Who Governs? Democracy and Power in an American City

Who Governs? Democracy and Power in an American City (ISBN: 9780300103922) was written by Robert Dahl in 1961, with a 2nd edition published in 2005.

Part 1

Sociology of New Haven

The larger part of part 1 is a demography, history, and survey of the persons of socioeconomic and governmental import in New Haven.

• Office holders are studied by their private occupations, how competitive their elections were, what socioeconomic environment they were born into, ethnicity, etc.

• The memberships of exclusive social clubs, urban redevelopment projects, and public education offices are tracked for Social Notables.

• C level executives of large (50 most valuable by local property assessment) corporations, owners of large (local property assessed at $250k+) wealth, C level executives of any bank or utility, and C level executives of multiple (3+) significant (total assessed at $250k+, or manufacturing with 50+ employees, or retail with 25+ employees) businesses are tracked for Economic Notables.

There was a patrician class formed by a few elite families with close connections to church ministers.

• All similarly religious and socially conservative.
• The approved nominees won annual elections without competition.
  • Upper chamber of general assembly elected by paper ballots, but support for candidates outside the approved 12 had to be made public.
  • Blank ballots were cast as protest.
  • State general assembly passed Stand Up law in 1801.

Social notables pursue professions, not offices. Economic notables tend towards self proclaimed apoliticism.

• Afraid of facing financial consequences of political stances.
• The two groups do not significantly overlap.

Turning point is Jefferson's recall of Adams' midnight appointees. The recalled collector of customs, Elizur Goodrich, was quickly made a professor of law at Yale. Merchants representing "more than seven-eighths" of New Haven port sent a letter to Jefferson in support of Goodrich. By 1818 the Patricians disappear from ballots. (p. 15)

• Democrats and Republicans nominate industrialists, entrepreneurs, etc., going forward.
• Secret ballots starting in 1826.

Theory of Ethnic Politics

Dahl also sneaks a theory of ethnic politics into part 1.

Ethnic politics as a "transitional phenomena" across three stages.

1. first stage: proletarian ethnic class dependent on politicians from other, more established, more assimilated ethnic groups
2. second stage: ethnic class diversifies economically and politically
   • key aspect of assimilation is new, cross-cutting social cleavages forming within the ethnic class

   • ethnic politicians cannot run purely on the virtue of their own ethnic identity

     • need to exploit new generation of proletarian ethnic classes
3. third stage: ethnic class is proportionally distributed throughout socioeconomic strata
   • ethnic political candidates are not compelling
   • class is now now "ex-plebes"
4. tracked in successive immigration waves of Germans, Irish, Russians ("mainly Russian Jews"), and Italians
5. key argument is that ethnic politicians face an eventual but certain loss of their support base

   • indivisible benefits (infrastructure, services) eventually win out over divisible benefits (jobs, contracts, nominations, grift).

Dahl specifically looks at the rise and fall of Irish ethnic politics in New Haven, which gave way to the rise and fall of Italian ethnic politics. He tracks proportions of candidates and office holders by race (imputed from family names).

Importantly, Dahl argues against theories structured on race and/or class.

• voting patterns do not follow economic class lines consistently, but they do follow ethnic classes consistently, so the two must be separate

  • ethnic politics specifically is not class politics in disguise

• failure of socialist parties in political competition, despite ideal environments in New Haven for socialist platforms
• divergence in parties between ethnic classes

  • Irish-Americans went Democratic, Italian-Americans went Republican

• rapid shift of Black Americans voting Republican to voting Democratic during Great Depression

Part 2

Theory of Organizations

Part 2 begins to introduce a theory of organizations.

On a fundamental level, how do individual people participate and influence democratic government?

• Only elected officials truly have a direct influence
• Of the registered voters, there is:

  • an apolitical stratum

    • do not have strongly held preferences on policy, or may not have consistent preferences
    • may not participate in politics at all
    • are motivated by the status quo and inertia rather than rational incentives

  • a political stratum

    • the aggregated opinions on policy spaces probably mirrors public opinion, because there is an electoral/social incentive to match and be in the in-group
      • not creating policy, just reacting to politics
    • politicians seem to pay attention to the approval/disapproval ratings of this group
• Politicians seem to have an incentive to manipulate the general public, especially the more influenceable apolitical stratum
  • diversionary issues to distract from real and inconvenient ones
  • fearmongering and muckraking to keep convenient issues relevant

Meanwhile, individual leaders have diverse motivations. No attempt to model these.

Dahl posits these axioms:

• a durable coalition depends on giving supporters some desired resource (money, prestige, fun, salvation, etc)
  • business leaders can generally get away with just money, but sociopolitical leaders need to obtain resources before they can be distributed. this generally requires government as a mechanism for coercing the collection of resources from non-supporters.
• there is no untouchable classes that are absolutely excluded from coalitions.
• legal methods are preferred over illegal methods
• in the public, leaders must be seen as supporting the prevailing national creed (esp. that democracy is the best form of government), even if that support isn't privately held or accurate in reality.

Leaders usually require subleaders

• Communicating with supporters
• Collecting specific information about supports, such as what resources they desire
• Especially for sociopolitical organizations that can only promise resources in the future, collecting resources
• Legitimizing the organization

Leaders can communicate different promises to subleaders as compared to the general public. Covert benefits vs overt benefits.

• Subleaders can and will defect if they don't get what they were promised.
• Incentive for leaders to align the covert and overt benefits, so as to keep the organization efficient.

Some organizations are vocational. Subleaders are professionals who get paid in money, regardless of what benefits are promised to supporters. Leaders have an obvious method for controlling subleaders in this case.

• The inverse is importantly true: leaders of volunteer organizations lack a powerful method of controlling subleaders.

Compared to the individual people, subleaders do seem to have a consistent and non-negligible (but indirect) degree of influence. Compared to leaders themselves, subleaders seem to have a consistent incentive scheme that can be modeled.

Theory of Political Organizations

An important specialization of the theory for political organizations, especially parties.

• Subleaders are absolutely needed.
  • Almost always avocational: political organizations have little or no way to generate profits; operate on donations and private funding
    • the 'professional politician' is even seen as derogatory
• Significant overt promises are generally necessary to build a competitive coalition.
  • Have to advertise a comprehensive platform.
• Benefits for subleaders and supports are always dependent on winning elections.
• Subleaders are usually sourced from the pool of supporters. The die-hard supporters seek opportunities to become more directly involved.
  • In addition to the incentives outlines above, this creates a further incentive to align overt and covert promises.

In summary, leaders of political organizations make significant overt and covert promises, but subleaders are a meaningful compliance and alignment mechanism. Abandonment by subleaders en masse would cripple the organization. Subleaders want alignment of overt and covert promises. Leaders should want to be efficient and align the overt and covert promises.

Examination of party nominations policy space. State party nominations had always been decided by committee in Hartford. In 1955 the state legislature passed a law requiring direct primary elections in some circumstances. Democratic Party membership in New Haven skyrocketed from 16,500 to 22,000 before the 1959 primaries. Still, less than half of the registered members turned out.

• Even before the legal change, leaders always espoused democratic ideals and advertised platforms that seem to have been strategic (predicting rivals' platforms, conservative in promises, etc).
• Party leaders changed behavior around the legal changes.
• Rivals did attempt to use the legal change to mount a takeover, but it was overwhelmingly a failure.

The theoretical framework suggests that party leaders aimed to use mass registration to drown out vocal minorities that had always been participants in the nomination process.

Dahl also introduces the idea of measuring the power of political leaders, especially factional leaders within the same party, by counting their subleaders that appear in democratic rituals.

Democratic ritual provides...

• legitimacy for leaders
• a normalized, regular opportunity for the public to join as supporters
• a normalized, regular opportunity for die-hard supporters to increase participation as subleaders
• a dispute resolution mechanism
  • this is the important detail for examining party nominations by this framework: there was a real dispute in what platform the party should push, and it was resolved internally (in the establishment's favor) before the general public became involved

Examination of urban redevelopment policy space.

• Origins in early 1900s but Federal Housing Act was passed and Henry Wells (prof of poli sci at yale, dem alderman of 1st ward) became a leading proponent
• Richard Lee ran for mayor unsuccessfully repeatedly on the Democratic ticket
• Republican mayor William Caletano and Democratic alderman/speaker Norton Levine brokered a deal to establish an urban redevelopment agency in 1952
• Richard Lee won mayors office in 1953 running on urban redevelopment platform
  • Subsequent election of 1955 gave significant popular mandate for the platform

The theoretical framework suggests urban redevelopment was introduced to politics by leaders but had latent, mass, non-partisan appeal; the leader that activated it first benefited politically. Persistence in messaging and strategic politics kept the policy space at the forefront of public opinion.

Dahl studies the introductions of motions, the outcomes of motions, the list of who voted in favor, and the list of who voted against to determine major influencers in the complex of agencies and bodies in that policy space. Restricts attention to 1950-1958.

• Mayor and development administrator were the major pushers.
• Economic elite were the effective veto power.
  • Twice was joined to a refusal to join or chair the Citizens Action Commission (CAC)

More on the CAC:

• 25 members, 6 committees, about 30 subcommittees, and over 400 members.
• Rarely opposed the mayor. For the most part, the members only contributed technical/professional services.
• Legitimized urban redevelopment as an open and nonpartisan policy space.

Given that the only effective veto power was an organization that had little actual, direct influence, suggests that indirect influence was at play.

• CAC was effectively a focus group of economic and social notables.
• A mechanism for avoiding conflict, rather than resolving conflict.
  • If the CAC opposed a motion, it would never pass the public opinion, so don't push it any further.

Dahl also examines public education, especially with regard to teacher's unions, school administration positions, and the way that those positions are appointed.

The point of part 2 is that subleaders seem to be marginal actors in policy spaces.

Part 3

Part 3 describes subleaders in New Haven.

As a group, subleaders are mostly older men, they are much more likely to have education beyond high school and own property, and they are the only group with any significant number of people having income above $10,000. They don't seem to be more successful, more ambitious, or have received a particular sort of education (parochial vs public). They just seem to have had a head start (i.e. fathers far more likely to have a white collar occupation). Subleaders are almost categorically politically active, even when controlling for income.

There is little overlap between leadership pools (leaders and subleaders taken together) across policy spaces.

• Specifically examines party nominations, urban redevelopment, and public education

Subleaders in the party nomination policy space are demographically similar to the voter pool. In contrast, subleaders in the urban redevelopment policy space are highly divergent.

• Income, occupation, significantly educational attainment, and especially which neighborhoods they live in.

• As explored earlier, there was little to no trend in who the effective 'pushers' were in each policy space.
  • Only consistency was the mayor and development administrator.
  • But there were meaningful trends to who the effective 'veto powers' were.
    • These powerful subleaders were specialized.
    • Effectively, there is dispersion and specialization of power.

Dahl posits that the difference between the policy spaces is prestige. Social and economic notables do not seek elected offices. Public education is a respected and well paid career. Urban redevelopment, as a 'created' policy space, had commissions designed with prestige and power in mind.

The historical trend in the structure of the political system seems to be:

1. separated policy spaces with separate leaderships, to
2. an executive-led centralization of the political system, to
3. rival coalitions (i.e. parties) competing for centralized control.

Power was significantly restricted to a patrician class as first. Power drifted to social and economic notables over time. Mayors became more powerful in the 1900s. But grand reform plans (like professionalization of public school administrators, as opposed to political appointment) did not pull through even when the mayor made overt promises to do so. The net effect was a weakening of separated, specialized leaders, not the accumulation of power into one organization under the mayor's leadership. Political competition tears down centralization in time.

Parts 4 and 5

The remainder of the book introduces a theory of political resources. There are many such resources that individuals can use to influence politicians. They vary in efficacy across policy spaces. They are all distributed unequally but without cumulative inequality. Everyone has some, few have an overwhelming amount. Specialization according to political resources leads to specialized leaders in each policy space; a fragmented political system. This is all built on the presupposition of the earlier theories.

My thoughts

There should always be some doubt on claims that covert behavior is rare. It is necessarily, obviously rare to observe.

The elephant in the room: Dahl's theory of ethnic politics fails to explain the experience of Black Americans.

I think the sociology is sound. I think the foundational theories (esp. the theory of ethnic politics, theory of organizations, and the theory of political organizations) are interesting and testable and can be built upon for further science.

I also think many of his conclusions ultimately rest on anecdotal evidence for which there are no (other) region studies experts to evaluate the evidence. This is a fundamental and recurring problem; we need region studies before we can have have social science. The doubt that Dahl selected only evidence supporting his conclusions for collection and reproduction cannot be cast off.

CategoryRicottone