After 20 years of mass unemployment:

Why we might wish for a parliamentary inquiry

Thomas Cool * December 18 1990


A synthesis of economic theory is presented, the solution to unemployment is restated, the intellectual need for a parliamentary inquiry is established, and as an example to such inquiry the performance of the Centraal Planbureau is evaluated.


In Holland, mass unemployment persists already for about twenty years, and will continue to do so for many years to come. Economists agree on the obvious solution, the reduction of labour costs. But for some reasons our decision making process doesn’t generate that decision. Policy measures that are taken, actually are troublesome, like the creation of a Centraal Bureau voor de Arbeidsvoorziening (CBA), or the recent ‘temporary and red tape’ ten percent subsidy on minimum wages (WLOM). The policymaking situation is analyzed In a more formal manner, to allow for more abstract reasoning. This requires a social welfare function, an income redistribution function, and a production function (for the unemployed cq. subsidized workers). In fact, we might attain the goals of high growth, price stability, full employment and a just income distribution, by means of monetary, fiscal and subsidy instruments. The conclusion however is that the present policy sclerosis derives from insufficient interest in and information about the form and location of those mentioned functions, and lack of interest In optimization itself; and this again may be caused by institutional weak spots. A review of the issue and of the policymaking process could be beneficial and actually logically needed. Among others, this would include a review of the Centraal Planbureau (CPB), that has not properly endogenised government behaviour in its models, projections and analyses. It is suggested that such review would be a task for parliament; and the logic for a so-called parlementaire enquete is compelling. Clarity on the issues is essential too for the European debate and our advice to the Eastern European countries.

*) The author is an econometrician at a government agency that has some involvement with the economic policy making process; the article expresses his own views only. This paper is adapted from a presentation at a parallel session at the conference in honour of prof. W. Albeda "The future of industrial relations in Europe" June 7-8 1990, Maastricht, The Netherlands

"There may be a communication problem. Using the words of Cairncross, again: 'Policymakers as a rule are slightly deaf: there is too much noise’. In other words, there is a need to raise the ‘signal-to-noise’ ratio. One cannot overemphasize the importance of the packaging — the simplicity and saleability of ideas and the need to pursue these in clear and non-technical language using simple diagrams, etc. Moreover, often the more important contributions of economic advisors are in the clarification of the most basic and simple (simple only for us, professionals) concepts (...)"

Michael Bruno (1990) p276
"In den eersten stoot pat."

Hans Ree (1979)

(That is: what is the use of shouting "you’re deaf!" to a deaf person ? Or, this paper might increase the noise level so that messages for the better are not heard.)



After 20 years of mass unemployment

4 Introduction

7 First insight: economics

11 Second insight: public choice

15 Third insight: an institutional management example

18 The three insights combined

22 Applying the synthesis

25 A Basic Income ? Not economically needed !

27 Intermediate conclusion

Why we might wish for a parliamentary inquiry

28 The policymaking process

32 First example: a failed austerity program

33 Second example: a parliamentarian turned minister

35 Third example: the Central Planning Bureau

42 Conclusion: the end of ‘laissez parler’ ?

46 Appendix: Some formal elements

53 Footnotes

56 Abbreviations

57 Literature

60 Summary of main propositions

(All translations from Dutchare mine.)


The basic argument can be fond in Cool (1990a). Since then, some other authors have refreshed our knowlegde about the problem area, like De Neubourg (1990), writing the ILO study on the Netherlands; OECD (1990); Graafland (1990b); Summers (1990); CEPS (1990); Schuyt (1990); Brittan (1990); Leontief (1990); the Social and Cultural Planning Bureau SCP (1990) reporting on ‘overproduction of policies’; De Kam & Nypels (1990) accusing labour and employer unions of sabotage; Roebroek (1990) calling Dutch social security a farce; the Albeda (1990) plea for a review of the SER (Social-Economic Council); VN (1990) interviewing De Vries, the Minister of Social Affairs end Employment, finding him acknowledgeing some of his former parliamentarian and majority party leader lack of perception; Pelkmans quoted (1990) as saying "Parliament Is generally incapable of making a correct economic analysis"; the CPB outsider Van Velthoven (1990) critical, and the CPB insider Don & Van den Berg (1990) rosy, discussion of the CPB. Just this October I got to read Piore (1987), who provides a beautiful and supporting analysis, that ‘unemployment’ is as much a concept and product of our social institutions as a result of economic hardship. Most of these studies have been taken along in the main text. In my opinion, these, and the 1991 government Budget and the saddening ‘Collective Agreement’ of thIs October and a collection of Ph.D. Theses recently published, strengthen my case.

I found Visscher (1990) explaining what kind of parliamentary inquiry I’m pleading for: "Up to thirteen years ago (this) had the purpose to enable parliament to gather its data for law-making."

This may remind us of the 19th century parliamentary inquiries into labour conditions. The more recent inquiries (on shipping and building subsidies, and on fishing) have provoked much political turmoil and scapegoating of those responsible for policy and its failure, but such is not at issue here. At issue is only that science tells us that without broad political will deriving from a proper understanding of economics, mass unemployment will continue.

But allow me one single comment on this aspect of policy responsibility. It is nowadays being widely understood that policies should be clear, controllable and fraud-resistant. I have only dared to present my own proposals after checking that they satisfy these demands. However, many who have been responsible for past policy mistakes are still in responsible position, and they are now likely to be the first to say that policies should be clear, controllable and fraud-resistant, and then to downgrade and reject my proposals on that account. This is not an invitation to lock me up for being paranoid, but just a short introduction into the magic of Public Choice, and a reminder that science thrives on unbiased Judgment.

An appendix on ‘the Arrow (1951) impossibility theorem’ was part of this paper originally, but it has been given a life of its own as Cool (1990b), when the existence of a SWF is established.

1 thank some reviewers for pointing out the more abstract and potentially more controversial arguments. Having done my best with these, I invite the reader to regard my reasoning as the sharp logic and intuition of an experienced meso/macro econometric modeller anno 1990. Though of course, here I present only an economic analysis, and not an econometrically developed study yet.

After 20 years of mass unemployment


In 1972 unemployment in the Netherlands almost doubled, to an alarming level of 116 thousand persons, about 3 percent of the labour force (CPB (1973)). Since then the phenomenon of mass unemployment has been with the Dutch. Projections for the near future show its persistence, with a large hidden unemployment in the disability scheme WAO and in subsidised agriculture; and there is a large potential female supply. Van Stiphout (1989), ESB 3739, the CPB projections, and my on experience with long term projections, suggest that present-day arrangements will not result into a solution. Thus: past, present, and future unite into a conservative estimate that there are at least twenty years of continual mass unemployment — only not to say 25 or even 35 years. Such round numbers invite some reflection; and it is useful In itself. 1)

A first question is whether the Dutch themselves are so happy with their state of affairs. A second question is how the Dutch experience can be of relevance to other nations.

On these points this paper will present a general analysis. The reference to Holland remains useful, since we can check how idiosyncracies fit Into the general framework, and also since we want to remain practical, i.e. don’t want to abstract from action.

Roughly the following answers will be defended below.

The Netherlands are a prosperous, highly educated and apparently rather generous nation. Holland has found a good solution for mass unemployment: It has settled in it, with a rather admirable show of adaptability; no sclerosis here !

Especially striking is the inflation in the notion of ‘mass’, where nowadays many perceive 8 percent of unemployment as no longer being ‘massive’. With mass unemployment being accepted as one of the unavoidable implications of the welfare state, it becomes a welfare item ‘to keep people meaningfully busy’. For example: to keep young people from the streets, to combat rising criminality. But at this point the economist has to advance the alternative analysis, namely that ‘business’ ought to be provided by the economy itself. Hence the Dutch may find themselves in a suboptimal equilibrium. In the end this welfare state might crack from the (forced) free rider overload.

How does this relate to other countries ? Well, very simply, those countries are not so rich. Especially in the Eastern European nations the change to a mixed economy appears to be difficult partially because of the threat of mass unemployment. It will obviously not do, to suggest that one could learn from the Dutch model as an example of the ‘paradise’ that is far ahead for everybody. We have to present a serious alternative.

The most relevant lesson of twenty years of mass unemployment in the Netherlands is, that the Dutch model gives some tensions of itself, and that the Dutch can only contribute to others, especially in Europe, when they get their act together. Others are interested, only to hear in what directions the Dutch are looking when trying to solve the matter in a more fundamental way; for example to redesign the welfare state along Swedish/Japanese lines. (See Standing (1988), Keizer (1990), ACB (1990), De Neubourg (1990).) And trying to get the Dutch act together, in my terms the saddlepointing of society, would mean something to be done by the unions of employers and employees, and also by Dutch parliament. Obviously, parliament enters our discussion, since we are discussing fundamentals. A prime suggestion to parliament would be, to review its past actions, hopes and policies, and to restate its position. Such review would also involve a critical look at the economic policy making process.

The following analysis derives strongly from my own experience with practical full-fledged meso/macro economic models applied to the longer run (cf. CPB (1990)). Auerbach & Kotlikoff (1988), and everybody else, clarity that the use of such models is ‘something special’. Economists do not have a lab, but in such models we finally have a tool for sharpening our intuition. The interdependencies in those models bring about thoughts, which specialists might overlook. At the same time, one is forced to a phone-call and a discussion with a specialist, when an ‘anomaly’ arises and a result is not warranted by blunt intuition. A multitude of projections has to be considered and rejected, before a wise selection is presented to colleagues and the body politic. And thus it was, in the course of such operations, noting the endurance of a very unfavourable inflation-unemployment ‘tradeoff’, that I found myself, apart from changing the model, also performing a Tinbergenian analysis of goals and instruments. And, in the context of the long run, this appeared to involve, logically though quite unfortunately, also a reconsideration of some rather established conventions of macro-economic policy making, and also a reconsideration of the Western welfare state. (I found: there cannot be a serious longer run projection, without some serious change of the system.)

First insight: economics

On one hand it must be admitted that the economic explanation of mass unemployment is not simple & straightforward. In the seventies we saw a renewed discussion between keynesians and classics, e.g. in the Netherlands the discussion about the Den Hartog and Tjan (Vintaf II) model. Among the various explanations for stagflation there were both the disequilibrium and the rational expectations approaches; though these fall into the keynesian/classics divide. In the eighties we saw more attention for supply and demand in the labour market itself; and other theories for a reduced government and strengthened market forces.

On the other hand the impression ought to be avoided that differences would be very large. Two insights appear to be accepted rather generally. The first is almost a litmus test for the true economist: that a job can only be maintained when real labour costs do not swallow up too large a part of the product for too long. 2) The second insight can likely be derived from the first: that unemployment will affect the least productive employees, whereby, if necessary, the more productive employees will take their places (‘crowding out’).

Requiring employers to pay more than people actually produce, generates unemployment, especially at the lower wage brackets. Indeed, various economists have argued regularly that present-day unemployment could disappear even at the present average wage level, when gross labour costs in the lowest wage brackets were reduced strongly. (And possibly, if net differences, between unemployment benefits, minimum wage and other wages, were increased, additionally.)

Note the three concepts involved: net income, labour cost, and productivity. These have a distribution across people, and these can be related to each other; e.g. an income distribution somehow reflects a (required) productivity distribution. Then the problem can be rephrased as a policy/instrument mismatch: the socially required income distribution is imposed as a cost distribution, and this appears to conflict with the actual productivity distribution.

References abound. Tinbergen (1956) already warned about the dangers of the minimum wage. Van Schaaijk (1983) is a gem. Meade (1985) writes: "I would remain a pessimist if I had to abandon all hope that the emphasis in the minds of the man and woman in the street and at the work-bench might ultimately be shifted away from wage-setting onto fiscal and welfare measures as the appropriate means for influencing the distribution of income." Layard (1986) explains about the selective reduction of labour costs that it is immaterial whether it is called a subsidy or a premium and/or tax reduction. (Indeed, as a rule one would subsidize people and tax products.) Drèze (1987) mentions it as a major point for the unemployment issue. Bakhoven (1988) has developed the case quantatively for the Netherlands. Bron (1989) describes a Rotterdam Community project. See HelliwelI (1988). CPB (1988), Van Reijn (1989), Van Eijk (1989), Van Opstal (1990), and OECD (1990). De Neubourg (1990) features labour costs in pp 140-141. CEPS (1990) p13 writes: "Previous reports of this group have called attention to the need for greater differentials in labour costs (...) if unemployment is to be kept at acceptable levels. In practice incomes policies, minimum wage legislation, the financing of social security, and the activities of trade unions all seem to have acted to squeeze differentials since the start of the seventies". I myself 3) have also tried to semantically unburden unemployment, by interpreting it as ‘subsidized search, but also a rather inefficient way of looking for work’; for in that manner it becomes a managerial problem of cost reduction of an ‘open-ended arrangement’, by finding more efficient ways of search. And Adriaansens (1990), of the Dutch Scientific Council on Government Policy: "But then, there is the recognition, that it may well be the socio-economic structure of our welfare state, which itself causes the impossibility of a larger participation in work. (...) Where unemployment is concentrated in the lower qualified categories, a structural redistribution of social security premiums would reduce labour costs for those categories and increase demand for that kind of labour."

For the UK, we find Minford & Rastogi (1989) pleading for workfare. Samuel Brittan (1990) points to a poverty and unemployment tax-trap in the UK, and: "A genuine market-based approach might involve a Basic Income for all, payable as a tax credit, which would be withdrawn through the tax system once people began to earn money from work. Although a full Basic Income scheme is not now affordable, partial moves are possible in that direction even under the present system." In my opinion a basic income so administered is less efficient than alternatives, but the reduction of labour costs is evident. (See also below.)

Incidently, OECD (1990), CEPS (1990) and De Neubourg (1990) present whole shopping lists of causes and cures, and the latter warns that there is no panacea or easy fix. The reader has to remind himself that these various measures concern various objectives, like growth incentives, flexibility in general, participation, etcetera. In a sense, these studies suffer from an overkill of objectives and analyses. This creates too much noise. It is far better to apply Occam’s razor, and select the item of labour costs — which item already in itself suffices for a complete overhaul of existing administrative procedures. Later, when unemployment is solved, and the authority or economists publicly enhanced, then we could discuss the marginal improvements from other measures, and in a much more favourable climate.

So, apart from details, the consensus about real labour costs is very strong. Among economists, unemployment has been quite solved for a long time. Since resistance against the solution has not been quite self-evident, especially where it concerns gross but not net wages, the viewpoint has arisen that we are dealing not with an economic but rather with a sociological problem —which might explain why most economists no longer study the subject. We have the paradox that a welfare system, that was meant to reduce the burden of unemployment, is a prime cause for it. The solution to this paradox then must be sought in institutional management. That this isn’t done, is a socio-economic phenomenon indeed.

Second insight: public choice

In an academic posture, we would be interested in trying to get a grip at the stated socio-economic phenomenon. This brings us to the body politic, and public choice theory. Note that our step to public choice is necessary. Logic forces us.

An analogy is the following. In our full-fledged economic model there occurs an exogenous variable called world trade. We can improve our model, by disaggregating and having variables like world trade by SITC group. But, it may also be that we don’t have to do this, since we are quite confident about how we have modeled the dependencies. Then, in order to improve our model, our attention shifts to really explaining that exogenous variable, by building a world model.

In the same way, we can say that we sufficiently understand the workings of the labour market, and now we must explain the policy input: how this market is created and managed.

An inroad would be to wonder who really cares about unemployment. There is no wide gap between economists and laymen. As said in the introduction, the Dutch society of laymen has adapted very well to life with mass unemployment. Also the employers and unions seem to get along in an agreement to disagree, and also quite silently, for, as soon as one party would utter ‘unemployment’, the other would utter ‘labour costs’, and after a while everybody has heard it and gets bored; and as long as wages in Europe keep rising, the Dutch can raise those too.

One gets the impression that the employer unions rather don’t want that the present income distribution is frozen by a more adequate tax and subsidy system; and likely they hope that the continued unemployment results in ever lower wages. The employee unions are against the recycling of unemployment benefits into work subsidies, calling it ‘false competition’: they apparently want to keep their own wages as high as possible by keeping out such competition. These positions curiously mirror each other: the different interests result into an opposite weighing of risks. One can obviously doubt whether properly understood selfinterest is the case here. For truely, what makes the competition false, is that people are banned from work by high tax & premium levies.

The descriptive term ‘hysteresis’ adds little to our understanding of what is actually happening: that the institutional framework has been shaken to its roots and then apparently is no longer able to satisfy the actual needs. The increasingly popular ‘insider-outsider’ theory (Lindbeck & Snower (1988), Gelauff e.a. (1990), Graafland (1990)) looses much of its novelty appeal when it is realized that this is just ‘employed-unemployed’ in other words. As for the econometrics of hysteresis: any test can only render the data it departs from: a specific path of unemployment. Rather than thinking that such simple statistics can provoke basic results, it is better to go beyond such data, and look at the causal chain of events, and at the system at work here. Thus, to really explain hysteresis, in addition to, or rather instead of, having partial labour market theories, we must explain the policy sclerosis of the third party involved, which allows for such dereliction. Indeed, government can provide itself with an alibi by allowing ‘free bargaining in the labour market’ to take the blame.

The only people in the Netherlands who still seem to care about the situation, are government and parliament. The difference in care derives of course from the division of labour between voter and representative. For an individual the welfare state is a blessing: with relatively few costs large risks are insured and less fortunate members of the family are being helped; and criticism is useless since it will not further change. Only government and parliament take the total view, they have to evaluate costs and benefits, and they have the power to act. Only they would be interested in activating the massive Dutch inactivity, the incentive being a large (though long term) benefit to the budget deficit and possible electoral gains. But here again the problem can be stated trivially in the boring form of high and partially conflicting aims and a limited budget. Another tautology: everything what parliament does, is well done. However, the law of the excluded middle tells us, it must be here that the care somehow dithers away.

As said, and seen again, we must turn to public choice theories, which tell us about the limitations of government. I am commonly hesitant about public choice theories, since they are too sceptical about honest government efforts, and hence they do not allow for the proper discourse with the democratically elected decision makers. In this I follow Tobin (1990). But eclectically, sometimes public choice is useful, and this case would be an example.

I can refer to the Bruno quote of Cairncross above, about the deaf policy makers. Of course it is more that just deafness. Pen (1989) writes: "politicians have more goals than those which they proclaim. There is a system in their hesitance about expounding their true preferences; e.g. the battle against unemployment receives a stronger verbal support than the for many politicians more dominating wish to maintain the prerogatives of the unions of both employees and employers."

After the war, the following items were officially proclaimed as the goals of economic policy: (1) as much growth as possible, (2) full employment, (3) price stability, (4) an external balance, (5) a just distribution of income. These can be found in any Dutch schoolbook on the subject. But this list obviously only expresses a small part of a quite complex whole. For example, what has not been mentioned, but which is of obvious importance, is the maintenance of the status quo — in terms of a certain institutional setting and the mentioned lobby prerogatives.

This relates to the Tlnbergen (1956) distinction for instruments of economic policy: quantitative, qualitative, and reform.

Thus, at issue is a (public choice) explanation why we don’t succeed in having a reform that allows the attainment of more generally accepted policy goals.

Third insight: an institutional management example

The Dutch system of unemployment insurance has been based traditionally upon the business cycle or conjunctural assumption. That is correct, as the argument above has shown that structural unemployment can be avoided. The Bakhoven suggestion of a wage cost reduction, referred to above, did not imply a change of system, but it was a policy measure in terms of the system; the suggestion was the very reason why the system was correct.

On the other hand, almost 20 years of mass unemployment were a fact, and it was decided that this implied the need for some action. There arose two main reasons for only partially adopting the Bakhoven suggestion. Of these two, the temporary increase in the budget deficit was not the basic drawback. 4) The basic point was that the high marginal tax rates, which the plan rendered, were thought to be a ‘poverty trap’.

Note that these reasons for rejection are problematic in themselves. Especially the ‘poverty trap’ is not convincing. As Adriaansens (1990) notes, in Holland the progressive tax system is offset by a regressive social security system: and the introduction of some steeper marginal rate might be more to the general liking. Also, dynamically, incomes grow, and this could be used. In general, governments have an incremental way of policymaking. They try the integral approach of using some encompassing system only occasionally; and when that is done, like with the Oort tax system review in the Netherlands, the result often is unfortunate for other reasons. So in this case, much can be said for the incremental policy approach, to simply have this ‘poverty trap’ for a few years, and mend it later, over the years. 5)

The Bakhoven suggestion however being rejected, the search for solutions has resulted into a change of the system (building upon earlier discussions). So nowadays structural unemployment can be fought more directly, as of January 1 1991, by means of a Centraal Bureau voor de Arbeidsvoorziening (CBA). 6) Two things are amazing. Firstly the Ministry of Social Affairs hides its failure of the last decennia by proclaiming a ‘success in decentralisation’. (Depressing reading is Van Bochove e.a. (1990), about the organisations GSD and GAB, both residing under that Ministry, who started to cooperate a few years ago... Depressing is also that these researchers deem it necessary to report as a major research finding, that money incentives work.) Secondly the CBA still operates within the context of the very welfare system, which causes that structural unemployment in the first place. In economic terms this means that the shadow costs of a more integral change apparently still are too large. In normal terms this means that one tries to prevent the sinking of the ship by pumping rather than by plugging the hole. To prevent a disappointment from too high expectations, two things then are needed. The first is to emphasize that pumping is very useful as pumping goes. The incremental way of policymaking has its advantages; in so far as creation of a new institution like the CBA is incremental. The second is to emphasize that one cannot expect from any organisation that it steps outside the limits of its definition or self-perception. One should not expect the CBA to succeed under present conditions, since success presupposes that it destroys those conditions.

(Indeed, while just getting started, some CBA officials already came up with the suggestion of a ‘labourless income’ for some elder unemployed, i.e. a benefit without the requirement of having to keep on applying hopelessly for a job. That means that our new labour market monitor turns into a benefit machine, contrary to law, intentions and official predictions. But of course, official statistics would show increased numbers of people being helped ... )

To complete the story, the discussion also resulted in a law (‘WLOM’) valid from April 1 1990 up to 1994, which gives a ten percent subsidy on exactly the official minimum wage (which is raised by 1.82 percent at the same time). However, only 7 percent of central bargaining contracts allow for that wage, and 93 percent set their own minimum higher. To qualify, there is red tape. At the end of the subsidy arrangement, in 1994, it may not be easy to fire someone, while the subsidy stops. Understandably, the WLOM proclaims a modest goal of 15000 jobs, i.e. 0.3 percent of the labour force.

The three insights combined

We may compound our view, on the areas of economic analysis, economic policy making, and institutional management.

At this spot it is useful to remind us of the old notion of the neoclassical synthesis. I my view, it still applies. The keynesian and classic views supplement each other, as expressed in Samuelson’s definition of ‘neo-classical’ and in the label ‘Vintaf’ (‘vintages’ for supply and ‘afzet’ for demand). This means that at the macro-economic level some policy remains needed to break the vicious circle of a recession, or other causes of mass unemployment. To this older synthesis would however be added a greater awareness of public choIce intricacies.

This idea of a synthesis is quite important, so there must be some more discussion of it. The Palgrave discussion of the neoclassical synthesis by Blanchard points to a ‘theoretical crisis’ from the 70s up to now. I tended to agree, but on close inspection do disagree, and let me explain why.

Two theories can be put together, only if they supplement each other, i.e. regard matters from a different but relevant angle. The neoclassical synthesis arose from the insight that micro-economics and macro-economics are different aspects of the same phenomenon, namely the economy. The same however holds for equilibrium and disequilibrium. When, for example, the Moon rotates around the Earth, one can describe this in terms of equilibrium (it stays there), but also in terms of disequilibrium (it wants to drop on us, but it continuously falls past the Earth). Similarly for the expectations angle, with the Moon falling where it is expecting to fall: still, with equal explanatory power, one can present at liberty Keynesian effects as the consequences of ‘disequilibrium’ spill-over effects.

In my perception, there is a clear continuation of the 1950s micro/macro synthesis into the later developments. The 1970s brought much economic havoc and subsequent theoretical disagreement, but the important strands of thought stayed within the traditional lines. Of course, there are many theories outside of this realm; there is a whole Zoo of economic theories and schools, for which there appears to be too little of thesis & antithesis to even consider the notion of synthesis. However, also a Kuipers & Van Ees (1990) review article concentrates on the mainstream diversity. Their conclusion that economics is not capable, as of yet, to decide whether an activistic macro-economic policy is useful or not, does not get my blessing, since the issue rather is optimization. But their selection of theories is very useful, for these are the very ones that can be synthesized.

An example may help. The CPB general equilibrium model (Gelauff e.a. (1990)) appears to contain disequilibrium notions like unemployment and the rate of capacity utilisation. So actually, the CPB multisectoral model Athena (CPB (1990)), with its extensive treatment of taxes etcetera, could be characterized as a ‘general equilibrium model’ too; and the differences with the Gelauff e.a. (1990) model can be traced to computer space limitations.

So let us stick with the neoclassical synthesis, and simply call it economics and let’s be honest and recognize that already Keynes was working in that tradition when writing the General Theory (he created the (possibility of the) synthesis). And to that synthesis. I want to add here a greater awareness that what labour economists call ‘rigidities’ in the labour market, is called ‘bureaucracy’ or worse by public choice theorists.

An earlier version of this paper (Cool (1990a)) has been criticized for saying that ‘economists agree on the obvious solution, the reduction of labour costs’ and that ‘economists would know the solution but only politicians prevent it’: this has been called ‘technocratic’, and ‘misplaced, since economists disagree’.

Admittedly, there is a problem, when people agree, but don’t see this themselves. Admittedly, it is quite a paradox, when I suggest agreement, and one starts to disagree with this itself. Admittedly, various economists seem to disagree, they have their pet theories, and their own choice of words. Admittedly, De Neubourg (1990), OECD (1990) and CEPS (1990) present whole shopping lists rather than the simple theoretical clarity which I propose here (though as a deliberate summary of the major item). And admittedly: people argue. Lindbeck & Snower (1988) p30: "The most convincing arguments of the contestants seemed to be their criticism of each other."

But I think that detailism is a trap to avoid. I’m willing to traverse many an avenue, but there is also the urge to look down and closely inspect what we’re actually walking on. For example: there is a gliding scale of concepts, going from the Den Hartog & Tjan concept of an occupation rate (the rate by which jobs are actually filled), to the efficiency wage theory, to a distribution approach: now, what is quite the difference ? — that is, ‘quite’ in terms of a practical full-fledged meso/macro model that we want to use to base policy conclusions on ! My conjecture is that we can be eclectic here, in other words synthetic, and select the ‘right’ elements and get the message across. For example, the vintage approach is a useful parable in some respects, but less so, when the consequence would be ‘that the less productive people would work only with the older equipment’: and then it is better to switch to another parable.

Admittedly, in the field of labour theory, there are various details like mentioned by De Neubourg (1990), OECD (1990) and CEPS (1990). But as already said above, we must distinguish clearly between measures to provoke growth and efficiency in general, and the specifics that affect the distribution, most notably, which affect the level of unemployment in the lower wage brackets. In that sense, there is no more technocracy involved as having your car repaired at a good garage. Is it always an ad hominem fallacy, to suggest, that economists failing the litmus test, are likely too much involved in the mentioned public choice problem ?

Admittedly, being synthetic about both labour economists and public choice theorists, gets one into conflict with both groups. My argument ‘we know enough about the labour market, so we have to switch to public choice’ may delight public choice theorists —until they see what havoc is wrought to their own pet theories: I call their theories ‘unpractical’ (e.g. in the sense that none such theory is incorporated into the CPB models).

Having discussed pets and Zoos in this quantity: there is also an elephant story, about the blind who stack detail upon detail, but who cannot determine the whole beast. To summarize these statements: I don’t like to argue with everybody, but the synthesis is just there. (Excluding my own blindness, here.)

Applying the synthesis

The synthetic theory leads to a description of the present state of the economy as likely suboptimal. Obviously, we cannot discuss the economy without thinking of the optimum. It would not be fair to simply and partially reject the CBA initiative, and leave it there. It may also be instructive to show how the old neoclassical synthesis suggests a way of returning quickly to the optimal high growth path; which would in fact involve a Swedish type of approach. (In Sweden much is done by external bureaucracy, in Japan the subsidies (apparently) are internal to the firms.)

The keynesian element is sufficiently obvious to neglect 7), so we look into the classical one. Specifically, we have to think of micro-oriented policies, which however are so general that they can be formulated at a macro level. According to the theory, growth can be enhanced by strengthening the market mechanism, so that relative prices reflect the (marginal) productivities, and so that interest, wages and prices are used for the allocation or scarce resources. It would be a good strategy to apply this principle also to the labour market; and tactics can deal with the limitations of the real world. (This is still in line with the ILO slogan ‘labour is not a commodity’; akin to Okun, Hicks (1989) and Laidler (1990); quite synthetic ! ) It cannot be avoided that companies go bankrupt or otherwise dispatch labour. What can be done is to reduce the costs of the process of reallocation and price and wage adjustment, not only the costs to society but especially to the people involved. Hence, as said within the framework of adequate macro-economic policy, an insurance system must be chosen, with a correct choice of base, rate and benefit.

One has to take account of free riders, moral hazard, fraud, and the like. But rather simple and strong rules seem at hand, that effectively exploit economic feedback. These fall under the general idea about institutional reform: that of saddlepointing society. (Contrary to earlier criticism: this paper is full of incentives. Earlier case-studies have been deleted.)

In particular two mechanisms suggest themselves. Firstly, by keeping cheap labour real cheap, one effectively combats crowding out, and the threat of unemployment is shifted to a more assertive but also more vulnerable/sensitive section of the labour market. Secondly, since the first might not be enough, witness the USA, subsidized search can be made more efficient by reducing costs: (a) by letting searchers be productive while searching. (b) by letting them succeed more quickly. Both are served with on-the-job experience.

Thus: making searchers a scarce commodity, induces competition among companies, and leads to a better allocation. There will normally be no need for additional mediation since the existing hiring procedures normally are sufficient; which makes it cost-effective. One may note that there is actually no big difference between both mechanisms: they both imply the need for subsidies. 8) The real difference is primarily in the duration of the subsidy. (This associates a bit with existing institutional arrangements. so that the difference may look larger than it actually is.) Likely, the higher paid unemployed need only temporary support, and the lower paid, like the present long term unemployed, pemanently.

It is useful to relate this approach to the older idea of an inflation-unemployment trade-off. Present policies may be described as ‘riding the Phillips-curve’, i.e. as either monetaristic indifference or as vulgar-keynesianic balancing (taking hysteresis in stride). What is best, however, is to reject the whole issue of a trade-off in these terms. Best is to restructure the labour market into a process of humane reallocation — in the same manner as the notion of unemployment was invented in the first place, see Piore (1987). It doesn’t strike one as fair, at least I think that this is the general feeling, that the people In the lowest wage brackets are also those suffering most from the dangers, risks, costs and vagarities of unemployment. As an alternative, a job guarantee for the lower paid would amount to making the Phillips curve more sensitive to unemployment, since then it would involve the higher incomes. But additionally, for those too, one can devise various systems of reallocation which are more efficient than unemployment. For there is no need to use unemployment as the instrument to battle inflation. A threat of some loss or income already suffices. In the simpler models, unemployment actually stands as a proxy for a loss of income, so a more general approach scratches unemployment. Next to such methods, one of course has competition policy and other measures to combat inflation, i.e. full employment anti-inflation measures.

Above we have refered to the major five economic policy goals as officially expounded in the Netherlands. To attain these, one might wish for as many policy instruments. Next to fiscal and monetary measures, this paper suggests the use or labour market subsidies.

A Basic Income ? Not economically needed !

Arguments like those above may remind one of arguments for a Basic Income. This is not required.

Brittan (1990) seems to advocate partial steps towards a basic income: "The underlying problems are ones of public attitude. The left and centre have to bite the bullet and realise that full employment means market clearing pay, which need have no relation to merit or need. The right has to appreciate that market-clearing wages are neither desirable nor possible unless there is an additional source of income on non-humiliating terms for those whose skills command a low value in the market. And throughout the political spectrum there needs to be a less puritanical attitude, in which the principle of a modest non-work income — long taken for granted by the old upper and middle classes — can be extended to all."

Schuyt (1990) pleads for a partial basic income for the present receivers of benefits.

My viewpoint is the following. Note firstly that in Holland we basically already have what one might call income security. What are at issue are the details of administration, thus, individuality, work-relatedness, and level. I associate the term ‘basic income’ with non-work-related, individual and substantial handouts. If a majority wants this, then I can live with it. (The formula presented In the appendix allows for such a system (b = 0 ).) But economically, a basic income means a reduction of the marginal proceeds from work, and hence this implies a reduction of real product. It would be better to administer such income, not via personal (negative) taxes, but via employer taxes. This means, there is a right to work at decent pay rather than a right for income. This also means that one can relate the new system with the presently existing one, since one can use existing channels leading to the labour market. People are already screened for applicability of work, ability and income related benefits, and this present system can be continued, whatever we do with the benefits.

The latter approach has the merit of aborting circumstantial discussions about the level of the basic income, the payments to housewives, and the general unaffordability of the whole scheme. And for this reason I deleted the term ‘basic income’ altogether in the earlier version of this paper.

Intermediate conclusion

In the appendix the matter is treated in a slightly more formal manner. Many arguments may disappear once we sit down and discuss formulas, as simple as those can be. I also outline a system of rules (a) to (f), which summarize the findings of many studies, and which ‘axiomize’ the market-oriented properties required for any system to work properly. I also give two formulas, which depict the general macro and micro situation. This exposition suggests that the matter is actually solved, in the same way as our monetary problems are solved by creating a Central Bank: that is, criticism is possible, but it is of a going-concern rather than a fundamental nature. (For example, within the confines of the theoretical synthesis, I am not afraid of, and in fact a proponent of deflationary policies: which force companies to compete by forwarding productivity increases into prices; and such might be discussed within an established banking system.) There is no need neither for a large bureaucracy, nor for strong involvement of labour and employer unions. The idea may remind us of the CBA institution mentioned above, but of course it is fundamentally more efficient than the present CBA. And thus, when working with (near) full employment (models), our attention may shift to other problem areas.

This neoclassical recipe of course implies some departure from present policies. As stated in the introduction: if it is going to have any chance of occurring, a review of policies is needed. This brings us towards Public Choice again. It is an essential part of our theoretical synthesis, and hence some analysis on this issue is required.

Why we might wish for a parliamentary inquiry

The policymaking process

The striking thing of the last two decennia is the policy making process. This does not involve only parliament, but also the earlier stages within the government body. For expository purposes, let us put the matter a bit sharp, and distinguish the following explanations:

1) The government wanted continual mass unemployment.

2) The government did not want it, but:

a) the advisers did not know it would happen;

b) the advisers did know it, but did not succeed in a convincing warning.

3) Any democratic mixture of the above: some wanted it, some did not know it, with the passing of time thoughts and majorities changed.

Of course we choose possibility 3. The other possibilities actually seem a bit unreal. The logic involved seems not sharp but blunt, since it cuts away all relevant detail. For example, the mass unemployment may have resulted in the first place from the baby boom, the bureaucratic backlog of ‘improvements’ in the welfare state, the two oil crises, etcetera: and one cannot expect politicians and their advisers to foresee all of these; while the events of the second place, given in our sections above, might be conjectured to be of secondary importance...

This is true up to a point: if one allows more detail, eventually all error could be smoothed away, which means, one ends up with the eternal misunderstanding between officials and politicians. But smoothing errors is not the issue. At issue are the costs of this kind of management. At issue is how to improve policymaking. Then one has to return to first principles. Only by looking at possibilities 1 and 2 we can understand the errors of the 3rd one.

This approach might look silly, but then, there seems to be no limit to the need to impress upon one’s fellows the necessity of being clear and unbiased in the dissection of government action. And we find ourselves in good company. Uno (1989) p181: "The role of government is being questioned the world over in such diverse areas as social security and R&D." Wolfson (1990) p120: "There is ample evidence that politicians and bureaucrats alike are not very keen on promoting policy analysis, let alone on divulging the uncomfortable facts it may bring to light."

To proceed, a description of the history according to category 3 thus may use the first categories to catalogue events, bearing in mind that these are mutually exclusive, so that one case would be followed in time by another. At the risk of oversimplification it is possible do some conjectures. These are correctly regarded as research hypotheses (but a simple majority vote might already confirm them):

(ad 1) As said earlier, policy makers were also ‘riding the Phillips curve’. This might also involve a theory that unemployment improves discipline and productivity.

(ad 2a) Actually at no moment there has been an early and systematic or at least a convincing analysis of the relevant possibilities and limitations.

(ad 2b) Stevers (1979) may be the exception that proves the rule: but apparently not convincing, since mass unemployment continued and got worse.

These propositions can be qualified.

— The rationality of the policymaking process mostly remained restricted to a horizon of one year or, if favourable, of four or five years, and then it was ad hoc and hardly systematic. One major policy finding of the seventies was (for those who believed in it before) that fine-tuning is as relevant for macro-economics as for elephants and supertankers; but it is striking that policymaking remained oriented to the shorter term (case 2a).

— Of course there was an occasional longer term outlook 11), and there were various institutional changes (which can be regarded as longer term policy making). But one can liken those to the example of the CBA. Not taking efficient measures is a clear case 2, with the suggestion of case 1. The fact of continual mass unemployment has some superior convincing power, here. It may be added that there was some luck with the international recovery, so the outcome could even have been worse.

— Of course, the welfare state remained in existence; at one occasion the unions and employers (the leaders Kok and Van Veen) reached an agreement. But this is a case 2 event, since unemployment still remained high afterwards; Van den Beld (1987) would make it a case 2a — an overestimation of the employment effects.

It is possible to test these conjectures, since the process is well documented. Various institutions and persons have reported regularly on the issues involved. Not only the Central Bank and the CPB, but also Den Hartog (1984), Knoester (1987) (summary in Pen (1987)), Driehuis (1988). Of course, no finding ought to be regarded as a personal criticism; there is rather cause to be grateful for the effort and the contributions to our own insight. While it is possible to review the matter, it of course remains a difficult task, for analytical reasons, and not only because at every point we want to remain sensitive to the persons involved and the difficult tasks confronting them.

It may be well be that the matters discussed above are sufficient example of the complexity of the case. But it will be useful anyhow to give three specific examples in more detail of the kind of review that might be needed for a correct assessment of the history of economic policy making in the last decennia. For, as indicated, in this matter overkill is hardly possible.

First example: a failed austerity program

Let us look back at the early 1980’s when the then Dutch Minister of Finance and now EC Commissionar Andriessen resigned from office for lack of support for his austerity program. Our chief economic adviser, the secretary general of the Ministry of Economic Affairs, Rutten (1989) writes: "The greatest disappointment at the end of the seventies and the beginning of the eighties was that it was impossible to win general support for a timely change in policy, even while it was obviously needed for the control of the budget deficit."

The point to be made is that the kind of proposed policy change of course affects the kind of support one gets. The austerity program likely had been approved generally when it had adopted the unemployment policy outlined above.

This analysis applies to the Stevers (1979) contribution. Stevers holds the original public choice contention, that spending departments crowd out the Treasury, and that this happens under any political regime, since every incumbent wants to be reelected. But this would imply that everybody is short-sighted; and in that case the result is a pure democratic outcome: an optimum which we cannot argue with... Rather, the Treasury fails to come up with the true optimal compromise, which would honour the basic longer term views and goals that people have. This has to do very much with good management and proper information. (Modern theory doesn’t exclude the possibility of mismanagement.) For example, the Treasury could inform the Ministry of Education that subsidies to a badly managed agricultural sector will cost education dearly in the future; and one can devise a contract that such will happen indeed.

Second example: a parliamentarian turned minister

In Holland, a new coalition government was formed after the general elections in September 1989. De Vries, leader of the Christian Democratic fraction up to then, became the minister for the department of Social Affairs and Employment. In a September 1990 interview, the VN weekly suggests that the minister has kept a too low profile.

VN comments: One gets the impression that the former leader of the CDA fraction has taken a sabbatical for the next four years.

VN asks De Vries: "What is it that you have learned up to now?"

De Vries: "You really go deep into problems. You analyse the statistics. You are confronted with puzzles: what can explain that — contrary to intentions and contrary to policies carried out — that long term unemployment, sick leave and disability are in such a bad state ? My amazement applies retrospectively. In the eighties I thought: once the economy recovers, then the long term unemployed will find jobs by itself. Now I have discovered that that judgement was premature. In the eighties, whenever someone signaled the danger of the creation of an underclass, then I spoke up against that. But nowadays I realise that there truely are vulnerable groups that have a hard time finding employment."

VN: "You really don’t want to suggest that you didn’t know that during your time as fraction leader ?!"

De Vries: "It sounds incredible, but you really have to study those statistics thoroughly before the full extent of events dawns on you. As fraction leader you traced the evolvement of unemployment in a succinct and general manner. But only this last year I fully understood how difficult it is: some groups will find jobs, but others don’t."

This present paper was written basically in June & July as Cool (1990a), and this later interview beautifully confirms a major contention. Not only parliament but also the supporting ministries can easily improve in insight and communication.

For, the following questions arise quite naturally: (a) As a fraction leader, did De Vries not have an assistent, to really study the statistics ? Or was such an assistent selected on proper grounds ? (b) Did one not have support from the department at that period ? (Since those signals about a developing underclass came from various distinguished sources, one would expect them to have been taken seriously.) (c) It is rather strange that the department doesn’t have at hand the analysis, which the new minister must make single-handedly, and which takes him a year to do.

One might add to this, that the immigration of hundreds of thousands of foreigners to Dutch society (especially from Surinam in the 1970s) would create problems of integration. The economics of integration are somewhat similar to the general employment problem (see footnote 2 and 8). I get the impression of a serious information gap, when these matters are not communicated to policy makers.

It should be clear that minister De Vries deserves our respect, in being so open-minded, and in turning out to be a good minister as well. I myself have been slow in producing this analysis. But, why not ask the tough questions ?

Third example: the Central Planning Bureau

In the Netherlands the official macro-economic projections and scenarios are coordinated at the Centraal Planbureau (CPB). Obviously it is only one of the many institutions in the field of economic policy making, and likely it is also one of the smaller ones. It is not directly involved in actual policies, only in providing projections and macro-economic reviews of policy proposals. It is chosen as an example here for five reasons. 12) Firstly it is the obvious institution to look at when one is interested in the macro-economic analysis of macro-economic policy. It is noteworthy how many times we have been referring to CPB publications up to now. Secondly politicians and the public attach a high value to the Bureau, so it would be a prime actor in one’s analysis anyway. The Bureau is generally regarded as being ‘relatively independent’, but it is useful to wonder what that is (‘a bit pregnant’ ?). Thirdly, sometimes it is said that the Netherlands could be an example to the world in its reliance upon quantitative methods for economic policy making: and that would be something to be tested. Fourthly, it wouldn’t do here to tackle parliament itself (apart from giving the De Vries example). Fifthly, suppose one would work at the CPB, would one not try to take some of the blame ?

As said, organisations seldomly transcend their own limits. The CBA has been mentioned. Another example would be a Central Bank that aspires a zero rate of inflation: It would seem austere and laudable, but in fact it would sanction any inflation of earlier periods. The CPB too cannot escape from its self-afflicted limitations. In particular this concerns the way it treats government.

Let us first describe the situation. In CPB models for the Netherlands the government is exogenous. The Bureau forms part of an interactive network between the ministries and agencies, and there it gets it data and policy input. If government and policy would be endogenous, the CPB theoretically would have no need for this network, since it could forecast the outcome independently. There is no criticism yet; it wouldn’t seem to matter much how a result is achieved; and when it does matter, it seems at first sight that the exogenous approach is more adequate. Indeed, Don & Van den Berg (1990) call such endogenization ‘rather useless".

This being the situation, the basic critical question arises from the following. Namely, the former is a fallacy. One cannot hold that, if one has a clear picture of government behaviour, then there would be no need to endogenize. One might also argue that we wouldn’t need macro-economic models, since they only render what also can be guessed at ! We have to endogenize, for the same scientific reasons: to objectify our findings, to expose them to criticism of colleagues and nature, and to coordinate expert opinion. And it is well-known in public choice literature (Van Velthoven (1990)), that not endogenizing government creates a bias. Up to now, the Bureau neglects public choice theory (even though Stevers has been an adviser for a long time).

One cannot expect too much from the CPB; the Bureau itself always puts much emphasis on this. In defence of the Bureau: it is known among policy making insiders that the short term forecasts are conditional upon policy success rather than on mere intentions — and in that sense the responsibility shifts from the Bureau to the policy makers.

Taking stock, critical questions arise on more counts: 13)

(a) To what extend is the CPB critical about the ‘information’ it receives within the network; and is it actually capable of it ?

(b) To what extend does the CPB transcend the interactive process ?

(c) Does this situation match outsiders’ perception of it ?

(d) Anyhow, who checks its scientific quality ?

(ad a)(1) Even though the final responsibility for short term forecasts is not with the Bureau, one still wonders about events In the intermediate stage, when other participants act upon preliminary output from the Bureau. It is well known that the Bureau once in a while initiates a proposal. Has this been done with sufficient pace, care and clarity ? When disagreement with policy makers arises, a way out is the publishing of a separate research study ‘under the author’s name’. Bakhoven and Van Schaaijk perhaps quality here — but that makes a low count.

(ad a)(2) It is very remarkable that the more independent Bureau studies, noteably the evaluations of political party election programs, adopt a similar approach: though much criticism is present, and likely internationally unique, fundamental critique and analysis are wholly absent. But such critique could easily be defended from the viewpoint of the celebrated major five economic policy goals which are widely accepted throughout Dutch Society. In other words, one would expect serious economic analysis of rigidities, policy failures and shadow costs in general, instead of a repetition of conditionality.

(ad b) There also have been some experiments with goal programming and control theory (Schrijver (1987)), and the Bureau may have very good reasons for not using those results for its standard practice. But in this respect it is interesting to note that Besseling & Brandsma (1988) do apply game theoretical techniques, albeit not to a close and dangerous Dutch government, but rather to the distant and more abstract world economy.

It will be useful to extend on the kind of review that one would wish and which the CPB has not furnished. For the matter is complex, and it cannot be understood enough that it is so.

In the appendix one finds a formal representation of the decision problem, involving three economic functions, including a Social Welfare Function (SWF). What we have not touched upon there, are the uncertainties attached to these. But, we don’t know the form and location of the functions involved. Some examples will show how important this is.

1. If we dislike people to work with a subsidy, this would not have to mean that this couldn’t be offset by more income.

(That it, it may be that SWF(Y,n) > SWF(Y’, 0) for n > 0. )

2. If we only appreciate income ( SWF(Y,n) = SWF(Y) ; horizontal indifference curves; no work-ethic), then:

a. If disincentive effects dominate employment effects, then the present situation would be optimal anyway, even though we are not ‘Capitalist’ and merely indifferent. (Y(n) is downward sloping, and SO = PS).

b. If employment effects dominate then we would have subsidized work. ( Y(n) has an initial upward section.)

3. It may be that we have a stronger work-ethic, and that we like to see more subsidized employment (e.g. in agriculture), notwithstanding even some reduction in income.

4. We might want to discuss how much the difference really matters between the present situation PS and the social optimum SO. (Cardinal SWF ?)

Obviously, how many and how much depends upon the functional forms. One can only determine exact forms and locations, when one has some inkling to look for those forms in the first place. The essential insight is, therefor, that endogenization of government behaviour would a prerequisite for proper analysis. One cannot reduce the problem to the difference between a good and wrong representation of government; and then suggest that exogenity might be wrong but not really bad. Endogenization (introduction of a SWF) is essential to start with. If you work for a big multinational, then you use optimization techniques. If you are an advisor to the government, and you are supposed to provide econometric support, then you endogenize. A consistent policy optimization framework is essential, for example in the case of the celebrated unemployment or poverty trap: such may exist in the static case, but a correct optimization framework will point to the dynamic irrelevance of such a trap. The CPB tries to approach these matters by providing ‘policy variations’, but that is wholly inadequate. The problem has to be analyzed in the proper manner, relevant data have to be gathered (noteably about subsidized work; especially about the efficiency of policy measures), and also the decision maker has to be tickled to give his opinion: only then one can securily feel that the decision maker has been provided with the relevant information. Don’t take me wrong, and don’t think that I would require a perfect optimization model before government even might exist: little things might already suffice as long as the methodology is right; and note: a qualitative rather than a quantitative analysis might already be sufficient to get a policy going. 14)

That the Bureau has failed its major task, has the following consequences:

— Government endogenization studies, being banned to the academic world, still seem quite experimental, though are not unattractive (Drissen & Van Winden (1990), Swank (1989), the recent Public Finance / Finances Publiques issues; Alesina/NBER).

— There is uncertainty about whether the present situation is optimal.

— One cannot simply use the evidence of the last twenty years to try to locate preferences. The method used by Schrijver (1987) breaks down here. Events cannot be taken at face value; the evidence is extremely blurred. The process of trying to apply the true SWF has been hindered by bottlenecks in the process itself. Perhaps that Delphi techniques are best now. 15)

To continue on the former points:

(ad c) There may be a serious misunderstanding between policy insiders and outsiders about the value of forecasts, when the Bureau does not take into account quite foreseeable policy errors, like failing austerity and labour market programs. The respectfully naive public would seem to expect forecasts conditional upon policy intentions rather than success.

(ad d) There are some technical issues, like the disregard of cointegration techniques and level effects in the substitution between higher and lower paid, or the difference between point and range elasticities, which influences decisions. See also footnotes 2 and 8. Apparently, unemployment insurance institutions have little need for proper analysis.

Evaluating generally: perhaps one could be satisfied with the role which the CPB played in bringing the issue of real labour costs to the frontier of the economic policy debate in the mid 1970’s. But when evaluating the last decennia, for example on the question ‘Has the CPB known it ?’, one could wonder, for example ad (b & d), whether it would not have been fruitful to endogenize government behaviour: from which mass unemployment would have been explained scientifically and could have been projected. Not really forecasted, since the ‘forecast’ could have resulted in policies to prevent it. Nowadays we have an actual experience with the phenomenon, and much appears possible to get rid of it: but would it not have been better if we had known in advance ?

The recent Don & Van den Berg (1990) paper, with its rosy view of the CPB, rather has the effect of emphasizing the negative aspects mentioned here. On endogenization of government it is wrong in general, and specifically so with the election programs. The ‘relative independence’ of the CPB comes out very much like ‘being a bit pregnant’. The paper does wrong to the Treasury, when this insisted upon obedience to the law, when in a particular incident the CPB directorate failed to inform the Treasury of its plans at an early stage. Also, this incident finally resulted into a compromise: and the paper presents this as an example of independence, while a truely independent agency would not need a compromise.

Conclusion: the end of ‘laissez parler’ ?

Our economy is in a bad state, and the cause is that its political head is not functioning well. Parliament discusses everything, from the size of classrooms to Europe’92: everything, except the real causes and cures of the bad state of the economy. So we plead for an end to laissez parler, laissez faillir.

Since our very analysis is that parliament is at disarray, there is obvious danger that this analysis will have no effect. None the less we feel we better conclude this paper by summarizing the compelling argument for a parliamentary inquiry into both the economy and parliament itself. For the Dutch there is the variety of a parlementaire enquete, albeit of the original kind — i.e. without scapegoatlng, witch-hunting and other popular substitutes for reason, which somehow have tainted some recent enquete’s.

There are two extreme possibilities. The first is that politicians are honest and rational beings. Then they themselves will see that our logic is compelling. But alternatively, politicians are the opportunistic creatures of standard Public Choice theory. Then circumstances will force them.

Nothing can stop the dissemination of the ideas of this paper. Without these ideas, the welfare state gets stuck, and looses from a Japan which incorporates these ideas. And mass unemployment in Middle and Eastern Europe threathens political stability. Widely different persons like Flip de Kam & Frans Nypels, James Meade, Samuel Brittan, Patrick Minford and Brian Reading (Financial Times December 18 1990, Letters) then reinvent sufficient elements. In time, and in line with the Lincoln quote, people in general will understand the analysis. and politicians will have to follow suit.

For the honest and rational politician, the logic is in a nutshell: (1) the costs of unemployment are enormously, in past, present and future, (2) economists are getting tired and disinterested, and few things are so costly as a culture of tired and disinterested economists, (3) logic points to public choice theory; all the evidence, like presented by the case of Minister De Vries, backs this up, (4) this is not to say that the matter is so simple as to make an inquiry redundant, (5) other measures than an enquete will not suffice.

(ad 1) Costs. One cannot escape from some worry. Twenty years of mass unemployment are not nothing. The Dutch not only suffer in income, competitive position and environmental needs, but there are many, many other costs. A rough estimate might put the costs at 5 percent of employment,i.e. a little less than 5 percent of income, roughly the size of the present budget deficit, and, this, for a period of 20 years — not counting the utility from accumulated income.

(ad 2) Getting tired. It may sound exaggerated when it is suggested that economists start loosing interest. On the other hand, CEPS (1990) repeats itself, it expressedly says it is repeating itself. Other studies, some referenced above, look into other problems than unemployment In itself — economists are merely human — and this results almost into a cacophonia of disdirected research and policy advice. And the young researchers at our academies start to get lost in a Byzantine accumulation of marginal problems and solutions. Now, what happened to Byzantium ?

(3) Logic & facts. All experience, mathematically, theoretically, econometrically, and the for many more accessible but circumstantial newspaper interviews, all lead to the conclusion: public choice theory explains many events. Thus it should be part of our models, and we must act accordingly.

(4) Complexity. Above analysis paints, purposefully, a bleak picture of the policymaking process. For example, it has been shown that for an evaluation of the economic policy making process it is important to know what is adequate information, and whether all participants in the process know what it is and actually give it. Instruments are an issue too, for example that one should not use gross minimum wage as an incomes policy instrument, or allow for it freely in collective bargaining. Likely, the problem is even more complex than the issues raised here.

(5) In the same vein. Far more important however is the need to find the politically acceptable implementation of the economic solution expounded here. This makes for a complexity which no study can tackle, and this obviously requires an intensive review from parliament. 16) Alternatives to parliament can be rejected: a general discussion (Brede Maatschappelijke Discussie), a WRR Scientific Council report, a parliamentarian committee hearing, a kitchen kabinet … These might be instrumental in providing a check on the analysis of this paper, but they cannot replace the need for a rational political analysis and choice.

Wishing for this Dutch phenomenon of a parlementaire enquete appears to be a sensitive issue. An enquete makes politicians nervous since it tends to put a strain on a ruling coalition. However, there is no need for a strain. The objectives are to force politicians to study a complex matter in great detail, to have them discuss details with their officials and external advisers and relations, and to arrive at conclusions across all political parties in full publicity: and I can think of only one process that satisfies those objectives. The logic here is compelling again. Economic analysis tells us that the decision machine got stuck, and thus it has to unwind itself.

Fortunately, the present coalition is extremely stable and solid. Just that makes it advisable to have this joint intellectual effort just now. A more occasional argument is, that the Dutch, not surprisingly, are presently engaged in a discussion on some social renewal, and it would be unreasonable to disregard the policymaking process itself.

There is more to this than simply solving the Dutch problem. There are obvious European and world dimensions. It is important to see that most economists adhere to some kind or another of the ‘Swedish model’. If Holland wants to be taken seriously in the international debate, it has to move into that direction. Holland and Europe must both move into that direction, in order to maintain social harmony and efficiency in the increasingly competitive international setting. Clarity on these issues is also essential for what we are going to tell the reconstructing European nations in the East. It is also essential for how this will affect their social harmony and efficiency.

Appendix: Some formal elements

It will be illuminating to state what we have done in a formal manner, since this allows for more abstract reasoning. What is needed is a full model, but, drawing on a standard model, we may concentrate on three ingredients:

- a production function, giving national income Y

- an income redistribution function, benefits U, premiums P

- a social welfare function, SWF(Y, n) with n subsidized workers

We will presume P = U. Hence Y = Y - P + U = (Y-Py) + (U-Pu) with obvious meanings, and net redistribution Py = (U -Pu). There will be a disincentive from Py on Y. It can be supposed that there are micro rules within the production function which determine the split in the ny people who pay Py and the nu who would benefit from it. This involves the difficult issue of labour demand and supply, and rules that define the ‘fitting’ job that determines the benefit; and it involves the issue whether premiums are levied on labour income or profits (via funds). But the people who receive an ‘unemployment benefit’ or subsidy can always be ordered in terms of the profitability of employing them with the stated subsidy; and this gives the production function of employing n subsidized workers: Yu(n; Py) with 0 £ n £nu. National income hence is dependent upon redistribution in two ways: premiums have a disincentive effect, and income subsidies could help employ more people: Y = Y0 - Yp(Py) + Yu(n;Py).

Including n in the social welfare function may be interpreted as giving us an expression of the work-ethic. That is, the SWF determines the split of the subsidized people into those who work (n) and those who simple sit at home (nu - n). This SWF can be very general. The difference between the Capitalist and the Swedish model, the dichotomy between n = 0 and allowing for n > 0, need not be depicted as a regime switch, since the Capitalist case can be given as SWF(Y, n) = 0 for n > 0.

Note that fomulating our problem as SWF(Y, n) is a choice. One might conjecture that another trade-off matters more. Some e.g. argue that we don’t change our present system since we have not studied sufficiently thoroughly what the consequences of change will be... But my conjecture is that this representation is better, and at the Albeda conference I got warm support from professor Meidner.

There exists a SWF under very general assumptions. Originally an appendix to this paper, now elsewhere, Cool (1990b) shows that Kenneth Arrow’s ‘impossibllity theorem’ does not hold.

However, without that, already in the older view, there were various possibilities for a SWF to exist, e.g. as a result of similar preferences of a dominant coalition, or simply ‘policy maker preferences’.

Now the redistribution. It would come fundamentally from the SWF, but then rather in terms of the net income distribution. While we need micro-oriented policies, they must however, as said, be so general that they can be formulated at a macro level. Therefor we select from the micro world a more general pattern.

This is specifically so, since the split of subsidized people into workers and non-workers can have consequences. We noted above how job search could give rise to a temporary subsidy only. Therefor, in the presently existing administrative circumstances, the rule is used: if you find (permanent) work, we reduce your subsidy. The major explicit purpose of this appendix is to show that it is exactly by such (extended) procedures that some society effectuates its SWF, i.e. when it wouldn’t want subsidized work. This namely clarifies that it is not some mysterious natural cause which creates unemployment, but it are the very procedures which we have chosen ourselves, which create a ‘kinked employment function’.

Thus we write U = U(n) and split it into Un and (U(n) - Un). and similarly for Pu, with the obvious interpretation that there are two kinds of subsidies, and the total depends upon the composition. (Actually, we might restrict our analysis to only those who are working, and drop the suffices; but as said it is more illuminating not to do that.) The reduction of the subsidy for workers also affects the production function Yu(n; Py). If we assume that such reduction is proportional then only the ordering of persons is not affected.

We will have a system and a rule. The system:

  1. There is a framework that establishes for a person the income replacement, and whether the following rules apply:
  2. The person has the duty to apply somewhere at some employer, and may choose himself where.
  3. Employers have the right to refuse to employ a person.
  4. The income replacement is paid to the person directly if he is out of work; otherwise it is paid to the employer.
  5. Premiums are paid, rather not on labour income but on value added or profits: and the balance of premiums and benefits would be discounted by the employer with VA or profit taxes. When given to the employer, the income replacement may be too large a sum, and by including it in the (employer) premium base, the actual payment may be reduced to a more adequate subsidy.
  6. Other working conditions are properly taken into account. 9)
These rules are claimed to summarize or ‘axiomize’ the requirements for workable systems. By this endeavouring in system-building, I might wonder, like Meade (1985), whether I have made a consummate ass of myself. My defence is not merely that someone, who would propose the present system of income tax, would be ridiculed and mocked to death. My defence is rather that the given rules are the economic ‘axioms’ — and they may be translated and implemented depending upon circumstance. And it is economic in the sense that it has not been elaborated here econometrically for a specific country. Though, for Holland, Van Schaaijk (1988) computes an increase of the standard tax-free-base in the income tax — which effectively is a subsidy. And for a country like Germany, one might consider a regional criterion, so that ‘unemployed’ originating from the former DDR would qualify for a high subsidy if they find employment in that region.

Indeed, it is useful to warn here, that much confusion can arise when this economic analysis is presented to others. Labour market jurists may regard one’s reasoning as sloppy, when one discusses job search subsidies first and then continues with tax base reductions, treating it as the same topic. So bear at least in mind that this paper is not a juridical discussion.

With other aspects combined, the following normed-profit premium rule catches many possibilities:

Pn º ((1 - b ).Y + b.Un) . (1 - min(LIQ, normLIQ)) . t

where L is labour income, LIQ ºL / Y the labour income quote; parameters are tariff t, base choice b and normLIQ, all at the macro level; and let Pu = Pun and also Pn = Un. 10) Note that normLIQ = 0 implies no profit based premium. Let normLIQ = .8 and Un = .03 Y. Then b= .99 implies a 25 percent subsidy:

(.01 Y + .99 Un) * .2 * t = Un => t = ((.01/.03 + .99) * .2)-1 = 3.75

=> .99 Un * .2 * 3.75 º Pun = Pu => Pu/Un = .75

At the individual level, the subsidy would rather be a function of individual replacement income. to prevent too much subsidy on too high incomes. Thus with company i and subsidized workers ij:

Pj = ((1 - b ). Yj + b . S i f(Uij)) . (1 - min(LIQj, normLIQ)) . t

To summarize our finding: it seems possible indeed to use some general rules and parameters, with only some small exceptions at the micro level, but such that those exceptions further the cause. The other impression one gets, is that the institutional committee that sets the parameters would not have to be large. This could have consequences for tripartite bargaining, e.g. one might dispense with much bureaucracy and insiderhood. (Let I indicate that my thoughts go into the direction of an Economic Supreme Court — but that’s all I can say now.)

We can now make our diagram of the (Y, n) plane. Firstly at the bottom the properties of the ‘unemployed’ are depicted, ordered by the potential profitability of employing them. There is an original production Y0n(n) which is negative since otherwise they would not need a subsidy; and there are benefits U(n) and withholding Pu (nonzero up to n). As said, the production function for the ‘unemployed’ is Yu(n; Py) and we can distinguish two parts in it. The net benefits Un -Pu accrueing to the employer should at least cancel negative original production, so that the first part Y1u(n; Py) ºUn - Pu + Y0n(n) ³0. The second profit of the operation consists simply of withholding Pu itself. These two can be added to give Yu(n; Py) = Un + Y0n(n).

Secondly, total national income is given at the top in the diagram (disregarding the reduction of disincentives). The process of social choice is depicted by drawing some indifference curves. PS is the present situation, SO the social optimum.


Optimal social choice, of income and subsidized work.



1) An argument to have this reflection in 1990, is that we will be too busy in 1992, commemorating Columbus 1492. It must however be doubted whether reflection will teach us much ... Our subject matter is old, and internationally so, and one must be quite creative in saying something new (at least giving the impression or doing so). This discussion therefor is subject to all kinds of provideds. Incidently, this paper appears to cover much of the same ground as the Tinbergen lecture of prof. James Tobin (1990); the reading of which is strongly recommended, though see also Cool (1990b).

2) In the Netherlands this consensus can be traced as the result or the mentioned Vintaf II debate, see Driehuis & Van der Zwan (1978). But actually, the matter is rather a tautology. We must take account of immaterials too. An employer might refuse to hire a disabled black woman e.g. ‘since be would have a problem to discuss her efforts and output in a manner he is used to’. Such things exist and must be taken into account. One theme in recent discussions is the need for schooling. There are obvious limits to this. In the example it would rather be the employer who has to learn how to deal with an employee of a third kind.

3) Unpublished discussion papers, "Neoklassieke knelpunten voor de Nederlandse economie voor de ontwikkeling op langere termijn", CPB November 1989, and "Werkloosheid als inefficiënt zoeken", January 1990. The relevant ideas are adopted in the present paper; examples of incentives and fraud-bashing however are deleted.

4) See also footnote 7. The temporary deficit might be offset by a similar tax increase as long as one does not think that this kills too much activity.

5) The choice of a tax regime is a welfare economics topic. One can be against a poverty trap for the well-being of the people in the trap. This however puts little value on the prospect of getting a job. And therefor, it is hard to distinguish from the more lacklustre bureaucratic argument for uniformity.

The poverty trap argument also neglects the dynamic situation, that general growth can provide sufficient incentives.

Incidently, Van Schaaijk (1988) presents data that show that the Oort tax-project eliminates an employment policy instrument.

6) ESB March 14 and April 25 1990; and some newspaper articles that highlight the discussion: NRC May 9 1990 Economisch Supplement. Volkskrant Februari 8 1990 p7. Trouw June 1 1990 p11.

7) The keynesian element has never been away in macro-economics. The term ‘multiplier’ may have disappeared from the vocabulary, but it returns in disguises. In the Netherlands the current account surplus indicates inadequate expansion. A world, where surplus countries do not want to expand and where deficit countries cannot, is in a suboptimal equilibrium. For the Netherlands the announcement now of a revaluation of the guilder next year would also allow a reduction of the rate of interest. It would create room for a wage reduction and would also force to it.

8) There is a Dutch Law (‘Vermeend-Moor’) that subsidizes particular kinds of long term unemployed. It is criticized, and correctly so, for that it leads to substitution with regular employment. This is caused by the limited applicability of the law. But if the law is widened and becomes a general scheme, the resulting competition for cheap labour will be beneficial to all.

The argument generalizes for immaterials and for continual disability. For example, look again at the case of the handicapped black women of footnote 2. One way of educating the employer is to expose him to practice: in effect to help the woman into a job by means of subsidies. Another example is, that a subsidy for a specialized truck control-cabin can be forwarded in the shape of reduced labour costs for the disabled driver. Note that there is a difficult econometric issue here. Some present-day empirical findings on production functions link capital outlays only to high-productivity workers (Gelauff e.a. (1990)). But this suffers from the present lack of observations on the alternative, since we have so few low-productivity workers actually working. Beware of imeorrect policy conclusions from bad econometrics !

There will remain a need for mediation, like in the CBA. We have a multicultural workforce and a more homogenous set of employers, and then friction will be higher. For some people there will be a need for mediation, not only for finding a job, but also for keeping it. Such professional mediation can be subsidized, for it can be compared to the truck control-cabin. In general, the subsidy is hidden in the reduced labour costs; and the market will find the best use of it. But for friction unemployment other than mobility etc. it is more efficient to make such subsidy subject to additional mediation; the notion of such friction is actually defined by this. Note also, that the need for mediation does not necessarily imply the present institutional form of the CBA; it could be privatized for a great deal.

9) Incidently, it must be remarked that the scetched system would need expert’s review when applied to the real world.

But drop the indices. It might be possible to link U to a reduction in working hours. For example a steward would earn 40000 guilders, becomes unemployed with benefit 30000, and would take a job as a porter, who would normally earn 35000: to preserve the latter’s hour wage, either the employer furnishes 6000 or the former-steward-and-now-porter works shorter time.

This increases the chances for unofficial payments though. We could link gross subsidy U/P to the replacement rate U/L:

1 ³ P/U ³ 2 - L/U

For, let working hours be reduced from h to h.U/L. Then the hourly wage L/h has to be lowered to P/(h.U/L), so that L/h > P/(h.U/L) or 1 > P/U. Assume next that the employer hands out an unofficial suppletion S = L-U up to the former wage (we disregard the possibility of S < L-U. partially since we can manipulate P ), then the hourly wage has to rise up to L/h < (L-U+P)/(h.U/L) or U < (L+P-U) or P/U > 2 - L/U, If the wage-benefit gap is L/U = 1.25 then the tariff could be P/U = .75. It means that an employer can get a wage subsidy of 25 percent for somebody who starts working 25 percent shorter time (perhaps not in another job).

10) That is, no redistribution within redistribution. Note that Pn differs from Pu since it also includes premiums based on Y.

11) In an interview (unfortunately no longer traceable), the former minister of social affairs, Jan de Koning, argued his larger sensitivity for the environment problem, in the spring of 1989, by refering to a CPB long tern projection, where the Netherlands would be almost twice as rich in 30 years to come.

12) Criticism at the CPB may be welcomed, because experience tells that this enhances its position. Politicians need one scientifically educated guess at economic truth.

13) This yet neglects the problem that the contribution of the government to GNP is measured by its wage bill.

14) Recently there is more information, but it is not sufficient, and one wonders whether it is useful to wait till it really is. Van Opstal’s (1990) micro & partial analysis of the 1979 and 1985 impact of minimum wage legislation finds that it created unemployment only in 1985, and then only for youngsters and women; the ‘only’ not implying that these are not people, though they may not be true insiders. This correlates with the average data, see CPB (1980). However, the model used doesn’t allow for labour hoarding in 1979; as Van Opstal notes: "The macro-economic situation has changed in the beginning of the eighties by the recession in such manner, that notwithstanding a large relative (partly absolute) reduction of the official minimum wage the employment effects are still larger." The study would also improve from including duration (job experience (loss), growing old in unemployment, crowding out), job matching, minorities, and labour relations (union minimum wages, disability aspects).

15) An aspect is, that government & parliament may not like seeing themselves reduced to a policy rule. A rational discussion could help to resolve this. Firstly, take the example that it is forecasted that on your next swim you will not be wearing a smoking: then your freedom of choice is affected in no way. This is a trivial example, especially chosen so, for it compares to a politician forecasted to dislike unemployment. Then, secondly, regard the political implications. Normally, when a statement is made, it is implied that there is some interest in it. If everybody were to discuss your swimming trunks, you would start regarding it as an issue yourself. Combining these two aspects: politicians are used to the latter phenomenon, and if there is some confidence that some officially published SWF is close to the true SWF, then the political implications of publishing could be handled.

16) This involves the topics mentioned above but also by Tobin (1990), like rules versus discretion, intertemporal consistency, and the political strength not to expand when times are good. A very small reform, to have hearings like in the US Congress, might already involve an amendment to the 1947 law on the creation of a CPB, Incidently, that law provides also for academic outsiders on the ‘central planning commission’, but in practice there are few (none) of these.


CBA tripartite organisation: Centraal Bureau voor de Arbeidsvoorziening
Intended to regulate the labour market, increasingly with Swedish types of measures; formerly under the Ministry of SZW, after January 1 1991 under a tripartite responsibiliy

CPB government: Centraal Planbureau
Instituted after the war, with first director Jan Tinbergen, to ‘coordinate a central plan regularly’ (the law is quite vague on this). In practice it makes indicative forecasts and performs policy studies.

GAB government: Gewestelijk Arbeids Bureau
Falling under the Ministery of SZW, regulating the labour market, regionally based.

GSD government: Gemeentelijke Sociale Dienst
Falling under the Ministry of SZW, regulating social security, locally executed by city councils.

SER tripartite: Sociaal-Economische Raad
A social-economic council, with employer and employee unions, and impartial members appointed by government, generally economics professors; members also are the director of the CPB and the president of the Central Bank.

SWF Social Welfare Function (an index of general optimality)

SZW government: Sociale Zaken en Werkgelegenheid
A ministry of social affairs and employment

WAO law: Wet ArbeIds Ongeschiktheid
Regulates disability for people with a work history. Natural handicapped fall under other rules. The general impression is that this law has been used as an easy way out for lay-offs, since benefits are higher than normal unemployment benefits.

WLOM law: Wet Loonkostenreductie Op Minimumloonniveau
Valid from April 1 1990, and valid for 4 years, for people aged 23-65 years, working at exactly the minimum wage of 2004 guilders a month. The subsidy is 800 guilders each quarter, so a generous 10 percent. There is red tape.

WRR government: Wetenschappelijke Raad voor het Regerings beleid
A scientific council for government policy. Independent advisers appointed by kabinet may study and give advise at great liberty.


ACB = Albeda Conference Book (1990), Concept, WRR, The Hague

Adriaansens (1990), ‘Naar een aangepaste verzorgingsstaat", Intermediair April 20, p17—23, adapted version 354th Dies, University of Utrecht

Albeda (1990), "Nieuwe studie naar functioneren SER is wenselijk", De Volkskrant 16/10/90

Arrow (1951), ‘Social choice and individual values", J. Wiley

Auerbaeh & Kotlikoff (1988), "Dynamic economic policy", CUP

Bakhoven (1988), "Een marktgerichte oplossing voor het werkloosheidsprobleem", ESB 3639, Januari 13

Van den Beld (1987), "Het Centraal Planbureau, un peu d’histoire", pp 213—225, in Knoester ed. (1987)

Besseling & Brandsma (1988), "Forecasting the world economy 1988— 1992", Occasional Paper 43, CPB

Van Bochove e.a. (1990), "Een opdracht, geeft niet hoe en met wie", Ministerie van Sociale Zaken en Werkgelegenheid, The Hague

Brittan (1990), "New roads to jobs", Financial Times October 18

Bron (1989), "Creatie van werk", 100, Den Haag

Bruno (1990), "Economic analysis and the political economy of policy formation", European Economic Review 34 273-301

CEPS (1990), "European labour markets: a long-run view", by Bean, Bernholz, Danthine, Malinvaud; CEPS paper 46, Brussels

Cool (1990a), "After twenty years of mass unemployment: why we might wish for a parliamentary inquiry", July 29, CPB library

Cool (1990b), "Why a social welfare (meta) function does exist: The Arrow Impossibility theorem for Social Choice resolved, A better analysis suggested", December 13, CPB internal note

CPB, Centraal Planbureau (1973, 80, 86, 88), "Centraal Economisch Plan", The Hague

CPB (1988), "Gevolgen verlaging minimumloonkosten, een verkenning", werkdocument 25

CPB (1989a), "Collectieve uitgaven, tekortreductie en lastenverlichting 1990-1994: een indicatie van de beleidsruimte", werkdocument 27

CPB (1989b), "Macro-economische effecten van de verkiezingsprogramma’s 1990-1994 van CDA, PvdA, VVD en D66", werkdocument 32

CPB (1990), ‘Athena, Een bedrijfstakkenmodel voor de Nederlandse economie", monograph 30 (‘A multisector model’)

Don & Van den Berg (1990), "The Central Planning Bureau of The Netherlands — its role in the preparation of economic policy", November CPB internal note 3 (D) / 40 (1)

Drèze (1987), in Lawrence & Schultze, "Barriers to European economic growth", Brookings 1987 page 90.

Driehuis & Van der Zwan (eds) (1978), "De voorbereiding van het economisch beleid kritisch bezien", Stenfert Kroese 1978.

Driehuis e.a. ed. (1988), "Challenges for macroeconomic modelling", North-Holland

Drissen & Van Winden (1990), "A general equilibrium model with endogenous government behavior", RM 9005, Amsterdam

ESB 3739: Economisch Statistische Berichten, January 3 1990

Van Eijk (1989), "Concerted action: a remedy against persistent unemployment ?". pp 62-71, in Muller & Zwezerijnen (eds), "The role of economic policy in society", Rotterdam UP

Gelauff e.a. (1990), "Towards an analysis of tax effects on labour market and allocation, a micro/macro approach", research memorandum 68, CPB

Graafland (1990), "Insiders and outsiders in wage formation: the Dutch case", CPB - paper for the EEA Lisbon convention

Graafland (1990a), ‘Persistent unemployment, wages and hysteresis", Doctoral Thesis, Erasmus University, December 20 1990 (Tradition requires that I do not comment on this yet.)

Den Hartog e.a. ed. (1984), "Toegepaste economie, grenzen en mogelijkheden", CPB

Hicks (1989), ‘A market theory of money", Oxford

Helliwell (1988), "Comparative macroeconomics of stagflation", JEL XXVI March pp 1-28

De Kam & Nypels (1990), "Het stuwrneer loopt over", Intermediair 45

Keizer (1990), "De Zweedse crisis", ESB May 9 1990

Knoester ed. (1987), "Lessen uit het verleden", Stenfert Kroese

Kuipers & Van Ees (1990), "Macro-economische theorie en economische politiek", ESB 3714 June 6, pp 32-43

Laidler (1990), "Hicks and the classics", J,. of Monetary Ec. 25 pp 481-489

Layard (1988). "How to beat unemployment", Oxford UP

Leontief (1990), "Some Soviet lessons". Challenge Sept-Oct. pp 14-15

Lindbeck & Snower (1988). "The Insider-Outsider theory of employment and unemployment", MIT

Meade (1985), "Wage-fixing revisited", IEA occasional paper 72

Minford & Rastogi (1989). "A new classical policy programme" In Britton (ed), ‘Policymaking with macroeconomic models", Gower

Muller & Zwezerijnen eds (1989), "The role of economic policy in society", Rotterdam UP

Muysken (1989), "Unemployment In the Netherlands: who cures? Who pays ?", Maastricht

De Neubourg (1990). "Unemployment and labour market flexibility: The Netherlands", ILO Geneva

OECD (1990). "Progress in structural reform; supplement to OECD economic outlook 47", Paris

Van Opstal (1990), "Minimumloon en werkloosheid; een empirische analyse of micro-data uit 1979 en 1985", Research memorandum 69, Centraal Planbureau

Pelkmans, Interview with, (1990), "Dr Jacques Pelkmans (Kroonlid): ‘Analyse Is voor de SER belangrijker dan politieke invloed’" Maatschappijbelangen nr 10 Oct. Haarlem, pp 246-248

Pen (1987), "Feestbundel", ESB November 18, pp 1099-1100

Pen (1989), "De economen en de politiek: een interessante relatie", pp 113—128 in Van Braband et al. (eds), "Openbare financien in drievoud", Thieme Zutphen

Piore (1987), "Historical perspectives and the interpretation of unemployment", JEL XXV 4 pp 1834-1850

Ree (1979), "In den eersten stoot pat", Keesing Amsterdam

Van Reijn (1989), "Van basisinkomen naar verlaging minimumloonkosten", SER-bulletin augustus/september

Roebroek (1990), "Sociale zekerheid in Nederland: Pappen en nathouden", Intermediair 41. October 12 pp 51-63

Rutten (1989), "De betekenis van Tinbergen en Zijlstra", SLA lecture, Dutch Ministry of Economic Affairs.

Van Schaaijk (1983), "Loordifferentiatie en werkloosheid", ESB, 21 september

Van Schaaijk (1988), "Overzicht gewaarborgd", Maandschrift Economie, pp 372-378, CPB reprint no 214 ("Microtax ‘88")

Schuyt (1990), "Het suddert voort, zolang er wat te halen valt", De Volkskrant October 6, het vervolg page 9

Schrijver (1987), "Analyse van macro-economische beleidspreferenties met behulp van optimale besturingstheorie", Research memorandum 35, CPB

SCP (1990), "Sociaal en cultureel rapport 1990", Rijswijk

Standing (1988), "Unemployment and labour market flexibility: Sweden", ILO (WEP), Geneva

Stevers (1979), "Na prinsjesdag in de Volkskrant", Stenfert Kroese

Van Stiphout (1989), "Unemployment components and employment policies: a reflection on the capacities of the economy", paper for the conference "The art of full employment" Maastricht 28—30 September 1989.

Summers (1990), "Understanding unemployment", MIT

Swank (1989). "The objectives of Democrats and Republicans In the United States, 1956—1984", Rotterdam

TInbergen (1956), "Economic policy: principles and design", North-Holland

Tobin (1990), "On the theory of macroeconomic policy", De economist 138 nr 1 p1-14

Uno (1989), "Measurement of services in an input-output framework", North-Holland

Van Velthoven (1990), "The applicability of the traditional theory of economic policy", J. of Economic Surveys Vol 4 no 1

Visscher (1990), "Het enquete-recht en de botheid van een minister", Trouw 18/9/90 p11

VN (1990): "Het grote heimwee van minister Bert de Vries", Ornstein & Van Weezel, Vrij Nederland 15 september 1990 pp 6-10

Wolfson (1990), "Towards a theory of subsidisation", De Economist 138, nr 2, p107-122

Summary of main propositions


Samuelson’s neoclassical synthesis is still valid. Micro/meso/macro and equilibrium/disequilibrium parable families all unite into what we might also plainly call economics. The unity shines clearly from our practical computer models and the way in which we use those. Also, the ordeal since the 1970s has the merit of making us much more aware of public choice intricacies, as an important element In the synthesis. This involves incentive and moral hazard spill-over effects from the public sector to the economy at large. What labour market economists call rigidities in the labour market, public choice theorists call bureaucracy or worse. In the synthesis, both are true; and most relevant is how the pain can be cured.


Unemployment arises from keynesian and classical causes, but of these the clash between the distribution of productivity and the socially imposed distribution of income is the most dramatic. This analysis actually is rather well known. Perhaps less well known is that there exists an alternative manner of administration which could greatly increase social welfare. This involves a job guarantee for the lowest paid, and temporary job search subsidies for the higher paid. This would increase Phillips curve elasticities, though of course a more intelligent approach is to get rid of the notion of ‘un’employment anyway; and at the same time, to use full employment measures to combat inflation.


Changing the formal structure of the economy however involves public choice rigidities. Entrenched lobbies and interests are a very part of our analysis of the problem. Having those in a society is not basically bad, but what counts is that the powers be distributed in such manner that the optimum optimorum is feasible. Our present system apparently has not been built with warning signals and self-repairing devices: so that trying to repair it now anyway, in fact saddlepointing society, would have to involve stronger measures. The very fact that my analysis of the causes of unemployment is old, points to the need for other measures. Serious study by parliament is required to evaluate the economic evidence and to create the political consensus to accept both the economic solution and the more arbitrary specifics of a particular arrangement. In other words, we would wish for a parliamentary inquiry.


Next to various other institutions, the Centraal Planbureau features strongly as a topic of inquest. It is supposed to provide an integration and a stronger empirical (scientifically warranted) base to economic policy. But it has neglected public choice theory. In its main model, government policy is exogenous, resulting in a major bias, which is well-known in the profession. The Bureau also does not give a proper place to policy optimization techniques (which would highlight the unemployment analysis); there are some technical issues, like the disregard of cointegration techniques and level effects in the substitution between higher and lower paid. There also may be a serious misunderstanding between policy making insiders and outsiders about the value of forecasts, when the Bureau does not take into account quite foreseeable policy errors, like failing austerity programs, labour market efficiency errors, and the like. Among the policy insiders, it is well known that the Bureau short term forecasts are ‘conditional upon policy succes’ — and In that sense the responsibility shifts from the Bureau to the policy makers. But then it is very remarkable that the more ‘independent’ Bureau studies, noteably the evaluations of political party election programs, adopt a similar approach; though some more criticism is present, fundamental critique and analysis are wholly absent. But such critique can easily be defended from the viewpoint of the celebrated major five economic policy goals which are widely accepted throughout Dutch Society.


An example of an immediately effective policy measure would be: to allow, as of tomorrow, existing longer term unemployed, at the lowest Collective Agreement pay level, to use 50 percent of their benefit as a subsidy for employment: and let them be free to look for a suitable employer. While this gets going, one may discuss in more ease, how this can be extended to a general system.


This solution to unemployment is obviously relevant for the process of European integration, for the reconstructing nations in Eastern Europe and the Soviet Union, and world mass poverty. It would help enormously when Holland would give the example of how things can be done properly.