STATISTICAL COMMISSION and ECONOMIC COMMISSION FOR EUROPE
CONFERENCE OF EUROPEAN STATISTICIANS

Forty-third plenary session
(Geneva, 12-15 June 1995)

CES/1995/R.18/Add.1
27 March 1995

• ANNEX I. EMPLOYMENT (STATISTICS CANADA)
• ANNEX II. INCOME STATISTICS IN HUNGARY
• ANNEX III. CHANGING FAMILY STRUCTURE (UNITED KINGDOM)

The political, economic and social upheaval in Europe in the 1990s has increased the need for information at all levels of society. First priority has been given to political and economic information, but it has become increasingly obvious that the social and developmental aspects of change are equally important.

This Declaration by the Bureau of the Conference of European Statisticians describes the need for statistical information in a democratic society and illustrates its importance by examples in the European and North American context.

The Conference of European Statisticians is a standing intergovernmental body whose members are the heads of the national statistical offices of the countries of Europe and North America. Its secretariat is the Statistical Division of the United Nations Economic Commission for Europe, and it is a subsidiary body of both the UN/ECE and of the United Nations Statistical Commission.

This is a revised revision of the paper that was presented to the Preparatory Committee for the World Summit for Social Development in draft form (Document A/CONF.166/PC/20/Add.17). It will also be submitted to the June 1995 plenary session of the Conference of European Statisticians.

INTRODUCTION

1. Statistics are the eyes and ears of policies, the non-partisan and science-based input into the eclectic process of decision-making, the lifeblood of a vigilant democracy. The provide the factual foundation on the basis of which public opinion can crystallize along the lines of possible alternative routes for development.

2. The Conference of European Statisticians addresses the World Social Summit in order to ask greater attention from policy makers and the public to the vast and largely unexplored potential role of statistics in this field, and at the same time offer its commitment and expertise for participation in the renewed efforts of national and international communities to enhance policy action for social progress.

3. The three annexes to this declaration illustrate some of the possible contributions to debate which statistics can make on employment (Canadian example) income (Hungary) and social integration (United Kingdom).

THE HISTORICAL RECORD

4. The past fifty years have witnessed a marked expansion in the contribution of statistics to social and economic policies.

5. To varying degrees and with the help of international organisations, many countries have well-developed national accounts, population censuses and agricultural censuses, income and expenditure household surveys and labour force surveys, as well as many other essential statistical data.

6. These sources are indispensable. The population census (or census-like data derived from population registers) has, over the decades, proved a great asset in physical planning, in social policy-making and in debates concerning population policies. For example, it has been used to derive indicators for small areas for the purposes of resource allocation to depressed urban and industrial areas.

7. Household income and expenditure surveys have been exploited to measure poverty, to assist in assessing the effectiveness of the social protection system and to simulate the revenue and redistributive implications of fiscal proposals.

8. Labour force surveys have been indispensable for manpower planning, for developing labour market policies and for the formulation of social policies concerning the unemployed. To a lesser extent, there have also been panels or longitudinal studies, either through administrative records or through household surveys, for the purpose of investigating dynamics, processes and transitions, and to shed light on those factors that appear to be associated with specific favourable or unfavourable outcomes. Data sources derived from administrative systems of governments have been useful in a wide variety of fields including social protection, health, education and crime.

9. However, these developments generally have not kept pace with the need for relevant timely, reliable and comparable statistics for policy formulation and monitoring. governments have often to navigate in the dark. Moreover, within the broad body of statistics, social statistics have tended to lag behind their economic counterpart. The best guess of the number of homeless in the European community is between 3 and 6 million persons in an age when the number of whales in the oceans have been carefully estimated and is closely monitored.

10. The preoccupation of governments with the economy has given a big boost to the development of economic statistics. Essential for sound economic management, they have received more attention and are more advanced than social statistics in most parts of the world. Indeed, sometimes social statistics tend to be seen as subservient to economic indicators.

11. But experience tell us that GDP cannot be the only basis for designing social policy instruments. Even economic goals cannot be pursued using solely economic instruments since economic problems often have their roots in wider social perceptions, people's attitudes and social psychology.

12. Even those social statistics that have become established can encounter difficulties in surviving. In many countries social statistics have sometimes suffered from cuts in social expenditures. Yet, the optimal allocation of scarce resources requires more information, not less, as a basis for assessing the relative merits of competing priorities. Moreover, the cost of social statistics, indeed all statistics, is minuscule compared with the cost, in terms of money and human suffering and welfare, of policy mistakes in the absence of hard information.

THE CHALLENGES AHEAD

13. The political and social contexts have altered drastically in Europe and, indeed, world-wide. Central and Eastern Europe is undergoing a structural transition. The Maastricht Treaty will present countries of the European Union (EU) with great opportunities and challenges. On a wider plane, there are the inexorable processes of globalisation and increasing international exchanges. At the same time, in many areas of the planet, uncertainty, fear and insecurity bring about an erosion of solidarity and collective identity. These radical changes are having far-reaching consequences, touching every corner of people's lives. Against this background, social statistics would lose touch with reality unless a new strategy is developed.

14. Economic and social realities are becoming increasingly integrated. It is widely recognized that economic performance is strongly affected by policies in the areas of education, health, design of social safety nets, and probably several others. Yet, much of this understanding is unquantified and heuristic - hence does not support the depth of understanding essential for the sound design and careful monitoring of policies.

15. Furthermore, most developed countries are struggling with deficit problems and are engaged in major policy reviews of social expenditures - typically the largest component of most governments' budgets. In developing countries and countries in transition, the struggle involves structural economic adjustment with a human face and the recognition that social spending constitutes an important investment in human capital.

16. Yet, it is a paradox of fundamental consequence to public policy, that while far-reaching efforts are underway to design or redesign social programmes, our collective empirical understanding in this area is at an elementary level. Few countries can claim to know what they are buying for their expenditures on education; what outcomes are we expecting our education systems to deliver; what "levers" are available to governments to improve these outcomes; what are the most cost-effective "levelers"? All these questions could be repeated in connection with heath policies and programmes, and they apply equally to the design of effective social safety nets.

17. With or without empirical support and understanding, budgetary and social pressures are forcing governments to undertake far-reaching social policy design changes. It is our deep belief that a better understanding would lead to quantum improvements in social policy and programme design. Such an understanding, if it is to be useful, must be based on quantitative information. Given the low level of development of social statistics in most countries, the rate of marginal return is likely to be very high indeed: a tiny fraction of program costs invested in the development of policy - relevant social statistics will yield a huge improvement in our collective understandings and insights.

18. The annexes provide some illustration of how statistics, which are currently available, can shed light on social issues. The illustrations are from Canada on employment, from Hungary on income and from the United Kingdom on changing family structure which reflect upon social integration.

CONCLUSIONS

19. Statistics should explicitly try to achieve two objectives of maximum public policy interest:

1. To monitor the achievement of outcomes of social policies and programmes (e.g. changes in unemployment levels, income distributions, population health); and

2. To identify those factors that appear to be associated with specific (desirable or undesirable) outcome and which lend themselves to intervention through social policies and programmes.

20. The task involves the development of policy relevant conceptual frameworks and corresponding measurement systems. this task is intellectually most challenging. A real collective partnership is therefore required among official statisticians, policy makers, public administrators working at various levels of government (including international and supranational) and external academic and research institutions.

21. The Conference of European Statisticians invites national and international organizations to pay priority attention to the achievement of these goals, and urges national governments to provide to the national statistical offices, on a evolutionary basis, the resources required to do so.

ANNEX I. EMPLOYMENT (STATISTICS CANADA)

1. In western industrialized societies employment has generally been viewed as participation in the waged labour force. The goal of the employee was economic security while the goal of the employer was a stable and available labour force. Gender was an important variable as generally men participated in the waged labour force while women provided domestic labour. Both often provided volunteer labour in the community. Demarcations between domestic labour and waged labour were relatively clear and they were primarily gender driven.

2. In the last quarter century in these societies, relationships between waged labour and unpaid domestic labour have changed significantly due to the shift to higher educational attainment of women (particularly post-secondary education) and the much greater participation of women in the waged labour force. This has led to demands for new services, such as day care, and new employee benefits such as maternity and other family related leave. Data suggest that women who enter the waged labour force tend to continue to provide disproportional share of the domestic labour in their families. This has led to demands for a more family-compatible work place in the form of flexible hours of work, part-time work and doing paid work at home (also known as telework). In addition, the market, quick to recognize opportunities, has responded with products and services designed to meet new needs (e.g., micro wave ovens, frozen and other quick to prepare meals, maid services, laundry services, etc.)

3. Employees, both women and men, have had to learn to juggle the demands of waged, domestic and volunteer labour.

4. While waged work continues to be the cornerstone of income security, the need to juggle competing demands for time has led to a concern with the notion of meaningful work. That is, that work (regardless of whether it is waged, domestic or volunteer) should have some intrinsic value in the sense that the worker has the belief that his or her effort results in some socially relevant good or service. In effect, there is the need to balance economic security against satisfaction with how one is employed within the totality of waged, domestic and volunteer activities.

5. Compounding this balancing act are concerns about having the right skills to compete in the waged labour force with the resulting need to devote some of the scarce remaining hours to training. Further concerns relate to the availability of social supports if waged work cannot be secured and to planning for economic security after retirement.

6. This, of course, is all taking place in the context of rapidly changing technology of production and shifts in world trading patterns, both of which have eliminated many employment opportunities while at the same time creating new opportunities for those with the right skills.

7. Rapidly changing technology has also impacted upon employers, as globalization (with its increasingly competitive markets, world-ride recessions) and mounting public debt. The response of employers, seeking a competitive edge, has been to re-engineer and restructure, contract out much specialized work and strive for just-in-time delivery and a right-sized and right-skilled labour force. Increasing employment in the service sector and a need on the part of employers for greater flexibility in scheduling work has led to more increasing demand for part-time workers. While part-time work sometimes meets the paid employment needs of those carrying heavy family responsibilities, part-time work also frequently comes with a reduction in, or elimination of, employee benefits and pensions compared to full-time wok.

8. Historically, through cross-sectional data we have generally had a good profile of the characteristics of the labour supply. The same cannot be said for the demand side where we have had precious little information. While through time-use studies we are beginning to gain some insights into domestic labour demands, we know relatively little about the demands for volunteer and waged labour. Nor do we well understand the effects of the shift from goods production to service production. We need to better understand the impact on waged labour demand of globalization and changes in automation, computerization and information handling.

9. The cross-sectional supply-side data, however useful they may have been, can no longer alone provide the information needed to make business decisions, formulate government policy or evaluate programmes in an economy and society that is changing so rapidly. Cross-sectional data reveal only the net change in what is being measured, they tell us nothing about the flows underlying the net changes. They portray the situation of workers at a point in time but do not provide information on the experiences of individual workers over time. We are thereby limited in our capacity to measure cause and effect. Longitudinal data are needed to facilitate an understanding of dynamics of labour supply and demand. They are needed to better understand relationships between the demands for waged, domestic and volunteer labour and their respective demands upon an individual's time. We need to better understand the trade-offs between an individual's economic security and satisfaction with their waged, domestic and volunteer activities. We need to better understand the relationship between human resource investments such as education, training, health care and other social supports and outcomes in terms of economic self-sufficiency, health status and quality of life.

10. Finally, we need to develop a model for incorporating our understandings of the relationships and dynamics of labour supply and demands in order to find the means for equitably distributing income, wealth and meaningful work. Only through such understanding can we hope to eliminate or reduce reliance on unemployment insurance and government provided welfare on the part of the able-bodied. And only through such understanding can we concentrate scarce resources on the support of those who are unemployable by reason of illness, infirmity or disability and the training of those who are unemployable by reason of inability.

ANNEX II. INCOME STATISTICS IN HUNGARY

1. Statistics on household income have two different approaches to measure its level, structure and, partly, its distribution: macro and micro. The macro statistics, which are an integrated part of National Accounts, traditionally are based on the following data sources: financial reports of enterprises; tax data; social security statistics; bank statistics; and central and local state budget.

2. In recent years there have been no major changes in the data sources, but the implementation of most of the recommendations of the 1993 SNA resulted in many changes being made to the conceptual framework. 1992 is the first year that the new SNA will be used as a basis for the compilation.

3. At the micro level, income distribution statistics have had some dramatic developments recently due to changes in the data sources used.

4. In Hungary survey based income statistics have a long history, for the first Household Income Survey (HIS) was conducted in 1963. Since then this survey has been repeated every five years on a regular basis until 1988, which was the last published HIS. The surveys have a large sample size, around 20 thousand households, and since the early 70s they have been based on a uniform sampling frame. The method of the survey was well suited to the economic characteristics of household income of that period, in that the two main important sources of income (wages and salaries, and income from household plots) were measured very accurately. As a result, data from these surveys provided highly reliable income statistics, even though they were conducted at only five year intervals.

Individual distribution of net per capita income in decile shares and selected inequality measures

5. In the mid-1980s there was a strong increased demand for more frequent distribution estimates related to the acceleration of economic and social changes occurring in the society. The government had a stronger interest in obtaining that kind of information on a regular basis. As a result, the Hungarian Central Statistical Office (CSO) set up the Hungarian Household Microsimulation System, which was part of the world-wide boom of the microsimulation development process. 6. In the short history of our microsimulation model almost all of the major application fields can be found, like tax simulation, projections, living conditions issues and distribution based analyses. For these applications, a micro income dataset was produced on a yearly basis, which made it possible to capture the distribution, inequality changes and poverty estimates. 7. The most recent HIS used 1991 as the reference year, instead of 1992 was originally planned. The results of the survey did not meet with earlier expectations, for fairly unreliable figures were obtained from it. Both the level and structure of the distribution were strongly biased. The unexpectedly low quality of the Income Survey results of 1992 and the analyses based on them led to conclusions being drawn concerning micro income statistics, and forced the CSO to set up a new strategic plan to handle this uncomfortable situation. 8. Tax data: The uses of tax files for statistical purposes is allowed by the Statistical Law which was passed in 1993. An access was built up to the administrative tax records, and a sample of 100 thousand tax units (individuals) was selected from the total of 4.5 million tax returns. Despite the problems in tax data (some important items of income are not subject to tax; some income items are taxed at the source; and in many cases a free choice can be made between the possibility of itemised deduction of costs or deductions based on presumption of the law), it seems a very useful supplementary source in the compilation of data on household income. The tax files are available every year and do not entail any extra resources, and the differences in conceptual terms could be narrowed with some additional effort. 9. The Panel Survey: Panel surveys have been introduced in recent years in many countries, and they have a special feature in periods of rapid economic and social change such as the one Hungary and other transition countries are experiencing. In this type of survey similar measurements are made on the same sample at different points in time. Sample elements are kept in the sample for the duration of the survey and follow all members of sampled households. Therefore, it is appropriate for capturing changes over time. The main analytical advantages of household panels are they make a crucial distinction between transitory and persistent characteristics; they can be used to study gross flows across important boundaries such as those defining the middle class; they can be used to conduct studies of intergenerational consequences of phenomena such poverty and dependence; and they estimate changes surrounding events of interest. 10. The Hungarian Household Panel Survey was designed with these purposes in mind, and the first wave was conducted in 1992 by the Social Research Informatics Centre and the Budapest University of Economics. The CSO has also been involved in this joint project since 1993, when the second wave was launched. 11. The Household Budget Survey: The Household Budget Survey (HBS) has the longest history among household surveys in the CSO. Although it has undergone major changes and improvements in recent decades, it still has the widest range of topics among household surveys. The principal focus of interest of this survey is the expenditure side feeding data for the consumer price index calculation, consumption pattern analyses, and estimation of household consumption in the National Accounts. However, in many countries this survey is also an important source of statistics on household income distribution. For the time being, the HBS has not been used for income estimation. Nevertheless, it could serve as a possible basis of estimation if post stratified HBS data were used. Therefore it is necessary to improve the importance of the annual income inquiry and to settle the system of adjustments, (basically, the system of reweighing). 12. The Labour Force Survey: The Hungarian CSO has been conducting a new household survey since January 1992, using the experiences of the pilot survey carried out in 1991, in order to obtain timely information on the labour force status of the Hungarian population. The Labour Force Survey (LFS) provides quarterly information on the non-institutional population aged 15-74. The aim of the survey is to provide information on employment and unemployment according to international statistical recommendations. Income related supplementaries are planned for the third and fourth quarter of 1994. ANNEX III. CHANGING FAMILY STRUCTURE (UNITED KINGDOM) 1. Divorce up, marriage down, births outside marriage increasing, more lone parents, more cohabiting, more working mothers the family structure in the United Kingdom has changed considerably in recent years. 2. The number of divorces in the United Kingdom has more than doubled in the twenty years to 1991 while the number of marriages has fallen (Chart 1 and Table 2). Nearly 10 per cent of marriages in the United Kingdom now last less than two years. 3. The number of children born outside marriage in the United Kingdom has increased dramatically since the early 1960s. In the ten years up to 1992 the proportion of live births outside marriage more than doubled to nearly a third of all births (Chart 3). However, there is some evidence that nowadays many more of these births may be occurring within stable relationships three quarters of births outside marriage in England and Wales in 1992 were registered by both parents. 4. The UK government have a target of reducing the rate of conceptions in girls under 16 by at least 50 per cent by the year 2000, which may play some part in reducing these figures. 5. Births outside marriage have been increasing in other countries as well all countries in the EC have seen at least a doubling in the proportion of births that are outside marriage in the last 30 years. 6. The proportion of dependent children living in oneparent families has more than doubled since 1972 18 per cent of children live with just their mother and a further 2 per cent with their father (Table 4). This probably reflects both the increase in births outside marriage and divorce. 7. In response to family structure changing in this way, the Child Support Agency was set up to ensure that when children are not brought up by both their parents appropriate financial contributions are made by the absent parent. 8. However most dependent children in Great Britain, four fifths, still live in a family with two parents, but they are now less likely to have two or more brothers or sisters than 20 years ago. 9. Just over 1 million dependent children, around one in twelve of all dependent children, lived in stepfamilies in Great Britain in 1991. It is estimated that over 5 per cent of all children would become stepchildren in married couple families and 7 per cent would become stepchildren in cohabiting couple families at some stage before their sixteenth birthday. 10. The number of adoptions in the United Kingdom has fallen considerably since its peak in 1968 there were just over 8 thousand adoptions in 1991 less than a third of the number in 1968. The decline in the number of adoptions may be due to it now being more socially acceptable for women to have a baby outside marriage than it was 20 years ago. As part of the UK government policy in relation to children, a white paper has been issued on proposed reforms on the adoption laws. 11. There has been a growing trend in the number of mothers who work. One in five of all mothers of children under 16 in Great Britain were working full time in 1993 and nearly two in five working part time (Table 5). 12. Part of the Children Act 1989 led to an extension in the framework for day care services for children. These services are particularly important to working parents and the new wider age range covered by the framework reflects the needs of parents of school age children for whom day care services may be required outside the school day. The government has also announced its intention to provide nursery education to all 4 year olds whose parents wish it. Chart 1 Table 2. Marriages: by type United Kingdom Thousands

Source: Office of Population Censuses and Surveys Chart 3 Table 4. Percentage of children1 living in different family types

Great Britain

                                    1972      1992



Couple with

1 child                             16        17

2 children                          35        38

3 or more children                  41        26



Lone mother with

1 child                             2         5

2 children                          2         7

3 or more children                  2         5



Lone father with

1 child                             0         1

2 or more children                  1         1



All dependent children              100       100

    1	Dependent children: children under 16 or aged 16 to 18
and in fulltime education, in the famil unit and living in
the household.

Source: Office of Population Censuses and Surveys

Table 5. Mother's economic activity status: by age of youngest dependent child, Autumn 1993