The seminal contribution of Roefie Hueting to economic science: 
Theory and measurement of 
Sustainable National Income
 
 

DRAFT

Thomas Cool, November 29 2000
Consultancy & Econometrics
http://thomascool.eu

Report TC-2000-11-01

JEL A00, Q00



 
 
 
 

"(...) my concern (... is ...) that the universally accepted compass of economic policy - changes in national income - is giving us the wrong signals about economic success."
R. Hueting, 2000
Summary





Roefie Hueting’s work on Sustainable National Income appears to constitute a seminal contribution to economic science. Aspects are: (a) Definition. (b) Freedom. (c) Environmental functions. (d) Using a model in statistics. (e) Numerous details. (f) Blockages. (g) Foundation .
 
 

Introduction





Roefie Hueting has been writing on economics and the environment since at least 1967. A reconsideration of his work leads to the conclusion is that it has a seminal quality and that Hueting may be counted as one of the major economists of our time - as worthy as each in Mark Blaug (1985)’s list. This paper gives an outline of Hueting’s work, and intends to clarify why we are witnessing an important achievement.

Hueting’s contributions concern the relationship of the indicators for National Income (NI) and Sustainable National Income (SNI) - where it may be observed that it was Hueting who defined that latter concept. It is important to see that Hueting’s work concerns economic statistics, both applied and with its theoretical foundations, so that there should be no confusion with economic policy making and future-oriented economic research. Hueting’s objective is to provide adequate information to the users of statistical data. These data are generally used in a future-oriented setting, but their value lies their immutable statistical quality.

After Roefie Hueting’s official retirement from Statistics Netherlands (CBS), a symposium was held in his honour in 1999, and the conference book will appear with Edward Elgar in 2001 under editorship of Ekko van Ierland, Steven Keuning, Jan van der Straten and Herman Vollebergh. It is the study of this material, in particular, that caused me to write this paper. It is my impression that other readers will arrive at a similar conclusion. 

It is useful to remember that Jan Tinbergen (Nobel Prize 1969) has always been a strong supporter of Roefie Hueting’s work. He wrote a foreword to Hueting (1974), and he once even tried to organise a supportive committee for a prize of the United Nations. We should think about such things indeed when we see what Hueting has accomplished.
 
 

Setting





To understand Hueting’s work, it is necessary to recall that national income accounting finds its raison d’ être in social welfare theory. This has been developed in the period 1930-1950 by economists like Jan Tinbergen (NP 1969), Paul Samuelson (NP 1970), Simon Kuznets (NP 1971), John Hicks (NP 1972), James Meade (NP 1977) and Richard Stone (NP 1984). The basic issue is to compare two points in time and to determine whether welfare has increased or not. Since the Bergson-Samuelson social welfare function (SWF) is not observed, income - that follows from the tangent hyperplane - can be used as a proxy, and observed market prices can be used to deflate to real values. The basic statistical challenge thus is not income per se, but the development of welfare. Observed market prices were used because of the assumption of optimality - whence tangency.

Especially since we are living in democracies, it would seem to be a safe assumption that the current allocation indeed is the optimum decided upon by society. This classical approach to national income accounting however runs into problems when one can suspect that the resources are not used optimally and income is not tangent to the SWF - as would be the case for the environment since the 1950s.

The reaction of ‘traditional statistics’ to this challenge has been along the lines ‘If people don’t act up on their beliefs, then we cannot measure it’. National Income is recorded at observed prices anyway, while separate indicators are provided on the state of the resources. Hueting’s answer has been to hold on to the classical notion, and to try to find the counterfactual tangent point. The difference in income then is a measure for the distance of the current economy from the optimal economy.

An analogy can help. Suppose that a medical doctor has been measuring the blood pressure of a client for some time, and has been indicating that things have been OK. However, at some point he notes that the client should do more exercises. The client objects, and says that the blood pressure still is OK, as it always was. The doctor then has to explain to the client that the proper concern always had been health in general, and that actually more aspects are relevant than just blood pressure. To help the client to make the switch, the doctor and the client henceforth can use both the blood pressure and the improved health indicator.

Indeed, modern economic agents and their parliaments and governments have appeared to be rather inconsistent in their opinions and actions on sustainability. They may state that sustainability could be a goal, but they don’t act in that manner. For example, a 1992 action programme of the European Union advocates "modification of key economic indicators, such as GDP, so as to reflect the value of natural and environmental resources in generating current and future incomes and to account for environmental losses and damage on the basis of assigned monetary values" (taken from Keuning (1992:9)). But the EU clearly is not adopting the required measures to achieve sustainability in reality, hence market prices are off, and hence the traditional statistical methods are useless for measurement of an SNI and for the correction which the EU asks. (1)

In this case the EU asked for an estimate, but as a good doctor, Hueting would measure the SNI regardless whether the client asked for it or not. In a sense the doctor is forced to make expert judgement anyway, since the client could not be fully informed or might be tempted to make biased estimates e.g. on the amount of exercise required. Hueting’s approach clearly causes the question how a Statistical Office can be the judge of social preferences. If our complex social fabric does not generate the proper information, how could statisticians do so ? The answer that Hueting has provided is that statistics basically provides information based on assumptions in all cases. The traditional measurement of national income at observed market prices assumes optimality - because otherwise it is not sufficient to decompose between price and volume effects only. The classical measurement would require assumptions about what other conditions would constitute optimality. The statistician thus always provides conditional information, and should be clear about those conditions. Hueting also concludes that publishing the two measures simultaneously would be best from an informational point of view, since this makes users more aware of the assumed conditions. This would indeed be the valid scientific approach.

National income accounting has appeared to be a sensitive political issue. The client has become preoccupied with the blood pressure count, and seems to have forgotten why national income was being measured in the first place. This has seriously complicated Hueting’s work, but he maintained an admirable integrity. The situation also caused him to be more explicit about why the assumption about current optimality is untenable and why the alternative assumption of sustainability would be an acceptable statistical yardstick.

Specifically, Hueting developed the notion of ‘blockages’ to show how statistics can deal with the situation. The idea is that sustainability can be defined objectively and could be estimated (by the techniques that he developed). Implied in the concept of ‘blockages’ is that people would respect the standard of sustainability - even if they nowadays don’t (are ‘blocked’). The resulting yardstick thus does not impose preferences, but provides information for the democratic process to be able to decide about actual adoption or not.

In policy making circles one can hear the argument that the benefits of calculating the SNI would not outweigh the costs of the exercise. I have collected some data on the actual costs of statistical measurement for NI, the Environmental statistics, NAMEA and SNI. These costs and the summary table are in the appendix. If we take the total costs of a national statistical bureau such as CBS - Statistics Netherlands as the yardstick, then the NI costs are 4.2% of its budget and the SNI according to Hueting costs 0.25% - thus a quarter of 1%. For comparison, the Consumer Price Index (CPI) costs are 1.3% of the budget. The reasons of these low estimates is that we are discussing high level statistical measures. These numbers do not contain all the surveys done and the low level statistics - which are produced anyway because of other reasons. Environmental statistics for example are already produced for reasons of health and agricultural policy. What NI, NAMEA and SNI do, is ‘only’ integrate the available data. A conclusion is that SNI indeed is costly, since it does only tell us what we already know, which is that NI is wrong and that the environment is doing badly. But if we want to know by how much NI is wrong then the price is only marginal. In that respect, this gives a situation where a small expenditure can cause lots of political upheaval, and perhaps this is a better way to understand the situation.

The above explains also how we could proceed to compare Hueting’s SNI with other indices developed by other economists. There are various such indices with the most prominent alternative being Herman Daly’s "Index of Sustainable Economic Welfare" (ISEW). (2)  Note that this current paper is not intended as a review of the literature on all indices, and certainly I would like to avoid the possible misunderstanding that this paper makes a value judgement about these alternatives. The work by other economists on indices on sustainability is relevant and deserves attention. It is also difficult to make comparisons when work on such indices has been based partly on the work of Hueting himself. Thus, to be sure, this current paper only intends to explain the contribution of Hueting’s work to economic science. However, the question about comparison arises naturally, and an indication remains useful how such a comparison could proceed. Hueting’s SNI has been grounded from the start in the system of national accounts as the base for national decision making, and he regards NI already as a welfare index, while other indices often call this system into question. Hueting tries to compare current NI with sustainability proper, and he thus excludes the income distribution and an issue like ‘work at home’ which topics generally are in ISEWs. Hueting’s SNI thus shares some properties with the alternative measures, but none has all properties, while some add more, and overall there remains a distinct difference. Let us then see what those properties are.
 
 


Hueting’s contribution





Hueting’s contribution consists at least of the following points.

  1. The development of the definition of ‘sustainability’ as a yardstick for economic performance, and the development of the ‘Sustainable National Income’ as the derived economic indicator.

  2. Note: ‘Sustainability’ had been longer around as a word and vaguer concept. Hueting & Reijnders (1998) refer to J.S. Mill 1876 for notions of stationarity. Hueting’s contribution however is the translation to modern economics.
    Note: This thus distinguishes clearly the scientific definition from possible acceptance as a policy goal. While sustainability appears to be imprecise since it does not clearly specify which species are crucial or which might become extinct because of natural causes anyway, it appears that the imprecision is statistically manageable, and that the yardstick can be applied in practice.
    Note: The SNI for Holland has been estimated by Harmen Verbruggen c.s. - to be published in Van Ierland (1999 draft, 2001 planned).
  3. The notion that it is freedom rather than income which is the relevant feature for sustainability. 

  4. Note: Amartya Sen (NP 1998, "Development as freedom" (1999)) has made the case for ‘freedom’ forcefully, but the idea has been with Hueting all along - and Hueting has both stated its theory and employed it in a practical statistical analysis.
  5. The development of the concept of ‘environmental functions’ and the statistical measurement of these. At Statistics Netherlands, Hueting has set a world standard of high quality statistics that uses the results of the natural sciences and biology and integrates those into an economic system.

  6. Note: These environmental functions are related to Von Neumann technologies, where one resource can be relevant for different activities. The standard Von Neumann model of course is linear, but with the natural sciences and biology there are all kinds of non-linearities.
    Note: Hueting has been the founding Head of the Dept. for Environmental Statistics, at Statistics Netherlands (the Dutch Centraal Bureau voor de Statistiek CBS). He has been the guiding force in setting up their world famous environmental accounts, explicity calling in for the expertise of the natural sciences and biology. Also the innovation of the NAMEA - the National Accounting Matrix including Environmental Accounts as originated by Keuning (1992) which approach is very useful and fortunately also very influential in statistics - has only been possible because of the results created by Hueting. The NAMEA though cannot replace the need for an aggregate indicator based on welfare - see point (g) below. The SNI calculated by Verbruggen op.cit. has been based on Hueting’s work.
  7. The notion that statistics and statistical observation of the past can be extended by the use of applied general equilibrium models to ‘backcast’ the distance in the past of the actual path of the economy from some optimal path. 

  8. Note: This is a major advance compared to the common thought that statistics is observation without theory and models. In the common view observations can be used to develop and test theories, but no more, while now theories and models are shown to be relevant in observation as well. There are precursors to this idea, for example in Robert Fogel and Douglas North (NP 1993 "For having renewed research in economic history by applying economic theory and quantitative methods to explain economic and institutional change."). Hueting’s advance is that he shows that this type of analysis is a natural part of the work that can be expected of a Statistical Office.
    Note: With an appeal to the ‘theorem’ of Ronald Coase (NP 1991 "For his discovery and clarification of the significance of transaction costs and property rights for the traditional structure and functioning of the economy") one can argue that environmental costs are already included in actual observations and current national statistics. According to Coase’s Theorem, property rights do not affect allocation, only the distribution of income. However, those ‘implied environmental costs’ are at current prices that do not reflect sustainable use. Thus a model is needed.
  9. Solving numerous details in actually implementing these issues. Valuation of non-market resources with reconstruction of ‘demand and supply’. The difference between ‘vertical supply’ and vertical standards. Problems of asymmetric bookkeeping. That environmental use enters as a cost and not as an addition to income, so that the SNI is lower than the NI. (3) Choice of the intertemporal welfare function - clarification that only the preferences of the current generation are relevant. Clarification on weak and strong sustainability. Identifying ecological risk factors, including the risks of assumptions on technology. Identification of the various points for sensitivity analysis.

  10. Note: Hueting’s treatment of technological growth shows how strikingly ‘statistical’ his approach is. When an SNI is being estimated for one year in the past, then his method accepts only the technology known in that year, since no other techniques have been statistically observed for that year. (Non-renewable resources, such as oil, however are allowed a path for substitution, otherwise they could not be used at all.) Hueting thus deviates from normal statistics in the use of a model and the issue of ‘blockages’, but remains a statistician in other respects. For other economists, whose frame of mind on policy making and technology is future-oriented, this is a crucial point to become aware of. 
    Note: Hueting’s (1996) Three Myths paper is a nice example of the clarification involved.
  11. The development of the notion of ‘blockages’ in the economic process and national decision making.

  12. Note: While a scientist easily runs the danger of stepping into the shoes of policy makers, Hueting can be admired for never having done so. In some of his texts he enlivens the discussion by telling about his personal motivation for example to become an economist and to deal with the environment, but he then clearly distinguishes this personal aside from the information generated for the decision maker. In fact, where other economists might be said to be rather lax with regards to the popular and political misconceptions about the NI indicator, Hueting sets a standard of scientific rigour for the quality of information.
    Note: James Buchanan (NP 1986 "For his development of the contractual and constitutional bases of the theory of economic and political decision making.") gives insights into the co-ordination problem, and John Harsanyi, John Nash and Reinhard Selten (NP 1994 "For their pioneering analysis of equilibria in the theory of non-cooperative games.") give insights into issues like the prisoners’ dilemma. Hueting’s contribution is to show that observation of such market failures can be a correct base for correcting statistical indicators.
  13. The development of the theory for the above and basing this theory on accepted notions of welfare analysis and the framework of national income accounting, and on Lord Robbin’s definition of economics itself as the allocation of scarce means over alternative ends.

  14. Note: It is important that a new contribution to economics can be related to basic theory. Hueting’s contributions do not diverge from the main stream but are directly in that main stream. This also makes for their powerful impact.
    Note: As said, the notions for national income accounting have been developed by for example Jan Tinbergen (NP 1969) and John Hicks and Kenneth Arrow (NP 1972 "For their pioneering contributions to general economic equilibrium theory and welfare theory."). Important have also been Wassily Leontief (NP 1973 "For the development of the input-output method and for its application to important economic problems.") and Richard Stone (NP 1984 "For having made fundamental contributions to the development of systems of national accounts and hence greatly improved the basis for empirical economic analysis.").
    Note: Hueting’s SNI can be found by imposing sustainability conditions upon a model that contains only reaction functions and no social welfare function. These reaction functions however could be integrated, at least in theory, and then give a social welfare function anyway. Alternatively, one assumes a social welfare function, and then derives the reaction functions. In both cases, the approximation of welfare by national income becomes superfluous, since now the social welfare function has been given. Nevertheless, there is much use in continued use of national income accounting.
     
     
     
Concluding remarks





Overlooking this list of achievements, one becomes aware of the involvement of so many other people and institutions with Hueting’s work. Statistics Netherlands has provided a crucial institutional setting, the Dutch government provided additional research funds and also funded the symposium in honour of Hueting’s work. Other researchers like Jan Tinbergen joined in at some moment, or provided criticism like Steven Keuning op. cit. or Wilfred Beckerman and Herman Daly and others in the symposium book, which criticism allows us to better understand the issues. And many more. Yet during all this time, it was Hueting himself who created this list of achievements.

It is useful to remark that this does not mean that all problems have been solved. Much research needs to be done. For example, where the research on the SNI according to Hueting has been done for one country only - Holland - the question now arises for other countries and the world as a whole. There is also the issue of the distribution of the resources of the globe to mankind as a whole, on which Hueting’s SWF draws a blank. There also is the question under what conditions societies will be willing or able to actually switch to sustainability. (4) For example, as more resources will become subject to ownership title, the environment will be included more and more into the normal economy. This however does not mean that the normal NI indicator would eventually become sufficient. Ownership does not imply sustainable use. However, this means that statisticians face the enormous task of separating the two uses, while society itself has to find ways, if it opts for sustainability, to find mechanisms that make the sustainable use also the optimal choice at the individual level.

But, whatever this new research, it is striking that it will build on all this work. As once stated by Edward Leamer: "The success of an economist perhaps should be measured by the amount of employment he or she generates for other economists." In that respect, Roefie Hueting is likely to have a good score as well.
 
 

Note: I have submitted this text to Hueting and he has indicated that, apart from my hyperbole, it gives a fair representation of the content of his work. It is important to note this, since there have been many misunderstandings about what this content actually is. Economists often have not understood the ecological aspects, the ecologists often have not understood the economics, while it also happened that science has been mistaken for politics. In addition, I would like to remark that I have found it an honour and privilege to write this paper, and I would like to thank Hueting for his kind attention and patience.
 


Appendix




One frequently heard argument is that it would be too costly to calculate the SNI, or that the benefit of calculating it does not outweigh those costs.

The costs are in Table 1. The Annual Report over 1999 by CBS - Statistics Netherlands gives total outlays of NLG 316 million, totalling the material costs and about 2300 FTE employees. We peg the exchange rate at the easy value of 1 $ = 2 NLG, and thus get a total cost of $70 thousand per FTE. The CBS Work Programme for 2001 provides detailed information per activity. I thank CBS for helping me retrieve all this information. I have averaged data where the Work Programme only gave Min and Max values. The SNI project by H. Verbruggen c.s. was budgetted at NLG 0.5 million, though it may well be that the researchers also used their own research time. Though dr. Hueting is retired and thus does not ‘cost’ anything, we count him in at average cost.

Measuring SNI costs about $ 390,000, or a quarter of 1% of total CBS outlays. To compare this with other statistics, CPI takes 1.3% and labour statistics take 3.3%.

The National Accounts Department requires 4.2%. This neglects all survey people and lower level statistical work, and just considers the work of integration - as we did for SNI. A similar work of integration like NAMEA and SAM takes about 0.5% - twice of SNI.

Given that the $ 390,000 tell us what we already know, i.e. that the NI is wrong and that the environmental situation is bad, the benefit/cost ratio indeed is low. But if you want to have an estimate of how wrong and how bad, the price does not seem to be so bad.
 
 

Footnotes




(1)  This holds even when the EU would adopt the measures gradually and would gradually approach sustainability: on the path of convergence the traditional statistical methods would be improper. These methods are improper, since they don't use a model to correct market prices.

(2)  “Friends of the Earth” at http://www.foe.co.uk/ allows you to manipulate an ISEW for the UK.

(3)   Keuning (1992:3) seems to contain this misunderstanding: “It can never entail that on balance something is substracted from NDP, as minimum enjoyment is zero.” If environmental input first had a price zero and then becomes scarce with a price, then nominally the new input cost should be substracted from the earnings attributed to it - while the real consumption point, which is the relevant issue, would be lower.

(4)   Here is my own topic of interest, see Cool (2000a) on stagnation in national decision making, in particular with respect to unemployment. Cool (2000b) discusses the SWF regime switch.
 
 

Literature


Blaug, M. (1985), "Great economists since Keynes", Harvester

Cool, Th. (2000a), "Definition and Reality in the General Theory of Political Economy", Samuel van Houten Genootschap, ISBN 90-802263-2-7, http://thomascool.eu

Cool, Th. (2000b), "The choice on sustainability: information or the meta-SWF approach 
to a shift of preferences", Report TC-2000-06-01/21, http://thomascool.eu/Papers/Environment/Sustain.html

John Hicks (1981), "Wealth and welfare" Collected essays on economic theory, volume I, Blackwell

Hueting, R. (1974, 1980), "New scarcity and economic growth", North-Holland

Hueting, R. (1996), "Three persistent myths in the environment debate", Ecological Economics 18, p81-88

Hueting, R. and L. Reijnders (1998), “Sustainability is an objective concept”, Ecologic Economic Commentary, http://www.atech.br/agenda21.as/ecologi (though many typos)

Hueting, R., and B. de Boer (2000), "Environmental valuation and sustainable national income according to Hueting", chapter for the book on the Hueting symposium - see Van Ierland.

Ierland, E. van, c.s. (1999 draft, 2001 planned ), "International Symposium Valuation of Nature and Environment", draft papers - to appear with Edward Elgar 2001

Keuning, S. (1992), "National accounts and the environment. The case for a system’s approach", occasional paper NA-053, Statistics Netherlands (CBS)

Samuelson, P. (1950), "Evaluation of real national income", Oxford Economic Papers NS 2

Sen, A. (1999), "Development as freedom", Knopf

Verbruggen, H., c.s. (1999), "Alternative calculations of sustainable national income according to Hueting" - see Van Ierland (1999).

Note: A discussion at UNEP can be found at http://www.unep.org/unep/products/eeu/ecoserie/ecos3/ecos33.htm


 

Table 1: Measurement cost of SNI compared to the cost of NI, Environment and NAMEA

1999-2001
FTE
$ million
% of CBS
All Costs / FTE = $ 70,000
1
0.07
 
       
CBS total
2300
158.00
100.00%
of which      
Labour statistics
76
5.19
3.28%
CPI
31
2.10
1.33%
National Accounts Dept.
96
6.56
4.15%
SAM, SESAME, NAMEA
12
0.82
0.52%
Environment
40
2.75
1.74%
of which SNI at CBS
1
0.07
0.04%
PM. Idem (SNI at CBS)
1
0.07
0.04%
PM. dr. R. Hueting
1
0.07
0.04%
SNI project by Verbruggen
4
0.25
0.16%
SNI total
6
0.39
0.25%

Sources: (1) CBS Annual Report 1999 and CBS Work Programme 2001, http://www.cbs.nl,
(2) SNI project, (3) 1 $ = 2 NLG