each branch in each canton; second, it would be interesting to know about the number of Swiss versus foreign employees by branch, and by canton; third, there could be statistics of the families of those employed in the individual branches (including, again, a differentiation by gender and national origin); and, finally, fourth, the majority of the counts essentially offered the possibility to break down the number of employees and dependents for the more important branches also by district and cities. The counts of businesses, performed since 1905 in irregular intervals, though receiving only rudimentary treatment in this chapter, occasionally contain numbers which would permit such a breakdown.
All these projects still remain to be performed. In terms of scientific currency, their payoff would consist in a much more precise quantification and description than has been possible to date of the Swiss population’s individual stops on its journey to the modern service economy.
Employment of foreign nationals
The development of the employed foreign national population has achieved growing significance since approximately the mid 1950s and thus gained increasing attention from statisticians. Using two complementary sources, the “Handbook of Swiss Social Statistics 1932–1971” (“Handbuch der schweizerischen Sozialstatistik 1932–1971” / “Manuel de statistique sociale suisse 1932–1971”) published by the Federal Ministry of Industry, Trade, and Labor and the magazine “The Economy” (“Die Volkswirtschaft” / “La vie économique”) published by the Swiss Department of National Economy, we attempted to depict important structural changes, which could be observed in this universe between 1955 and 1991 in the form of time series. We were able to facilitate this by combining individual characteristics of the entire employed foreign population, such as immigration status, gender, national origin, and branch of employment. Those numbers are then shown either annually or for selected years.
Using data published in the Statistical Yearbook of Switzerland, the monthly number of those seeking full time employment, and their composition by branch, can be reconstructed back to the years 1917 and 1920, respectively. At the canton level, those statistics commence at the end of the 1920s, but full unemployment data is carried only since 1938. We divided these statistics into three parts, of which in the table section we show only the first, covering the Great Depression of the 1930s, and the last, pertaining to the 20-year period 1973–1992. For the period 1931–1938 our tables depict the monthly number of job seekers registered with the unemployment offices of the cantons and a few of the larger cities; and for the years 1973–1992 at the canton level, the annual number of job seekers, the number of fully and partially unemployed persons, and the number of job openings. The job market performance of the cantons, the crisis susceptibility of certain branches, and the development of overall unemployment during the Great Depression can be determined even more precisely by the inclusion of data collected by the major unemployment insurance institutions in those dramatic times. This data consists of monthly statistics that show the number of those insured and the degree to which a given insurance institution was taxed by the partial or full unemployment of its members, all strictly divided by public, “private one-sided”, and “private parity” type of institutions. We adopted those categories and processed the data printed in 1931 in the “Economic and Social Statistical Bulletin” (“Wirtschaftliche und sozialstatistische Mitteilungen” / “Rapports économiques et Statistique sociale”) and after 1932 in “The Economy” in a manner that allowed the formation of two summary indices populated with monthly values for the period 1931/32–1938, one showing the number of those insured with public and private carriers and the other the portion of those fully and partially unemployed for those same institutions.
Weekly working hours in industry were dealt with in the factory inquiry of 1895, but it was not until 1942 that the Federal Office for Industry, Trade, and Labor began publishing annual averages. Until 1968, these values were based on notifications by factories about paid weekly working hours; in the years 1973–1983 they had their basis in records of (individual) working hours of accident victims insured with the Swiss Accident Insurance Company (SUVA). Since 1984, the revised employment statistics provides the pertaining working hours.
In addition to the numbers published by the statistical offices, we also have the retrospective surveys performed under the National Fund project “Real salaries of Swiss industrial workers from 1890 to 1921” (“Reallöhne schweizerischer Industriearbeiter von 1890 bis 1921”). The participants in this project did not only analyze the printed factory statistics and reports of the federal factory inspectors, but also processed a series of registers, contemporary reports, and scientific studies of unions and workers’ conditions. Detailed data on branch-specific working hours was found in a reference book authored by Erich Gruner and Hans-Rudolf Wiedmer.