Select Page

[ Fred E. Foldvary, Dept. of Economics, Santa Clara University, and School of Management John F. Kennedy University, California. Bradley Fellow, Center for Study of Public Choice, 1989-91. ]

In the Calculus of Consent: Logical Foundations of Constitutional Democracy, James Buchanan and Gordon Tullock analyzed “the calculus of the rational individual when he is faced with questions of constitutional choice” (p. vi). A key problem in governance, recognized by James Madison and the other Founding Fathers of the US Constitution, is the rise of factions or special interests, which, as stated by Buchanan and Tullock, “try to use the processes of government to further their own differential or partisan interests” (p. 25). These interests engage in seeking transfers or rents which not only divert resources from the public but, as Tullock (1967) has demonstrated, the transfer-seeking process is itself a waste of resources.

The three main methods used in the US and State constitutions to limit the danger of factions have been federalism, the division of government, and constitutional restraints. Federalism divides government into the federal and State governments (and also Indian-national or tribal governments) which in principle were to have separate parallel sovereignty. The division of governments at both federal and State levels into three branches (legislative, executive, and judicial) provides for separate, but interlinked, powers at each government level. Constitutional constrains further limit the powers of each government.

John Arthur (1989, p. 12) notes that the Federalists recognized that “majoritarian government (which the framers refer to as ‘democratic’) is unable to protect liberty or to promote the general welfare.” Despite the presence of all three power-limiting methods, governance in the United States of America has become ever more centralized, and the trend has been for both federal and State governments to grow as a proportion of GDP (see, for example, Stansel and Moore, 1999). The federal structure has been circumvented by concentrating taxation in at the federal level, and then using conditional “revenue sharing” to dictate state expenditures, as well as imposing mandates on the States. The division of powers has been undermined by the dependency of the courts on federal appointments (especially so during the Great Depression when president Roosevelt threatened to pack the Court) and, as argued below, by a common structure of mass democracy which inherently induces transfer-seeking by factions and special interests.

Constitutional constraints have been continuously undermined by the exploitation of ambiguous and vague constitutional language, accommodated by a judiciary which for the most part has not sought to challenge legislation that pushes the constraints. A well-known example is the loose interpretation of the interstate-commerce clause into a rationale for any kind of regulation of commerce.

As James Buchanan (1980) noted, “rent seeking” itself creates barriers to reform, and so its reform requires constitutional changes, rather than just operational reforms under the status-quo constitution. The direction of such reform was indicated by Buchanan and Tullock in their conclusion that that “where possible, collective activity should be organized in small rather than large political units” (pp. 114-5). Mancur Olson (1971, p. 63) had also theorized that in large groups, the incentive is lacking to avoid transfer seeking, but that incentives could be present in a federal grouping, i.e. with smaller groupings.

This public-choice and constitutional-economics theory points to the need for a more fundamental constitutional change if special-interest and factional rent or transfer seeking is to be minimized. The very structure of governance and social choice needs to be decentralized and more radically federated to change mass democracy into a cellular democracy where all voting takes place only in small groups. (As a matter of terminology, in what follows, the term “small group” will refer to the generic structure of voting in small groups which are in some federated system, while “cellular” will refer to a small-group structure with a particular set of rules which, as discussed below, would minimize transfer seeking.)

Small-group governance and voting

In the cellular structure described below, the “calculus of consent” would shift away from rational ignorance and apathy. The small group, in multi-level federation with other small groups, would alter the incentives and thus recalculate the elements of consent. The principles of the calculus (e.g. methodoligical individualism) would not change, but the parameters would, hence the outcome of a recalculated consent.

In cellular democracy, a jurisdiction such as a county or city is divided into neighborhood districts with a population of about 500 persons, with about 100 to 200 households. In some communities, the neighborhood boundary would be obtained from the institutional structure, such as a condominium or residential association or a voting precinct. In others, boundaries would be drawn, which could be altered later if the communities so desired.

The voters in the neighborhood district would elect a council. The population size, 500, is a suggestion, based on having a community small enough so that the voters could personally know the candidates for the council. With a voting base that small, candidates can avoid the expense of mailings and media advertising. They can easily cover all the voters door to door and host local meetings. Those who wish to be candidates may do so at little cost. And once elected, the candidates are easily accessible, both because they are local and because the constituent base is small.

Mass advertising could still take place if there are political parties which candidates belong to, and so in addition to the local personal campaigning, the parties could try to sway voters to vote for their candidates. But this does not detract from the face-to-face nature of the local election process. With mass democracy, the thousands and millions of voters do not personally know the candidates, and the candidates then must project an image over the mass media. The need to finance the media messages is met with a supply of funds from the special interests, who are later rewarded with transfers and privileges. While political parties might still find it worthwhile to have media messages, special interests would find much less reward in doing so. Personal campaigning could overcome the influence of party propaganda.

Each neighborhood district would elect representatives plus one alternate to its council. The alternate participates in the council but does not vote unless the regular representative is absent. The neighborhood shall be designated as the level-one community, level zero being the individual voter. The council members could be removed whenever the voters desired. The small size of the neighborhoods would make it feasible to circulate a petition to recall a council member and then hold an election to replace the representative.

A region containing about ten to twenty neighborhoods would then have a level-2 council. Each level-1 council elects a regular representative and an alternate to the level-2 council from its own regular membership. The alternate that was elected to the level-1 council would then replace the regular member now representing level-1 at the level-2 council. The representative to the level-2 council may also be recalled at will by the level-1 council and replaced by another member.

The suggested number of level-1 councils represented at level-2, about ten to twenty, is intended to be small enough so that the level-2 representatives would be personally known to the level-1 electors. Interestingly, the description by Alex Comfort (1990) of the decentralized governance of the agricultural Santal tribal people of India includes the observation that a “group of between ten to 20 villages constitutes a territorial federation” (p. 101).

Several level-2 districts would then belong to the next higher-level council. The council of level 3 is elected by the level-2 district councils, each level-2 council again electing a regular and an alternate representative to level 3. The level-2 representative elected to the level-3 council is then replaced at level 2 by his alternate from the level-1 neighborhood. By electing alternates, each lower level council thus retains representation at the next higher level council when it elects a member to a higher level council.

Ever higher-level councils are elected up to the highest level council for that jurisdiction, designated as level h. This concept of bottom-up multi-level governance is not new, but the idea has not penetrated to academic or popular discussion. A similar multi-level governance structure extending to a continent or the world was envisioned by Spencer Heath (1957) for proprietary communities, although he did not present any details. Comfort (1990, p. 111) writes that “The Spanish [anarchist] movement was essentially Bakuninist in that it favoured an organisation of society into localised collectives which would federate into local federations, and in turn form broader federations.” Hence this is an old anarchist idea in which society is not a chaotic mass of atomistic individuals, but a hierarchy of small organizations, each in voluntary union with the others. What I have done here is to apply public-choice and constitutional economics to the idea and add some particular institutional rules to give it specificity and provide greater immunization against the disease of democracy, the seeking and taking of transfers, rents, and privileges.

Bryan and McClaughry (1989) have proposed an electoral reform for Vermont, where many small towns still exist, which would partially implement this structure. In their system, much of state governance would devolve to a “shire” (derived from the old English shire) one level above the small town or city neighborhood. They emphasize that genuine representative democracy depends on having a vital direct democracy at its base. In their plan, however, the citizens still directly elect the state legislature, whereas with small-group democracy described here, the state legislature would be elected by the next-lower governance level.

The level-h council elects from its membership an executive and other officers. Alternate representatives replace the executive and possibly other officers. The level-h council members have all been successively elected by lower-member councils down to level 1, and are in effect on leave from the lower-level councils. Hence, if any lower-level council removes its higher-level representative, he is then removed from his higher-level council. A level-h representative could thus be removed from level h by the level-1 council that originally elected him.

Public finance and expenditure

Additional constitutional rules would strengthen the decentralist tendencies of the bottom-up multi-level structure. On the expenditure side, “service substitution” would enable any level-n council to substitute its provision of a service for that provided by the level n+1 level, and deduct the average expense from its tax liabilities to level n+1. For example, suppose that schooling is provided by the level-4 governance, such as a city, and a level-3 district council found the educational level to be unsatisfactory. It could withdraw its schools from level 4 and create its own level-3 school district, and deduct the average cost per student from the tax liabilities to level 4. This partial withdrawal or service substitution would be an alternative to an attempt to reform the level-4 service. The exit option would supplement the voice option to limit the power of the next higher level, as well as to provide a greater incentive for the higher level to serve the lower level more efficiently.

Such a secession possibility is practiced to a limited degree in St. Louis, where a local neighborhood is able to privatize the streets and take over the maintenance. With such local control, the neighborhood association can close off one end of a street as well as to organize better surveillance. Condominiums and residential associations are other examples of local governance and provision of services that are more typically considered governmental public goods (Foldvary, 1994). Selective secession or service substitution also provides greater scope for competition among communities, as analyzed in the theory set forth by Charles Tiebout (1956).

The secession of governance itself could take place in the cellular structure by enabling any level-n councils to secede from a level n+1 council (except for level h) and create an alternative level n+1. For example, a minority which felt itself to be discriminated against could create its own higher-level council from those level-n districts which chose to affiliate with it. Thus, the membership in the higher-level councils, other than the highest level, would be flexible and this broader exit option would act further to check inefficiency and favoritism.

On the revenue side, the bottom-up voting structure is reinforced if the public finances also flow from bottom to top. The basic constitutional rule for public revenues under cellular democracy is therefore that a level-n council obtains its general revenues from the council at level n-1. A level-n council has no authority to tax the councils at lower levels, and only the level-1may directly tax the households and individuals at level zero. There can, however, be some special higher-level sources of revenue, such as royalties from oil and mineral extraction rents, and pollution charges on the sources of pollution, but no general taxes such as based on income or sales.

With each level-1 government being able to select its own source of public revenue, competition among the communities for enterprise and residents would discourage the imposition of taxes on incomes or sales. Property taxes would be most likely. But while existing buildings are rather immobile, new construction can avoid being located where it is highly taxed. Hence, with very localized competition, the property most amenable to taxation without flight or evasion is land sites, which are both immobile and incapable of being expanded (the use of the space can be changed, but the space itself is fixed by nature).

Community competition would thus in most places lead to site rents as the basic source of public revenues. These revenues would originate with the level-1 communities and be passed up to ever higher levels. There could be some inter-level institutional arrangement to assess sites so that localities do not have the incentive to under-assess.

Besides being suitable to local collection without an excess burden, evasion, or flight, the taxation of site rents has been recognized as an effective capture of the externalities of a community. The Austrian economist Friedrich von Wieser (1927, p. 340) stated that “Urban rent is that part of the rental which is paid as a premium for the advantages of the better location.” The Swedish economist Knut Wicksell (1896, p. 113) noted that “the general economic development of the community” increased the value of its land, and he proposed taxing such increases (p. 114). There is a literature on the “Henry George theorem” with the proposition that the optimal amount of public goods is that which maximizes the land rent and is paid for by that rent (Stiglitz, 1977), Henry George (1879) having theorized that land rent is the most efficient and least intrusive source of public revenue. Buchanan and Goetz (1972) found that externalities would be internalized if communities are proprietary and competitive: “Tax shares would have to be related to the size of the locational rent component in individual income receipts” (p. 35).


Russell Hardin (1993, p. 130) stated that “The real magic of liberal democracy often lies in its tendency – sometimes overcome – to decentralize decisions.” Mass democracy, however, has tended to centralize power due to the lack of individual political power by the lowest unit, the voter. The “calculus of consent” falls under the sway of the special-interest “market for legislation.” Bottom-up multi-level cellular democracy offers the structural incentives to maintain decentralist governance and minimize transfer-seeking by achieving the proposal of Buchanan and Tullock (1965, p. 114-5), as noted above, that “where possible, collective activity should be organized in small rather than large political units.”



Arthur, John. 1989. The Unfinished Constitution. Belmont, California: Wadsworth Publishing.

Bryan, Frank M., and John McClaughry. 1989. The Vermont Papers: Recreating Democracy on a Human Scale. Chelsea, Vermont: Chelsea Green Publishing.

Buchanan, James. 1980. “Reform in the Rent-Seeking Society.” In Towards a Theory of the Rent-Seeking Society. Eds. James Buchanan, Robert Tollison, and Gordon Tullock. Texas A & M University Press, College Station, pp. 359-67.

Buchanan, James M., and Charles J. Goetz. 1972. “Efficiency Limits of Fiscal Mobility.” Journal of Public Economics 1: 25-43.

Buchanan, James M., and Tullock, Gordon. 1962 [1965]. The Calculus of Consent: Logical Foundations of Constitutional Democracy. Ann Arbor: University of Michigan Press.

Foldvary, Fred. 1994. Public Goods and Private Communities. Aldershot, UK: Edward Elgar Publishing.

George, Henry. 1879 [1975]. Progress and Poverty. New York: Robert Schalkenbach.

Hardin, Russell. 1993. “Liberalism: Political and Economic.” In Liberalism and the Economic Order. Eds. Ellen Frankel Paul, Fred D. Miller, Jr., and Jeffrey Paul. Cambridge, UK: Cambridge University Press, pp. 121-44.

Heath, Spencer. 1957. Citadel, Market and Altar. Baltimore: Science of Society Foundation.

Olson, Mancur. 1971. The Logic of Collective Action. Cambridge, MA: Harvard University Press.

Stansel, Dean, and Stephen Moore. 1999. The State Spending Spree of the 1990s. Policy Analysis 343 (May 13), Cato Institute.

Stiglitz, Joseph E. 1977. “The Theory of Local Public Goods.” In Economics of Public Services. M. S. Feldstein and R. P. Inman, eds. London: Macmillan. Pp. 274-333.

Tiebout, Charles M.. 1956. “A Pure Theory of Local Expenditure.” Journal of Political Economy 64: 416-24.

Tullock, Gordon. 1967. “The Welfare Costs of Tariffs, Monopolies, and Theft.” Western Economic Journal 5 (June): 224-32.

Wicksell, Knut. 1896 [1958]. “A New Principle of Just Taxation.” In Classics in the Theory of Public Finance. Eds. R. A. Musgrave and Alan T. Peacock. London: Macmillan & Co. Pp. 72-116.

Wieser, Friedrich von. 1927 [1967]. Social Economics. Trans. A. Ford Hinrichs. New York: Augustus Kelley Publishers.