Studienbeschreibung:
Study Description (Purpose, Content):
The growth of the public sector in Germany since the creation of the “Reich” in 1871 is examined until 1958. The study of Suphan Andic and Jindrich Veverka follows the well known study of Alan T. Peacock und Jack Wiseman: The Growth of Public Expenditure in the United Kingdom (N.B.E.R., Princeton 1961). “Government expenditure is defined by public authorities of a non - financial character on services which are not sold against specific payments to the private sector” (Andic, S./Veverka, J., 1963/64: The Growth of Government Expenditure in Germany since Unification, in: Finanzarchiv, Neue Folge Vol. 23,1963/64, p. 228). The growth of government expenditure is discussed against a background of a changing population and of a growing output.
„The purpose of this paper is to measure the growth of the public sector in Germany from its period of unification until today [1958]. The problem of measuring secular changes in economic aggregates can be divided into three parts: firstly, the determination of the concept; secondly, the collection of the data; and thirdly, their interpretation. We have abstained from a general discussion of the concept of government expenditure as we have adopted that of other studies (A. T. Peacock and J. Wiseman, 1961) … The size and the structure of government expenditure is subsequently explained by reference to the ideological evolution and the nature of the political process, as well as to secular changes in the average size of the production and the consumption unit … The growth of the government expenditure will be discussed against a background of a changing population and of a growing output. Since territorial delimitation of Germany has been altered twice as a result of the upheavals of war, our estimates are not directly comparable between the various periods of different geographical size. It is not only government expenditure that was affected by these changes, but also our twp other time series” Andic, S./Veverka, J., 1963/64, p. 169 - 171). The conceptual problems discussed in the first part of the paper are mainly limited to those arising out of particular conditions specific to Germany. In this section Andic/Veverka consider in more detail four separate points which are of special importance in the German case, namely (1) the definition of government and the distinction between trading services and non-trading services; (2) the comparability of the estimates before and after 1913; (3) the effects of the territorial changes and per capita estimates; (4) the elimination of price changes and the productivity assumption. The subsequent parts of the paper deal with the interpretations of the statistical findings; aggregate expenditure is considered first and the changes in the compositions later. The empirical evidence is then discussed against Wagner’s “Law of Increasing Extent of State Activity” as an example of a general theory of government expenditure, and the size and structure of government expenditure is subsequently explaining by reference to the ideological evolution and the nature of the political process, as well as to the secular changes in the average size of the production and the consumption unit.
The statistics presented in the Appendix fall into four groups: demographic statistics, price index, estimates of Gross National Product, and finally estimates of government expenditure. The three former categories are of a subsidiary character, giving the framework for the discussion of the estimates of government expenditure on which our effort has been concentrated. The remaining tables give the estimates of “total government expenditure”. The aggregate expenditure is broken down according to three criteria, namely the “economic category”, the “function”, and the “level of government”.

Untersuchungsgebiet:
Place and Time Period Covered:
German Empire (Deutsches Reich), 1872 – 1938. German Federal Republic (Bundesrepublik Deutschland), 1950 – 1958.

As far as possible, all series cover the actual territory of German at each and any period. The old German “Reich” remained unchanged during its whole existence, and thus the estimates until 1913 refer throughout to the same geographical area. The inter-war estimates refer to the reduced territory that existed in 1925, allowed being made for the inclusion of the Saarland after 1935. After the Second World War, the statistics are limited to that part of Germany which constitutes the German Federal Republic. The exclusion of Berlin and the Saarland from the post-war estimates was dictated by the lack of statistical sources. Where possible, estimates for the same year have been given for a territory before and after the changes.

Time Period: Our study covers the period from the unification of Germany until today. The first year in the series for which data exist is 1872, and the last covered by the estimates is 1958. We have attempted to provide estimates for each individual year with the exception of those during the war and post-war disturbances. The two gaps extend from 1913-1925 and 1938-1950. Very little data are available for those years, and their interpretation poses difficult conceptual problems which are beyond the scope of this study.

All the estimates refer to calendar rather than to fiscal years, with the exception of the estimates of government expenditure for single "benchmark" years prior to the First World War when conversion was impossible. For the remaining years we have followed the usual but imperfect method of convert¬ing fiscal into calendar years based on the assumption that expenditure is spread evenly over the whole year.

Verwendete Quellen (ausführliches Verzeichnis):
Data Sources (bibliography / citation of sources):
(1) Demographic Statistics:
Statistisches Bundesamt (Hrsg.), versch. Jg. 1952-1958: Statistisches Jahrbuch für die Bundesrepublik Deutschland, 1952 – 1958. Stuttgart.
Statistisches Bundesamt (Hrsg.), 1958: Bevölkerung und Wirtschaft. Langfristige Reihen 1871 bis 1957 für das Deutsche Reich und die Bundesrepublik Deutschland. Stuttgart.
Statistisches Bundesamt (Hrsg.), 1958: Statistik der Bundesrepublik Deutschland, Bd. 199, Tabelle 1 und 3, S. 10-13.

(2) Price Index:
Institut für Konjunkturforschung, 1925: Sonderheft No. 37. Berlin.
Statistisches Bundesamt (Hrsg.), 1958: Bevölkerung und Wirtschaft. Langfristige Reihen 1871 bis 1957 für das Deutsche Reich und die Bundesrepublik Deutschland. Stuttgart, Tabelle 2, S. 84.

(3) Estimates of Gross National Product:
Statistisches Reichsamt (Hrsg.): 1932: Das deutsche Volkseinkommen vor und nach dem Kriege. Einzelschriften zur Statistik des Deutschen Reichs Nr. 24. Berlin.
Statistisches Bundesamt (Hrsg.), 1954: Die langfristige Entwicklung des Sozialprodukts, in: Wirtschaft und Statistik, 6. Jg., Heft 2.
Hoffmann, W.G./Müller, J.H., 1959: Das deutsche Volkseinkommen 1851-1957. Tübingen.

(4) Government Expenditure:
Most, O., 1908, 1910: Die Gemeindefinanzstatistik, in: Schriften des Vereins für Socialpolitik, Band 126 und Band 127. Leipzig.
Statistisches Reichsamt (Hrsg.), 1925ff: Statistisches Jahrbuch für das deutsche Reich. Berlin.
Statistisches Reichsamt (Hrsg.), versch. Jahre: Einzelschriften zur Statistik des Deutschen Reichs, ausgewählte Bände. Berlin.
Statistisches Reichsamt (Hrsg.): Statistik des Deutschen Reichs, spezielle Publikationen.

Denkschriftenband zur Begründung des Entwurfs eines Gesetzes, betreffend Änderungen im Finanzwesen, in 4 Teilen, in: Verhandlungen des Reichstages, 1907-1909, Band 249, 250, 251. Die Schätzungen sind aus Teil I, 1908, entnommen: Das Finanzwesen der öffentlichen Körperschaften in Deutschland.

Statistisches Reichsamt (Hrsg.), 1930: Die deutsche Finanzwirtschaft vor und nach dem Kriege nach den Hauptergebnissen der Reichsfinanzstatistik. Einzelschriften zur Statistik des Deutschen Reichs, 14. Berlin.
Statistisches Reichsamt (Hrsg.), 1930: Finanzen und Steuern im In- und Ausland. Ein statistisches Handbuch. Berlin.

Kaiserliches Statistisches Amt (Hrsg.), 1902ff: Die Finanzen des Reichs und der deutschen Bundesstaaten, in: Vierteljahreshefte zur Statistik des deutschen Reichs. Berlin. Die Schätzungen beruhen auf veröffentlichte Tabellen aus: 2. Heft 1911, 2. Heft 1913, 4. Heft 1922.

Nathan, O., 1944: Nazi War Finance and Banking, N.B.E.R., New York.
Recktenwald, H.C., 1962: Die Entwicklung der öffentlichen Ausgaben in der Bundesrepublik Deutschland, in: König, H. (Hrsg.), 1962: Wandlungen der Wirtschaftsstruktur in der Bundesrepublik Deutschland, Schriften der Vereins für Socialpolitik, Neue Folge, Band 26. Berlin. (Recktenwald berücksichtigt in den Gesamtstaatsausgaben nicht die Ausgaben für die Sozialversicherung, daher die Differenzen zu den Angaben in der vorliegenden Studie, die die Sozialversicherung einschließt!).
Länderrat des amerikanischen Besatzungsgebiets (Hrsg.), 1949: Statistisches Handbuch von Deutschland 1928 – 1944, München, S. 546-555.
Statistisches Bundesamt Wiesbaden (Hrsg.), versch. Jg. 1952-1961: Statistisches Jahrbuch für die Bundesrepublik Deutschland, 1952 – 1960. Stuttgart/Mainz.
Statistisches Bundesamt Wiesbaden (Hrsg.), 1950ff: Statistik der Bundesrepublik Deutschland, Bd. 59 (S. 31-39), Bd. 227 (S. 20, 21, S. 36-39); Bd. 259 (S. 33, S. 44-47).
Statistisches Bundesamt Wiesbaden (Hrsg.), 1957: Statistik der Bundesrepublik Deutschland, Die öffentliche Finanzwirtschaft in Bund, Ländern, Gemeinden und Gemeindeverbänden in den Rechnungsjahren 1948 bis 1954. Stuttgart/Mainz.
Statistisches Bundesamt Wiesbaden (Hrsg.), 1952-1959: Staatliche Finanzen. Rechnungsjahre 1949 – 1957. Stuttgart/Mainz.
Statistisches Bundesamt (Hrsg.), 1961: Der Staat als Teil der Volkswirtschaft 1950 bis 1959, in: Wirtschaft und Statistik, Heft 3.

Anmerkungen:
Annotations: for the entire study; for single (selected) data tables:
(See also the attached pdf – document with detailed methodical informations: Chapter „II.: Conceptual and Statistical Problems“, p. 172- 178 and Chapter VI.: Statistical Appendix”, p. 223 – 237).

“The growth of government expenditure will be discussed against a background of a changing population and of a growing output. Since the territorial delimitation of Germany has been altered twice as a result of the upheavals of war, our estimates are not directly comparable between various periods of different geographical size. However, we have assumed that government per head of population or alternatively the proportion government expenditure to total output is directly comparable between different periods. The changes in territory may, of course, bring about an increase or a reduction in the government expenditure per head, or in the proportion of government expenditure in the General National Product (G.N.P.). The effects of changing territory on government expenditure are not analysed in this study as the conceptual and statistical difficulties are insurmountable.
The deflation of current estimates of government expenditure for price changes presents formidable difficulties of both conceptual and statistical character. The conceptual difficulties are due to the absence of market valuations of government output”. The necessity to value this output at some values observed in the private sector introduces an element of ambiguity about the actual amounts of services provided by the public authorities. This difficulty is an additional to the general index number problem which is not to be neglected in comparisons over such long periods. The statistical difficulties are due to the paucity of price information, especially in the earlier years … The use of a single index reflects rather the availability of information than any conceptual consideration” (Andic/Ververka, 1963/64, p. 171).

“Each of the expenditures is to be included only once in the aggregate, transfers between different public authorities and between different accounts of one public authority have to be eliminated. Any expenditure item is allocated to that public authority which is responsible for the provision of a given service whatever the source of finance. In the terminology of the Financial Statistics, our concept is that of “unmittelbare Ausgaben” and not of “Eigenausgaben”. The elimination of internal transfers within the public sector has already been carried out in the Financial Statistics, and no adjustment was, therefore, necessary for the period covered by them. Until 1913, not all transfers could be excluded and the aggregate includes double counting to that extent. We have excluded non-specific transfers (“Überweisungen und Matrikularbeiträge” between the central government and the State governments but not specific grants (“Zweckgebundene Vergütungen aus der Reichskasse”) which were negligible As we have included in the public sector total expenditure on social insurance, that part which was financed out of the general revenue of public authorities had to be excluded” (Andic/Veverka, 1963/64, p. 228).

Before the reader is presented with the statistical findings we propose to discuss briefly a limited number of problems encountered in the evaluation of government expenditure in Germany.


(1) The comparability of the estimates before and after 1913
The lack of comparability between the periods before and after 1913 has both a) statistical and b) conceptual reasons. As will be discussed in more detail in the Statistical Appendix, there is a break in the available statistical sources around 1913. The estimates until 1913 include items which should not pro¬perly appear there, such as some double counting and a considerable amount of current expenditure on trading services. These items could only be removed with great effort, if at all. But there is an additional problem introduced by the change in the definition of public sector around 1913 which is, moreover, complicated by a reduction in the territory. Prior to 1913, most of the trading services provided by public authorities were not separated from their budgets. For example, the railways and postal services were included in the budgets of the state governments which were also responsible for a great many other pro¬ductive enterprises. The first set of estimates, given in Appendix Tables A. 6 to A. 30, includes, therefore, all capital expenditure by state governments prior to 1913 which amounted to a considerable proportion of capital forma¬tion. After the First World War most of the public enterprises became autono¬mous and their accounts disappeared from the budgets and from the Financ¬ial Statistics. But the change in most cases was purely formal and did not affect the decision-making process. We have, nevertheless, not attempted to add capital formation by trading enterprises to the inter-war estimates, one of the reasons being the difficulty of obtaining data. On the other hand, we have computed "adjusted" estimates of government expenditure for the years until 1913 which exclude all capital formation by trading enterprises. These estimates, given in Appendix Table A.35, seem to us to reflect better the long-term trend and we have therefore adopted t m in the main text. Had we been able to exclude the other items mentioned above, i. e. double counting and some current ex¬penditure on trading services, such series would be directly comparable with the remaining estimates. As this is impossible we have assumed that the per capita expenditure in 1913 on the total territory was equal to that on the 1925 territory. We have therefore increased the post-First World War estimates for 1913 which refer to the reduced territory for the additional population of the old Reich. The total obtained in this manner still appeared substantially lower than the estimate of government expenditure, exclusive of capital formation by trading enterprise, based on pre-1913 sources. We have then re¬duced the estimates for all the years prior to 1913 by the same proportion. The assumption of constant expenditure per head of population in 1913 for both old and reduced territory brings us to the next problem of territorial changes and of the meaning of per capita estimates.


(2) Effects of territorial changes and per capita estimates
The major difficulty in the discussion of the secular growth of public sector in Germany is introduced by the territorial changes. The break in the statistics around 1913 is complicated by the reduction in the territory, and the Second World War brought an even greater change. Neither time nor the necessary data were available to analyse the effects of the territorial changes on the level and composition of government expenditure in Germany.
Andic/Veverka try to eliminate the change by expressing it in terms of population or alternatively of national income on the assumption that the expenditure is distributed on the territory before the change in the same way as population (or national income). In the absence of any alternative, the interpretation of estimates proceeds on the assumption, that the territorial changes have not affected per capita estimates.


(3) The elimination of price changes and the productivity assumption
Current estimates of government expenditure reflect the real amount of public services only imperfectly as the unit of measurement changes over a time. The elimination of these changes comes up against two sets of problems, namely the statistical problem of the availability of the appropriate price indexes and the index number problem.
Further, in the case of government expenditure, we encounter an additional problem introduced by the absence of any direct valuation of the output of public services.
Most of the study of government expenditure uses either a cost-of-living index or an index of wholesale prices. This is a very crude method an a certain degree of refinement can be obtained by using separate indexes for individual components of government expenditure. This method was used by Peacock and Wiseman (1961). Our choice of the price index for the period until 1925 was dictated by the absence of any price information other than for a limited number of commodities, with foodstuffs predominating. The index reflects the secular trend of the price of government output only approximately. The index used for the period after 1925 may be expected to be better at least in this respect, producing fewer oscillations. For that period we have used the index implied in the current and constant estimates of national income.

„Until 1925, it represents the movement of prices of selected commodities, and after this date it is based on the index implicit in the official estimates of the national product at current und constant prices. The first part of the index, that which covers the individual years until 1913 and the price changes between 1913 and 1925 is divided into two. During the 1870’s, it represents an index of wholesale prices with shifting weights (this Index was calculated by the Institut for Konjunkturforschung, Sonderheft No. 37, Berlin), and after that date it is an unweighted geometric average of the former index and a retail index of good prices (for the index of retail prices see: Statistisches Bundesamt (Hrsg,.), 1958: Bevölkerung und Wirtschaft, “Statistik der Bundesrepublik Deutschland”, Band 199, Table 1 and 3, pp. 10-13). The resulting index has the advantage of damping to a certain degree the price fluctuations typical for wholesale prices without giving undue emphasis on price changes of foodstuffs” (Andic/Veverka, a.a.O., S. 224f).
As far, as the long-term trend is concerned, it will be correct if the productivity and the factor rewards in the public and private sector have changed to the same degree. Assuming that competition has kept the factor rewards more or less equal in both sectors and that the factor proportions have not changed, the relative prices of the two respective outputs would remain constant. In such a case the price index of private output would reflect the prices of the public output if such output were valued in the market. All studies, using an index of private outputs for deflation government expenditure, implicitly or explicitly assume identical productivity changes in the private and public sector. Has this assumption any empirical basis? If not, we must adjust the price index for the productivity assumptions of constant productivity in the public sector, such an adjustment price index would be identically equal to the price index for government inputs, e.g. wages in public employment. Strictly speaking, we should take into account prices of all inputs, yet statistically it is very difficult to obtain price indexes other than for wages. As an approximation such an index is good enough, as been half and three quarters of government expenditure consist of wages and salaries. That the productivity in the public sector is a purely conceptual quantity does not diminish its importance. Later on we shall argue that the relative productivity changes have accounted partly for the secular growth of the public sector.


(4) Gross National Product
“The historical statistics of G. N. P. are very incomplete in spite of the fact that the Imperial Statistical Office ("Statistisches Reichsamt") published a national income monograph as early as 1932. On the other hand, as a result of the relatively early use of income tax or similar taxes in the German tax system, the income statistics on which the first estimates of national income are based are available long before 1872 when our enquiry starts. Systematic estimates of national product, based on either the production or expenditure method, are only available after the Second World War. The concept consider¬ed most suitable for our purposes is that of the Gross National Product at Factor Cost, but it is only since 1950 that it is directly available from the offic¬ial national income estimates. Until then, official statistics gave estimates of National Income rather than of Gross National Product, and the underlying concept was different. The minor conceptual differences can be ignored as they are quantitatively negligible. The major deviation from the concept of nation¬al product as it is accepted today is the treatment of government. While today government is usually treated as a final consumer, then all expenditure which added directly to individual welfare was considered as expenditure out of private income. Such expenditure is, however, already included in private in¬comes in the form of direct taxation. It is only that part of expenditure on behalf of the consumer, financed out of indirect taxes, which has to be added to total aggregate income. The adjustment of such a concept of National In¬come to that accepted today consists of reducing the aggregate to the original level. The estimates of G. N.P. are then obtained by adding depre¬ciation. From 1925 onwards, it is possible to use official national income estim¬ates, since both adjustments have been carried out by the Federal Statistic¬al Office. The series has been given for the pre-war years, alternatively for the territory of the 1925 Reich and of the Federal Republic. For the period prior to 1925, two alternative series of national income estimates are available. The 1932 monograph gives estimates from 1891 onwards. A slightly different series going as far back as 1851 is given by Hoffmann and Müller (1959). Our estimates, given in Table A. 5, are based on the latter series, which is preferable in view of its longer run of estimates and of a higher degree of sophistication in its method. As no estimates of capital consumption in those years are available, we have assumed that the capital depreciation in 1913 was equal to 11 % of the national product, i. e. the same proportion as observed in the inter-war period. We have further assumed that between 1891 and 1913, the proportion of capital depreciation in G.N. P. declined by the same relative amount as the share of capital formation in G.N. P. For the years prior to 1891, we have extrapolated the estimates of G.N. P. at Factor Cost on the basis of those of national income” (Andic/Veverka, 1963/64, p. 225f).



(5) Government Expenditure
“The remaining tables in the Statistical Appendix give the estimates of “total government expenditure”. The aggregate expenditure is broken down according to three criteria, namely the “economic category”, the “function”, and the “level of government”; and the procedure of obtaining these divisions will be explained in connection with the relevant tables.
The break-down of total government expenditure by different levels of govern¬ment is often implied in the presentation of statistics. In many countries the only available financial statistics are the separate accounts published by the different government levels. This was the case of Germany before World War I, while for subsequent years we can rely on the consolidated accounts of all levels of government. The discussion of sources and methods will, therefore, fall into two parts, namely that dealing with statistics prior to 1913, and that dealing with the years since then.

The lack of published consolidated accounts prior to 1913 meant that first of all, we had to compute estimates for the different levels of government and combine them afterwards. The estimates for
1. the central government
were relatively easy to obtain as the accounts of the “Reich” government were published regularly.
The other two levels of government, namely
2. the governments of the states ("Bundesstaaten" before World War I, "Länder" afterwards) and
3. the local authorities ("Gemeinden" and "Gemeindeverbände") each published their own accounts.

Governments (‘öffentliche Verwaltung’) include the governments of the ‘Reich/Bund’, those of the states (‘Bundesstaaten/Länder’), and the local authorities (‘Gemeinden’ and ‘Gemeindeverbände’)”, Andic/Veverka, a.a.O., S. 227).
Levels of Government:
(a) Reich/Bund (‘Central Government’); Reichsausgaben („Central government expenditures”); Reichsausgaben nach Funktionen (“Central government expenditures by functions”).
(b) Bundesstaaten/Länder (‘Governments of the States’); Bundesstaatliche Ausgaben (“State government expenditures”).
(c) Gemeinden und Gemeindeverbände („Local Authorities“).
(d) Gesamtausgaben (“Total government expenditures”); Gesamtausgaben nach Funktionen (“Total government expenditures by functions”).

We have not attempted to estimate government expenditure for each separate state, although the accounts of the state governments were published separately. It is impossible to do so in the case of the thousands of local authorities, even though in most of the states some data on local govern¬ment expenditure were published. The first attempt to present a complete picture of the financial transactions of all public authorities was made in a report accompanying the proposal for finance reform in 1907. Soon after the turn of the century, the central statistical office ("Kaiserliches Statistisches Amt") started to publish at irregular intervals summary financial statistics of the central and state governments. These summaries, together with the "Denkschriftenband" 1908, represent the main source of information for the period until 1913. Whenever we have used additional sources, they will be mentioned in the discussion of individual tables.
For the period from 1913 onwards (i.e. for the year 1913 and the years 1925 and after), we have used statistics prepared by the Imperial Statistical Office and published in the current statistical sources such as the Statistical Yearbook or in special publications ("Einzelschriften zur Statistik des Deutschen Reichs" or part of the "Statistik des Deutschen Reichs"). After World War II, they have been published by the Federal Statistical Office ("Statistisches Bundesamt"). They cover all public authorities, with one exception, namely that of social insurance funds. These are not treated as part of the public sector and so their accounts have been published separately. For our purpose we have used the tabulations given in the Statistical Yearbook.
Table A. 6 gives expenditure of total government for selected years be¬tween 1881 and 1913 and annually for the years covered in our study. Current expenditures are then deflated for price changes and both current and constant estimates are given on a per capita basis. Public authorities are defined in the usual manner, including governments and bodies carring out govern¬ment functions. Governments ("öffentliche Verwaltung") include the government of the "Reich" ("Bund"), those of the states ("Länder"), and the local authorities ("Gemeinden" and "Gemeindeverbände"). All these are included in the Financial Statistics. One category of public bodies, namely those carrying out certain economic or social functions on the territory of several local authorities ("Zweckverbände"), are omitted from the Financial Statistics and consequently from our estimates. As they provide services in part to the local authorities against payment, they are nevertheless included to that extent under local government expenditure. The omission is more im¬portant insofar as public capital formation is concerned; such expenditures are excluded. The only group of non-governmental character included is the semi-public bodies administering the German system of social insurance, which has already been mentioned.

Government expenditure is defined as expenditures by public authorities of a non-financial character on services which are not sold against specific pay¬ments to the private sector. Each expenditure is to be included only once in the aggregate, transfers between different public authorities and between differ¬ent accounts of one public authority have to be eliminated. Any expenditure item is allocated to that public authority which is responsible for the provision of a given service whatever the source of finance. In the terminology of the Financial Statistics, our concept is that of "unmittelbare Ausgaben" and not of "Eigenausgaben". The elimination of internal transfers within the public sector has already been carried out in the Financial Statistics, and no adjust¬ment was, therefore, necessary for the period covered by them. Until 1913, not all transfers could be excluded and the aggregate includes double counting to that extent. We have excluded non-specific transfers ("Überweisungen und Matrikularbeiträge") between the central government and the State governments but not specific grants ("Zweckgebundene Vergütungen aus der Reichskasse") which were negligible. As we have included in the public sector total expenditure n social insurance, that part which was financed out of the general revenue of public authorities had to be excluded. These contributions, which were entirely the responsibility of the central govern¬ment prior to the Second World War, and, to a minor degree, of the states after that, were deducted.
The existence of inter-governmental payments complicated the problem of excluding ‘Berlin’ from the estimates. The whole amount of expenditure by the Berlin local authorities ("unmittelbare Ausgaben") should not have been excluded since part of it is financed from the resources of the other public authorities in Germany. However, neither the published data nor the information made subse¬quently available by the Federal Statistical Office give enough detail to warrant its inclusion.

The next step in obtaining the estimates of government expenditure was to eliminate expenditure items of a purely financial nature. Lending, for example, is not generally considered part of government expenditure; neither is the purchase of second-hand assets. We have, nevertheless, included the substantial loans ("Gewährung von Darlehen, Inanspruchnahme aus Bürgschaften") by all levels of public authorities, especially by state governments. The majority of these went towards housing and represented capital grants to housing associations. The fact, that the provision of housing facilities by the public authorities in Germany was carried out in this way rather than by a direct system of grants or by a direct provision of housing facilities, is, from our point of view, irrelevant. The other major point in this respect is the treatment of loan charges borne by the state and local authorities. As most of the debt has been incurred for productive purposes, we have assumed that the loan charges measure approximately the flow of real services derived from the ownership of capital assets by these authorities. The same assumption cannot be made at the central level. Besides the imputed stream of services, we have the actual payments n debt service at all levels of government. Of this, only the interest has been included, as the repayment of the principal repre¬sents a purely financial transaction” (Andic/Veverka, 1963/64 p. 226 – 229).


(6) Inclusion of the German System of Social Insurance (“Soziaversicherung”):
Inclusion of the social insurance expenditures: The only group of non - governmental character included in this study is the semi – public bodies administering the German system of social insurance.

For the years until 1948:
(a) „Soziale Krankenversicherung“; (b) „Gesetzliche Unfallversicherung“; (c) „Invalidenversicherung“; (d) „Angerstelltenversicherung“; (e) „Knappschaftliche Rentenversicherung“; (f) „Arbeitslosenversicherung und Arbeitslosenhilfe“ (since 1928).
For the period 1950 until 1958:
(a) „Gesetzliche Krankenversicherung“; (b) „Gesetzliche Unfallversicherung“; (c) „Rentenversicherung der Arbeiter“ (named „Invalidenversicherung“ until 1956); (e) „Knappschaftliche Rentenversicherung“; „Rentenversicherung der Angestellten“; (f) „Arbeitslosenversicherung und Arbeitslosenhilfe“; „Familienausgleichskassen“ (since 1955).

The contributions to social insurance have been considered as taxes.


(7) Classification of Expenditure by Functions
“The classification of expenditure by functions, given in Tables A. 18 to A. 29, which is the second of our three divisions, came up against the same difficulties as those met by the previous one. A complete breakdown is possible only since 1913 and the Reich expenditure in the 1930's could only be classified incompletely. In the years for which a full classification was possible, we have divided all expenditure into eight main categories with the Subdivision of social services into their individual components. We have followed the classification pattern adopted in other studies (such as those for Britain by Peacock and Wiseman), in so far as the details given in the Financial Statistics made this possible. In some cases, and this is especially so in the case of economic and environmental services, the categories given there differed from those in our classification and we had to adopt them as they were.
The first category, “Administration and other”, includes the cost of general administration other than the administration of a particular service, such as education, and other minor items which cannot be allocated to any specific function. It also includes expenditure on overseas services which are often classified separately, as well as the cost of fiscal administration and of debt management. The second category, “Law and Order”, represents the cost of maintaining an internal system of peace and lawfulness. The cost of the Federal Border Police ("Bundesgrenzschutz") is not included here as this function is usually carried out within the framework of military defence. It is, there¬fore, included in the next category, “Defence”, which also includes the contributions towards the cost of Allied Forces in Germany. The main expenditure is, of course, the cost of armed forces. This category does not include those expenditures which can be classified only indirectly as defence, such as econ¬omic measures increasing the self-sufficiency of the country, or those which were deliberately financed under the non-military sections of the budget. The first problem is that of definition and it can be argued that only those expend¬itures which serve direct defence purposes should be included. During a pre¬paration for war, defence expenditure so defined must underestimate the actual military effort. The second problem is that of availability of data. It is impossible to separate direct defence such as construction of airfields or forti¬fication from other public work programmes in the 1930's. To that extent, our data underestimates the actual defence expenditure. The next category is that of “War damage compensation” which is better described by the German term of "Kriegsfolgelasten". It includes the burdens arising out of war, other than those of a social nature, the most important among them being the pay¬ment of reparations after World War I. It includes a compensation to Israel as well as internal payments for damages caused during the war. The occupation costs which are sometimes classified under this category have been treated as defence expenditure. “Social services” include social insurance, social assistance whether related to the war or not, education, housing and health. The last category has not been given separately in the inter-war period from 1929 onwards. The category of “Economic services” includes the provision of services directly assisting economic activities, including here the provision of trading services. On the whole, these are provided in two forms, either by economic enterprises, such as utilities or transport enterprises, or by the ownership of real assets such as building ground ("Erwerbsvermögen" and "Wirtschaftsunternehmen"). As we have pointed out before, both these forms are included in go¬vernment expenditure only to the extent of capital formation or deficits on operating accounts. The non-trading economic services consist of services to agriculture and industry. An important element in this category is direct production subsidies. The Financial Statistics also include under services to agriculture the cost of the provision of dams, coast protection, etc., which is usually treated as an environmental service. The provision of transport facili¬ties such as roads, bridges, airports, harbours and so on has been considered as directly promoting economic activity and is included in this category. The next category comprises “Environmental services”, i. e. the collective services necessary to communal life. There is no clear-cut distinction between this and the previous category. For example, the provision of roads can be classi¬fied under both these headings, since they can be looked on as providing direct service to consumers and to enterprises, indistinguishable from other inputs, or as a part of the general framework necessary for the functioning of the economic system. Environmental services consist of the cost of pro¬vision of basic urban services such as sewerage, cleansing and lighting of streets, fire brigade, markets, cemeteries, and similar installations which appear in the Financial Statistics as "Kommunale Anstalten und Einrichtungen". Since the Financial Statistics do not give any break-down of this category, the services of clearly economic character, such as markets and municipal savings banks, could not be transferred to the economic category where they should belong. The last category is “Debt service”. This includes interest payments on the debt of the central and state governments, and that part of the local loan charges which are not allocated to any particular func¬tion. As already mentioned, the cost of debt management has been classified under Administration and other. This is the only case when the cost of mana¬ging a given function was considered part of the general administrative ex¬penses and not as the cost of a specific service. The category of debt service does not include the amount of capital repayments, which are considered purely financial transactions.

For the years prior to 1913, a complete functional break-down for all levels of government is impossible. Even the published statistics of expenditure of the Prussian local authorities, which we have used in estimating local expenditure during that period, are not detailed enough to separate individual functions. More details are available for the expenditure of the Reich and of the state governments, yet the break-down for the two levels given in Tables A. 22, A. 24, A. 25, A. 26 is in no way comparable with the classification for 1913 and after. The category of Administration and unallocated (the term used prior to 1913) is thus much wider than the corre¬sponding category for 1913 and after. It included all services that were not allocated to any specific function. The expenditure out of the French reparations in the 1870's have been kept as a separate category. They included items classified as military defence, (fortifications, military pensions, etc.) which would otherwise fall into the separate category of "Kriegsfolgelasten". The category of Social insurance in Reich expenditure represents the expenditure by social insurance funds, computed separately, which we have added to the expenditure of the central government proper.

A different arrangement of functional categories is presented in Table A.30 which gives separately all expenditures related to war and then relates them to total government expenditure and Gross National Product. The cost of defence, the interest on dept of the Reich which was mainly incurred in financing defence expenditure, and the obligations of the government arising directly out of the war effort, all fall into this group” (Andic/Veverka, 1963/64, p. 234 – 236).


The Treatment of Trading Activities
„The final, and most difficult, problem has been the treatment of trading activities. Over the period as a whole, the trend has been towards their separation from other governmental activities. Prior to 1913, they were mostly not separated from the provision of other services and thus, they represented part of the public sector. They are included in our estimates, as given in the Tables A.6 to A.34, to the extent of expenditure on capital account. Table A. 35 as well as all the tables in the text, is based on the “adjusted” series” (Andic/Veverka, a.a.O., S. 229).
„The estimates until 1913 include items which should not properly appear there, such as some double counting and a considerable amount of current expenditure on trading services. These items could only be removed with great effort, if at all. But there is an additional problem introduced by the change in the definition of public sector around 1913, most of the trading services provided by public authorities were not separated from their budgets. For example, the railways and postal services were included in the budgets of the state governments which were also responsible for a great many other productive enterprises. The first set of estimates given Tables A.6 to A.30, includes, therefore, all capital expenditure by state governments prior to 1913 which amounted to a considerable proportion of capital formation. After the World War most of the public enterprises became autonomous and their accounts disappeared from the budgets and from the budgets and from the Financial Statistics. But the change in most cases was purely formal and did not affect the decision-making process. We have, nevertheless, not attempted to add capital formation by trading enterprises to the inter-war estimates, one of the reasons being the difficulty of obtaining data. On the other hand, we have computed “adjusted” estimates of government expenditure for the years until 1913 which exclude all capital formation by trading enterprises. These estimates, given in table A.35, seem to us to reflect better the long-term trend and we have therefore adopted them in the main text. The test tables also reproduce the “adjusted” estimates. Had we been able to exclude the other items mentioned above, i.e. double counting and some current expenditure on trading services, such series would be directly comparable with the remaining estimates. As this is impossible we have assumed that the per capita expenditure in 1913 on the total territory was equal to that on the 1925 territory. We have therefore increased the post-First-War estimates for 1913 which refer to the reduced territory for the additional population of the old Reich. The total obtained in this manner still appeared substantially lower than the estimate of government expenditure, exclusive of capital formation by trading enterprise, based on pre-1913 sources. We have then reduced the estimates for all the years prior to 1913 by the same proportion” (Andic/Veverka, a.a.O., S. 174f).

Sachliche Untergliederung der Datentabellen:
List of data tables in HISTAT and topics of the tables:
(A = Data tables from the appendix)

A.01 – A.03 Population:
A.01 Population (1872-1958)
A.02 Distribution of population by density (1871-1958)
A.03 Population and area (1871-1958)

A.04 – A.05 Price Index, Gross National Product:
A.04 Price index (1872-1958)
A.05 Gross national product at current and 1900 prices (1872-1958)

A.06 – A.14 Total Government Expenditure:
A.06 Total government expenditure at current and 1900 prices (1881-1958)
A.07 Total government expenditure by levels of government, at current and 1900 prices (1872-1958)
A.08 Total expenditure per head of population, by levels of government, at current and 1900 prices (1881-1958)
A.09 Expenditure on social insurance, by economic categories, at current prices (1885-1958)
A.10a Total government by economic categories, at current prices (1881-1913)
A.10b Total government expenditure by economic categories, at current prices (1913-1958)
A.11a Total government expenditure by economic categories, at 1900 prices (1881-1913)
A.11b Total government expenditure by economic categories, at 1900 prices (1913-1958)
A.12a Total government expenditure by economic categories per head of population, at 1900 prices (1881-1913)
A.12b Total government expenditure by economic categories, per head of population, at 1900 prices (1881-1958)
A.13 Total government expenditure by economic categories, as percentage of the total and of Gross National Product (1913-1958)
A.14 Total government expenditure on transfers and subsidies, at current prices (1913-1958)

A.15a – A.17b Central, State and Local Government Expenditure:
A.15a Central government expenditure by economic categories, at current prices (1872-1913)
A.15b Central government expenditure by economic categories, at current prices (1913-1958)
A.16a State government expenditure by economic categories, at current prices (1881-1913)
A.16b State expenditure by economic categories, at current prices (1913-1958)
A.17a Local government expenditure by economic categories, at current prices (1881-1958)
A.17b Local government expenditure by economic categories, at current prices (1913-1958)

A.18 – A.15 Total, Central, State and Local Government Expenditure by Function:
A.18 Total government expenditure by function, at current prices (1913-1958)
A.19 Total government expenditure by function, at 1900 prices (1913-1958)
A.20 Government expenditure by function, per head of population, at 1900 prices (1913-1958)
A.21 Total government expenditure by function as percentage of total expenditure and gross national product (1913-1958)
A.22 Central government expenditure by function, at current prices (1872-1913)
A.23 Central government expenditure by function, at current prices (1913-1958)
A.24a Central government expenditure by function, per head of population, at 1900 prices (1872-1913)

Download weiterer Texte zu dieser Studie im PDF Format (Forschungsberichte, Publikationen, Materialien zur Studie)