Keywords

JEL Classifications

A central place is held in economics and social science in general by principles and forces making for order or equilibrium in economic and social systems. Disorder and disequilibrium are then understood as resulting from some malfunction of these principles or forces. Explanations of order–disorder or equilibrium–disequilibrium have typically been discipline-bound, dealing with either the political or the economic world. Since the two are interrelated it would be useful to have a construct that bridges them. Such is the claim of the exit–voice perspective. It addresses the changing balance of order and disorder in the social world by pointing out that social actors who experience developing disorder have available to them two activist reactions and perhaps remedies: exit, or withdrawal from a relationship that one has built up as a buyer of merchandise or as a member of an organization such as a firm, a family, a political party or a state; and voice, or the attempt at repairing and perhaps improving the relationship through an effort at communicating one’s complaints, grievances and proposals for improvement. The voice reaction belongs in good part to the political domain since it has to do with the articulation and channelling of opinion, criticism and protest. Much of the exit reaction, on the contrary, involves the economic realm as it is precisely the function of the markets for goods, services, and jobs to offer alternatives to consumers, buyers and employees who are for various reasons dissatisfied with their current transaction partners.

The exit–voice alternative was proposed and explored in Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States (Hirschman1970, henceforth EVL). Attempts to apply the book’s perspective were made over many areas of social life. In the following, the basic concepts will be recapitulated and, where necessary, reformulated. Subsequently some major applications of the exit–voice polarity will be reviewed.

Basic Concepts

Exit

Exit means withdrawal from a relationship with a person or organization. If this relationship fulfils some vital function, then the withdrawal is possible only if the same relationship can be re-established with another person or organization. Exit is therefore often predicated on the availability of choice, competition, and well functioning markets.

Exit of customers (or employees) serves as a signal to the management of firms and organizations that something is amiss. A search for causes and remedies will then be undertaken and some plan of action designed to restore performance will be adopted. This is one way in which markets and competition work to prevent decay and to maintain and perhaps improve quality.

Exit is a powerful but indirect and somewhat blunt way of alerting management to its failings. Most of the time, those customers and members of organizations who exit have no interest in improving them by their withdrawal, so that exit does not provide management with much information on what is wrong.

Voice

The direct and more informative way of alerting management is to alert it: this is voice. Its role is, or should be, paramount in situations where exit is either not available at all or is difficult, costly, and traumatic. This is so for certain primordial groupings one is born into – the family, the ethnic or religious community, the nation – or for those organizations one joins with the intention of staying for a prolonged period – school, marriage, political party, firm. With regard to buying and selling, voice should take over from exit when competition is weak or nonexistent as in the case of goods and services being produced under oligopolistic or monopolistic conditions, or when exit is expensive for both parties as in certain interfirm relations.

Unlike exit in the case of well-functioning markets, voice is never easy. It can even be dangerous. Many organizations and their agents are not at all keen on being told about their shortcomings by members and the latter often expose themselves to reprisals if they utter any criticism (Birch 1975). Even in the absence of reprisals, the cost of voice to an individual member will often exceed, in terms of time and effort, any conceivable benefit from voicing. Frequently, moreover, any effective channelling of individual voices requires a number of members to join together so that voice formation depends on the potential for collective action.

In spite of these problems, voice exists or, rather, it has come into being. Its history is to a considerable extent the history of the right to dissent, of due process, of safeguards against reprisal, and of the advance of trade unions and of consumer and many other organizations articulating the demands of individuals and groups who once were silent. Similarly, the history of exit is the history of the broadening of the market, of the right to move freely, to emigrate, to be a conscientious objector, to divorce, etc. Being two basic, complementary ingredients of democratic freedom, the right to exit and the right to voice have on the whole been enlarged or restricted jointly. Yet, there are important instances of unilateral advances or retreats of either the one or the other response mechanism (Rokkan 1975; Finer 1974).

Interaction of Exit and Voice

As noted, exit is paramount as a reaction to discontent in some circumstances and voice holds a similarly privileged position in others, but frequently both mechanisms are available jointly. In such situations they may either reinforce or undercut each other. The availability and threat of exit on the part of an important customer or group of members may powerfully reinforce their voice. On the other hand, the actual recourse to exit will often diminish the volume of voice that would otherwise be forthcoming and, should the organization be more sensitive to voice than to exit, the stage could be set for cumulative deterioration. For example, after an incipient deterioration of public schools or inner cities, the availability of private schools or suburban housing would lead, via exit, to further deterioration – a turn of events that might have been prevented if the parents sending their children to private school or the inner city residents who move to the suburbs had instead used their voice to press for reform. In their aggregate effects, the individual exit decisions are harmful – an instance of the ‘tyranny of small decisions’ – also because they are likely to be taken on the basis of a short-run private-interest calculus only and do not take into account the ‘public bad’ that will be inflicted, even on those who exit, by decaying inner cities and segregated education (Levin 1983; Breneman 1983).

These kinds of situations are sufficiently numerous and important to be of interest not only as curious paradoxes showing that under some circumstances the availability of exit (that is, of competition) could have undesirable effects. In this connection, EVL stressed the value of loyalty as a factor that might delay over-rapid exit. Loyalty would make a member reluctant to leave an organization upon the slightest manifestation of decline even though rival organizations were available. Provided it is not ‘blind’, loyalty would also activate voice as loyal members are strongly motivated to save ‘their’ organization once deterioration has passed some threshold.

The difficulties of combining exit and voice in an optimal manner are in a sense ‘problems of the rich’: they relate to situations and societies where exit and voice are both forthcoming more or less abundantly, but where, for best results, one would wish for a different mix. Historically more frequent are cases where exit and voice are both in short supply, in spite of many reasons for discontent and unhappiness. There is no doubt, as many commentators have pointed out, that passivity, acquiescence, inaction, withdrawal, and resignation have held sway much of the time over wide areas of the social world. This is largely the result of repression of both exit and voice – a repression that has flourished in spite of the fact that all human organizations could put to good use the feedback provided by the two reaction modes.

Problems in Voice Formation

The development of voice among customers of firms or members of organizations poses a number of problems that were not fully explored in EVL. Critics have asserted that, in its endeavour to present voice as a ready alternative to exit, the book understated the difficulties of voice formation. In examining this issue it is useful to start with the extreme no-voice case: the authoritarian state which is dedicated to repressing and suppressing voice. This situation has given rise to a useful distinction between horizontal and vertical voice (O’Donnell 1986). The latter is the actual communication, complaint, petition, or protest addressed to the authorities by a citizen and, more frequently, by an organization representing a group of citizens. Horizontal voice is the utterance and exchange of opinion, concern and criticism among citizens: in the more open societies it is today regularly ascertained through opinion polls revealing the approval rating of presidents, prime ministers, mayors, etc. Horizontal voice is a necessary precondition for the mobilization of vertical voice. It is the earmark of the more frightful authoritarian regimes that they suppress not only vertical voice – any ordinary tyranny does that – but horizontal voice as well. The suppression of horizontal voice is generally the side-effect of the terrorist methods used by such regimes in dealing with their enemies.

The distinction between vertical and horizontal voice is relevant to the ‘free ride’ argument in relation to voice formation (Barry 1974). For vertical voice to come about, that is, for members of the organization to engage management in meaningful dialogue, it is frequently necessary for members to forge a tie among themselves, to create an organization which will agitate for their demands, etc. But the hoped-for result of collective voice is a freely available public good; hence, so goes the critical argument, self-interested, ‘rational’ individuals may well withhold their contribution to the voice enterprise in the expectation that others will take on the entire burden. Important as it is, this argument has its limitations. First of all, it is addressed only to vertical voice which it mistakenly equates (as EVL did) with voice in general. Horizontal voice is not subject to the strictures of the free-rider argument: it is free, spontaneous activity of men and women in society, akin to breathing. As just noted, extraordinary violence has to be deployed if it is to be suppressed. Under ordinary circumstances, horizontal voice is continuously generated and has an impact even without becoming vertical: in many environments managers of organizations cannot help noticing and reacting to critical opinions and hostile moods of the members, whether or not organized protest movements break out. That the planned economies of Eastern Europe function to the extent they do has been explained on precisely this ground (Bender 1981, p. 30).

Another limitation of the free-rider argument lies in its assumption that individuals will always act instrumentally. Just because the desired result of collective voice is typically a public good – or, better, some aspect of the public happiness – participation in voice provides an alternative to self-centred, instrumental action. It therefore has the powerful attractions of those activities that are characterized by the fusion of striving and attaining and can be understood as investments in individual or group identity (Hirschman 1985).

Some Areas of Application

Trade Unions

In economics, the major application of the exit–voice theme has been the analysis of trade unions as collective voice by Freeman and Medoff in their book What Do Unions Do? (1984). Instead of looking at unions as a monopolistic device raising wages for unionized workers beyond the ‘market-clearing’ equilibrium level or – much the same zero-sum interpretation in different language – as a tool in the class struggle serving to reduce the degree of exploitation, the book finds that a major function of unions is that of channelling information to management about workers’ aspirations and complaints. Collective voice, in the form of union bargaining, is more efficient in conveying information about workers’ discontent – and in doing something about it – than individual decisions to quit, as voice carries more information than exit. The presence of union voice is shown to reduce costly labour turnover. Moreover, the fringe benefits, workplace practices, and seniority rules which unions negotiate often result in offsetting labour productivity increases.

Markets and Hierarchies vs. Exit and Voice

Renewed attention has been given in recent years to the question why some kinds of economic activities are carried on through many independent firms while others, to the contrary, are tied together through bureaucratic and hierarchical relations. In accounting for hierarchy, one approach has directed attention to such matters as uncertainty about the evolution of the market and the technology and in particular to asymmetric availability of information to buyer and seller, creating opportunities for deceitful behaviour (Williamson 1975). Hierarchy is then seen as superior to markets whenever there is need for a sustained and frank dialogue between the contracting parties. Critics of this position have argued: (1) relations between independent firms, such as contractors and subcontractors, are often quite effective in discouraging malfeasance; (2) correlatively, hierarchy frequently leads to characteristic patterns of concealment and control evasion (Eccles 1981; Granovetter 1985); and (3) industry structure varies substantially from one country to another as well as within the same country over time: in Japan, for example, subcontracting is much more widely practised than in the West and in Italy subcontracting has become more widespread in the last 10–20 years.

A formulation in terms of exit–voice is helpful here. The characteristics which are said to justify hierarchy – incomplete information, considerable apprenticing of one firm by the other, openings for ‘opportunistic’ (i.e., dishonest) behaviour, etc. – all make for situations in which there is need for voice: the firms contracting together must intensively consult with, and watch over, each other. But the need for voice does not necessarily imply that hierarchy is in order. Whether voicing is done best within the same organization or from one independent firm to another is by no means a foregone conclusion. Moreover, when the two parties are independent and resort a great deal to voice, the possibility of exit from the relationship often looms in the background. The implicit threat of exit could carry as much clout as that of sanctions in hierarchical relationships.

The argument for hierarchy in cases where voice has an important role to play may arise from thinking of market relationship only in terms of the ideal, anonymous market where voice is wholly absent. But most markets involve voice: commerce is communication, and is premised on frequent and close contact of the contracting parties who deliver promises, trust them, and engage in mutual adjustment of claims and complaints – all of this was implicit in the eighteenth-century notion of doux commerce (Hirschman 1977, 1982). Adam Smith even conjectured that it was man’s ability to communicate through speech that lies at the source of his ‘propensity to truck and barter’. How odd, then, that the need for frequent and intensive communication should be adduced as a conclusive argument for hierarchy.

Public Services: Education, Health, Others

The organization of public services represents a privileged area for the application of exit–voice reasoning – significantly the exit–voice idea had its origin in the analysis of a public service in trouble, the Nigerian railroads (EVL, Preface). Public services are typically sold or delivered by a single public or publicly regulated supplier, for various well-known reasons.

With the production of most public services being thus deprived of the ‘discipline of the market’, problems of productive efficiency and quality maintenance arise necessarily. An obvious way of mitigating these problems is to attempt to reintroduce market pressures in some fashion. For example, when certain categories of goods and services are to be made available either to all citizens regardless of their income or to some deprived social groups, the state and its agencies can sometimes refrain from producing or distributing these goods directly, and instead issue special purpose money or vouchers enabling the beneficiaries to acquire the goods or services through ordinary market channels. In this manner the voucher system reintroduces the market and the possibility of exit. A particularly successful example of the voucher system is the distribution of Food Stamps to low-income persons in the United States. Instead of creating and administering its own food distribution network the state hands out vouchers (food stamps) which the beneficiaries can then use at existing, competitive commercial outlets.

In part because of the success of this programme and in part because of the belief in ‘market solutions’ as the remedy for all that ails government programmes, voucher schemes have been proposed for a large number of other public services, from education to low-cost housing to the supply of certain health services. Voucher systems are appropriate primarily under the following conditions (Bridge 1977): (1) there are widespread differences in tastes and these differences are recognized as legitimate; (2) individuals are well informed about quality and different qualities are easily compared and evaluated; (3) purchases are recurrent and relatively small in relation to income so that buyers can learn from experience and easily switch from one brand and supplier to another.

These conditions are ideally present in the case of foodstuffs, but much less so in the case of, say, health and educational services. Hence the development of voice constitutes here an important alternative strategy for assuring and maintaining product quality. In other words, the beneficiaries of certain public services should be induced to become active on their own behalf, individually or collectively As always, development of voice is arduous because of apathy and passivity of the members, but also because it will often be resisted by the organizations that have been set up to deliver the services. A number of proposals and attempts have been made to introduce more voice into the administration of both health and educational services (Stevens 1974; Klein 1980).

EVL had insisted on the see-saw character of exit and voice interventions in these fields. Education and health systems seemed particularly exposed to the danger that premature exit – of the potentially most influential members – would undermine voice. The opposite relation may also occur, however, for the opening up of the exit perspective could serve to strengthen voice: parents who have been wholly passive because of feelings of powerlessness and fear of reprisals may feel empowered for the first time once they are given vouchers that could be used ‘against’ the schools currently attended by their children, and will be more ready than before to speak out with regard to desirable changes in those very schools.

Spatial Mobility (Migration) and Political Action

Another substantial area of exit–voice applications opens up when exit is taken in the literal, spatial sense. Here exit–voice boils down to the familiar flight or fight alternative. While often institutionalized among nomadic groups (Hirschman 1981, ch. 11), this alternative is not necessarily available in sedentary societies. Here the traditionally available choice is fight or submit in silence. The option of removing oneself from an oppressive environment has become available on a massive scale only in modern times, with the advances in transportation and the uneven opening up of economic opportunity, religious tolerance, and political freedom. Where the option has existed, the interaction of exit and voice has been on display in three principal types of migration: (1) that from the countryside to the city, the oldest and no doubt largest of the modern migrations; (2) the migration from the city to the suburbs, which was most intense in the United States during the fifties and sixties, owing to the spread of the automobile and also to the large-scale migration of blacks and Hispanics into the cities; (3) finally, of course, international migration with its numerous economic and political determinants and constraints. Under this rubric, the international movement of capital also deserves attention.

Looking at the varieties of exit-voice interplay in these diverse settings, it is possible, on the basis of the numerous studies now available, to distinguish the following patterns:

  1. (1)

    In accordance with the basic hypothesis of EVL, exit-migration deprives the geographical unit which is left behind (countryside, city, nation) of many of the more activist residents, including potential leaders, reformers, or revolutionaries. Exit weakens voice and reduces the prospects for advance, reform, or revolution in the area that is being left.

    Something of this pattern can be observed in all three types of migration. Massive rural-urban migration could obviously reduce the potential as well as the need for land reforms which the voice of the countryside might otherwise have precipitated (Huntington and Nelson 1976, pp. 103ff.). The large outward migration from Europe to the United States in the 19th century up to World War I probably functioned as a political safety-valve for the rapidly industrializing European societies of that period, as has been shown for Italy (MacDonald, 1963–1964). In a similar vein, the possibility of westward migration within the United States has been invoked as an explanation for the lack of a militant working-class movement in that country. Finally, the city-to-suburbs migration in the United States has led, at least initially, to cumulative deterioration in the urban areas affected by out-migration in spite of, and in some cases because of, reduced density. At times, the voice-weakening effect of exit is consciously utilized by the authorities: permitting, favouring, or even ordering the exit of enemies or dissidents has long been one – comparatively civilized – means for autocratic rulers to rid themselves of their critics, a practice revived on a large scale by Castro’s Cuba and, on a more selective basis, by the Soviet Union.

  2. (2)

    But the basic see-saw pattern – the more exit the less voice – does not exhaust the rich historical material. The mechanism through which voice is strengthened rather than weakened as a result of exit is distinctive in the case of migration. In some societies the accumulated social pressures could be so high that authoritarian political controls will only be relaxed if a certain amount of out-migration takes place concurrently. This is what happened in the fifty years prior to World War I when the franchise was extended in many European states from which large contingents of people were departing. In other words, the state accommodated some of the pressures toward democratization because it could be reasonably surmised, in part as a result of out-migration, that opening the door slightly to voice would not blow away the whole structure. A similar positive relation between exit and voice may exist today with regard to such southern European countries as Spain, Portugal, and Greece: here the large-scale emigration to northern Europe may also have eased the transition to a more democratic (more vociferous) order.

  3. (3)

    Exit–voice theory posits remedial or preventive responses to any large-scale out-migration on the part of the entity that is being left. A firm losing customers or a party losing members will normally undertake a search for the reasons of such declines in fortune and then determine upon a strategy for recovery. For out-migration such reactions are not easy to identify. In the case of massive rural–urban migration, for example, there is usually no organized entity such as the ‘countryside’ that registers the flight from it and can undertake corrective action. With regard to migration from the city to the suburbs, the situation is not too different. Here entities exist – city administrations – but they have generally been ineffective in modifying the individual decisions of millions of people to move into their own homes in the suburbs.

The analogy to the firm is – or should be – most applicable when the geographic entity losing residents is the State, which is after all a highly organized, self-reflective body with considerable means of action. There is, of course, the already noted possibility that out-migration relieves economic or political stress in a country, is therefore welcome, and may even be encouraged by the state. But massive emigration is at some point bound to be viewed as dangerous. Just like a business firm, the state may then take measures to make itself more attractive to its citizens. One example of this reaction is the national plan for economic recovery and industrialization adopted by Ireland in 1958, in the midst of very high levels of emigration, mostly to England (Burnett 1976). It has also been shown that the pioneering welfare state measures of the late 19th and early 20th century, starting in Bismarck’s Germany in the eighties and then spreading to the Scandinavian countries and Great Britain, were all taken in countries with high rates of overseas migration. These measures can be seen as attempts of states to make themselves more attractive to their citizens (Kuhnle 1981).

The international movement of capital was first commented upon from the exit perspective in the 18th century. Montesquieu and Adam Smith both thought that the threat of exit on the part of movable capital could play a useful role in preventing arbitrary and confiscatory measures against the legitimate interests of commerce and industry. The threat of exit or exit itself was expected to function, like the customer’s exit, as a curb on misconduct, this time on the part of the state. While this relationship is still pertinent, exit of capital often plays a less constructive role today. In the more peripheral capitalist countries the owners of capital have become fully alive to the possibility of removing part of their holdings to the United States or other reliable places in case they become unhappy about the ‘investment climate’. In this manner, capital exit (or flight) will often be practised on a large scale as soon as the state undertakes some, perhaps long overdue, reforms with respect to such matters as land tenure or fiscal equity. Instead of preventing arbitrary and ill-considered policies, exit can thus complicate and render more hazardous certain needed reforms. Moreover, exit undercuts voice: as long as the capitalists are able to remove their patrimony to a safe place, they will have that much less incentive to raise their voice for the purpose of making a responsible contribution to national problem-solving. Capital mobility and propensity to exit may thus be a major reason for the instability of states in the capitalist periphery (Hirschman 1981, ch. 11).

Political Parties

Two principal propositions were put forward by EVL with regard to the dynamics of political parties in a democracy:

  1. (1)

    In a two-party system, the tendency of the parties to move toward the nonideological centre in order to capture the (allegedly) voluminous middle-of-the-road vote is countered by those party members and militants who are on the parties’ ideological fringes, have ‘nowhere else to go’, but just because of that are maximally motivated to exert influence inside the party, by forceful uses of voice.

  2. (2)

    In a multi-party system, with the ideological distance from one part to the next being presumably shorter than in two-party systems, dissatisfaction with party performance is more likely to lead to exit than in two-party systems; in the latter, voice will play the more important role as switching to the other party requires too big an ideological jump. One inference is that parties in two-party systems may be expected to exhibit more internal divisions, but also more internal democracy and less bureaucratic centralism than parties in multiple-party systems.

The first of these propositions has been strongly supported by events subsequent to the publication of the book. At that time, only the nomination of Barry Goldwater to be the standard bearer of the Republican Party in 1964 could be cited in support. Since then, additional evidence has accumulated: from the nomination by the Democrats of George McGovern to contend the Presidential elections in 1972 to the increasing power of the more radical wing of the Labour party and the ascendancy of Margaret Thatcher within the Conservative party and of Ronald.

Reagan among the Republicans. The theory that in a two-party system the two parties would increasingly converge toward some middle ground has been amply disconfirmed.

The second proposition on political parties which was deduced from the EVL framework has undergone several qualifications. For example, in democracies with old cleavages along ethnic, linguistic, and religious lines, the distances between the several parties rooted in ethnic, etc., identities could actually be wider than that between the parties of two-party systems. Under these conditions, the exit voice logic would in fact predict that member participation (voice) in parties of multi-party systems would also be vigorous and exit infrequent (Lorwin 1971; Hirschman 1981, ch. 9).

A more serious complication is being stressed in a work by S. Kernell still in progress. In two-party systems, exit is a particularly powerful move for dissatisfied members as by casting their vote for the other party they are doubling its impact, something they cannot be sure of in multi-party systems. Hence, in case of disappointment with the performance of one’s own party, there could arise a special temptation in two-party systems to switch to the other party so as to punish one’s own. Such a preference for exit is likely to come to the fore primarily when a party in power is perceived as having seriously mishandled its mandate. Under the circumstances, the prospect of being able to punish that party retrospectively could overcome party loyalty and past ideological commitment. This constellation was an important factor in the sharp defeat of the Democratic ticket in the 1980 Presidential elections in the United States.

The Family: Marriage and Divorce

Modern marriage is one of the simplest illustrations of the exit–voice alternative. When a marriage is in difficulty, the partners can either make an attempt, usually through a great deal of voicing, to reconstruct their relationship or they can divorce. The complexities of the interplay between exit and voice are well in evidence here. Just as the threat of strike in labour-management relations, so is the threat of divorce important in inducing the parties to ‘bargain seriously’; but as exit becomes ever easier and less costly (and perhaps even profitable to one of the parties – see Weitzman 1985), its availability will undermine voice: rather than being an action of last resort, divorce could become the automatic response to marital difficulty with less and less effort made at communication and reconciliation.

This is exactly what appears to have happened in the United States during the last fifteen years, i.e. since EVL stated that ‘the expenditure of time, money and nerves’ necessitated by complicated divorce procedures serves the useful, if unintended purpose of ‘stimulating voice in deteriorating, yet recuperable organizations which would be prematurely destroyed through free exit’ (p. 79). In 1970 California adopted a new ‘no-fault’ law on divorce which spread, though often in attenuated form, to most other states (Weitzman 1985). The California law drastically altered divorce procedures: instead of requiring proof that one of the parties was guilty of some specific type of behaviour constituting grounds for divorce, the new law permitted divorce when both or just one of the two parties asserted that the marriage had irretrievably broken down. The possibility of a unilateral decision, of just ‘walking out’, is symbolic of the way in which the California law undercuts the recourse to voice.

With the new regime, the pendulum has swung quite far in the direction of facilitating exit and of thereby weakening voice. It was of course a reaction to the many abuses of the older fault-based system which required costly and degrading adversarial proceedings, and in effect discriminated against the poor. But the framers of the new legislation probably did not realize the extent to which the earlier obstacles to divorce indirectly encouraged attempts at mending the so easily frayed conjugal relationship and how much the new freedom to exit would torpedo such attempts, with the results that one of every two new marriages now ends in divorce.

The Family: Adolescent Development

This is another family situation for whose analysis a formulation in terms of exit and voice has been found useful (Gilligan 1986). Adolescent development has often been portrayed as a process through which the ‘dependent’ child becomes an ‘independent’ adult through progressive ‘detachment’ from the parents. Freud saw this as ‘one of the most significant, but also one of the most painful psychic accomplishments of the pubertal period … a process that alone makes possible the opposition, which is so important for the progress of civilization, between the new generation and the old’ (1905, p. 227). Here is a celebration of exit; Freud’s statement neglects a complementary aspect and task of adolescent development which is to maintain and enrich the bond with the older generation through continued, if conflict-ridden, communication. In other words, voice has an important role to play in transforming the adolescent’s relationship to the parents. The peculiar poignancy of the adolescent–parents conflict resides in fact in the impossibility of relying wholly on voice in resolving it: given the closeness of the relationship, a full accord that would be the outcome of successful voicing risks ending up in incest, as the ‘meeting of minds would suggest a meeting of bodies’ (Gilligan 1986). It is because of the incest taboo that exit must be part of the solution, but different generations of adolescents are likely to achieve emancipation by practising very different characteristic mixes of exit and voice. Moreover, as Gilligan stresses, the balance of exit and voice differs according to gender. Girls place a greater value than boys on continued attachment to the family, and are therefore less attracted to the masculine ideal of independence-isolation. Hence they experience a greater tension between exit and voice.

With this imaginative use of the exit–voice concept, the outer limits of its sphere of influence may have been reached.

See Also