Contribution
It extended the exit–voice framework in three directions the book had neglected: when voice is a benefit rather than a cost, when exit is costly or uninformative, and how organizations, including states, manage the two responses for their own purposes. It also surveys early uses of the framework in political science and urban studies.
Key insight
Exit and voice are not only members’ options; they are feedback channels, and an organization’s survival depends on keeping each between a floor, below which it learns nothing about its decline, and a ceiling, above which it falls apart. Those who run the organization have their own reasons to set the levels.
Assumptions
- The book’s framework: members or customers of a deteriorating organization respond by leaving or by complaining, and both responses can prompt recovery.
- Exit is market-like and voice political; the paper aims to show economists that voice matters for economic phenomena.
Key results
- Voice as benefit (pp. 432–434). Participation can become “a highly desired end in itself”. Problems with a public-interest side, such as a health hazard or car safety, call forth voice where matters of taste call forth exit. Shareholders who long sold their stock instead began to use their influence when public-interest issues arose.
- Information (pp. 437–438). Voice carries detailed information; exit may not even show that customers are discontented, when they churn between equally bad suppliers. Voice suits situations of “ignorance and uncertainty, shared by consumers and producers”, and dominates in periods of general loss of confidence in the traditional system.
- Exit’s costs (p. 439). The book took exit as costless except where loyalty was present; switching costs between firms make it costly anyway, and vertical integration and marriage institutionalize voice.
- Floors and ceilings (p. 441). Every organization “navigates between the Scylla of disintegration-disruption and the Charybdis of deterioration due to lack of feedback”. Managers may encourage the response “least dangerous to its perpetuation in power” (p. 442).
- A hedged theorem (p. 444). A state can control only two of exit, voice and repression; Hirschman does “not wish to make too much of” it.
- Parties and cities (pp. 444–451). Voice is used with special energy by those with nowhere to exit to. Where loyalty is stronger, as in two-party systems, dissatisfied members voice rather than exit. In cities, exit to the suburbs often wins narrowly, so small improvements in voice could change much. Defenders of the National Health Service want potential exiters kept inside as critics.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Voice can be experienced as a benefit, not only a cost | moderate | argument and examples |
| C2 | Voice carries richer information than exit, and exit can signal nothing | moderate | argument; competition-as-collusion case |
| C3 | Organizations need floor levels of exit and voice for feedback and cannot exist beyond ceiling levels | moderate | argument, after Rokkan’s history of European state-building |
| C4 | Managers manipulate exit and voice to protect their position | weak | asserted from the book; illustrated |
| C5 | A state can control only two of exit, voice and repression | weak | proposed with examples; hedged by the author |
Concepts
- exit — leaving: ceasing to buy, resigning, emigrating, moving away.
- voice — trying to change the organization from within by protest or participation.
- loyalty — not redefined here; it appears as what makes exit costly and voice more likely.
- exit-fatigue — after one exit, the inclination to use voice in the new setting (after Orbell and Uno, p. 449).
Connections
It restates and revises the book (LIT-748). It draws on Rokkan for the history of state-building and on Williams for urban mobility as “a substitute for formal politics”. In the record’s reading, voice as feedback is close to Longino’s uptake requirement (THEORY-151): criticism matters only where there is a channel through which it can change the community.
Bearing on the record
Source of THEORY-152.
For the owner’s essay: §6.3’s capture from below and extraction from above have here the members’ side, what they can do when the organization serves them worse, and management’s interest in steering them to the least threatening response. §6.4’s platform case fits the exit that signals nothing, though platforms are not mentioned. §7.2 (acting through one’s own organization) is what voice is. §8.4 has the observation that floor levels of exit and voice are a condition of the organization learning about its own decline. The record’s application to platforms is its own.
Limitations
Informal and partly a survey of others’ work. The theorem is hedged and the party model admitted to be “primitive”. Loyalty is presupposed from the book, not treated. No systematic evidence is offered; examples carry the argument.
Open questions
What determines the floor and ceiling levels for a given organization, and whether members who are locked in, with exit made costly, voice more or simply comply.