The national accounts as a tool for analysis and policy; past, present and future

An overview is provided of the history, logic, merits and limitations of the national accounts. Past In the second half of the seventeenth century the first estimates of national income were made. These estimates served clear purposes, like demonstrating that a revision of the English tax system could raise sufficient resources for waging a war with Holland or France. The number of estimates and their frequency gradually increased, in particular since the First World War. Major innovations, like the development of the sector accounts, input-output tables and the arrival of the first guidelines took place in the thirties and forties. These innovations were stimulated by the Keynesian revolution and the development of macro-econometric model building. Since the Second World War national accounts statistics have become institutionalised and standardised, i.e. they are regularly compiled by national statistical institutes, Central Banks or Ministries and are based on one universal set of multi-purpose concepts and classifications. About a decade ago, the third generation of international guidelines on national accounting was introduced. In comparison to the first guidelines, the scope was drastically extended, e.g. by the inclusion of prices and volumes, balance sheets and input-output tables. However, the basic concepts, like the production boundary have hardly been changed in fifty years. Since the collapse of communism, no separate guidelines and concepts exist for the (formerly) communist countries. The European Unification has stimulated a revolution in national accounting, in particular with respect to the development of jurisprudence and the improvement and harmonization of the estimates. The European Union was not only aware of the possibilities of the national accounts as a tool for European policy, but acted also as a critical consumer. The latter was vital for generating this revolution in European national accounts practice. Present National accounts statistics are a miracle come true: all over the world, very incomplete, imperfect, heterogeneous and partly outdated data are transformed into a complete, consistent, internationally standardised and up-to-date overview of the national economy and its major components. What is the magic behind this miracle? The universal model National accounts statistics are estimates of a universal accounting model for describing, analysing and managing national economies. This universal model is not a neutral description of economic reality: it is focused on what can be readily observed in monetary terms, it contains substantial transformations of what can be observed and is based on a specific way of labelling economic reality. Different choices would have resulted in a different picture of economic reality. The major biases of the universal model merely reflect the natural focus of a regular economic statistic, i.e. a focus on what can be readily observed in monetary terms. This explains why the economic importance of unpaid household services, leisure time, pollution and tax expenditures is ignored. Including such major analytic elements in the basic concepts would seriously endanger the statistical purpose of the universal model. Furthermore, it would also drastically decrease the relevance of the universal model and its major aggregates for important other data needs, e.g. those of budgetary and monetary policy. The substantial transformations of what can be readily observed are required in order to look –on behalf of analysis and policy- through the complex, chaotic and many different economic and institutional realities. The specific concepts used are the result of many implicit and explicit considerations with respect to relevance, reliability and comparability. They are also influenced by the need to agree on one set of concepts, even if arguments are not sufficient to settle the score. The universal model incorporates two types of perspectives on the national economy. Firstly, it describes the national economy in terms of its major components (sectors/industries, various types of flows and stocks and several economic processes). This is the general perspective of the universal model. However, the universal model describes also each major component in a macro-economic context and in relation to the other major components. These are the specific perspectives incorporated in the universal model. Seven major specific perspectives can be distinguished: 1. Non-financial corporations (business accounts); 2. Financial corporations (monetary policy, financial markets); 3. Government (budgetary policy, government finance); 4. Households (personal income, wealth and consumption); 5. Rest of the world (balance of payments); 6. Industries (production, employment and input-output analysis); 7. Other (e.g. the environment, human capital and the welfare state). The old and most recent history of the national accounts demonstrates that these specific perspectives, and in particular that of the government and its relations to the national economy, are a major motivation for compiling national accounts statistics. The universal model is an ingenious and very practical product. The various perspectives demonstrate that it is a synthesis of various types of applied economic analysis, e.g. business accounts, Keynesian type of analysis, input-output analysis and index number theory. The universal model also reflects three hundred years of experience in compiling national accounts statistics. Nevertheless, the universal model can still be improved in various ways. The universal model stresses the importance of flexibility, but does not make an explicit link to the seven perspectives. For each of these perspectives, standard supplementary concepts, like government expenditure and revenue, entrepreneurial income after tax and net worth to the owner, are very relevant and can be easily derived from the basic concepts of the universal model. For each of these perspectives also the importance of prices, volumes, real values and key-ratios should be stressed. Also some changes in the basic concepts are proposed e.g.: - Other non-market output, e.g. of the government, should be valued including a net operating surplus by amount of an opportunity interest on the capital invested. - The purchase of consumer durables by households should be recorded as capital formation instead of as final consumption expenditure. Imputed services of owner-used consumer durables should be recorded by amount of its consumption of fixed capital plus an opportunity interest on the capital invested. These imputed services are consumed as part of the final consumption expenditure of households. These changes increase the relevance of the national accounts from an economic theoretic point of view. Furthermore, also various changes in the presentation of national accounts statistics are proposed, e.g. a more systematic distinction between actual and imputed flows and the limitation of the number of accounts. The national measurement process The universal model can not be estimated directly. It should first be translated into an operational model for a specific country during a specific period of time. This involves interpretation of the universal model in view of the national economy and further specification of the concepts, detail and scope. For example, which units belong to the sector government en to what extent can price- and volume measures take account of changes in quality? The operational model decides to a substantial extent what is actually measured. Differences in national operational models are therefore a serious threat to international comparability. The operational model is estimated by combining very heterogeneous and incomplete sets of data; the latter include national accounts estimates for previous periods and frames of reference for grossing up and combining data, e.g. a business register or a population census. The major estimation tools are accounting identities, plausibility checks and assumptions. The estimation process is influenced by environmental factors like skills (e.g. skills in combining data and making plausible assumptions), resources (e.g. resources for compiling good price-statistics, for maintaining a reliable business register or for compiling national accounts statistics) and policy (e.g. a mixed strategy of continuity or a preference for prudence and stability). Uses of the national accounts National accounts statistics are important for economic policy and analysis. Four different roles are played by national accounts statistics: 1. description and object of analysis; 2. tool for analysis and forecasting; 3. tool for communication and decision-making; 4. input for alternative accounts, budgetary rules and estimates. As a description and object of analysis, national accounts statistics are unique. They define and measure the national economy and its major components. They make the sizes and developments in national economies all over the world visible and put them into quantitative terms. As a consequence, the world economy, the national economies and their major components can be monitored and analysed. Not all descriptions are suited as an object of analysis. National accounts statistics are partly built on assumptions. Assumptions are essential in combining and completing the basic set of data. Plausible assumptions are even to be preferred above unreliable data. The more encompassing, up-to-date, detailed and reliable the basic data set, the smaller the role played by assumptions can be. By changing the definitions of the universal model, the role of assumptions can be increased or decreased. This also changes the usefulness of national accounts statistics as an object of analysis. For a proper analysis of national accounts statistics, sufficient information should be available about the operational concepts underlying the national accounts statistics, their reliability and the role of major specific events and institutional circumstances. Furthermore, users should have sufficient knowledge of the logic, merits and limitations of national accounts statistics in general. As a tool for analysis and forecasting, national accounts statistics are built on three very useful stocks of knowledge: the universal model, the operational model and the national compilation skills. Ignoring the national accounts as tool for analysis and forecasting can result in serious conceptual and statistical pitfalls. However, as a tool for analysis and forecasting, the national accounts have also clear limitations. For a proper use, national accounts statistics should often be rearranged or be supplemented with alternative concepts, alternative data and equations describing economic behaviour. As a tool for communication and decision-making, national accounts statistics are unique. They serve as the universal facts and language for thinking and communicating about national economies and their major components. They provide new opportunities for decision-making by providing information about major macro-economic developments, by providing explicit targets for many types of policy and by providing price-indexes for inflating contracts and agreements in real terms. However, for a proper and optimal use, knowledge of the merits and limitations of national accounts statistics is essential. National accounts statistics serve also as an input for alternative estimates, accounts for non-national accounting purposes and policy targets. As an input for alternative estimates, official national accounts statistics serve as a very cheap, well-designed, universal semi-manufactured product. These alternative estimates may reflect fundamentally different perspectives on the national economy, e.g. welfare-measures. However, some of the major alternative estimates are best labelled as non-official national accounts statistics, e.g. by providing much longer time series. National accounts can also serve as a benchmark or source of inspiration for accounts for non-national accounting purposes (e.g. for the bookkeeping systems of municipalities) and policy targets (e.g. in defining the budgetary ceilings for state expenditure in the Netherlands). In this way, the national accounts actually extends its scope as a tool for communication and decision-making. The future On the brink of the twenty-first century the world is undergoing dramatic changes. Four trends (globalisation, regionalisation, automation and more-market oriented government) are changing the data needs and possibilities of the national accounts. Globalisation and regionalisation will increase the political use of national accounts figures. This reinforces requirements on international comparability and standardisation as evidenced by the European experience. Globalisation and more-market oriented government will pose serious difficulties for the quality and completeness of the statistics and administrative data sources used for compiling national accounts figures. A pro-active response is essential for statisticians. The possibilities for national accountants may be increased due to automation, putting minimum standards on the inputs for the national accounts statistics to increase their international comparability and advances in national accounts compilation techniques. More-market oriented government can stimulate the development of more efficient, effective and attractive national accounts statistics that appeal to a wide range of data users. However, it can also result in cutting down the resources for national accounts statistics and its major inputs below a minimum-level. National accounts statistics will then be trapped: resources are not enough to meet a minimum standard of reliability, to make national accounts statistics more attractive and to find new users; the potentials of the national accounts statistics are then trapped. In order to meet these challenges and dangers of the future, the efficiency and the national accounts as a product should be improved. This can be done in various ways, e.g. by: - Developing an international long term strategy for improving national accounts compilation techniques; - A better balance between the compilation process and the outputs; - More modules for specific purposes. Some of these could contain internationally standardized supplementary concepts, like entrepreneurial income after tax, government expenditure and revenue or household income in cash. They may also contain modelling results, as modelling is often essential for a balanced view on the national economy. - An international, EU-like, program for increasing the reliability and comparability of major national accounts variables, e.g. Domestic Product and the volume growth of Domestic Product. - Investigation of user practice and the lessons that can be learned from that. - Supplementing national accounts statistics with information about their meaning and reliability, e.g. about the operational concepts, the data sources, compilation methods, the size of the differences between successive estimates and the results of various types of sensitivity analyses. In order to clarify the value-added of national accounting and to fight wide-spread illiteracy in national accounting, marketing and education should be taken up seriously, preferably by an international long term strategy and by making use of all the possibilities of internet.