Commentary to the table section of this chapter is limited to a discussion of our estimates of the nominal and real value added for the years 1851–1913. With the exception of the value added by branch series prepared for the years 1960–1990 by the St. Gall Center for Future Research, which were made available to us in the form of computer printouts, all other tables represent compilations from the Statistical Y earbook of Switzerland, various special publications by the Federal Statistical Office (later the Federal Office of Statistics), and a supplement to the October 1985 edition of the journal “The Economy” (“Die Volkswirtschaft” / “La vie économique”).
The first major social product estimates in Switzerland were performed in the middle of the 1920s by Paul Mori and Julius Wyler. Even though both estimates concern the same year – 1924 – they vary by more than 25%. This is not the place for a detailed review of the methodologies employed by Wyler and Mori for the construction of their estimates; we leave it at mentioning that, following the presentation of their results in the “Journal for Swiss Statistics and Economy” (“Zeitschrift für schweizerische Statistik und Volkswirtschaft” / “Journal de statistique et Revue économique suisse”), the two authors did not publish for quite some time. However, after Wyler was charged with the preparation of annual income statistics by the Federal Statistical Office, it was finally possible to present an official estimate of the Swiss national income in the late 1930s.
The estimation method used by Wyler, used also for the retrospective calculation of national income in the years 1929–1937, was retained until 1960. Three years later, there was a first attempt at calculating the social product according to the principles of modern economics. But very soon, the estimation of macro-economic values had to be abandoned again “due to insufficient source data in the area of economic statistics”. A new attempt led to the 1977 publication of a sizable volume entitled “Revised Series of the National Accounting of Switzerland”, which contained homogenized estimate series for the entire time period 1948–1976. A follow-up, published in 1983, covered the years 1977–1981; the development of subsequent years is described in a series of statistical reports on the Swiss economy entitled “The national accounting in Switzerland”, which began in 1985. It must be noted, though, that Switzerland does not have a set of comprehensive (i. e. computed at the balance-of-payment level) national economy financials before 1983. Those interested in summaries of this kind can find them in supplements to the journal “The Economy” and to the monthly report of the Swiss National Bank.
The first two-page spread in the table portion of this chapter reports on the major results of the value added calculations performed in the National Fund project “Money supply and economic growth in Switzerland 1851–1913” (“Geldmenge und Wirtschaftswachstum in der Schweiz 1851–1913”). The methods used to estimate the value added of the first and second sector, and the service branches tourism, transportation, banking, insurance, and public sector, can be seen in the commentaries to the respective chapters in this volume. Remaining to be discussed are the rough estimates of the value added in the branches “personal services” and “commerce”. After that, we will briefly touch on the quality of the nominal overall value added series. Finally, a few comments to probably the most delicate aspect of the whole estimation process, the construction of the deflator, lead into the table section of this chapter.

Rough Estimate of Wholesale and Retail Value Added

We did not have sufficient data to reliably estimate the value added of this important branch, but nevertheless had to attempt such an estimate unless we risked failing our goal of presenting a value added series aggregated from absolute values. Faced with this difficult situation, we resorted to a rather daring construction method, using the difference between the consumer and wholesale price indices as the development indicator. Using the national census results, we determined the number of persons employed in the commerce branch for the years 1860, 1870, 1880, 1888, 1900, and 1910, and created a continuous employment series through interpolation and extrapolation. Assuming that the sfr. 5000 figure for the value added per employed person, which we found in the “reports and calculations” of the Basle General Consumer Union (“Allgemeiner Consumverein Basel”) for the years 1891–1895, were applicable for the entire commerce branch, we set this value at sfr. 3000 for the year 1850 and sfr. 6000 for the year 1910, and connected the two values with a straight line. The data points so gained for the years 1850, 1860, 1870, 1888, 1900, and 1910 could now be connected with one another through the development indicator. The end result was a series of estimates whose values at first sight appeared grossly inflated compared with the value added estimates of the other branches. However, the estimate for the 1960s of the St. Gall Center for Future Research also assigned a rather high weight contribution to the commerce branch for the total value added, specifically 18.0% for 1960, 18.3% for 1965, and 17.3% for 1970, whereas our estimate for 1910 indicates a contribution of 15.2%. In his analysis of the Swiss consumers cooperatives, Hans Müller wrote in the mid-1890s that while nothing was known about the number and distribution of small retail shops, such “‘mom and pop shops’ were ubiquitous wherever there was not yet a consumer cooperative, which indicates that for every few hundred souls, there was one family that made its living from intermediate trade”. (Hans Müller: Die schweizerischen Konsumgenossenschaften – ihre Entwicklung und ihre Resultate [The Swiss consumer cooperatives – their evolution and results]; Basle 1896 [p. 440]). This observation makes it conceivable that commerce – which apart from private retail and consumer cooperatives, also included all wholesale businesses – may indeed have been by far the most important branch in the service sector in the second half of the 19th and the beginning of the 20th centuries.

Rough Estimate of Value Added in the Branch “Personal Services”

Based on the census results for the period 1860–1920, and with the help of “Furrer’s National Economy Encyclopedia” (“Furrer’s Volkswirtschaftslexikon”, vol. I, p. 225), it was possible to create a continuous time series reporting on the percentage of service personnel from the total of all employed. Using the index of industrial wages, depicted in chapter G., and information provided by the Swiss Farmers Office (“Schweizerisches Bauernsekretariat” / “Secrétariat des paysans suisses”, Bulletin no. 30: “Agricultural worker issues”, p. 100) on the annual income – consisting of cash salary, room and board – of a chambermaid at the turn of the century (sfr. 702), we estimated the wage sum of all maids, farmhands, and servants, i. e. of all service personnel, and simply assumed it to be equal to the value added of this branch. A control calculation for the year 1888 merely showed a sfr. 20 deviation between the annual wages reported for a maid (sfr. 571) by the Swiss Farmers Office, and our value added estimate (sfr. 551) originating from the industrial wage series. Nevertheless, a value added series arrived at in this manner has the characteristics of a rough estimate at best.

Quality of the Aggregated Nominal Value Added Series 1851–1913

Since some of the value added series we created at the branch and sector levels are marred by serious weaknesses, the overall result of our research also does not represent a high quality product. We were, however, aware of this from the start; anyone who has an idea of the precarious data situation will view the mere existence of an overall total as a success. Perhaps it is defensible to go further yet and dare to state the conjecture that our final series looks quite plausible, at least from a bird’s-eye view. We cannot here delve into a comparison with the value added estimates for other European nations, but note with a certain relief that our estimate of sfr. 4000 million for the year 1910 only varies insignificantly from the estimate by Ulrich Zwingli and Edgar Ducret for the same year (sfr. 4324 million). It must be pointed out, though, that Zwingli and Ducret estimated not the gross domestic product, but the net social product at market prices, a value which, in contrast to the gross domestic product, also includes the foreign net income but not the national economy depreciations. The difference between the two estimates increases if one omits for the calculation of the value added for the year 1910 the foreign net income, but assumes a value for depreciations significantly different from zero. On the other hand, it must be considered that for the years 1910 and 1913, estimates considerably lower than those of Zwingli and Ducret also exist. Calculations by Franz Ritzmann, Ernst Laur, Julius Landmann, and Paul Mori for the years 1910 or 1913 yielded values of sfr. 4100, 4000, 3500, and 3400 million. Traugott Geering even assessed the Swiss national income for the last year before World War I at only sfr. 2500–3000 million. With the help of replacement indicators (bank balance sums and import values), Zwingli and Ducret computed a net social product of sfr. 2500–2800 million for the year 1900, while W. Geiger, F. Kneschaurek, and W. Winkler indicate a value of sfr. 2230 million. (All numbers originate from an essay by Zwingli and Ducret listed in the bibliography). Our own estimate shows a value of approximately sfr. 2500 million for the year 1900, and thus agrees with the lower value of the estimates by Zwingli and Ducret. This, however, only holds true if commerce value added was indeed as high as we estimate.

Quality of the Aggregated Real Value Added Series 1851–1913

It is obvious that the value added given in nonadjusted prices does not yet represent a useful prosperity indicator. Because economic prosperity is measured primarily by consumer buying power, we needed a real value that includes overall price level progression in order to comment on this buying power. Our task therefore was to design an aggregated price index or deflator. For this purpose, the staff member of our team assigned to this project, Hildegard Muff, first constructed a consumer price index (CPI) for the years 1851–1890, which could easily be linked to the price index “Food, beverages, heating and lighting” generated under the National Fund project “Real salaries of Swiss industrial workers from 1890 to 1921” (“Reallöhne schweizerischer Industriearbeiter von 1890 bis 1921”, see commentary to the table section of chapter H.). One flaw is that the aggregate, which Muff computed according to Laspeyres’ formula, merely indicates consumer price evolution in the cities of Zurich and Berne, while the index starting in 1890 has a much broader geographic basis. If the series were to be revised, it should primarily reflect the fact that there are credible price statistics for the Frenchspeaking part of Switzerland as well. After estimating the CPI, Muff expanded her index series into a so-called consumer deflation index that, apart from the groups food, beverages, heating and lighting, fed on three additional sub-indices, those being rent, clothing, and construction cost series. Since we are not entirely certain that the quality of the deflator really benefited from this inclusion, for caution’s sake we applied not only the consumer deflation index, but also the CPI, to the nominal value added series that we had estimated. At this point we are unable to comment on the quality of the two resulting series.

SOURCE: «National Accounting» in Ritzmann/Siegenthaler, Historical Statistics of Switzerland, Zürich: Chronos, 1996, 859-864