Jon Elster

  • Essays in Trespassing: Economics to Politics and Beyond by Albert Hirschman
    Cambridge, 310 pp, £20.00, September 1981, ISBN 0 521 23826 9
  • Shifting Involvements by Albert Hirschman
    Martin Robertson, 138 pp, £9.95, September 1982, ISBN 0 85520 487 7

In Anglo-American social science Albert Hirschman occupies a position at once central and peripheral, or at least anomalous. Of his centrality there can be no doubt. As one of three permanent members of the School of Social Science at the Institute for Advanced Study in Princeton, he has a unique vantage-point for gathering and influencing scholars from all over the world. His Strategy for Economic Development (1958) had an immense impact on development economics, by introducing the notions of backward and forward linkages, and by preaching the virtues of unbalanced growth. The short and incisive Exit, Voice and Loyalty (1970) was an instant classic of political science no less than of economics. In that book he looked at ‘exit’, i.e. taking one’s business elsewhere, and ‘voice’, i.e. protest against the leadership, as alternative ‘responses to decline in firms, organisations and states’. This way of rephrasing the distinction between economic and political behaviour proved very useful, although the main virtue of the book lay elsewhere. Drawing on an extremely broad range of empirical knowledge, sifted, purified and rearranged by an acute and imaginative analytical mind, it generated a state of almost intolerable intellectual excitement in the reader. For many of us it provided a durable model of what a work in the social sciences should be like: free of jargon, devoid of abstract theorising, wide-ranging in application, yet with an intense analytical focus. Among current practitioners of the social sciences only Thomas Schelling and Amos Tversky come to mind as demonstrating to the same extent Hirschman’s quality of controlled imagination.

Yet Hirschman is not a part of the social-science establishment. This has nothing to do with his European origins. Many of the other scholars who left Europe in the late Thirties or the early Forties went on to become pillars of the American academic community. Rather his outsider status – as I perceive it – is due to the faint aura of the amateurish surrounding his work. To achieve power in the social sciences, one must have something teachable to impart, be it a formal theoretical approach or techniques for collecting and analysing data. Hirschman can have no pupils, only kindred spirits. It would be incorrect to say that he is a jack of many trades, but master of none, for in his ‘trespassing’ from one discipline to another he is indeed a master of the inventive analogy. He is a professional all-round amateur: his task is that of perceiving connections, and connections between connections, rather than that of grinding out theorems and correlations. Writers of his ilk are at the pinnacle of the profession, in the triple sense of being supremely gifted, having the ability to survey a variety of disciplines and being parasitic on what goes on at the lower levels. It is good that there are some people like him, and good that not everybody is like that.

Sadly, the two books under review do not come up to the high standards to which we had become accustomed. To those of us who want him to be perfect, they come as a disappointment. To put it bluntly, but in a friendly spirit, Hirschman appears to have been spoilt by success. There is an element of self-indulgence, sometimes of self-congratulation, that prevents him from achieving the same rigour and clarity that characterised his earlier work. The self-indulgence shows itself in a predilection for unsubstantiated speculation. The self-congratulation crops up on the numerous occasions when he feels obliged to tell the reader that some observation or other is novel, that a certain connection ‘has not been duly noticed’, ‘has remained invisible’, is ‘unfamiliar’, ‘surprising’, ‘overlooked’, ‘has hardly been noted’. These are mere irritants, and it might appear needlessly uncharitable to mention them. The matter is more serious when Hirschman, in the opening chapter of his Essays in Trespassing, states that his earlier work on unbalanced growth was a forerunner of Herbert Simon’s work on ‘satisficing’, when the chronology tells us that Simon’s famous article preceded Hirschman’s book by four years. I also think it detracts from his achievements when he goes out of his way to pat himself on the back, as in the following, not uncharacteristic passage from an essay on policy-making in Latin America: ‘the fact that I was led right into [these speculations] from seemingly innocuous and abstract reasoning about the characteristics of policy-making sequences shows how studies in the field can serve to illuminate not just how tax or education policy is made – an important enough endeavour in itself, of course – but can contribute something to the understanding of the most important, baffling and tragic events in recent Latin American history.’ Yes, perhaps, but why not leave such reflections to the reader?

In Hirschman’s recent work, the substantive analysis of social phenomena is increasingly embedded in, and at times almost submerged by, a variety of meta-inquiries. There is, first, the taxonomy of social theories with an eye to showing how his own approach relates to others. This is of course a legitimate procedure, but Hirschman sometimes is too carried away by the delights of pigeonholing. There are too many phrases like the following: ‘It is at this point, among others, that the linkage approach and the staple thesis make contact with the development-of-underdevelopment thesis.’ Secondly, there are numerous excursions into intellectual history. In his previous work on The Passions and the Interests (1977), subtitled ‘Political Arguments for Capitalism before its Triumph’, Hirschman explored at some length, and with great success, the 18th-century idea that le doux commerce had a beneficial impact on political stability, over and above the consequences for economic growth. In Shifting Involvements the historical perspective is also important, and generally very illuminating. The historical sketches offered in the various Essays in Trespassing are less interesting, and, more often than not, only preliminaries to the construction of typologies. Some inaccuracies occur. Thus Marx is said to have denied ‘any claim of mutual benefit from trade between capitalist and “backward” countries’ – Marx in the Theories of Surplus-Value makes the point that the former exploit the latter even though both parties gain from exchange. Also, the reference to John Rawls is misleading. His statement in A Theory of Justice that ‘a departure from the institutions of equal liberty ... cannot be justified or compensated for by greater social or economic advantage,’ was explicitly made within the context of ideal theory, neglecting the problem of how to get from here to there. Hence Rawls is prepared to accept, as are those accused by Hirschman of ignoring him, that second-best considerations might justify sacrifice of liberty in an imperfect society. Yet Hirschman’s slightly cavalier attitude to the classics is not overly disturbing. Even those who know little of the history of ideas would not take Hirschman for an authority, and even those who know much can profit greatly from some of his insights.

Thirdly, and less acceptably, he engages constantly in a highly speculative sociology of knowledge. Not content with comparing his views to those of others, and tracing the genesis of the latter, he also tries to explain the emergence – and non-emergence – of theories in terms of social or intellectual ‘needs’. These explanations in some cases have a good deal of intuitive plausibility. It may well be true that the doctrine of the Invisible Hand ‘assured those who had been brought up with the permanent injunction to serve the public weal, yet somehow found themselves absorbed by money-making activities, that they had by no means betrayed their calling,’ and that this, in fact, explains the extraordinary success of Adam Smith’s theory. Yet some kind of specific evidence would be needed before this explanation could be seriously entertained. If the doctrine succeeded because it relieved the guilt feelings of the conquering bourgeois, then presumably these feelings could be identified in their unrelieved state, and their ‘relief’ observed as an event rather than as a mere absence. Hirschman might well reply that in the intellectual division of labour, his is the task of devising hypotheses, not of testing them. The problem, however, is that in the sociology of knowledge it is all too easy to invent plausible-sounding explanations of theories in terms of the needs they serve.

In many cases, moreover, Hirschman’s explanations do not sound particularly plausible. Discussing the early hopes for development economics, he argues that ‘the focus on rural underemployment was sufficiently similar to the Keynesian concern with unemployment to give the pioneers a highly prized sensation of affinity with the Keynesian system, yet it was also different enough to generate expectations of eventual independent development for our fledgling branch of economics.’ Well, yes, maybe. It would have been useful had there been some evidence that this delicate balance exists outside Hirschman’s imagination. In his well-known essay on the ‘tunnel effect’ Hirschman speculates about why this tendency had not been dealt with earlier. He suggests that the explanation may be that ‘social scientists live in an intensively competitive atmosphere in which envy and “relative deprivation” are far more prevalent than hopefulness caused by someone else’s advance’ – the latter being the key mechanism behind the effect in question. This is not the only occasion on which he feels called on to explain the non-emergence of a theory. Discussing Hegel’s safety-valve theory of imperialism, he asks why Marx did not pay any attention to it, and suggests ‘the hypothesis that Marx repressed this thought (and the Hegelian passage which he knew) because of his will to believe in the immediate prospects of the proletarian revolution.’ In Shifting Involvements he argues that the success of Mancur Olson’s Logic of Collective Action (1965) ‘owes something to its having been contradicted by the subsequently evolving [wave of public movements, marches, protests, strikes and ideologies]. Once the latter had safely run their course, the many people who found them deeply unsettling could go back to The Logic of Collective Action and find in it good and reassuring reasons why those collective actions of the sixties should never have happened in the first place.’ I leave it to the reader to determine whether these speculations are worth assigning to print.

At the level of substantive social analysis, one hallmark of Hirschman’s writing is a preoccupation, almost an obsession, with ironical, ‘dialectical’, paradoxical and counterintuitive findings. To some extent this is of course true of any social scientist worth his salt. His heart beats quicker at a chance of undermining the conventional wisdom. Yet Hirschman goes beyond the call of duty in his tendency to turn coincidences and mishaps into dialectical twists. It is no doubt true that ‘spending of the tax proceeds for ... “unproductive” purposes can be a lesser evil than their expenditure for supposedly productive investment projects that turn out to be failures’ – but so what? At one point Hirschman himself seems to be aware of his excessive enthusiasm for such perverse connections, when he notes that ‘the lesson implicit in these examples is not that erosion is good for development – although it can work out that way if fertile lands remain to be opened fairly close to the eroded ones.’ Observe the key use of the modal ‘can’ in these passages – a characteristic feature of Hirschman’s reasoning. He explicitly defends his emphasis on possibilities rather than certainties, commenting critically that ‘social scientists often consider it beneath their scientific dignity to deal with possibility until after it has become actual and can then at least be redefined as a probability.’ Yet the combination of dialectical and possibilist reasoning makes for tame conclusions: evil can turn into good, good can give rise to evil.

At one point Hirschman criticises Latin American analysts for their excessive interest in the negative side-effects of economic policies. Temperamentally, he is much more of an optimist, as reflected in the title of an earlier book of essays on the Latin American experience, A Bias for Hope (1971). This leads him to be more intrigued by the curious ways in which evil turns into good than by the opposite mechanism, being in that sense closer to Mandeville or Merton than to Marx. The affinity with his somewhat older contemporary Merton is also evident in the way he tends to use the unintended, positive side-effects to explain the behaviour from which they derive. As in Merton, the functionalist mode of explanation is rarely made explicit, but there is a strong suggestion that the beneficial side-effects provide the meaning and hence the explanation of the phenomena in question. One striking example is the Principle of the Hiding Hand, proposed in his earlier Development Projects Observed (1967) and alluded to in several places in the books under review. This is the fortunate tendency in human beings to underestimate the difficulty of the tasks they set for themselves – fortunate in that it offsets another deep-seated tendency to underestimate one’s own creative powers. Hirschman never states, yet in my view strongly suggests, that the tendency is to be explained by this offsetting consequence. When it occurs, it does so because it has that consequence, i.e. serves that function.

Here are some scattered examples of covert functionalism from the Essays in Trespassing. When he writes that ‘one historical function of the rise of development economics was to inspire confidence in the manageability of the development enterprise,’ the word ‘function’ may carry no excess meaning beyond ‘positive consequence’, yet the invocation in the same passage of the Principle of the Hiding Hand suggests otherwise. Later he argues that the inflationary technique for carrying out import-substituting industrialisation in Latin America was a ‘brilliant social invention to get around structural obstacles’ – in fact, ‘the invention was the more brilliant the less conscious or planned it was.’ Read in isolation, the passage might be taken as rhetoric, but this becomes less plausible in the light of the many other similar statements found in his work. In an essay on inflation, for instance, he also refers to the ‘latent benefits’ it brings, describing it as a ‘device’ and, once again, an ‘invention’. When in that essay he turns from the explanation of inflation to its social and political consequences, he notes that inflation has the useful consequence of helping the government resist political pressures towards more spending, and then adds that this ‘helps explain the strange incapacity to reduce inflation to zero of the authoritarian, stability-oriented governments that have come to power in various Latin American countries in the wake of hyperinflation’. In other words, although the essay is organised around the distinction between explaining inflation and tracing its consequences, the latter are also invoked to explain it, in true functionalist vein.

My complaint with this procedure is not that consequences never explain their causes: they often do, by a variety of feedback and reinforcement mechanisms. I am only arguing the need for a mechanism, since otherwise functional explanation becomes a frictionless, arbitrary speculation. Hirschman’s ingenuity and powers of observation allow him to come up with a number of instances in which erosion, inflation, self-deception, imperialism and, of course, unbalanced growth have had positive net consequences that are easily obscured by the more obvious negative short-term or local impact. Yet I wonder whether he does not tend to fall in love with his paradoxes, endowing them with a stability and a significance which they do not always possess. He is caught, therefore, on the horns of a dilemma. Either the counter-intuitive results are mere coincidences, hardly worth the careful attention accorded them in his work, or else they are intended to have some explanatory power, and then he is open to the charge of indulging in unsubstantiated speculation. Neither possibilism nor functionalism seems very promising as an avenue of social research.

Essays in Trespassing comprises 14 chapters, written between 1973 and 1980. With one exception they are organised around five of Hirschman’s earlier books. The exception is the opening essay, on ‘The Rise and Decline of Development Economies’, which tries to explain why development economics has by now slowed down and reached the pace of the object of its study, the economic development of the Third World. He finds the explanation in two sets of causes, one related to ideology and one to world events. To bring out the ideological defeat of development economics he constructs a typology of approaches to development, based on two dichotomies. First, theories can be distinguished according to whether they accept or reject the claim that trade between industrial and underdeveloped countries brings a benefit to both parties. Secondly, some theories assume that the same economic tools can be brought to bear on backward economies as have been used to explain advanced countries, whereas others argue that backwardness requires a different set of concepts and theories. Development economics claimed that there are mutual benefits to be derived from trade, and also that one had to create new economic categories to explain backwardness. Increasingly, however, this claim was attacked by an unholy alliance of orthodox neoclassical economics and Neo-Marxist theories, the former rejecting the need for new theories and the latter the mutual-benefit postulate. If development economics did not recover from this criticism, it was because its practitioners had by then become disillusioned by the turn of events in the underdeveloped countries. To the extent that aggregate economic growth had taken place, it had as its concomitant either political oppression or economic inequality or both, vitiating the easy assumption that ‘all good things go together.’ The essay makes for a good story, but is, perhaps, somewhat less plausible as an explanation. In particular, could one not argue that the lack of economic development, rather than its negative side-effects, was the main cause of disillusionment?

The second essay is a brief comment on Hirschman’s first book, National Power and the Structure of Foreign Trade (1945). He makes the point that by virtue of this book he is ‘the frequently unacknowledged founding grandfather’ of dependency theory – perhaps he means forerunner, which carries no implication of causal influence. He then goes on to criticise his own original effort and later dependency theorists for having neglected the possibility that dependency might contain the seeds of its own destruction – a characteristically dialectical and possibilistic argument, to which my earlier comments apply.

Next follow three essays around ‘The Strategy of Economic Development’. The first looks at ‘the changing tolerance for income inequality in the course of economic development’, introducing the ‘tunnel effect’. This explains why in the process of economic development people may for some time put up with inequality because they believe that the benefits of growth will ultimately trickle down to themselves, much as drivers in the left lane of a two-lane tunnel will put up with a jam because there is some movement in the right lane that holds out a hope for themselves. This, Hirschman argues, is the opposite mechanism of the ‘relative deprivation effect’, whereby an individual experiences a loss of welfare by observing the advances of others. Yet the two phenomena are not strictly comparable. Relative deprivation – and its strict opposite, altruism – arise because one individual’s welfare is an immediate source of utility, negative or positive, for another. The tunnel effect, by contrast, depends on one individual’s welfare having diagnostic value as to the future welfare of another. This, at least, seems to be the main idea, although in one place Hirschman also suggests that the welfare of one person may causally affect that of another, if the second has grounds for believing that the first will share some of his wealth with him. Having produced some rather impressionistic evidence for the existence of the tunnel effect, Hirschman suggests some explanatory uses to which it might be put, and discusses various determinants of its strength. Two main conclusions are drawn. First, ‘if growth and equity in income distribution are considered the two principal tasks facing a country, then these two tasks can be solved sequentially if the country is well equipped with the tunnel effect.’ Next, a distinction is made between two kinds of ‘development disasters’: those which occur in highly-segmented societies where groups do not identify themselves with each other to the extent necessary for the tunnel effect to operate, and those which occur because the leaders ‘fail to realise that the safety valve, which the effect implies, will cease to operate after some time.’ The essay is crammed with ideas. Not all of them are perfectly controlled, yet there is no doubt that this is vintage Hirschman.

The same cannot be said of the two essays that follow, ‘A Generalised Linkage Approach to Development, with special reference to Staples’ and ‘The Turn to Authoritarianism in Latin America and the Search for its Economic Determinants’. The First of these relies on an excessive generalisation of the notions of forward and backward linkage, first extended to cover ‘consumption linkage’ and ‘fiscal linkage’, and then restated as a distinction between ‘inside linkage’ and ‘outside linkage’. Although Hirschman, as always, manages to say a great many interesting things within this framework, he does so in spite of it, not because of it. The ‘reference to staples’ concerns the importance of rice, sugarcane, coffee, tobacco, cocoa etc in economic development. Hirschman is able to show that the physical characteristics of these staples have important economic consequences, yet spends too much time forging highly tenuous links between this observation and the vaguely related theories of Kelvin Lancaster and Marx. The essay on authoritarianism has an even higher ratio of meta-analysis to substantive discussion. His own contribution, drawing heavily on the earlier work on the tunnel effect, lies in a discussion of the interplay between the entrepreneurial function and the reform function in the course of economic and political development. Yet it appears to me that this is at best a conceptual proposal, not a set of theoretical propositions.

Hirschman’s Journeys toward Progress (1963) looked at Chile, Colombia and Brazil as different instances of economic development. Three of the Essays in Trespassing relate to this work. One is very explicitly ‘a return journey’, a second look at some of the problems and explanations offered in the earlier study. It is useful reading, but not, by the nature of the case, very original. The brief note on ‘Hegel, Imperialism and Structural Stagnation’ records his surprise that Hegel in 1821 sketched a theory of imperialism that can be read as a forerunner of Hobson and Luxemburg. I believe he fails to understand the place of this argument within The Philosophy of Right: Hegel was concerned to explain why civil society had to be kept in check by the state, by pointing to some consequences which – hypothetically – would ensue were it not. Moreover, the typology of development theories to which the reading of Hegel inspired Hirschman is too briefly presented to be persuasive, or even suggestive. The essay on ‘The Social and Political Matrix of Inflation: Elaborations on the Latin American Experience’, although in places highly speculative, is a more solid piece of work. I particularly liked the notion of a ‘political multiplier effect’ that causes inflation by requiring the government to grant favours to many groups in order to put together a viable political coalition for a project of its own choice. I was not convinced by the criticism of the Prisoner’s Dilemma model of inflation, nor by his argument about risk aversion. By trying to cover too much ground, the analysis becomes rather patchy.

Four essays cluster around Exit, Voice and Loyalty, making further distinctions and suggesting further applications. Two of them are ‘return journeys’ without a clearly specified focus, which nonetheless make several valuable contributions. A central theme is the idea that the costs of voice, i.e. of political participation, may turn into benefits if such activities come to be seen as ends in themselves. The idea is also incorporated into Shifting Involvements, and is discussed below in connection with that work. Conversely, he makes the point that exit may be more costly than he assumed in his book: viz. if there are transaction costs associated with switching from one supplier to another. Two other essays apply the exit-voice apparatus to the theory of the state and the problem of European integration. The first of these argues the important point that the availability of exit damages the capability of contemporary capitalism to reform itself, since capital flight is a constant possibility. Hirschman adds, in a characteristic flash of insight, that this may apply more to the nations in the capitalist periphery than to those in the centre, suggesting ‘a key to the old puzzle why anti-capitalist revolutions have consistently broken out at the periphery rather than at the centre of the capitalist system’. By contrast, in earlier stages of capitalism labour migration provided a form of exit that was uniformly stabilising – preventing revolution and by that very fact permitting gradual reform. The essay on integration is somewhat more strained, with a meta-theoretical rather than a substantive focus. Here we find the interesting suggestion that the creation of a supra-national community might hold out a hope for the solution of regional problems within the individual nations, by giving the regions a new interlocutor not burdened by the heritage of past conflicts. Generally speaking, this group of essays forms the most interesting part of the book, although none of them taken by itself makes a contribution comparable to the tunnel effect.

Finally there are two relatively slight pieces related to The Passions and the Interests. The note offering ‘An Alternative Explanation of Contemporary Harriedness’ introduces an idea further elaborated in Shifting Involvements: participation in public life often demands more time than anticipated, so that one becomes overcommitted and, in fact, harried. The essay on ‘Morality and the Social Sciences’ is engaging, with an apt quote from Hölderlin to explain why there is so little talk about morality in works of social science.

Hast Du Verstand und ein Herz, so zeige nur eines von beiden,
Beides verdammen sie Dir, zeigest Du beides zugleich.
(If you have reason and a heart, show only one of them.
If you show both at once, they will both condemn you.)

Yet the essay is rather rambling, and soon forgotten.

Shifting Involvements is a much more ambitious enterprise than any of these essays, with the exception of the study of the tunnel effect. The ratio of meta-analysis to substantive discussion is lower, the focus is better defined. Compared to his two previous monographs, it is less rewarding, no doubt because much of the ‘evidence’ offered is so impressionistic and even speculative. Yet I should emphasise that it fails only when measured against this high standard. It raises important and deep questions, and in some cases goes a long way towards resolving them. The subtitle of the book is ‘Private Interest and Public Action’, and its thesis that Western societies tend to go through a cycle of collective behaviour, with the concern for private interest giving way to action in the public interest, and a subsequent swing back to the private sphere. The book is organised in eight chapters. The first introduces the idea of ‘disappointment’ as the key dynamic element of the cycle. Chapters Two and Three look at the disillusionment with material wealth, while the two following chapters argue that the disappointed consumers are then led into the public arena. Chapters Six and Seven demonstrate the frustrations of public life, while Chapter Eight studies the ensuing privatisation which makes the cycle start all over again. The immediate source of inspiration clearly was the contrast between the private life-style of the Fifties, the turn to public involvement in the Sixties, and the reprivatisation which occurred in the Seventies, but Hirschman suggests that the cycle may be a more permament feature of our societies.

Hirschman’s task is related to what some economists and psychologists have studied under the heading of ‘endogenous preference change’: preferences lead to action, and action leads to new preferences. Whereas orthodox economists tend to assume that preferences are stable over time, this alternative approach starts from a premise stated in italics: ‘The world I am trying to understand in this essay is one in which men think they want one thing and then, upon getting it, find out to their dismay that they don’t want it nearly as much as they thought, or don’t want it at all and that something else, of which they were hardly aware, is what they really want.’ This, however, is somewhat misleading in the context of shifting involvements. What the quoted passage describes is a process of learning, which converges towards the ‘real’ wants of the individuals. The book, however, deals with something much less rational – a cyclic change of preference that does not allow us to talk of what the individual ‘really’ wants. The persons Hirschman describes are rather like someone who, when in London, prefers Paris over London while, when in Paris, prefers London. Yet there is also a difference. Hirschman is concerned with explaining preference change in terms of the ‘disappointment potential’ of certain specific objects and activities – not in terms of particular people for whom the grass is always greener on the other side of the fence. Hence the cycle is not inevitable, in the sense of following from an inner compulsion, but depends on the contingent nature of the objects to which the disappointed individual turns. Hirschman does not exclude that a learning process might occur and converge towards some more durable satisfaction, but it is no part of his purpose to examine this possibility.

He finds two sources of disappointment in material consumption goods (and services). The first is related to their specific qualitative properties, the second to their sheer novelty. With respect to the first mechanism, Hirschman singles out consumer durables and social services as having a particularly high disappointment potential – the former because they tend to hang around even after our need for them is exhausted, the latter because our expectations are so vaguely defined and often excessive. The ‘general hostility toward new wealth’, on the other hand, arises because of two double binds that are described as follows. First, ‘if [new material wealth] filters down to the masses, the conservatives are aroused because it threatens the social order. If it does not filter down, the progressives are appalled by the widening disparity in consumption standards.’ Next, ‘some new products could have the dubious distinction of being simultaneously the targets of ... two apparently contradictory critiques ... they could be “trifling” as innovations and conveniences, yet have most serious environmental or ecological side-effects.’ Examples of these double binds are, respectively, automobiles and household sprays. I must confess that I found the two chapters in which Hirschman details the frustrations of consumption quite insubstantial and almost boring – a rare quality indeed in his work. One may easily find oneself yawning when Hirschman asks: ‘Could it be, incidentally, that Robespierre, with colifichet preying on his memory, came up with chétif, a simple rearrangement of the last two syllables of Rousseau’s favourite term?’

Next, Hirschman turns to the explanation of why consumer dissatisfaction should lead into the public arena. As he notes, this is a case in which exit leads to voice – it is exit from the way of life in which exit is the predominant reaction to disappointment. Two roads into this arena are described. The first explains the change of interest by endowing the consumer with a hierarchy of preferences, along lines recently explored by Amartya Sen and others. On Hirschman’s account, the change in second-order or ideological preferences may precede and facilitate the change in the first-order preferences that actually guide action. Thus consumer dissatisfaction first brings about a change in the consumer ideology, and only later a reversal in the preferences for consumption v. public life. The former is the more fundamental change, yet requires some precipitating event to lead to the latter. This, Hirschman claims, ‘corresponds to the essence of the human condition – “neither angel nor beast”, in Pascal’s unequalled formulation. To explain important turns in our lives and societies exclusively in terms of precipitating events would downgrade us to mere playthings of change; to attribute such turns only to autonomously occurring changes in volitions would on the contrary make us appear as more noble and capable of self-determination than we really are.’ These are deep waters, both philosophically and psychologically, and I did not feel that Hirschman’s treatment yielded much insight, a fact perhaps reflected in the extreme paucity of empirical examples in this chapter.

The second mechanism by which consumers turn into citizens is more specific, and more interestingly described. It is in a sense Mancur Olson’s theory of collective action turned on its head, by postulating that such action may be desirable rather than costly for the participants. According to Olson, the main obstacle to collective action is the omnipresence of the ‘free rider’, the person who wants the benefit of a public good such as a pollution-free environment or an orderly society while not being willing to incur the costs of setting it up. And, in fact, why should he be? If everyone else contributes, then he can get a free ride – and if no one else does, his contribution would have no impact. Hirschman undermines this argument by turning the contribution into a supplementary benefit – the ‘happiness of pursuit’ – that may in many cases prove more important than the achievement of the official goal being pursued. Hence ‘the benefit of collective action for an individual is not the difference between the hoped-for result and the effort furnished by him or her, but the sum of these two magnitudes!’ The joys of public life, of participating in a common effort for the common good, come in addition to the good itself, and in fact remain of value even if the good is not achieved.

I am not quite sure how this argument is to be taken. If read only as an explanation of why many people get involved in public life, I think it is an important and correct observation. If, however, it also offers such ‘in-process benefits’ as good reasons for going public, I think it fails – and, I would add, disastrously. Consider the following passage: ‘the very act of going after the public happiness is often the next best thing to actually having that happiness (and sometimes not even the next best thing, but much the best thing of the whole process ...).’ This may be a true statement ex post, but could it ever motivate the individual ex ante? Could a person enter into a movement which in his view had no chance of success, merely for the sake of enjoying the benefits of participation? The answer surely is yes. In the various consciousness-raising activities over the last decades, many people have done exactly that. I would deny, however, that they could rationally count on succeeding in achieving happiness through this kind of bootstrap-pulling. To pursue the happiness of pursuit is self-defeating unless there is some substantive first-order goal in which one believes. What brings happiness is the activity of striving with seriousness and a sense of purpose – with the eye firmly fixed on some external goal, not on one’s own character development. Hence I would question Hirschman’s algebraic formula. The hoped-for result and the effort furnished by the individual do not contribute separately and additively to the benefit of collective action. The former is the only independent variable, although it contributes both directly and indirectly, by enhancing the value of participation. The process may be more important than the goal, in the sense of bringing greater benefits, yet it is logically parasitic on the goal.

I am not saying that Hirschman embraces the narcissistic view of politics, yet he sails close to doing so. For one thing, in the following chapters on the frustrations of public life he does not list the staleness of participation for its own sake as a source of disappointment. For another, the confusion between ex ante and ex post is a characteristic feature of Hirschman’s thinking in other contexts, notably in the Principle of the Hiding Hand. The most charitable, and I believe plausible, reading of this chapter of Shifting Involvements is simply as an analysis of why collective action may be easier to achieve than Olson had made us believe. Yet adherents of the view that politics is mainly about self-realisation may well think that Hirschman lends support to their theory.

The next two chapters discuss the frustrations of public life in terms of, respectively, overcommitment and underinvolvement. The first is explained through initial ignorance about the demand public action makes on one’s time. Hirschman quotes Oscar Wilde to the effect that socialism couldn’t work, because it would take too many evenings, and Rousseau to the effect that men must devote themselves either wholly to the state or wholly to their private pursuits. Moreover, the problems of dirty hands in politics lead many to recoil in distaste. On the other hand, of course, the arcane delights of wielding power may also become addictive, as Hirschman recognises. Yet there is such a thing as voluntary disintoxication in politics, he argues: i.e. acting on second-order preferences to get rid of the first-order addiction to politics. The more familiar theme of forced underinvolvement is discussed mainly with reference to voting, a rather tame way of participating in public life, especially after the introduction of the secret ballot. Most of what Hirschman says about this source of frustration is true and interesting, but it is not obvious that it fits well into his general theory of cyclical behaviour. He notes himself that the frustrated voter may turn into an activist, rather than retiring to the sphere of privacy. In fact, voting is public only in the sense of being concerned with the public good: it is not public in the sense of being part of the public life. Hence the frustrated voter cannot retire to the private sphere, since his frustration derives from being there already. As noted above, a source of disappointment more specifically located in public life is the disillusionment of the activist who finds that it is futile to set up as a goal of behaviour what can only be a by-product, such as self-realisation or consciousness-raising.

The chapter on privatisation points to corruption as a main factor pushing people out of public life, and to the market ideology as a main element pulling them back to privacy – easing the transition by making the pursuit of private interest appear as a public-spirited action. The chapter is not only brief but also quite slight, as are the few pages of conclusion. Summing up for myself the impact of the book, I found it in numerous small observations and in one big idea: participation in public life may offer private benefits, thus facilitating the exit from the private sphere. Hirschman might have explored more fully the ambiguities inherent in this option, but he should certainly be credited with having opened up an important field of study. The grander theory of cycles in collective behaviour left me unimpressed: the mechanisms are too heterogeneous, the evidence too scanty. Perhaps the most promising insight occurs on the penultimate page of the book, where he notes that the debate does not concern one but rather two dichotomies. On the one hand, there is the opposition between the private and the public. On the other hand, the conflict between the instrumental and the expressive sides of politics. Too much of the book assumes implicitly that these two coincide, yet I would strongly argue that they are distinct. Politics, in my view, should be public in nature, instrumental in purpose – thus avoiding both the autistic model of politics embodied in the voting system and the narcissistic view that the only goal of politics is the process itself.