Database on historical statistics
Via the database-system histat we offer since 2004 a product, which was developed and promoted by the GESIS-staff-members Jürgen Sensch, Gabriele Franzmann, and Rainer Metz. Histat serves as access-platform of time series data collected in the framework of economic and social historical research. Two important purposes were to be served by the development of a data-access-platform: On the one hand the access of the data stock should be uncomplicated and therefore user-friendly, on the other hand the database should meet scientific standards and offer time-series together with a detailed documentation of their context (as well as annotations on data-sources). These objectives have been successfully implemented via the histat-database. This year we will attain more than 3000 registered user from the whole world. The data base encloses meanwhile about a quarter of million time series from more than 360 studies. During the last year more than 70.000 time series has been downloaded. Monthly several studies are imported into the database. In addition histat has got an organizational solid basis. The closing of the Center of Historical Social Research (Zentrum für Historische Sozialforschung, ZHSF) and the implementation of a new team structure within GESIS ensures that the Histat-database is supervised and developed permanently by the team Data service Historical Studies. The team currently consists of ten employees. Three employees of this team are dedicated to the continuous care and expansion of histat’s data stock. In doing so they are supported by student assistants. In 2011 GESIS provided funds for an extensive relaunch. With these financial means the surface has been reworked basically and a consistent user interface across all functions was created.
Examples of studies that have used histat-data:
Klemann, Hein A. M. / Kudryashov, Sergei, Occupied Economies: An Economic History of Nazi-Occupied Europe, 1939-1945, Oxford [u.a.]: Berg 2012.
Lampe, Markus / Sharp, Paul, Tariffs and Income: A Time Series Analysis for 24 Countries, Discussion Papers on Business and Economics No. 17/2012, Department of Business and Economics Faculty of Business and Social Sciences University of Southern Denmark.
Acemoglu, Daron / Cantoni, Davide / Johnson, Simon / Robinson, James A., The Consequences of Radical Reform: The French Revolution, in: American Economic Review, 101/7 (2011), pp. 3286-3307. DOI: 10.1257/aer.101.7.3286.
Gehrmann, Rolf, Infant Mortality in Germany in the 19th Century, Comparative Population Studies – Zeitschrift für Bevölkerungswissenschaft 36/4 (2011), pp. 839-868.
Uebele, Martin, Die Identifikation internationaler Konjunkturzyklen in disaggregierten Daten: Deutschland, Frankreich und Großbritannien, 1862–1913. Jahrbuch für Wirtschaftsgeschichte / Economic History Yearbook 52/1 (2011), pp. 19-44. DOI: 10.1524/jbwg.2011.0002.
Demeulemeester, Jean-Luc / Diebolt, Claude / Jaoul-Grammare, Magali, The growth of aggregate wage earnings in Germany, 1810–1989, in: Applied Economics 43/21 (2011), pp. 2657-2669. DOI:10.1080/00036840903299813.
Because Histat was designed from the very beginning as a system of a heterogeneous data stock, the number of the time series per study as well as the covered time span of the different studies varied substantially. About 20 percent of the studies comprise up to 50 time series, other 13 percent of the studies consist of about 100 time series, and further 40 percent comprise 101 and 500 time series. Ca. ten percent of the studies contain 500 to 1.000 time series and the data section of further 10 percent of the studies comprise 1.001 to 5.000 time series. Nine studies (id est. three percent of all studies in histat) enclose more than 5.000 time series, the biggest ones of them comprise only a total of 61.274 time series. The following figure shows the number of time series per study (in logarithmic scale) as well as their starting and final points. It is shown that a whole series of studies deal with very long periods of several hundred years, including studies containing a large number of time series. Nevertheless, the main focus (dark colour) is unambiguously on the time period between 1850 and 1950 and between Studies containing about 100 and 1.000 time series.
All together, all in the database Histat contained time series enclose about five and a half million values. From the year 1800 at least thousand values for nearly each year are available, from the middle of the 19th century there are already more than 10.000 values on hand via Histat, and with the foundation of the German Empire (in 1871) there are more than 20.000 values available per year. The highlights are the years at the beginning of the 20th Century: for several years there are about 60.000 values available via histat. But in particular for the forties of the 20th century the number of available values drops to less than 2.000. For the post-war period up to the present there are between 20.000 and 30.000 values in histat. In comparison with that the so-called prestatistical period (i.e. the period prior to 1871) is rather modest. However, also here we can offer since the 16th century predominantly more than hundred values per year, and since the mid of the 17th century at least for every tenth year more than thousand values.
Since the offer of the online-database histat via the GESIS-Web a yearly growing number of registered users can be attained. In 2010 the compulsory registration for the database call was cancelled and relocated to the histat-menue of final data-download. Users of the data-base are now allowed to research the study descriptions and study contents in histat without registration. Therefore, the data-base calls have been more than doubled within two years. This year we will attain more than 4.000 registered users.