German census taking before 1871, R Gehrmann

Tags: population, Prussia, lists, census, censuses, Napoleonic era, German states, registers, population registers, Historical Demography, improvements, GCU, population statistics, Statistisches Reichsamt, enumeration, counts, census material, Heimatverein Menslage, Deutschen Staaten, Hamburg Bremen, Carl Friedrich Wilhelm, Berlin, population counts, demographic data, German Customs Union, house to house, working paper, Max Planck Institute for Demographic Research Laboratory, Rolf Gehrmann, Max Planck Institute, census methods, statistics, Monatshefte zur Statistik des Deutschen Reichs, administrative areas, working papers, census results, Menslager Hefte
Content: MOSAIC Working Paper WP2012-001 SEPTEMBER 2012
Max Planck Institute for Demographic Research Konrad-Zuse-Str. 1, 18057 Rostock - Germany +49 381 2081-190 (secretary's office)
German census-taking before 1871 Rolf Gehrmann
This working paper has been approved for release by: Mikolaj Szoltysek ([email protected]), Deputy Head of the Laboratory of Historical Demography. © Copyright is held by the authors. mosaic Working Papers receive only limited review. Views or opinions expressed in working papers are attributable to the authors and do not necessarily reflect those of the Institute.
Max Planck Institute for Demographic Research Laboratory of Historical Demography Working Paper German Census-Taking Before 1871 Rolf Gehrmann
Contents Preliminary remarks ............................................................................................3 Introduction ........................................................................................................4 1. Old administration statistics ............................................................................5 2. Administration statistics in the Napoleonic era...............................................7 3. Restoration, continuity, and new impulses......................................................8 4. The breakthrough of modern census methods ..............................................12 5. Epilog ...........................................................................................................14 Conclusions .....................................................................................................15 Commented bibliography ..................................................................................16 Appendix: Examples of census lists..................................................................18 2
Preliminary remarks This paper is a condensed version of various reports on German states that were prepared for the Laboratory of Historical Demography at the Max Planck Institute for Demographic Research in 2008/09. The purpose of these overviews (in German) was to find out how exactly population was counted, where these counts were documented, and, whenever possible, to locate the original census lists. Future publications will provide inventories of the scattered census household listings that survive around Germany. The purpose of this paper is to provide a general overview of the evolution of the German census system in the 19th century, to the extent that it is possible to do so. There are some smaller states for which further enquiries about sources are necessary, a task which is beyond the scope of this research project. But I am confident that new details will not seriously change the general conclusions presented here. I would like to express my gratitude to Barbara Zuber Goldstein, Joshua R. Goldstein, and Rembrandt Scholz for their cooperation. Berlin, June 2009 R. Gehrmann 3
Introduction A continuous documentation of the whole German population from censuses does not exist before 1867 (Volkszahl 1879, 1). That does not mean that censuses and census lists did not exist before that time. But the Statistisches Reichsamt had to fall back on several estimations and interpolations in order to get a complete time series in three-year periods from 1816 onwards (Volkszahl 1879, 54). The original data were not corrected, even though it was known that they were not fully reliable before 1834. While this provisional reconstruction has the status of an official one, and is widely used in research, it could be improved through a more critical evaluation of the censuses on which it is based. The history of these censuses has not yet been written. Beukemann (1911) only considers practices since 1867, about which he provides valuable information. For the German Confederation (Deutscher Bund) and its member states, the forewords of the publications on census results, which also contain remarks on former censuses, are still the most important sources. The annotations in the synthetic publication of the Reichsamt (1879) are based on them, and Kraus (1980) and Michel (1985) have drawn most of their information from the compilation of the Reichsamt. The first two publications focus on a presentation of the general results and discuss the question of who was counted, but they neglect the problem of how the counting was conducted. Only the primary sources can help to resolve this problem and to address the issue of whether or not the census was based on a complete listing of household members. While states with statistical offices issued volumes on censuses with prefaces that often explain former procedures, smaller states lack descriptions in a printed form, so that a comprehensive history can only be written with the help of archival sources. A broader methodological discussion of the procedures can be found in Engelґs description of the Prussian censuses and his proposals for improving them (Engel 1861). The general history of statistics by John (1884), otherwise still a reference, is not helpful here since it only looks at the evolution up to 1835. Meitzen (1903) only gives some census years, but does not indicate sources and does not discuss them. By using the word "census" (Volkszдhlung) for all kinds of population counts, he does nothing to clarify the situation. Nevertheless, his "census" references are often cited, particularly when the question of which is the oldest is raised (Bulst/Hoock 1981). So, what can be called a "census," and what cannot? Unlike the German "Volkszдhlung", the English term "census" is precise ­ when it is employed correctly. It is reserved for a "written enumeration of all individuals inhabiting a well-defined area" (Willigan/Lynch 1981, 79). All other kinds of procedures which are designed to yield a population total by using lists are usually called "enumerations." When there is no proof that there were lists involved, the term "count" is the most appropriate. The history of German "Volkszдhlungen" before 1871 is made up of enumerations and censuses, and, in earlier years, perhaps even of counts. In order to distinguish the cen- 4
sus and enumeration methods by their form and by their results, the following classification will be used here: Distinction by the product: 1. Anonymous table (if nothing else was noted, then the underlying procedure can be called a count), 2. Nominal listing of the household heads with only numeric information about the members of the household (the underlying procedure can be considered to be an enumeration), and 3. Census list with individual names and data for all persons. Distinction by the procedure of the collection of information: a) Extract of a population register (Registerzдhlung), b) Transcript of the statements of convened household heads (Protokollzдhlung), and c) Collection of the data from house to house (Naturalzдhlung). For example, the American censuses until 1840 were listings of category 2, while France moved to the category 3 in 1836, and England moved to category 3 in 1841. The Belgian census of 1846 set standards for 3c, while Sweden continued to produce statistics based on registers (category a) until 1860. The main question will be when censuses of the Third Order appeared in the different German states. Apparently, every census of this order before 1840 must be considered exceptional. Such exceptions exist, but they are rare. The Canadian census of 1666 and the Icelandic census of 1703 are the oldest examples of that kind. In the 19th century, the censuses of the type 3c were the most reliable ones. They were preceded by other forms that often had their roots in the 18th century. Because they influenced later evolution, censuses of this type should be described first. An exhaustive presentation of them is, however, not intended here. 1. Old administration statistics Population counts conducted for specific purposes of the state by civil servants or by clergymen were not new in modern history, but they became more frequent and more detailed in the 17th century, and especially after the Thirty Years' War. This development, which was connected to the evolution of the early modern state, will not be outlined in detail here. But because some of the old procedures were maintained until the Napoleonic era and even beyond, they are worth mentioning. What kind of data were administrators interested in collecting, and what type of product ­ a count, an enumeration, or a census ­ could, therefore, be expected? a) In the centuries following the Reformation, church authorities were interested in counting their sheep, and in knowing who went to school and who attended Holy Communion. Like the "Husfцrhцrslдngder" in Sweden, in some German regions "Seelenregister" developed into complete registers of the whole population. In Thuringia, they were introduced in the second half of the 17th century. In the 19th century, these registers were still in use, and they later served as the basis for censuses in some small states. They contain all the information of a census, and they can be regarded as equiva- 5
lents, provided the date of the creation of the lists is clearly defined, and the whole population is registered. This was the case for the Status animarum of Mьnster, in which the whole population of more than 200,000 people was registered1. b) The chief aim of the fiscal administration was to establish the number of taxpayers. Poll tax lists are of little interest per se, but in some regions they turned into registers from which the "census" data were drawn. They not only distinguished between households and adults; they registered practically the whole population. This was the case in Prussia in the 19th century ("Klassensteuerlisten"). Underregistration was more or less systematic in these lists, and they were continuous registers without a defined qualifying date. Unlike the Seelenregister, they cannot be considered to be a kind of census. Moreover, these poll tax lists have not survived. Occasionally, the levying of special war taxes could have served as an excuse for counting the whole population, as was the case in Osnabruck on four occasions between 1793 and 1801. c) The growing interest of the early modern state, and especially of enlightened absolutist rulers in demographic measures as measures of the wealth of the state, led to a total count of the population. As these counts were considered part of the yearly balance sheet of the state, local officers were asked to furnish yearly numbers. The principal features of these counts were not demographic, but economic. Adults and children were distinguished, but more attention was paid to the distinction between different kinds of land property and the number of tradesmen. In Prussia, the figures furnished by the so-called "Historische Tabellen" were in most regions fairly reliable, and they were sometimes obtained from individual census lists compiled by local authorities. These original lists or notes have not been filed. The principle of yearly counting was maintained for some time after 1815. d) Military needs led to the distinction of the age group of young men. Normally, one does not expect such counts to include the whole population. But this was sometimes the case. The figures that were referred to as the sum of the "Mannschaft" were indeed the population totals in the Gottorf part of Holstein in the 1770s, for example. The Bavarian government even ordered an enumeration of the whole population in 1803 (Kleindinst 1914, 10) in order to hide its real intentions, so that a military count was included in a general one. The advantage of these kinds of administrative statistics is that they include data on age groups, and thus can furnish demographic information. e) A certain distinction by age groups was also made in the Saxon consumersґ lists ("Konsumentenverzeichnisse") after 1772. The crisis of the same year also provided an opportunity to the government of Osnabruck to 1 This source has been neglected for a long time. Publication has begun only recently ( In Russia and in Austria the "census data" were obtained by periodical revisions of such lists of people or subjects. 6
organize an enumeration (category 2b or 2c); however, except in Saxony, it was not renewed in the following years. For the 18th century no census list from category 3 is currently known. The best chance of finding them is in the domain of the Seelenregister. The desire to know the exact number of people existed, but the limited number of variables the administration wanted to know did not make it necessary to produce state-wide census lists. However, some scientists, like Oeder in the Danish kingdom, proposed comprehensive population registers. Administrators in territories under French rule or influence tended to go in this direction after 1800. 2. Administration statistics in the Napoleonic era In the first decade of the 19th century, a "recensement" was often no closer to the definition of census than the term "Volkszдhlung". In the dйpartements which were directly under French rule, the first lists only named the head of the household, and thus must be called enumerations of individuals over age 12 (1799). But in the following years, lists became more detailed, and eventually censuses of type 3 were taken2. From west of the Rhine come the first preserved census lists in Germany (1801) 3. They were contemporary with the Danish (1801) and Schleswig-Holstein censuses (1803). Details and circumstances of some of the recensements varied, and the prescriptions were not very clear. For example, the prefect of the dйpartement of the Ems announced as the result of a recensement in 1807 figures that were simply copied from the Prussian tables of 1802 and 18044. For Recklinghausen, a brief description exists that tells us how numbers were obtained by a procedure that we can call "category b," that is, by questioning the household heads in the local police office5. In principle, counting had to be done annually. Thus, in 1810-12, serious efforts were made to establish population registers, as in Ems supйrieur 18116 and in Lьbeck 18127. In the Kingdom of Westphalia, the population registers of 1809 were first conceived as household registers (category 2), as is shown by the example of Witzenhausen (Eckhardt 1968). But it seems that, for 1811 and 1812, some real census lists exist. Did the other states of Germany follow this example? In general, this was not the case. In most of the states, as in Saxony or in Baden, population counts were performed using old methods. In Hessen 2 The history of the censuses of the Napoleonic era cannot be studied in detail here. More research has to be done. In the archives of the Rhineland a certain number of lists has been preserved, and some of them are published. A first census list can found for 1801 in Rhin-et-Moselle. Cf. Dupвquier 1977, 115. 3 Lists in Rheinland-Pfдlzisches Landeshauptarchiv Koblenz, Best. 261, 612 and 655.132. 4 The examples from Northern Germany before 1840 are more detailed in my Bevцlk- erungsgeschichte Norddeutschlands (2000). 5 Similar procedures were admonished by the prefect of the dйpartement Rhin-et-Moselle in 1810 (LHAK, Best. 714, Nr. 8513). 6 Original lists of some villages in StAOs, Rep. 240, Nr. 124. 7 Partially preserved in StAHL, Senatsarchiv, Interna, Volkszдhlungen 1 und 2. 7
(GroЯherzogtum), progress was made in 1804 as instructions were given to the local priests on how to perform the enumerations, and a form for enumerations was provided (category 1). In Sachsen-Meiningen, numbers were extracted from the Seelenregister that had been renewed in 1791. Some of these registers still exist (Moritz 1985). Such registers, called "Familienregister", were introduced in Wurttemberg in 1807. They served as sources for the annual counts of the priests (1a). For Bavaria, regular collection of demographic data was an important innovation of the Napoleonic era, and in 1809/10 and 1811/12, the so-called Montgelas statistics brought together the reports of all the regional officers (Landkommissare) for the whole kingdom. Governments needed more information about population than before, but the sources of the given figures are difficult to verify, in part because conservation of the files from these years is in generally bad8. This particular problem cannot be addressed here. In Prussia, the Historische Tabellen were abandoned in the reform period, and were replaced by a collection of demographic data with age groups (0­6, 7­15, 16­44, 45­59, over 60). Locally, a real census was performed (Dieterici 1844, 190), but, as is nearly always the case in Prussia, no lists have been preserved. The precision of the data was obviously greater than in the years after 1815. 3. Restoration, continuity, and new impulses Compared to the Napoleonic era, there was a certain lack of interest in the collection of exact demographic data in the 10 or 15 years following the restoration of the old regimes. The SьЯmilch impetus that had motivated the collection of Vital Statistics had disappeared, and censuses were not regarded as a necessity, either for the computation of life expectancy ­ which was based on the assumption of a stationary population ­ or for the creation of population registers that could have been used for different purposes. Even in Denmark-Schleswig-Holstein, no new census was performed until 1835. Statistical efforts were made in the direction of general descriptions of countries and regions, including their geographical features, in which population was no more than a figure. In order to obtain this number, the old administrative procedures seemed to suffice. Budgetary restrictions did the rest: even the rare statistical offices counted no more than a handful of persons ­ in contrast to the Napoleonic state of Wьrzburg, for example. Where annual reports on the number of people existed--as in Prussia, Bavaria, Hessen-Darmstadt, Baden or Wьrttemberg--a drive to reduce administrative efforts led to the establishment of intervals of three years or more. Intervals were introduced in Prussia and Hessen-Darmstadt in 1822 (three years), in Wьrttemberg in 1823 (10 years), in Bavaria in 1825 (three years), in Braunschweig in 1831 (three years), and Baden in 1833 (three years). Some small states--like Nassau (until 1866), Waldeck (until 1837), Gotha (until 1874), Weimar (until 1844), and other Thuringian states-- 8 In Hesse, nearly all the files of the former central administration of the Kingdom of Westphalia were destroyed at the beginning of the Restoration. 8
continued with annual counts, even after joining the German Customs Union. Mecklenburg-Schwerin proceeded likewise with its annual church statistics ("Martinilisten") until it was obliged to take a census in 1867. The impetus of the Bundesmatrikel The establishment of the "Bundesmatrikel" in 1818 ­ quota of participation on the activities of the German Confederation­ only rarely inspired governments to improve their population statistics. But it made it necessary to determine at least the total number of inhabitants in states where population had not recently been counted. Thus, the years 1818 and 1819 marked an important date in the demographic history of otherwise poorly documented states ­ with the exception of Hanover, where data of the Napoleonic era were complemented by counts in the new provinces. The three princes of Anhalt ordered an enumeration (category 2) that was conducted with care in May 1818. Subsequently, Anhalt-Kцthen updated the results until 1846 without counting the people again. For the city of Lьbeck, where French authorities had brought remarkable progress previously, the count done for the Confederation in 1815 was to be the last one for 40 years. In Waldeck and Pyrmont, the same type of enumeration as in Anhalt was performed in March 1818, but in Waldeck the tradition of annual counts persisted for another 20 years. The census of Frankfurt of 1817 was in conjunction with the military contribution to the confederation, since it had to serve as a platform for the organization of the Landsturm. The listings have not been preserved, and only one other count was done before 1834. In two other states, the impetus of the "Bundesmatrikel" also led to procedures that can certainly be classified as fully developed censuses (category 3), though they were sometimes performed over one month, instead of on a single day. They took place in 1819 in Mecklenburg-Schwerin and Fьrstentum Lьbeck. For Mecklenburg, the census would remain the only one conducted until 1867, while the principality of Lьbeck continued with censuses (3c) in 1840 after the accession to the Holstein customs system. The motivation for the census of the 15th of February 1819 in the principality of Lьbeck stemmed from the need to know the number of young men aged 18 to 24 who could be recruited for the confederation army (Bundeskontingent). It is probable that the lack of a count of the population by state authorities and military necessities also motivated officials in Mecklenburg to conduct a census in the same year, even though the annual Martinilisten established by the church had already been transformed into a count of the whole population with the inclusion of children under age five in 18189. The preserved listings of the two 1819 cen- 9 The German Confederation certainly could not "order" (Manke 1999, 650) such a census, but state authorities may have remarked that counting the population could be useful "zu anderen statistischen Zwecken" (Verordnung vom 18.6.1819, cit. by Manke). But there was no incentive to take steps to prevent underregistration, since contributions to the charges of the Confederation would increase with population. That is perhaps the reason why Mecklenburg did not communicate the results of the census to Frankfurt, and preferred to be classified by an older and thus lower number. 9
suses are the only state-wide sources of this kind in Germany outside of Schleswig-Holstein in the first half of the 19th century. 10
Changes introduced by the German Customs Union Though the Statistisches Reichsamt was rather confident of the accuracy of the figures obtained by the population counts established for the GCU in 1834, some doubts remain10. There was no institutional control of the counts, and the sole protection against underregistration was a desire for accuracy. It was of no benefit to the states to underestimate their population, since the redistribution of the custom duties depended on it. The potential gain in income for the treasury was an argument employed by Engel when he defended his new and more expansive census methods in Saxony. But Engel was an exceptional personality, and he already belonged to the following period. In 1834, the states that founded the GCU (Prussia, Hessen-Darmstadt, Hessen-Kassel, Bavaria, and Wьrttemberg) agreed to let officials simultaneously count the population of the member states every three years in December, as Prussia had done previously. Until 1843, the procedure could last four weeks, then the period was three days; it was not until 1858 that one single day (the third of December) for the count was designated. The aim was to determine the number of people who were permanently present, and were thus consuming ("Zollabrechnungsbevцlkerung"). That was not precisely the same as the effective population, since persons who were present for a short period of time or were traveling ("guests" in 1846) were not counted. It also was not the resident population, since residents absent for a longer period of time were not counted either. Rather, the main concept was closer to the second than to the first, which later determined German censuses up to 1925, and created a certain degree of overregistration for other reasons11. All persons, including foreigners, were counted as inhabitants. Military forces were counted separately, but soldiers who did not serve most of the year (Beurlaubte) were classified as civilians in 1846. The amount of information that the GCU asked for was not large enough to create an incentive to improve methods of census-taking. Only two variables which together produced four features had to be collected. Even an enumeration was not prescribed, so that simple counts of the number of Girls and boys under 14, of male and female adults, and of the number of families were possible. At the same time, however, there was an incentive to avoid bad results. So enumeration often seemed the best way to reconcile the de- 10 The argument of the Reichsamt focuses on the date of censuses, which is not the only and not the most important argument, and on the assumption that Frenchmen were still behind in their census techniques: ,,dass sie sich an Sicherheit der Ergebnisse etwa mit den heutigen franzцsischen messen kцnnen, wo das Prinzip der Ortsanwesenheit nicht festgehalten, ein bestimmter Aufnahmetag nicht vorgeschrieben, vielmehr den Maires mehrere Wochen (1876: 2 Monate) Zeit gelassen ist, um das Zдhlungsgeschдft ins Werk zu setzen" (Volkszahl 1879, 2). 11 Beukemann (1911) critized this principle, which diverged from most of the international censuses, since intuitively temporarily absent people were counted as being at home, and thus potentially twice. If he was right, the overregistration was at that time about 400,000 people, or more that 0.6 %. It was also criticized for its impact on the representation of household structure. 11
sire for reliable information and the wish to avoid the effort of collecting individual information on each member of the household (see the first figure in the appendix). The states were free to determine the procedure and to collect additional data, which sometimes was done. But, in general, the minimalist prescription of the GCU led to minimalist censuses, even where states had been steadily making improvements previously. The main advantage of the GCU "censuses" consisted of a homogenization of dates and criteria, and in the obligation ­ which was not always respected, as we have seen for example in Anhalt-Kцthen ­ to do regular counting. The shortcomings of the system did not go unnoticed, but the members of the GCU had to be persuaded to alter their procedures, and no one could be forced to make changes. In order to improve the accuracy of the data obtained, simple extracts of registers were unanimously criticized in 1843, and a real count of all individuals from house to house was prescribed (Volkszдhlung 1879, 8). That did not mean a census as it is strictly defined, but an evolution in that direction was implicitly encouraged by this requirement. Big states like Prussia and Saxony went ahead with improving their counting procedures. Censuses as remedies against inaccuracy In Prussia, the annual population statistics were combined with the poll tax registers in 1820. From 1822 onwards, the collection of the data was spaced out and only had to be done every three years, a rhythm that was later adopted by the German Customs Union. Until 1834, the tables asked for specific information for each village, including the number of buildings, the number of inhabitants by age (under age 14 years, from age 14 to 60, and over age 60) and by religion, the number of persons eligible for military service ­ and the number of cattle. The age groups became a bit more differentiated in 1837 (0-4, 5-6, 7-13, 14-15, 16-44, 45-59, over 60), but only the age groups of interest for military service became really detailed (age limits 20, 25, 32, and 39 for the male population; in 1858, the same categories were also applied to women, probably for reasons of symmetry). In 1858, the feature of nationality was added. We are far from having tables that would provide an age distribution that could be used to illustrate life expectancy, for example! All these operations were based on counts in police or tax registers, and officials soon became doubtful about whether the data were complete (Dieterici 1844, 190). This was probably the main reason why census lists were introduced in 1840. These lists showed that underregistration had occurred in the previous enumerations, especially in eastern provinces. The problems concerning Berlin were not resolved until 1846. In 1840, the census lists were probably still extracted from existing registers of inhabitants (category 3a), but in 1843 a real counting operation was prescribed. In the following census years, this was controlled more strictly: in 1846 the use of inhabitant registers was formally forbidden, and counting from house to house (3c) was unambiguously prescribed (Engel 1861, 24ff.). In 1852 it was discovered that in Minden household heads still were called to assemble at the 12
local office to provide information (3b) and this procedure was admonished12. The desire to get more precise figures that led PRussian authorities to establish real censuses did not prevail everywhere in Germany. Generally, the impetus for change came from the outside, as local statisticians rarely had enough influence or were not sufficiently aware of possible improvements to push for change. Saxony is, however, another positive example. Saxony was in a similar situation as Prussia. The "Konsumentenkonsignationen" had eventually led to an underregistration of about 10% of the population. As the Restoration period came to an end in 1830, a statistical association was founded, and the population statistics were reformed. For the census of 1832, house lists were prescribed, but obviously they were not compulsory. It was not until 20 years later that household lists were introduced, and thereafter became obligatory. Though there was a relatively long period of improvements, the main step had already been taken in 1832, earlier than in Prussia. Unfortunately, centralization of the statistical works in Dresden led to a centralization, and then to the destruction of all materials. 4. The breakthrough of modern census methods The famous Belgian census, which became widely known and was favorably commented on by German statisticians, was taken in 1846. The creation of the German Statistical Association (Verein fьr deutsche Statistik) in 1846 by von Reden, the revolution of 1848/49, and the creation of an ephemeral Central Statistical Office (Statistisches Zentralbureau) in Frankfurt were expressions of a strong tendency in liberal civil servant circles to see statistics as a means of improving knowledge of the home country--that is, the whole of Germany--and thus to contribute to unification. Improvements and greater homogenization were the slogans in population statistics, as well as in other domains. In order to prepare a general census, as mandated by the National Assembly (11/6/1848), the Undersecretary of State in the Department of Commerce, Fallati, sent a circular (11/14/1848) to all German governments asking about the real state of population statistics. Not all governments answered, but it became more and more evident that old habits could not continue. Moreover, 1846 was the year in which real censuses became reality in many states. In theory, the decision was already made in 1843, but it took some time before appropriate methods were really used everywhere, and the lack of original sources makes it rather difficult to determine an exact date for each state. It seems that, in 1843, proper censuses were rarely conducted outside of Prussia. At the GCU conference in 1843, the proposal of the Bavarian representative that the real intentions behind the decisions of 1834 in 12 In the Regierungsbezirk Koblenz, the government wrote to the district administrator that the census of 1843 would be easier to realize since the lists from 1840 could be updated ("es daher gegenwдrtig nur darauf ankommt, das neu aufzunehmende VerzeichniЯ bei der Zдhlung nach den seitdem vorgekommenen Verдnderungen zu berichtigen") (LHAK, Best. 655.069, Nr. 320) 13
specifying that censuses had to be done from house to house13 be "clarified" was accepted unanimously, and all participants offered assurances that they were already conducting proper censuses, or that they would start doing so from that point onwards. Following the decision to do real counts from house to house, German census statistics became reliable. So 1846 must be established as the year of reference, and not 1834, as the Reichsamt later suggested ­ contrary to the statements made by Boeckh, who pointed out the problem of underregistration before the 1840s. The last step was done in the period of unification. With the census of 1864, the system of household forms that had been tested by Engel in Prussia in 1861 became compulsory for all the member states of the GCU. By this measure, Germany applied the decision of the International Statistical Congress in London (1860) to conduct a census at least every 10 years, using household lists filled out by household heads and checked by civil servants to count the effective population, and to use the household (and not the family) as the basic unit. (Engel 1861, 16). In the meantime, most of the states had introduced lists with individual designations for each member of the household that were established by periodical real counts (category 3c): Holstein (Tabellen 1836), Fьrstentum Lьbeck (Statistisches Bureau 1857, 24), Prussia (Engel 1861), Anhalt14, Birkenfeld (Statistisches Bureau 1857, 28), Baden, but without asking for age (Schaab 1971, 177), Coburg15, Homburg16, Saxony17, Hanover18, Oldenburg (Statistisches Bureau 1857, 5), Weimar-Eisenach (Beitrдge Sachsen-Weimar-Eisenach 1:II), Gotha19, Wьrttemberg20 and Hessen-Darmstadt21. Bavaria22, Nassau,23 and the rest of the Thuringian states then followed (see table). Thus, the first common census of 1864 marked the end of an evolution that had begun 18 years before in the greatest part of Germany ­ except for the states that did not belong to the GCU at that time. That is why Mecklenburg-Schwerin did not conduct its first regular census until 1867, after a 13 "Die Volkszдhlungen .... grьnden sich auf eine, von Haus zu Haus vorzunehmende Aufzeichnung aller einzelnen Einwohner mit Namen, Stand oder Beschдftigung, Geschlecht und Alter." (Verhandlungen 1843, 122). The Bavarian representative in particular pointed out the difficulties in towns (cf. note 21), and he said clearly that the main reason for introducing census lists was that experience had shown that they were the best means of assuring accuracy. 14 Proved for Anhalt-Bernburg and probable for Kцthen (lists found for 1849) and Dessau. Landesarchiv Sachsen-Anhalt, Dessau. 15 Coburg archives. 16 Landeshauptarchiv Koblenz, Best. 386 for Meisenheim. 17 Recommended in 1832, generalized in 1852 (Engel 1855, 149). 18 Statistik Hannover 1855, see also: 19 Statistik Thьringen 1:94. Since 1829 the soul registers were established in the form of census lists. 20 Recommended in 1846, generalized in 1861 (Schaab 1971, 174 / Rьmelin 1863, 307). 21 Volkszahl 1879, 14. Curiosly, "Volkslisten" already existed in 1852 and in 1855 in the village of Rossdorf (Staatsarchiv Marburg, Best. 331). 22 In towns, census lists were introduced in 1846. 23 Census lists from 1861 still named only the household head (HStA Wiesbaden, Best. 240). For Hessen-Kassel there is no evidence at all, so one must conclude that there was no census before 1864. 14
break of 48 years following the Bundesmatrikel census. For Lьbeck, Bremen, Hamburg, and Mecklenburg-Strelitz, 1867 was the first (category 3) census year ever. The 1867 census has to be considered the first in the history of the Reich. Not only did all states participate, but the population counted was also no longer the "Zollabrechnungsbevцlkerung" as defined by the GCU, but the effective ("faktische" or "ortsanwesende") population, which was a little bit lower (-0.3% for Prussia, -0.23% in total)24. For Prussia, census lists were established at same time as the last census. (Table) Regular census-taking with household forms (type 3c)
1835 1840 1843 1846 1849 1852 1855 1858 1861 1864 1867
Ho ls tein Fst. Lьbeck Prus s ia Anhalt Birkenfeld Baden Coburg Homburg Saxony Hanover Oldenburg Weimar Gotha Wьrttemberg H.-Darms tad t Bavaria Nas s au Rest of Thuringia H.-Kas s el Mecklenburg-Schw. Mecklenburg-Str. Lьbeck Hamburg Bremen
(1835, 1840, 1845, 1855, 1860) (1840, 1845, 1850) (recommended) (recommended) (in towns)
5. Epilog The year 1871 marks the beginning of the series of censuses that take place first in four-year periods (1867, 1871, 1875) and then every five years, until this pattern was interrupted by war and crises. The member states of the Kaiserreich had to furnish information on certain variables to the Statistisches Reichsamt, but were free to collect more information. The sole condition was that the census had to be conducted from house to house, and using household or individual sheets. The core variables in 1871 were: 24 That means that some temporarily absent persons were counted twice ­ a problem also found in the censuses of the Reich up to 1925. 15
- name (Name) - position in the household (Stellung im Haushalt) - sex (Geschlecht) - place of birth (Geburtsort) - year of birth (Geburtsjahr) - marital status (Familienstand) - religion (Religionsbekenntnis) - profession (Beruf oder Erwerbszweig der ьber 14 Jahre alten Personen (Haupt- und Nebenbeschдftigungen) - national status (Staatsangehцrigkeit) - place of residence of temporary guests (Wohnort bei nicht zum Haus- halt gehцrenden Personen). Absent persons were to be named and their positions in the household were to be indicated. In most of the states, the exact date of birth was noted. In many states, physical and mental deficits were noted, and in Prussia, the literacy of persons over age 10 was recorded. This was omitted in 1875, and information about handicaps was still collected in only a few states. The Reichsamt no longer asked for the place of birth, but this information was collected in some small and medium-size states. Like in Saxony in 1852 and in Bavaria, in 1867 all the Prussian household lists were collected and analyzed centrally in the capital ("das gesamte Aufnahmematerial an Haushaltslisten in Berlin sammelte und bearbeitete" (Beukemann 1911, 225)). In the following census (1871), no more household sheets were used in Prussia. They were replaced by individual sheets (Zдhlkarten), which were not preserved. That is why it is impossible to find original census lists after 1864 ­ only in exceptional cases were copies or compilations from the individual forms made for local purposes. Original census files for Prussia can cover a maximum period of 24 years. But the registers of inhabitants ("Bьrgerrollen" or "Hausbьcher;" later "Melderegister") are sometimes organized like household forms, so that they can be used like census lists for demographic research25. Not all states went over to counting by individual sheets. In 1871, only Braunschweig and Hamburg followed Prussia; while in 1875, Hessen, the two Mecklenburgs, Oldenburg, Schaumburg-Lippe, Lippe, Bremen and Elsass-Lothringen also made the shift. Some states used individual sheets just for a period of time, and then returned to household forms. But the centralization made it impossible to preserve the originals. As Wietog has shown, census material was only kept until the following census and old paper was recycled (Wietog 2001, 57). The statistical office of Wьrttemberg made an exception. In July 1933, it gave all the sheets of the census of 1925 to the archives of Ludwigsburg because they could be of some value for future family histories (Wietog 2001, 58). Conclusions 25 For Hцhscheid, for example (Stadtarchiv Solingen). 16
The German census system developed rapidly in the middle of the 19th century, and it attained high quality standards in international comparisons. But the political structure of Germany made it nearly inevitable that each improvement created a gap between forerunners and laggards. So it was not until 1867 that the same standards were adopted everywhere. Historical demography can cope with this. Two other limitations of possible studies based on census lists are more restrictive ­ the limitations in duration and the disposable sources. Census lists cover a relatively short period, and, outside of some Regional Clusters, they are relatively rare in archives. Mainstream census material documents the period from 1843/46 to 1864/67, which is not even the length of one generation. Two remarkable exceptions exist nevertheless: Holstein (1803 to 1864) and MecklenburgSchwerin (1819 to 1867). Together with Fьrstentum Lьbeck (1819) and Hanover (1852), most of the big clusters are from the north, which was not the most dynamic part of Germany. The rest of Germany is also of interest. In the period from 1846 to 1867, the speed of change accelerated. Crisis, revolution, emigration, and beginning industrialization affected the demographic structure in ways that can be studied in using census material. In the form of case studies, it seems to be possible for some towns, particularly when the census returns were used as a basis for systematic population registers. In order to gain insight into the population structure of the past, censuses of the Napoleonic era can be taken into consideration. In some rare cases, census-like material on a relatively wide scale existed even for the 18th century. The exceptional Status animarum of Mьnster (1749) must be cited here. But such insights will be possible only for isolated parts of Germany. The second limitation is imposed by the loss of original census material, and the difficulties in gaining access to the lists that have been saved. In general, they are kept in local archives, and the central archives ­ and there is often more than one in each federal state ­ are not always informed of their existence. More research must be done in this field. We can, however, say that, for 1846 and for 1852, samples of tens of thousands of people can be drawn from outside the big northern clusters (Holstein and Hanover), and future discoveries will perhaps help to ascertain a better regional repartition. Then the crucial question will be whether the final outcome will have the quality of a random sample. When the question of coverage in space and in time is raised, it is useful to recall that there are other sources of information on age, household, and professional structure for any given point in time. Some of them are even more precise, since they also provide information on population movements. The Seelenregister of Thuringia, the family registers of Wьrttemberg, and other forms of population registers are sources for historical demographic research that are not of lesser value than censuses. Thus, further enlargements in the field of observations are possible, and in this sense German census results can be seen as the starting point for a wider examination of the population structure of Germany before 1871. 17
Commented bibliography Beitrдge zur Statistik des GroЯherzogthums Sachsen-Weimar-Eisenach, 1-2 (1864-1865). [Population of the country since 1816 and census methods] Beukemann, Wilhelm 1911. Methode und Umfang der deutschen Volkszдhlungen, in Zahn, F. (Hg.): Die Statistik in Deutschland nach ihrem heutigen Stand. Mьnchen, Berlin: Schweitzer. (1), 197­235. [On German censuses 1867-1910] Bevцlkerungsliste der Stadt Wesel mit Feldmark und der Bьrgermeisterei Obrighoven-Lackhausen 1858 2000. Wesel: Historische Vereinigung Wesel e.V. [Example for a census list published by local historians] Bulst, Neithard & Hoock, Jochen 1981. Volkszдhlungen in der Grafschaft Lippe, in Bulst, Neithard & et al (Hg.): Familie zwischen Tradition und Moderne: Studien zur Geschichte d. Familie in Deutschland u. Frankreich vom 16. bis zum 20. Jh. Gцttingen: Vandenhoeck & Ruprecht. [Example for an arbitrary use of the term "census"] Dieterici, Carl Friedrich Wilhelm 1844. Statistische Uebersicht der Stadt Berlin. Berliner Kalender auf das Schalt-Jahr, 159­260. [Early censuses in Berlin] Dupвquier, Jacques 1977. Quelques йtapes dans le dйveloppement des publications dйmographiques au XIXe siиcle, in INSEE (Hg.): Pour une histoire de la statistique. Paris: INSEE, 111­123. [Useful overview] Eckhardt, Karl A. 1968. Populationsregister der Stadt Witzenhausen von 1809. Witzenhausen: Selbstverlag. [Example for a population registers in the Kingdom of Westphalia] Engel, Ernst 1855. Die Volkszдhlung im Kцnigreiche Sachsen 1855. Zeitschrift des Statistischen Bureaus des Kцnigl. Sдchsischen Ministeriums des Innern 1, 148ff. [Census methods in Saxony] Engel, Ernst 1861. Die Methoden der Volkszдhlung, mit besonderer Berьcksichtigung der im preussischen Staate angewandten. Berlin: Oberhofbuchdruckerei. [Propositions to improve Prussian censuses and general remarks on census methods] Fallati 1850. Stand der administrativen Statistik in Deutschland im Jahre 184849. Zeitschrift fьr die gesamte Staatswissenschaft 6, 727­795. [Overview given by a member of the administration of the provisional German government in 1848/49] Hanssen, Georg 1849. Ьber die beabsichtigte allgemeine deutsche Volkszдhlung. Archiv der politischen Oekonomie und Polizeiwissenschaften NF, 8, 337­384. [Interesting remarks about the German censuses, presents Schleswig-Holstein and Belgium as models] John, Victor 1884. Geschichte der Statistik. Stuttgart: Enke. [General evolution of statistics before Quйtelet] Kleindinst, Josef 1914. Geschichte der neueren bayerischen Statistik. Mьnchen: Lindauer (Beitrдge zur Statistik Bayerns 86). [Gives an overview of the history of censuses in Bavaria and furnished some tables extracted from the Montgelas statistics, for example infant mortality in 1810/11] 18
Kraus, Antje 1980. Quellen zur Bevцlkerungsstatistik Deutschlands. Boppard. [Census results and methodological comments, drawn from Volkszahl 1879 and the publications of the statistical offices of the German states of the 19th century] Manke, Matthias 1999. Die Historische Bevцlkerungsstatistik in MecklenburgSchwerin bis zur ersten Hдlfte des 19. Jahrhunderts. Genealogie 48, 641­658. [About the Mecklenburg census of 1819] Meitzen, August 1903. Geschichte, Theorie und Technik der Statistik. Stuttgart, Berlin: Cotta. [Of little use for the history of censuses] Michel, Harald 1985/2. Volkszдhlungen in Deutschland. Die Erfassung des Bevцlkerungsstandes von 1816 bis 1933. Jahrbuch fьr Wirtschaftsgeschichte, 79­91. [Overview based essentially on Volkszahl 1879 and Kraus 1980] Moritz, Marina 1985. Die lдndliche Familie in der Periode des Ьbergangs vom Feudalismus zum Kapitalismus: Volkskundliche Studien zu den Familienverhдltnissen und zur Familienlebensweise werktдtiger Klassen und Schichten zweier Sьdthьringer Kirchspiele (Effelder und Mengersgereuth) vornehmlich in der Zeit zwischen 1789 und 1815. Diss. AdW der DDR. [Example for an utilization of Thuringian soul registers] Rьmelin, G. 1863. Das Volk, in Kgl. Statistisch-Topographisches Bureau (Hg.): Das Kцnigreich Wьrttemberg: Eine Beschreibung von Land, Volk und Staat. Stuttgart: Nitschke, 307­425. [Wьrttemberg censuses] Schaab, Meinhard 1971. Die Herausbildung einer Bevцlkerungsstatistik in Wьrttemberg und in Baden wдhrend der ersten Hдlfte des 19. Jahrhunderts. Zeitschrift fьr Wьrttembergische Landesgeschichte, 30, 164­200. [Censuses in Baden and Wьrttemberg] Statistik des Hamburgischen Staats, 2-4 1869-1872. Hamburg: Meissner. [Underregistration in the traditional ,,Bьrgermilitдrzдhlungen"] Zur Statististik des Kцnigreichs Hannover, H. 4 (1855). Hannover: Culemann [About the Hanoverian census 1852] Statistik Thьringens, 1 (1867). [Fundamental data and information about Thuringia 1834-1864]. Statistisches Bureau 1857. Stand der Bevцlkerung im Grossherzogthum Oldenburg nach der Zдhlung vom 3. Dezember 1855. Mit einer Einleitung, betreffend die Ergebnisse der Volkszдhlungen seit 1815. Oldenburg: Stalling. (Statistische Nachrichten ьber das Grossherzogthum Oldenburg, Bd. 2). [History of Oldenburg censuses] Tabellen ьber die, nach der Allerhцchsten Resolution vom 24sten Mai 1834, in den Herzogthьmern Schleswig und Holstein am 1sten Februar 1835 vorgenommene Volkszдhlung. Auf Veranstaltung der Rentekammer herausgegeben 1836. Kopenhagen. [First census results for SchleswigHolstein published officially] Verhandlungen der Sechsten General-Konferenz in ZollvereinsAngelegenheiten 1843. Berlin: Decker. [Agreement to establish census lists house by house] 19
Volkszahl 1879. Die Volkszahl der Deutschen Staaten nach den Zдhlungen seit 1816. Monatshefte zur Statistik des Deutschen Reichs 37(7), 1­84. [Population of all German states and greater administrative areas since 1816] Volkszдhlungen 1996. Menslage: Heimatverein Menslage. (Menslager Hefte). [Transcription of a Hanoverian census list from 1852] Wietog, Jutta 2001. Volkszдhlungen unter dem Nationalsozialismus: Eine Dokumentation zur Bevцlkerungsstatistik im Dritten Reich. Berlin: Duncker & Humblot. [Censuses in the Third Reich and preservation of census material] Willigan, J. D. & Lynch, Katherine A. 1982. Sources and Methods of Historical Demography. New York etc: Academic Press. [Basic definitions] 20
Appendix: examples of census lists Filled GCU-form (enumeration, category 2), Coburg 1840 21
Census list (category 3c) 1846: Hцhscheid (Prussia) 22
Census list 1852: Menslage (Hanover) 23
Copy of a census list 1858: Obrighofen (Prussia) 24

R Gehrmann

File: german-census-taking-before-1871.pdf
Author: R Gehrmann
Author: Rolf Gehrmann
Subject: German census-taking before 1871
Published: Fri Oct 26 13:42:46 2012
Pages: 25
File size: 0.67 Mb

, pages, 0 Mb

, pages, 0 Mb
Copyright © 2018