Computer-Assisted Plotting and Analysis of Village Returns in German National Elections, 1893-1912

In the past five years, an impressive amount of new research has been published on the history of German elections prior to the First World War. For all the virtues of this research, one key aspect of politics has been ignored ? how party politics actually functioned at the village level. Historians have not neglected local political history because it is unimportant, but rather because it was thought that voting results for individual villages did not exist. During my own research on German agrarianism, I have come across material that should significantly reduce this deficit. This material will also enable us to recognize more accurately the social and economic bases of agrarian radicalism in Germany between 1893 and 1914.

.01. THE STATE OF GERMAN ELECTORAL HISTORY

The past 5 years have witnessed a renaissance of work on German elections in the time period 1871-1914. No less than 4 major works have appeared on the topic. The contributions of Jürgen Schmädeke, Jonathan Sperber, Brett Fairbairn, and Margaret Anderson although differing in interpretation, have all greatly advanced our knowledge of the dynamics of German politics in a time ofhave all written massive works that, although differing in interpretation, have greatly advanced our knowledge of the dynamics of German politics in a time of massive change and upheaval. [1] In discussing the critical decade of the 1890s, however, all these works share one shortcoming: they do not address the local dynamics of change and continuity as they were negotiated at the local, village level. Schmädeke and Sperber's interpretation depends upon statistical analysis, combining election results and census data. Frequently, this means comparing apples and oranges, since they use data collected at dissimilar administrative levels. None of this data, moreover, reaches down to the level of the village. Fairbairn and Anderson, on the other hand, rely primarily upon newspaper and archival evidence. Anderson's analysis is carried out in a very complete and sophisticated manner, but the same problem remains - political life in the villages is only sporadically addressed. All four scholars were hindered by the fact that village election returns are thought either not to exist or to exist only in isolated instances. This is particularly troubling when one considers the key role usually ascribed to local élites in the countryside as they struggled to cope with sinking commodity prices, popular mobilization, and the increasing democratization of the political system. [2]

Of the four authors cited above, only Schmädeke indulges in historical mapping - the second volume of his work consists of a boxed set of 32 digitally produced color maps - but the very process of unfolding and refolding 32 separate maps each displaying only two variables limits the impact of his presentation. In short, they lack the dynamic, interactive dimension that historical GIS analysis can provide. If historical GIS were to be utilized for purposes of German electoral studies, would it reveal anything noteworthy that other methodologies have missed? [3]

As Schmädeke's work indicates, the use of historical GIS in understanding Imperial German elections is much less advanced and is far behind what the British have accomplished with the British Isles Historical Survey. Andreas Kunz at the Institute for European History in Mainz and Norbert Winnige, at the Forschungsinstitut für die Geschichte Preussens in Berlin are both undertaking the creation and dissemination of digitized maps. According to reports made at a workshop in Florence, Italy in June 2000, both have suffered funding problems, but Kunz has succeeded in creating and posting a series of excellent historical maps. [4] Christian Stögebauer, a doctoral student at the Institute for Economic History in Munich, has demonstrated the value of historical GIS in his "Weimar Voting" project. In a dissertation funded by the Volkswagen Foundation, Stögebauer created a beautifully dynamic map of voting which charted four variables - confession, class, Nazi and Communist votes - for the last four elections in the Weimar Republic. Unfortunately, Stögebauer's work has suffered the typical response from historical traditionalists towards those moving into historical GIS. [5]

On a more positive note, a series of sites employing historical GIS have sprung up in the wake of the German federal elections of 1998. The best of these, which displays 50 years of election results, was prepared by the Friedrich-Ebert-Stiftung in Bonn, which is allied to the Social Democratic Party. This is a graphically pleasing site, which I have used as a template for my own work. [6]

.02. PROBLEMS FACING STUDENTS OF GERMAN ELECTORAL HISTORY

Both traditional historians and practitioners of the new school face a series of obstacles. Foremost among them is the federal nature of the old German Empire. While the Empire possessed a national parliament - the Reichstag, consisting of 397 members elected through a system of universal, secret, adult male suffrage - it was also a federal state. The 36 federal states not only exercised significant power over broad ranges of governmental functions, but also elected legislatures through different forms of franchise and kept data along different geographic units.

This level of federalism makes performing statistical comparisons most difficult. The Statistical Handbook of the German Empire, published by the Imperial Statistical Office and records the votes cast in every national election for the candidates of different political parties down to the county (or Kreis) level. [7] The same Statistical Office also conducted a decennial occupational census for the Kingdom of Prussia, but published the results not by county but by Prussian administrative district (or Regierungsbezirk). And not to be outdone in the apples and oranges game, it published the results of its decennial population census by village.

A second problem faced by electoral historians is the archive-centered nature of graduate training. When I went off to examine local elections, I read through documents such as this linked PDF file, a report from the president of the administrative district of Stade (the Regierungspräsident) to the provincial supervisor (Oberpräsident) in Hanover. [8] This particular report, written in August 1900, is very important to the project that I will be discussing, but it contains no statistics. In fact, it is a rare instance to find village results in the archival record, and such instances are usually only for one or two villages which were of obvious concern to the authorities.

.03. DIEDRICH HAHN AND GERMAN AGRARIANS

While researching a biography of prominent agrarian politician and functionary, Diederich Hahn, I have stumbled across material that, I believe, allows us to extend our understanding of the German electoral politics into the villages themselves. Who was Diederich Hahn? [9] Hahn was first elected to the national parliament - the Reichstag - in 1893. He represented the 19^th Hanoverian Reichstag constituency, which extended in a crescent along the North Sea coast between Hamburg on the Elbe and Bremen on the Weser. The 19^th electoral district was one of three such districts (the others being the 17^th and 18^th) contained within the Prussian administrative district of Stade. [10]

Hahn was an extraordinarily innovative - and extraordinarily noisy - politician. While serving in parliament, he was also the director of Germany's largest, best-funded interest group, the Agrarian League. [11] As the director of the League, Hahn was a participant in all the major political controversies of the Wilhelmine era, especially those regarding trade and tariffs. This prominence was reflected in the political cartoons attached below. [12] In the first, favorable representation Hahn is shown in the center of the cartoon, with his arm upon the Chancellor's shoulder as he begs for a better deal for agriculture. In a second, and far more typical depiction, Hahn is portrayed as Don Quixote, preparing to lead his legions of supporters in yet another fruitless crusade against the forces of modernity. Most frequently, sport was made with Hahn's name ("Hahn" meant "rooster" in German). All indicate that Hahn was a household name in the German political world.

In the spring of 1997, I happened to be reading the coverage of the 1893 national election in Hanover's province-wide newspaper, the Hannoversche Courier. As I was reeling the microfilm forward to see if the Courier contained a post-election response to Hahn's successful bid to win a parliamentary mandate, I saw a list which gave the votes cast for each candidate in every village within the 19^th electoral district. This was terribly exciting since, as I have stated earlier, in my earlier archival work I had found only limited village data. I quickly put all my other work aside and scanned through the issues subsequent to the elections of 1898, 1903, 1907, and 1912. To my disappointment, I found only very incomplete village returns for 1898 and 1907, with none for 1903 and 1912. I was, however, able to find such returns by using newspapers published within the constituency held at the Institut für Zeitungsforschung in Dortmund. [13] Using these local newspapers, I was able to compile a complete list of results, from every village and polling place, for each of the five national elections between 1893 and 1912. Moreover, it became clear that newspapers in the administrative district of Stade regularly reported polling place results from all three Reichstag constituencies within the Regierungsbezirk's borders for the five national elections between 1893 and 1912. [14] Why historians have previously overlooked this data, I can only guess. Perhaps the data is peculiar to the North Sea coast. Perhaps people have failed to look at the back pages of local newspapers, in the section equivalent to the want ads.

.04. THE PROJECT

The first step in this project was compiling polling place data. Using these newspapers as sources, I arrived at a list of 91 polling places for the entire 19^th electoral district in the first four elections, with twelve additional polling places added in 1912. [15] While doing this I made a series of methodological notes explaining how I arrived at the list. [16] I then assigned each polling place a number form 1 to 91 based on an alphabetic listing. (Those twelve added in 1912 were assigned 92 to 103 alphabetically.) As the attached files show, I created a numbered and alphabetized list, which I could then use to record election results. [17] I originally created these charts in Microsoft Word and posted them to the web site on German agrarianism that I created. I have since transferred the results data into an Excel spreadsheet so that I can move them into a GIS format.

Once I had organized the data, I next set about creating maps so that I could plot the election results by polling place. To create a map displaying the 91 polling places, I scanned a map of village boundaries prepared by the Lower Saxon government. [18] Using Adobe Photoshop 4.0 and the list of polling places that I had reconstructed, I revised this map to create a map of Wilhelmine-era polling places. [19] The polling places on these maps are identified by scroll-overs and alt-tags, but I have not been able to get those features to work when uploading the maps onto the web. Using the template thus created, I was able to plot the results for the five national parliamentary elections between 1893 and 1912, showing which political party (Agrarian, National Liberal, German-Hanoverian or Social Democratic) carried each village or polling place. [20]

Once this mapping was complete, I was able to create maps to plot cumulatively those villages always or usually carried by any one party. [21] The attached map plots strongholds (Hochburgen) that a party carried in a given election (in this case 1893) with 65%of the vote or more, those that a given party carried with a majority of the vote, and those which were carried by a plurality. [22] I then prepared yet another map plotting each village by whether it was always (or nearly always, four of five times) was carried by the same party. [23] Using the same data, I also prepared charts listing each village according to the stronghold/majority/plurality criteria both for each individual election and cumulatively over the entire period. [24] Using the same base polling place map, I have also been able to plot other variables such as land tax rates, population density and confession.

At the same time that I created the polling place map, I also created a soil map of the administrative district. To do so, I scanned a gray-scale geological map displaying soil composition, water and field usage. Using Adobe Photoshop 4.0, I colored this map and placed over it a layer consisting of the polling place map. By coordinating a map displaying the types of soil found between the Elbe and Weser rivers with the map of village boundaries, the strong correlation between soil fertility and the success of Agrarian candidatures is readily apparent. [25] Marsch areas are very fertile and highly productive. The marsch lands along the North Sea Coast paid the highest land tax — which was calculated on soil fertility — in the Prussian kingdom. Geest, on the other hand is very poor, sandy soil that produces very low cereal yields and is best used as pasturage.

In addition to creating maps to plot election results and various other data, I decided to create a "home page" for each village or polling place. [26] I created the original home pages using Netscape Composer. At the top of each home page, I listed the name and number of the place, as well as its county (of which there are six in the constituency). Using the Gemeindelexikon der Provinz Hannover, which was based on figures collected in the 1905 census, I listed the area and population of each place (from which I calculated population density), as well as average land tax rate and tax revenue. Dividing the tax revenue by the population, I derived a per capita wealth figure. [27] Each of these six categories was then ranked, from 1 to 91 and 1 to 103. The census data also lists Evangelicals, other Protestants, Catholics and Jews, from which I have calculated each faith's percentage of the village population. On page two of each home page, I have created charts showing the votes each party received in the village and various analyses that have been undertaken with the data. On page three, I have listed whether my newspaper sources report the village as the scene of a Hahn political rally. Lastly, I have listed, by year, the individuals from that place who signed a petition supporting Diederich Hahn's election. Each polling place home page is accessible in two ways, either by double-clicking on an individual polling place in any of the elections maps (they were each created with hotspots imbedded), or by choosing it from a menu list. For the second generation of these pages, created in Dreamweaver 3, I have separated out the data into distinct pages, accessible through menu lists. [28]

Just as the polling place home pages are now in their second generation, I have migrated the original result maps into a new format built from frames using Dreamweaver 3. [29] In this new format, I have expanded the study to include data that I have collected from the entire administrative district of Stade. As is clear, this new format reflects the design of the Friedrich-Ebert-Stiftung "Wahlatlas." Because of the difficulties that I have experienced with the display window of the "Wahlatlas" - it is run as an applet and has a tendency to crash - I have created a fresh page for each display. There are still bugs that I have not worked out - the frames format tests successfully on both Internet Explorer and Netscape according to the <preview in browser> selection in Dreamweaver, but when I access the site independently on Netscape, the frameset varies and the menu lists do not display properly.

.05. PROVISIONAL CONCLUSIONS

The conclusions that can be drawn from this project are still provisional. I have expanded the scope of the project to include the other two electoral districts in the administrative district of Stade. [30] I continue to add data into my database. I have begun to construct an interactive map for the project similar to the National Atlas web site produced by the United States Department of the Interior. [31] Even so, what does the initial attempt at computer-assisted historical mapping tell us?

• The mapping of local results enables us to see clearly the geographic distribution of the Agrarian electorate.
• When combined with a map of the soil composition of the 19th electoral district, it confirms contemporary reports that the Agrarian constituency was in well-to-do farming areas, the areas with higher land tax valuations.
• The "normal" Agrarian village was average in area and population and as a rule less densely populated than average. It was also in the upper third of places for land tax, land tax revenue and revenue per capita.
• The Agrarian turnout was relatively stable - Hahn carried 35 polling places in every election and a further 14 in four of five elections.
• Although the Agrarian base was relatively stable, overall voter turnout was low. This pointed to potentially successful National Liberal and Social Democratic strategies if those parties could bring non-voters to the polls.
• "Marsch" communities enforced enormous political conformity. In the 1893 election, for instance, Hahn carried 26 of 91 polling places with more than 65% of the votes cast. In Krautsand, he received 244 of 254 votes cast (96%). In neighboring Dornbusch he won 91%, in Drochtersen 90%, in Großenwörden 86%, and in Neuenkirchen/Oste 89%.
• Civil and political society was most open in the county of Neuhaus, least open in the counties of Kehdingen and Hadeln.
• When ballot secrecy was increased in 1903, socialists were able to make inroads into Hahn villages.

.06. FUTURE RESEARCH AGENDA

I plan to effect a series of technical improvements to the site. As indicated above, I intend to include data from the 17th and 18th electoral districts so that the site will be a complete local electoral history of the Stade administrative district from 1893 to 1912. When doing this, I will need to work out the bugs, such as non-working hotspots and browser compatibility issues. I am trying to find a way to include two thumbnail maps on the polling place home pages: one to highlight the polling place upon a map of the electoral district, and the other to use current topographic features such as those created by the Lower Saxon State Survey Office. [32]

A more substantial improvement will be to create a companion GIS to be accessible on the Internet. It will be similar to the National Atlas of the United States, allowing cross-referencing of population, income, tax, vocational and confessional data with the local election results. I will continue to cross-post this data in the Dreamweaver 3 format that I have devised for the current project and as hyper-linked text so that the material is accessible across various platforms without the importation of new software. Once the new GIS format is in place, I will supplement it with the petition data that I have collected.

These technical improvements, when linked with the insights that I have already garnered from the project, will facilitate a whole new series of new and revised research questions. Four come immediately to mind:

• How unique was Diederich Hahn and the local political machine that he created? By expanding the scope of my analysis to the neighboring constituencies in the Regierungsbezirk Stade, I will perhaps get a better sense of the interplay between environmental determinism and personal agency.
• What was the link between levels of indebtedness and Agrarian activity? Historians of the 1890s are divided as to whether the Agrarian upsurge was the result of large cereal-producing landowners' anger at new tariffs favoring industry or the result of a crisis of commodity prices and wide-spread indebtedness. This project has already shown the correlation between high-value farming areas and Agrarianism. Were those areas also more indebted than others?
• What was the relationship between Agrarianism and anti-Socialism? Hahn was prolifically anti-Socialist, an opposition that he more frequently based upon religious/ethical rather than economic grounds. Did the Agrarian cause benefit more from the increasing presence of Social Democrats (in terms of increased bourgeois turnout)?
• Similarly, what was the relationship between Agrarianism and anti-Semitism? Hahn was a virulent anti-Semite from his student days, but anti-Semitism was neither widespread nor popular in the Stade region when he began his political career. Did it prove a useful tool or a stumbling block as he sought to create a winning political base? Were Jews at all present in those areas where Hahn had his strongest support, or was Hahn's success a precursor to the phenomenon of "anti-Semitism without Jews"?
• Derived from the prior two questions, was the Agrarian League, as Hans-Jürgen Puhle insisted, a dress rehearsal for fascism? [33]

Without the local election results detailed and plotted in this article, none of these questions can be definitively answered. With the advent of computer-assisted historical research, perhaps we will come closer than before to answering these questions as well as offering more wide-ranging explanations of German political behaviour in the first half of the twentieth century.

.07. NOTES

1. Jürgen Schmädeke, Wählerbewegung im Wilhelminischen Deutschland (Berlin: Akademie Verlag, 1995); Jonathan Sperber, The Kaiser's Voters. Electors and Election in Imperial Germany (Cambridge, Cambridge University Press, 1997); Brent Fairbairn, Democracy in the Undemocratic State. The German Reichstag Elections of 18998 and 1903 (Toronto, University of Toronto Press, 1997); Margaret Anderson, Practicing Democracy. Elections and Political Culture in Imperial Germany (Princeton: Princeton University Press 2000).

2. Even in 1912, by which time Germany had become Europe's second leading industrial power, the countryside and rural élites retained enormous power. Of the Reichstag's 397 constituencies, in 1912, 104 (26%) possessed a majority earning its livelihood from agriculture and 49 (12%) possessed a significant minority that did so. Statistik des Deutschen Reichs, volume 250, (Berlin, 1912/13), pp. 4-68. Also see Thomas Kühne, "Das Deutsche Kaiserreich 1871-1918 und seine politische Kultur: Demokratisierung, Segmentierung, Militarisierung," Neue politische Literatur 43/2 (1998): 206-264.

3. The reader is directed to Social Science History, volume 24/3(2000), that is entirely devoted to works of historical GIS presented at the SSHA meeting in Fort Worth, September 1999.

4. Winnige's project can be accessed at http://www2.hu-berlin.de/fgp, Kunz's at http://www.geog.port.ac.uk/hist-bound/papers/germany.htm. The progress reports for both projects can be accessed at http://www.geog.port.ac.uk/hist-bound/. For Kunz's site, see http://www.inst-euro-hist.uni-mainz.de/for/iegmaps.htm.

5. See http://www.weimar-voting.de.

6. See http://www.wahlatlas.de.

7. Vierteljahreshefte zur Statistik des Deutschen Reiches 1893, 1898,1903,1907, 1912.

8. 8.gif

9. For background on Hahn see George Vascik, "Agrarian Conservatism in Wilhelmine Germany: Diederich Hahn and the Agrarian League," in Between Reform, Reaction and Resistance. Studies in the History of German Conservatism from 1798 to 1945, eds. Larry Eugene Jones and James Retallack, (Providence: Berg, 1993), pp. 252-256, and Alfred Vagts, "Diederich Hahn - ein Politikerleben," Jahrbuch der Männer vom Morgenstern , 46 (1965): 155-192. Further Hahn material can be accessed at http://www.ham.muohio.edu/vascikgs/agrarianism.html.

10. The 19th electoral district: 10a.gif. Location of the administrative district of Stade within the German Empire: 10b.gif.

11. Thomas Nipperdey, Deutsche Geschichte 1866-1918, 2nd volume, (Munich: C.H. Beck, 1995), p. 583. The essential texts on the League are Sarah Tirrell, German Agrarian Politics after Bismarck's Fall. The Formation of the Farmers' League (New York: Columbia University Press, 1951), and Hans-Jürgen Puhle, Agrarische Interessenpolitik und preussischer Konservatismus im wilhelminischen Reich 1893-1914. Ein Beitrag zur Analyse des Nationalismus in Deutschland am Beispiel des Bundes der Landwirte und der Deutsch-Konservativen Partei (Hanover: Verlag für Literatur und Zeitgeschehen, 1967).

12. 12a.gif, 12b.gif. Hahn 12c.gif. is also important in that he embodied the new sort of professional politician that emerged in Germany at the turn-of-the-century - what Max Weber called the "new political man". Such politicians, in Weber's formulation, lived both from and for politics.

13. The newspapers that were consulted were the Stader Tagblatt, the Neuhaus-Ostener Zeitung, the Otterndorfer Wochenblatt (later the Nordhannoversche Landeszeitung), the Nordwestdeutsche Zeitung and the Provinzial-Zeitung.

14. In addition to this electoral data, I also found extensive lists of names appended to petitions supporting the various candidates. These petitions exist in greater or lesser length for all five elections and list over 2200 individuals. Many of these individuals, who were all listed by village, also gave their occupation. I have presented this data in another forum, and will be publishing it in a forthcoming issue of German Studies Review.

15. 15.php

16. See a discussion of methodology: methodology.php

17. http://www.ham.muohio.edu/vascikgs/agrarianism.html

18. Topographisches Atlas Niedersachsen und Bremen, Niedersächsische Landesverwaltungsamt (Hanover, 1977).

19. 19.jpg

20. These sites will view best if accessed with Internet Explorer! Results for 1893: http://www.ham.muohio.edu/~vascikgs/H191893.html.

• 1898: http://www.ham.muohio.edu/~vascikgs/H191898.html.
• 1903: http://www.ham.muohio.edu/~vascikgs/H191903.html.
• 1907: http://www.ham.muohio.edu/~vascikgs/H191907.html.
• 1912: http://www.ham.muohio.edu/~vascikgs/H191912.html.

21. By way of introduction to electoral politics in Hanover see Bernhard Ehrenfeuchter, "Politische Willensbildung in Niedersachsen zur Zeit des Kaiserreiches," (Ph.D. diss., Georg-Augustus Universität Göttingen, 1951) and Johannes Flathmann, Die Reichstagswahlen in der Provinz Hannover 1867-1896 (Hanover, 1897), and Manfred Hamann, "Politische Kräfte in ther Provinz Hannover am Vorabend des Ersten Weltkrieges," in Dieter Brosius and Martin Last, Beiträge zur niederächsischen Landesgeschichte. Zum 65. Geburtstag von Hans Patze, (Hildesheim: Lax, 1984), pp. 421-453.

22. 22.tiff.

23. 23.gif.

24. baseanalysis.php

25. 25.gif.

26. http://www.ham.muohio.edu/~vascikgs/H1910pp.html

27. Gemeindelexikon der Provinz Hannover (Berlin, 1908). I have not yet entered the data from a prior Gemeindelexikon reflecting the 1885 census. As I migrate my work into a GIS format, the population shifts reflected by these two censuses should provide further information as to the context of agrarian activism.

28. http://www.ham.muohio.edu/~vascikgs/H191ppm.html.

29. http://www.ham.muohio.edu/~vascikgs/H191893.html.

30. I presented an early version of this work at the October 1998 meeting of the German Studies Association in Salt Lake City, Utah. (See http://www.ham.muohio.edu/~vascikgs/GSApaper.html.) I will present the conclusions garnered from these further steps in a paper to be delivered at the international conference of the International Association for History and Computing in Posnan, August 2001.

31. http://www.nationalatlas.gov/.

32. Landesvermessung + Geobasisinformation Niedersachsen, Top 50. Niedersachsen/Bremen. Amtliche Topographische Karten. See http://www.lgn.de

33. Hans-Jürgen Puhle, Von der Agrarkrise zum Präfaschismus. Thesen zum Stellenwert der Agrarischen Interessenverbände in der Deutschen Politik am Ende des 19. Jahrhundert (Wiesbaden: Steiner, 1972).

George S. Vascik: [email protected]

8.gif

10a.gif

10b.gif

12a.gif

12b.gif

12c.gif

POLLING PLACES IN THE NINETEENTH HANOVERIAN ELECTORAL DISTRICT

METHODOLOGICAL NOTES

The village statistics used in this study are derived from Gemeindelexikon für die Provinz Hannover (Berlin, 1908), which were themselves based on the Imperial census of 1905. The voting data was compiled from the following newspapers: Hannoversche Courier, Neuhaus-Ostener Zeitung, Nordhannoversche Landeszeitung, Provinzial-Zeitung, and the Stader Tageblatt.

The number at the top left corner (#1 - #91) indicates the alphabetic ordering of the polling places in the 19th constituency. Polling places that were newly created in 1912 are numbered alphabetically beginning with #92. Exceptions: I have not listed the individual polling places in Lehe and Geestemünde, as they expanded over the period with little coherence from one election to the other.

Polling places that were divided will list the subsidiary polling places under "other places" in italic print. The statistics for such polling places, will be listed in plain print for the total district, and in italic print for the 1912 election when the subsidiary polling places are established. Some villages are listed separately in the 1905 census, but are never given their own polling place. These places are listed under "other places" in plain print. The statistics from such polling places reflect total figures drawn from the 1905 census. The figures for "land tax" in such places is arrived at by separately multiplying area times land tax for each area, then adding them together as tax/area. The tax/area figure is then divided by the combined area figure to arrive at a notional estimated land tax for the entire polling place. This is less problematic than it seems, since the "land tax" figures were themselves only official estimates.

There are several exceptions to this general polling place dictum. #1. The polling place Mittelstenahe-Nordahn-Varrel was at Nordahn in 1893, but in Mittelstenahe in 1898-1907. Nordahn became autonomous in 1912, while Varrel remained under Mittelstenahe. For simplicity' sake, I have recorded the polling place's 1893 as Mittelstenahe. #2. In the election of 1893, Voigtding residence voted in Cadenberge, but voted in their own town thereafter. I have kept Voigtding as its own unit for statistical purposes and polling place numbering. In all maps concerning 1893, therefore, Voigtding and Cadenberge will be shown together under Cadenberge's number, with no constituency identified separately as Voigtding. #3. In the elections of 1893, 1898 and 1907, residents in the small villages of Lanhausen and Welle voted in the larger polling place of Nesse. In 1903 and 1912, they voted in Lanhausen and Welle respectively. I did not separate out their votes in 1903, which would have led to a unique 1903 map, but instead figured them into the Nesse total. For 1912, I made them independent, since they were part of a whole series of new polling places. I decided to call the polling place Lanhausen in that year rather than Welle to minimize the chance of confusion with the permanent Geestemünde polling place (in the 18th Wahlkreis) of Wellen.

Several polling places proved especially problematic. #1. The Neuland/Neuenfelde/Neuenwalde problem. #2. The Drochtersen-Dornbusch problem.

The ranking of polling places is given twice. That in plain type reflects the 91 polling place model, the italicized figures represent the 1912 election.

Calculation. I have rounded up all values of 0.5 or greater, rounded down 0.4 or less. In calculating the percentage value parties received in elections, I have calculated all save the lowest vote-winning party on the round-up basis. The lowest vote-winning party is then assigned a whole number to make 100%. In some cases, this will result in slight inequities and variation from officially reported percentage results.

Calculating voter participation. I have used the figures from the 1905 census to calculate the notional number of eligible voters for each polling place. The number of notional eligible voters was derived by multiplying the number of inhabitants by .26, the average percentage of the constituency's population eligible to vote in each election. When that figure was ascertained, the total number of actual votes cast was divided by the notional number of eligible voters to arrive at an estimated ("notional") percentage of participation for each polling place. Because the population of the constituency's 4 predominately rural counties remained static during the period under study, I was able to arrive at a notional participation rate for those polling places for each election. For the growing counties of Lehe and Geestemünde, I have restricted this notional calculation to the elections of 1903 and 1907.

Calculating 100 index for the number of votes that Hahn received in each polling place at each election. To calculate each 100 number, the actual number of votes cast in a polling place was divided by the average number of votes cast and the average percentage of votes cast.

19.jpg

22.tiff

23.gif

25.gif