Measuring Social Reporting Intensity in Industrialized Countries: An Exercise in the Sociology of Knowledge

Since the beginning of the 90s, a new wave of social reporting has appeared in those European countries which did not participate in the first wave of the 1970s. Nevertheless, large differences remain with respect to reporting intensity, regularity and models of social reporting. Societal characteristics still influence the way and degree of national social reporting.

The Four Main National Approaches to Social Reporting

The activities of international organisations have strongly influenced the implementation of social reporting in several European countries. Nevertheless, there are tremendous differences between and variations of the forms of national social reporting in Europe. These national differences have also been shaped by the historical requirements existing in these countries when they developed national social reporting, by differing statistical traditions, and by differing ways of solving problems of statistical organisation: in France, for instance, the statistical office was given an economic research institute and COLOR="#000000">later on a sociological research institute as well; the same is true for Luxembourg, concerning the first point; in Norway, representative of Scandinavia, the statistical office has taken over nearly all tasks of social reporting and publishes a regular social report; in Sweden, a division for social reporting was established within the SCB. It must be said here that national styles of social reporting certainly exist in the individual countries, but often other styles of social reporting coexist. In any case, it is possible to construct some models or types of national social reporting within Europe: first, the resource approach, second, the quality of life approach, and third, the living conditions approach. The typology of social reporting, as presented in Table 1, is similar to Esping-Andersen’s (1990) typology of welfare states, as it is possible to distinguish three types of welfare research in Western Europe which not completely, but to some degree, fit into the welfare state regime types. Nordic level of living research is thus related to the social-democratic welfare state regime, continental quality of life research partly fits into the social insurance type of welfare state, and the living conditions approach fits into the liberal model of welfare state or the Southern "residual" model of welfare state. A clear co-variation of both dimensions can only be found in the Nordic countries, while the variations are higher for the other two types. Living standard research is a fourth type of social reporting. It is assumed that this is the predominant research tradition in Eastern European countries, where social reporting centred on the basic needs of the citizens, such as food, housing, work and health. It is assumed that this tradition will also continue to exist during the transition phase of these countries.

Table 1: Typology of Social Reporting

(1) "Level of Living" Research in the Nordic Welfare States

There is a close relationship between the development of the Scandinavian type of social reporting and the welfare state development in Scandinavia. The Nordic welfare states developed late but very fast under social-democratic governments and belong to the "universalistic model". The most characteristic feature of the Scandinavian approach is the importance placed on individual resources, the things a person should have. Individual resources should be distributed as equally as possible. In this respect, special importance is placed on vertical redistribution, and less to horizontal distribution between different forms of living or family types. The goal of equality between the sexes is of utmost importance in the Scandinavian approach.

In order to measure living conditions effectively and to evaluate such a governmental welfare policy it is inevitable to introduce a comprehensive social reporting system as a public task. The introduction of a public system of social reporting was furthered by the long Scandinavian tradition of official statistics - regular statistical records were introduced as early as the 18^th century - and the positive attitude of the Nordic people towards official statistics. Thus protests against personal identification numbers are unknown in Scandinavia; they allow for statistical analyses of data stemming from different administrative registers, and the money that would otherwise be needed for data collection can be saved.

It is typical for the organisation of social reporting in Scandinavia that the statistical offices are the organisational focus of social reporting. They carry out the surveys on living conditions, analyse the data and publish the results in their own statistical series. The close co-operation between the statistical office and the sociological research institute (mostly the only one) is typical (see Sweden and Denmark).

Combined with a rather well established system of special statistical surveys, such as surveys on time use, income, household budgets, family formation and fertility, health, etc., which are also organised by the statistical office, it becomes possible to regularly publish national social reports. This system is for instance predominant in Norway. In Sweden, the statistical office publishes a series on social reporting. In Denmark, Danmarks Statistik collaborates closely with the national social research institute, which is under the tutelage of the Ministry for Social Affairs, and is at the same time the main actor in the field of social reporting. In Finland – where the statistical system is strongly decentralised - Statistics Finland publishes the results of the Survey on Living Conditions (Vogel 1997; Uusitalo 1994).

(2) Quality of Life Research on the Continent

Based on the model "United States", a tradition of quality of life research was established in Germany, Austria, Switzerland and the Netherlands, which combines objective and subjective elements and thus emphasises both dimensions. Besides objective living conditions, subjectively perceived evaluations of individuals with respect to their objective living conditions are taken into account. The subjective reflection of objective living conditions on the individual level is thus incorporated into this research tradition. Subjective well-being is not only assessed by using a summary measure as the overall life satisfaction, but also with respect to satisfaction regarding specific living domains, as the income situation or health status. Thus, by combining the subjective and objective dimension, a theoretical understanding of the dynamics of well-being is enhanced. In those European countries where it was possible to establish quality o f life research, a certain division of labour between official statistics and university research evolved, with the former concentrating much more on objective living conditions, while the latter predominantly dealing with subjective well-being or the perception of objective living conditions.

(3) Living Conditions Approach

The third approach can be called "living conditions approach". The underlying theoretical concepts neither come from the Nordic resource approach nor from research on subjective well-being. The main point of orientation is rather the component approach, that means that mainly objective data stemming from different statistical sources are presented in a systematic arrangement of several living domains. The main actor in the field of social reporting in those countries is the statistical office, therefore subjective data are normally not included in official social reports. Social reports are necessary for several reasons. One reason is to present results from often very numerous social surveys. The second one is to organise statistics which partly come from numerous public bodies in a coherent form for being used by government departments themselves and the wider public.

This living conditions approach is the most widespread approach in Western and Southern Europe. While Western European countries, such as Great Britain and France, developed continuous reporting systems very early, the development in Southern Europe was retarded. Therefore a distinction can be made between Western Europe and the Latin Rim concerning this approach.

a) Western Europe

In France, social reporting has a strong position. Again, this is related to the state organisation. France as an extremely centralised state introduced a centralised social reporting, which is carried out nearly exclusively by national institutions. The central actor in France is the national statistical office (INSEE) which has combined since its foundation in 1946 statistics and economic research in order to enhance the scientific use of the statistics collected. Economic research at the INSEE was supplemented in 1980 by a division for sociological research (Begué 1981). In France it has therefore been possible to put economic and sociological research under one roof together with official statistics. The advantageous impacts of this organisational strategy are demonstrated by the main product of French official social reporting, the "Données Sociales". The studies included in this publication have a very ana lytic approach. They are oriented towards research questions and are innovative in methodical and material terms.

In Great Britain, the statistical system was strongly decentralised – at least until recently. Not only the individual parts of the country of the United Kingdom, such as Wales, Scotland, Northern Ireland and England, have their own statistical offices (territorial differentiation), but the statistical system is also fragmented with regard to functions. Until the formation of the "Office for National Statistics", two main institutions were responsible for official statistics, the Central Statistical Office (CSO) and the Office for Population Censuses and Surveys (OPCS). Furthermore, the part of statistics remaining with the ministries is still rather large. The introduction of the main product of official social reporting in Great Britain, the "Social Trends", was inter alia a result of the decentralisation and thus intricacy of the statistica l system, with the intention to present a synthesis of the most important statistical information for state administration and the public. "Social Trends" was also published for Wales. The characteristics of the statistical systems made a stronger integration of social science and official statistics possible. This becomes obvious not only with respect to "Social Trends", but also in other fields of statistical reporting, where a strong influence of the social sciences prevails (see "Population Trends", e.g.).¹

The "consociational democracies" of Belgium, the Netherlands and Switzerland do not reveal coherence concerning their social reporting systems. If the thesis that social reporting and "welfare stateness" do co-variate closely is correct, then coherence in social reporting cannot be expected in these countries. Due to the fragmentation of these countries concerning religion and language and the strong federalisation, centralised social reporting cannot be expected to have a strong position. The Netherlands seem to be an exception to this rule. Due to their strongly developed welfare state they have established a strong position of social reporting. They have implemented a regular social report (at the Social and Cultural Planning Office (SCP)), they conduct a living situation survey (Leefsituatiesurvey) and have a large publication system on social reporting. The Netherlands, however, represent a spe cific type of welfare state which is - in contrast to the Nordic type - not service-oriented, but provides overwhelmingly fiscal welfare and direct cash transfers. This could be the reason for the relatively big importance of surveys substituting administrative statistics.

In Belgium, national social reporting could not be established, neither at the national statistical office, nor at the Planning Bureau. Early attempts turned out not to be permanent. The further federalisation of the country has obviously made national social reporting more difficult; on the other hand it made reporting for the language communities easier. In the Flemish Community social reporting on the family and population (C.B.G.S.) and social reporting on social policy impacts (CSP, Antwerp) could be established. On the level of the nation state the national statistical office holds a rather weak position due to the strongly decentralised Belgian statistical system. Thus, essential parts of official statistics remained at the ministries, the numerous social security bodies and other organisations. This probably renders national social reporting highly difficult.

In Switzerland, social reporting developed rather late and has not been institutionalised so far. In the early 1980s a system of social indicators was developed; however, the project was cancelled due to resource restrictions, and social reporting was not continued for some years. A Swiss national social report has not been published by the statistical office, and a national level of living survey or a quality of life survey have not been implemented in Switzerland. Nevertheless, in the last years specialised social surveys, such as the Labour Force Survey or a health survey, were introduced. Switzerland is now on its way to modernising data production by introducing regular social reporting through social surveys, a social indicator system and regular social reports. The late and slow development of social reporting in Switzerland can be attributed to the specific type of liberal welfare state, keeping the state sector low, where social expenditures a nd therefore also government social services play only a limited role. In the system of welfare production associations, markets and families play a comparatively important role. Social reporting in such a liberal welfare state probably faces other organisational problems than it does in a developed welfare state. Furthermore, due to undeveloped public services, which could be monitored by administrative statistics, it would have been necessary to rely mainly on surveys in order to measure informal welfare production by the private sector.

In Luxembourg, social reporting developed by the end of the 1980s. First results of this institutionalisation became evident in the 1990s. On the organisational level the statistical office (STATEC), the main social science research institute (CEPS/Instead) and the General Inspection for Social Security (IGSS) co-operate in order to improve information on social conditions in Luxembourg. These three institutions together have published since 1994 resp. 1995 two information bulletins on the demographic and social development in Luxembourg. With respect to data production, the situation has improved a great deal due to the introduction of the Luxembourg panel (P.S.E.L.L.). This panel for the first time allows for analyses on the income situation, income inequality and poverty in Luxembourg. The panel furthermore allows for a comprehensive analysis of the living conditions of households and families in Luxembourg. This way Luxe mbourg has joined the group of countries with a well-established social reporting system. As a third element of a well-established social reporting system, a first social report was published in 1997, which was to give a "global and synthetic view on the society of Luxembourg". In 34 contributions seven living domains are analysed: population, human resources, employment and work, income, type of living and household budgets, "Cadre de vie" (housing) and social policies (CEPS/Instead 1997).

b) "Latin Rim"

In Southern Europe, social reporting developed only late and weakly. This peculiarity has several causes. First, the dictatorships in Spain, Portugal and Greece made an early development of the social sciences impossible. After the transition of these countries to democratic governments, social sciences made very fast progress; with the late establishment of the social sciences, social reporting was also slowly established. All in all, the take-off of social reporting in these countries (with the exception of Italy) only started in the 1990s. This demonstrates that the introduction of social reporting requires a tremendous infrastructural input which often requires a previous modernisation of the statistical system.

Thus, official statistics of Portugal was basically reformed in 1986 in order to comply with the statistical requirements after Portugal’s EC membership in 1986.² Attempts to establish official social reporting were only made after the statistics reform of 1986 – a first social report "Portugal Social" appeared in 1992 (INE 1992). The Institute for Social Sciences of the Lisbon University (Instituto de Ciências Sociais (ICS)) published in 1996 a "Report on the Social Situation, 1960-95" as book and disk (Barreto 1996). The time series data can also be accessed via the internet homepage of the institute and will be regularly updated. Time series are presented for 10 living domains.³

In Italy, private foundations, as CENSIS, e.g., were the first ones to engage in social reporting. CENSIS (1995) has published annual (!) social reports since 1967. Additionally, a shorter English version is produced. The national statistical office (Istituto Nazionale di Statistica - ISTAT) has since the late 1970s produced various titles which are similar to a social report, but the main input only came by the end of the 1980s. The probably most important project of the last years was ISTAT's implementation of the "Indagine multiscopo sulle famiglie". Social reports also exist for some provinces or for specific sectors, such as demography. The Italian situation – according to a national expert – seems to be rather fragmented.

In Greece, social reporting has only developed in the 1990s. Traditionally, official statistics in Greece (National Statistical Service of Greece (NSSG)) has a relatively weak position and suffers from too little acceptance in the population, low wages for employees, and therefore shortage of staff and insufficient productivity and "timeliness". Under these conditions official statistics was not and still is not in a situation to take the lead in social reporting in the country. Thus, the main statistical instruments for social reporting in Greece, such as well-established social statistics with a plurality of social surveys and comprehensive surveys, does not exist. The national social research institute (EKKE) could not and cannot fill this gap. After the reorientation of research priorities in the 1990s, EKKE is now systematically establishing social reporting. Thus the statistical compendium "Recent Social Trends in Greece&quo;t; was compiled (EKKE, forthcoming). Furthermore, EKKE plans to produce a data base with social indicators. While social research in Greece was dominated by an anthropological research tradition in the past, the focal point has meanwhile shifted to the more "modern" fields of social reporting (Diamandouros 1997).

In Spain (see box), the first "Panorámica Social" dates from 1975, and as late as 1994 (INE 1994) a second edition appeared. In 1991 the first edition of "Indicadores Sociales" was published (INE 1991). However, private foundations, such as the Fundación FOESSA or research centres, have published sociological reports since 1966. The decentralisation of the country as a result of the introduction of the autonomous communities ("Comunidades Autónomas") has not only led to the production of social reports for various of these autonomous communities, but also to separate statistical offices in some of these. Social reporting also spread in the big cities, as municipal social reports for Madrid and Barcelona, e.g., prove.

Ireland did no develop regular social reporting, although in 1979 a prototype of a social report ("Towards a social report", by the National Economic and Social Council (NESC)) was published. But this project was discontinued. Thus, a series of regular social reports or a socio-statistical compendium is not produced in Ireland, although a great deal of research on topics related to social reporting is undoubtedly carried out, mainly by the Economic and Social Research Institute (ESRI), the Institute of Public Administration (IPA) and the Central Statistical Office (CSO). The CSO, for example, did not organize a level of living survey, but organized a statistical database storing socio-economic macro data and indicators.

Summary: The late and – altogether – weak development of social reporting in Southern Europe can on the one hand be attributed to the belated democratic development in three of the four countries considered here. Additionally, the weak development of the welfare state is important (when measured by the proportion of the GDP, social expenditures in Southern Europe are below the average). Other actors in the field of welfare production, such as the family and relatives or other intermediary associations, play a more important role by comparison. Reporting on the contributions of these actors of welfare production to the population’s welfare is much more difficult in comparison than to produce administrative statistics based on a publicly administered social security scheme. The resistance of citizens to the investigation of familial and other transfers in these residual welfare states will probably be much more difficult to overcome.

The Southern European countries did not develop an explicit ideology of welfare research, be it politics of redistribution, or be it extensive research on subjective welfare. The approach of the Southern European countries depends far more on available data and therefore on the material provided by official statistics, because official statistics normally only provides information on material living conditions due to its orientation towards objective statistics. Only empirical social research has led to a partial correction of a predominantly objective social reporting.

(4) Living Standard Research in East European Transition Countries

In the East European transition countries the living standard research which was created during the socialist phase was more or less maintained. Since 1945, the East European countries have been permanently confronted with basic economic problems to guarantee the basic standard of living of the population. For this reason this type of social reporting will be labelled living standard research here, which differs strongly from the concepts of level of living research and quality of life research. Living standard research is oriented only towards elementary and objective living conditions, and not towards topics such as the subjective measurement of life satisfaction or the measurement of vertical redistribution ("distribution of shortage"). In the socialist countries of Eastern Europe, until the collapse of the socialist systems, quality of life research in the Western sense and level of living research of the Scandinavian typ e could not evolve. Social reporting developed mainly and nearly exclusively in the National Statistical Offices on the basis of only few main social statistical investigations. These main surveys were: regular surveys on household budgets and surveys on time use. Besides the population censuses, essential elements of social statistics were developed from administrative statistics, as the whole labour statistics, e.g.

Explaining Quantitative Differences in Social Reporting

Concerning the relationship between welfare state regimes and type of social reporting in European welfare states, the assumption could be made that welfare stateness and the importance of social reporting are interrelated. The hypothesis can be formulated that the higher the degree of welfare stateness, the higher the degree of institutionalisation of social reporting. In order to test this hypothesis I have tried to collect data on the introduction of four types of social reporting activities in Table 2. They are:

1. Year of introduction of a social report or of a socio-statistical compendium
2. Year of introduction of a comprehensive level of living or quality of life survey
3. Year of introduction of a special journal on social well-being
4. Year of introduction of a social indicator system.

Table 2: Introduction of Social Reporting and Welfare State Regimes

Notes: NA=not available. – est.=estimated.

Furthermore, countries have been grouped according to their welfare state type (cf. Esping-Andersen 1990). One can distinguish between five welfare state models:

1. The liberal model with UK, USA, Canada, Ireland and Australia.
2. The Scandinavian social-democratic welfare state model with Norway, Sweden, Denmark and Finland (Stephens 1995).
3. The continental corporatist social insurance model with Germany, France, Belgium, the Netherlands, Austria, Switzerland and Luxembourg.
4. The "Latin Rim" of Southern Europe, which by some authors, such as Ferrera (1996), is conceived as a special type of welfare state, but by others as, e.g., Katrougalos (1996) included in the continental model.
5. The former socialist East European countries which still are in the process of transition.

Table 3: Construction of Intensity Index of Social Reporting

Unfortunately, it was difficult to collect information on the years when these types of activities in the field of social reporting first started, and – far more frequently – such activities did not exist at all. Thus, only for social reports it is possible to calculate the mean year of introduction for the country cluster. The results are: in the countries belonging to the liberal model a social report was introduced earliest, that means in 1973. This is understandable, as the Anglo-Saxon countries were the pioneers in social reporting. Both the Scandinavian and the continental countries introduced a social report in 1976, on the average three years later than the Anglo-Saxon countries. The South European countries introduced a social report only in 1985, on the average ten years later than the Northern European countries or the continental countries. It must me said here that Spain and Italy – although belonging to Southern Europe – introduced a social report as e arly as 1975. In Eastern Europe several countries have not published a social report to date, but it can be estimated that the mean year of introduction will be after 1990. Therefore, these results do not support the above mentioned hypothesis that a relationship exists between welfare stateness and the early introduction of social reporting.

In order to construct a better index of intensity of social reporting, the following procedure was found. An intensity score of social reporting is defined, consisting of the three dimensions: (1) period of introduction of a social report, (2) availability of all four types of social reporting as in Table 3, and (3) regularity of social reporting. The first dimension "period of introduction of a social report" is subdivided into 6 time periods with scores ranging from 1-6; the earlier the introduction of a social report the higher the score. The second dimension "availability of all four types of social reporting" has scores ranging from 0-4, from none of these types to all of these types of social reporting. The third dimension "regularity of social reporting" has only three values, 1 for "irregular", 2 for "annual" and 3 for "several times a year". The summary score is built by adding the sc ore value of all three dimensions for each country.

The results of this measurement of intensity of social reporting can be seen in Figure 1. The results have some plausibility, as we see that Sweden scores highest, but Germany comes second, which is a result of the fact that all types of social reporting are produced in Germany. But all in all a North-South divide also with respect to social reporting intensity becomes evident, with the Nordic countries having the highest intensity, most of the continental countries being in the middle field, and the South and East European countries having the lowest social reporting intensity. As can be seen, there is regularity but also much variation.

If the question is posed again whether any relationship between social reporting intensity and degree of welfare stateness exists, we can correlate both dimensions, "welfare stateness" measured by the proportion of social protection expenditures in per cent of GDP, we receive Figure 2. Now, in contrast to Table 2, we find a clear relationship between both dimensions in Europe. Welfare stateness and social reporting are therefore interrelated, and we can predict from this result that if welfare stateness proceeds in Southern Europe, social reporting will probably proceed there as well.

Social Reports in Spain

(a) The Fundación FOESSA was founded in 1965 with the aim to present systematic information on the social situation of Spain. To this end the Foundation published, beginning in 1967, 4 sociological reports on Spain (Informes sociológicos, in 1967, 1970, 1975, 1980-83). The fifth report was published in 1994 (Juárez 1994). These reports give a comprehensive picture of Spanish society and can be considered to be the "main" social reports on Spain.

(2) The Centro de Investigaciones Sociológicas (CIS) was established in 1977 as an autonomous state agency attached to the Office of the Presidency and replaced its predecessor, the Instituto de la Opinión Publica (IOP), founded in 1963. The main task of CIS is to study the Spanish society and its evolution, mainly by conducting public opinion surveys. Thus far the Centre has conducted over 1,200 surveys which have been organised in a data base that is open for potential users. CIS organises such surveys at all stages of the research process, from questionnaire design to publication of the results. The results are published in series of monographic studies. Furthermore, CIS publishes the main social science journal in Spain, the REIS, introduced in 1978. CIS functions as a national representative in international research programmes and conducts the national surveys, e.g. the Spanish Family and Fertility survey of the United Nation s ECE.

(3) The Centro de Estudios Avanzados en Ciencias Sociales (CEACS) was established in 1987 and forms part of the Instituto Juan March de Estudios e Investigaciones. Its main task is to carry out postgraduate research and teaching in the field of the social sciences. Fundamental research is undertaken on the institutional framework for economic modernisation, the sociological and political implications of regionalisation and internationalisation, and problems regarding the redefinition of the welfare state. Research concentrates on Europe as a geographical and cultural entity. The Centre publishes Estudios/Working Papers and Theses.

(4) The Institute for Advanced Social Studies (Instituto de Estudios Sociales Avanzados (IESA)) was created in 1988 as a research centre within the Spanish National Research Council (Consejo Superior de Investigaciones Scientíficas (CSIC)). Its main aim consists in undertaking research in social and political sciences in a comparative perspective. Research focuses on 4 topics: social transformations and the welfare state; interrelationships between science, technology and society; problems of political organisation (convergence, federalisation, etc.); analysis of social participation and economic development. The institute publishes working papers and monographs.

(5) The Centro de Investigaciones sobre la Realidad Social (CIRES) was founded in 1990. Its main objective is to stimulate and multiply the quality of sociological investigation at Spanish and foreign universities and research centres. CIRES organises each year several investigations and distributes the data on diskettes to national and foreign research centres. Each year a voluminous report on the social reality in Spain (La Realidad Social en España) is published. CIRES has introduced its own social indicators system.

Notes

1. This point is made by Desrosières 1989.
2. The situtation is similar in Spain. Only in Greece a basic reform has still to be awaited.
3. 1 Population, 2 Health, 3 Education, 4 Employment and Work Situation, 5 Product, Income and Living Conditions, 6 Housing, Comfort and Welfare, 7 Social Protection, 8 Culture, 9 Justice, 10 National Budget and the State’s Social Function.

References

Barreto, António (organizer) 1996: A Situação Social em Portugal 1960-1995. Lisboa: Instituto de Ciências Sociais (ICS), Universidade de Lisboa.

Begué, Jean 1981: La création de la division Études sociales à l’INSEE, Courrier des Statistiques no. 20, 14-15.

Centro Studi Investimenti Sociali 1995: 29^o Rapporto Sulla Situazione Sociale del Paese 1995. Roma: CENSIS.

CEPS/Instead 1997: Recueil d’Etudes Sociales 1996. Luxembourg: CEPS/ Instead (=Document PSELL No. 100).

Desrosières, Alain 1989: Les spécificités de la statistique en France. Une mise en perspective historique. Courrier des Statistiques no. 49, Janvier, 37-54.

Diamandouros, Nikiforos 1997: EKKE – National Centre for Social Research, Athens. EURODATA Newsletter, no. 5, Spring 1997, 11-15.

EKKE forthcoming: Recent Social Trends in Greece, 1960-1993. Frankfurt/New York: Campus; McGill-Queen’s University Press.

Esping-Andersen, G. 1990: The Three Worlds of Welfare Capitalism. Cambridge: Polity Press.

Ferrera, Maurizio 1996: The ‚Southern model‘ of welfare in social Europe, Journal of European Social Policy 6, 17-37.

INE - Instituto Nacional de Estadistica 1991: Indicadores Sociales. Madrid: INE.

INE - Instituto Nacional de Estadistica 1994: Panorámica Social de España. Madrid: INE.

INE - Instituto Nacional de Estatística 1992: Portugal Social 1985-1990. Lisboa: INE.

Juárez, Miguel (dir.) 1994: V Informe Sociológico Sobre la Situación Social en España. Sociedad Para Todos en el Año 2000. Madrid: Fundación FOESSA.

Katrougalos, George S. 1996: The South European welfare model: the Greek welfare state, in search of an identity, Journal of European Social Policy 6, 39-60.

Stephens, John D. 1995: The Scandinavian Welfare States: Achievements, Crisis and Prospects. Geneva: UNRISD (=UNRISD Discussion paper 67).

Uusitalo, Hannu 1994: Social statistics and social reporting in the Nordic countries. In: P. Flora, F. Kraus, H.-H. Noll, F. Rothenbacher (eds.), Social Statistics and Social Reporting in and for Europe, 99-120. Mannheim/Bonn: InformationsZentrum Sozialwissenschaften (=Europe in Comparison, Vol. 1)

Vogel, Joachim 1997: Summary: Living Conditions and Inequality 1975-1995, 627-658. In: Statistiska Centralbyrån/Statistics Sweden 1997: Välfärd och ojämlikhet i 20-årsperspektiv 1975-1995. Stock-holm: SCB (=Levnadsförhållan-den, no. 91).