Econometric Contributions to Public Policy

Econometrics, as an organised discipline within the wider field of economics, has not yet celebrated its fiftieth anniversary: the Econometric Society was established at the end of 1930 and its journal Econometrica began publication in 1933. The aim of the Society, as stated in its constitution, is to advance economic theory in its relation to statistics and mathematics. The meaning of this phrase was elaborated by the editor, Ragnar Frisch, in the first issue of Econometrica. He pointed out that ‘Experience has shown that each of these three view-points, that of statistics, economic theory, and mathematics, is a necessary, but not by itself a sufficient, condition for a real understanding of the quantitative relations in modern economic life. It is the unification of all three that is powerful. And it is this unification that constitutes econometrics.’ A similar point of view was expressed by the Evaluative Committee for Econometrica (Paul Samuelson, chairman, Tjalling Koopmans and the present writer) whose report was published in Econometrica for April 1954. This committee was concerned with the kind of journal that Econometrica should be, its size and the kind of editorial organisation it should have. As a prelude to making recommendations on these issues the committee defined econometrics as ‘the quantitative analysis of actual economic phenomena based on the concurrent development of theory and observation, related by appropriate methods of inference.’ The committee went on to suggest that ‘editorial policy should be aimed even more, and by additional means, at increasing the proportion of space devoted to empirical studies meeting a high standard of quality and relevance.’

When econometrics first emerged in the inter-war period, one of its earliest concerns was to help improve economic policy. Now, almost fifty years later, econometricians can take pride in the success of their subject, which no economist worthy of the name can overlook today. But they have to admit that the impact of their work upon the conduct of economic policy is still somewhat limited. This is one of the reasons why the programme of this conference is of such particular interest.

The measurement of any excess of the wage rate of union workers over nonunion workers no longer stimulates much controversy in the US. The measure-ment methods have been standardised and there is now a broad consensus on how these union wage effects have moved over time and even how they differ as between black and white or male and female workers.² This end to con-troversy has undoubtedly resulted for a variety of reasons: for one thing, as George Johnson [6] has remarked, the mere existence of trade unions is no longer a serious question of public policy. In addition, the quality of the measurement devices and of the microeconomic data available to researchers has made it possible to eliminate some of the ambiguity of measurement present in the earliest studies.³

Ragnar Frisch, the spiritual father of the generation of Norwegian economists to which I belong, had a favourite idea which he returned to often in the early post-war years. As he saw it, scientifically based economic planning in Norway — and Frisch was very much for scientific planning — would have to develop as the result of joint efforts by three centres: university economists would provide the theoretical tools, economists at the Central Bureau of Statistics would organise data to fit the theory and turn the theoretical tools into workable models, and finally, the group of economists within the Ministry of Finance responsible for planning would put the models to practical use. As we shall see, this vision by Frisch proved to be a fairly accurate anticipation of what was later actually to happen. Even today, macroeconomic planning in Norway is organised pretty much according to the pattern he foresaw.

This paper is about the development of a large-scale multisectoral dynamic model of the British economy and its potential use for economic planning.¹ Such a model is intended as a tool for managing the economy to achieve goals of both a short-term and a medium-term nature. This is economic planning in its broadest sense, involving taxation policy, the planning of government expenditures, monetary policy, the operating rules for nationalised industries and direct controls over resources. A model is useful at several points in the planning process: at the start, it can be used to explore alternative goals and the alternative paths towards them; when the choice is narrowed, the model provides the detailed implications of a particular strategy as regards industrial output, employment and investment for example; and as the plan is carried into action, the model can be used in the monitoring process in order to take into account changes in relationships and exogenous variables and to revise tire plan for these changes.

The formulation of public policy appropriate to a country’s economic circumstances depends on an accurate assessment of that country’s economic prospects. Because of the growing interdependence of economies today, it is of increasing importance that such assessment allow for the transmission of economic fluctuations from one country to another.

The model of foreign trade described in this paper is part of a system of models being developed at the University of Maryland. The models, under the general title INFORUM — a contraction of INterindustry FORecasting at the University of Maryland — are being used for input-output forecasting and simulation. The original INFORUM model is of the United States. It is described in [1] and recent additions to it are described in [2].

The distributional consequences of price changes and of government policy on the prices of consumer goods have received considerable attention in the press in the last few years. Not all unpopular price policies have caused the extremity of reaction which occurred in Poland in 1970 and June 1976. In both cases proposed food price increases caused political upheaval which in 1970 led to the resignation of Party Chief Gomulka. To be sure, these price increases, actual or proposed, had not only distributional consequences but also an effect on average purchasing power. However, in Poland in 1970, the policy package would have added only around 2 per cent to the official price index and included other sweeteners. Therefore, the distributional consequences were the dominating ones. Another example is the increase in the prices of fuel, electricity in particular, in 1975–76 in the UK. A great deal of concern was expressed on behalf of pensioners and other low-income groups — see National Consumer Council (1976) and Committee on Nationalised Industries (1976). Under pressure, the government in July 1976 decided to extend the means-tested benefits available to households eligible for Supplementary Benefit and Family Income Supplement to include fuel bills.

A renewed interest in the theory of optimal taxation has emerged in recent years. The theory of optimal taxation generally has followed two disparate paths: the theory of optimal commodity (or indirect) taxation and the theory of optimal income (or direct) taxation. Only occasionally have attempts been made to integrate or compare these two approaches. While each has an interesting history in economics, much important work is of recent origin and, I shall argue below, much remains to be done if we are to use optimal tax theory as a guide to the formulation of tax policy.

A major economic fact of the last fifty years is the increasing importance of the state. In the case of the Canadian economy, the total government budget grew from 10 per cent to 46 per cent of gross national product. Two technical consequences which one cannot over-emphasise stem from this stage of affairs: The first consequence has pushed economists to revise their views on optimal taxation: a modern state cannot limit itself to uniform taxes and lump-sum taxes. Because of the second consequence optimal taxation formulae cannot be derived from partial equilibrium analysis but rather from a generalised Pareto optimum framework.

A prevalent feature of social and economic planning in most developing countries, until quite recently, has been the particular emphasis on growth maximisation and optimum resource allocation. The rationale behind this approach to development planning has been the view that the maximisation of the rate of growth of output not only entails more material well-being for the nation as a whole but also results in a better distribution of income in the long run.

I have been asked to discuss certain aspects of some recently proposed methodologies for project appraisal. At the same time Professor Stone did not fail to remind me of the title of this conference. Accordingly, my task, as I see it, is to present an account of a few issues that have been raised in recent writings on project evaluation, and to pay particular attention to some of the measurement problems that appear rather naturally to arise in the course of conducting a project appraisal exercise. This will then lead me to discuss several approximation methods that have been suggested for measuring the social values of various goods and services involved in an industrial project.

This year is a triple anniversary. With great fanfare the United States is celebrating the 200th anniversary of its Declaration of Independence. With considerable solemnity but less fanfare economists all over the world are celebratinglhe 200th anniversary of The Wealth of Nations. And with almost no fanfare whatsoever benefit-cost analysts are celebrating the 40th anniversary of their craft, which had its official inception in the Flood Control Act of 1936.¹

Even though national planning in France began in the immediate post-war period, it was only with the IVth Plan (1962–5) that a genuine regional dimension was introduced with the establishment of regional allocations for public investment (franches opératoires régionales).¹

In the first part of this paper we shall develop some theoretical and methodological considerations to clarify the role of econometric analysis in regional and urban policy. The shortcomings of the theoretical models will be outlined. In the second part we shall propose two interconnected models for regional and urban planning (the second model being in fact a general framework for building operative models). A few considerations on the flexibility requirements of urban planning will conclude the paper.

In this paper we are concerned with the contribution of econometrics to environmental policy. The contribution of econometrics to public policy lies in the fact that it provides decision-makers with the possibility of conducting experiments before taking political measures. Of course, during the phase of model-building and testing of hypotheses much insight is gained into the mechanisms of the economic system, but the ultimate purpose of econometrics is simulation.

According to the announcement for this conference the general subject area of this paper is ‘Environmental Policy and Public Goods’. The term ‘public good’ is, of course, not a synonym for ‘public welfare’, but rather an antonym for ‘private good’. And what is a ‘good’? In the sometimes peculiar jargon of economics, the word connotes nothing of inherent virtue: basically, a ‘good’ is anything tangible that conveys a ‘service’ to a ‘consumer’. A consumer, by the way, does not physically ‘consume’ anything except air, drinking water and food. For the rest, consumption is essentially a process of exhausting the utility of goods by extracting services from them. The material object, itself, is discarded.

At budget time in the US Federal Government’s Office of Management and Budget in the Executive branch, and in the Congressional Budget Office in the legislative branch, the question is probably asked: is the investment in applied social research (concerned with major social problems) worth it? Should we spend more, the same, or less? What should we cut, what should we augment?

Japan’s rapid economic growth since the mid-1950s has brought about remarkable improvements in living conditions and reduction of wage differentials,but it has also been pointed out that it has caused imbalances such as excessive concentration of population, external diseconomies such as pollution, traffic congestion, and the like.