Forty years ago, in Capitalism, Socialism and Democracy, Joseph Schumpeter argued that the expectations for democracy bequeathed to us by the classical tradition were hopelessly unrealistic. He offered an alternative bare-bones definition, one actually satisfied by many large-scale nation-states. ‘The democratic method,’ he said, ‘is that institutional arrangement for arriving at political decisions in which individuals acquire the power to decide by means of a competitive struggle for the people’s vote.’ This minimal definition has since been treated by many as a pale shadow of the full-bodied democratic idea. After all, élites may ‘compete’ for office in elections without any very meaningful or widespread participation by the electorate, without majorities being satisfied on particular issues, or without the groups that care most intensely about particular issues getting what they want. While this may be the kind of democracy realised in many nation-states, the question naturally arises whether so pale a version of the real thing is worth having.

We may grant that Schumpeter’s ‘competitive’ model (the mere requirement that politicians compete for elective office) is pale and uninspiring as an ideal, but the conundrum facing democratic theory is that there are serious reasons to be distrustful of each of the more full-bodied versions – at least as any of them might be instituted at the level of the large-scale nation-state. Of course, some issues might be decentralised and turned over to small, highly participatory sub-units – a possibility explored in Robert Dahl’s new book Dilemmas of Pluralist Democracy: but if we restrict the discussion to those issues handled at the national level, the burden of proof is on those who would go beyond Schumpeter’s minimal and uninspiring position.

There are four major ideals of democracy that seem to have a hold on the popular imagination. Let us see what ‘more democracy’ might mean in each of these four senses. First, suppose that by ‘more democracy’ we mean increasing the fit between majority opinion and public policy. In all the major Western democracies there are serious impediments to majority rule on specific policy issues. The framers of the American Constitution were self-conscious in their design of checks and balances, many of which stand as impediments to majority rule. The separation of Executive, Legislative and Judicial branches, the territorial design of the election system, the bicameral character of the legislature, the evolution of judicial review, all provide impediments that have frustrated majority opinion on major policy issues.

A more general impediment to majority rule on particular policies, one not special to the American system, results from the simple fact that electoral competitions typically select parties or office-holders, not specific policies. As a result, the process of building coalitions in constructing party platforms, or in forming a government from various parties (in multiparty systems), will typically consist in piecing together groups with intense interests in particular areas of policy. Such coalitions of intense minorities produce packages or platforms of policies, which, on balance, may have majority support. But each particular policy in the package, if submitted to a referendum in isolation, will often be opposed by a majority of voters. In other words, each group is willing to go along with the overall package because it gets its intensely valued policy. But no particular policy is in the package because it has majority support. This pattern, which Robert Dahl dubbed ‘minorities rule’ in a classic analysis twenty-five years ago, is commonplace throughout the Western democracies.

But would a more majoritarian system necessarily be a good thing? In considering majoritarian reforms, it is important to demystify majority rule of its special aura. Empirical studies of American schoolchildren show that they learn at quite a young age to decide controversies by majority rule. It is a device for collective decision that they adopt quite naturally. Yet no adequate account of democracy would simply equate it with majority rule. Democracy also requires restraints and impediments to majorities. For majorities, like other winning coalitions, can ride roughshod over losing coalitions. Majorities may be tyrannous and, in that way, undemocratic. When Idi Amin expelled the Asians from Uganda that action was apparently resoundingly popular among black Ugandans. However, few of us would have regarded it as justified if Amin had simply held a procedurally fair election or a plebescite on the issue. Even within the confines of democratic theory, majority rule seems an insufficient principle to justify such outcomes. While the definition of those minority rights that might properly restrain majorities has always been an especially controversial part of democratic theory, few would deny that there are some things that majorities may not do to minorities. Hence reforms designed to correct the fit between public policies and majoritarian opinion are not necessarily an improvement.

We currently lack any adequate criteria for the appropriate limits on majority rule within democratic theory. The American founding fathers considered the problem in the context of an 18th-century consensus on certain God-given ‘natural rights’, and for them it was self-evident that majorities ought to be forbidden from depriving any persons or groups of those natural rights. In the 20th century, however, this consensus on God-given rights has evaporated in a climate of secular scepticism and religious pluralism. One result has been that democracy has increasingly come to be identified with majority rule itself and any impediments to majority rule have, increasingly, come to be regarded as antidemocratic.

An entire school of constitutional theory, for example, has developed around the notion that because of scepticism about rights, an activist judiciary should be limited to perfecting the process of political competition and representation – see, for example, John Hart Ely’s Democracy and Distrust. Judicial activism in the interest of protecting substantive rights not required (or previously decided) by the political process itself would, on this view, be undemocratic. It would be undemocratic because it would involve some judge’s opinions about rights standing against the will of the people.

If, however, a judge had intervened to protect the Asians from being expelled in Uganda, would that have been undemocratic? Or would it have been an invocation of the ban on tyranny of the majority – a ban compatible with the democratic tradition? The difficulty, of course, is that when we move from the obvious and extreme case of Amin’s Uganda to more routine public policies, we lack adequate mechanisms for identifying majority tyranny. Is it majority tyranny for governments in both Britain and the US to fight inflation through toleration of increased unemployment? Why are those who lose their jobs as a result of such policies not like Uganda’s Asians: after all, they lose drastically from policies that produce, at the very best, much smaller benefits to a larger number (the population who benefit from a moderation in inflation)? To answer such questions we would need a clearer and more serviceable definition of majority tyranny. Because of the lost consensus on natural rights, any such definition is bound to be controversial. But unless the problem of majority tyranny is dealt with somehow, we will lack any adequate criteria for deciding how majoritarian a democratic system ought to be, even in theory.

Consider now a second major ideal of democracy. If the problem is that winning coalitions may do dreadful things to losing ones, the unanimity rule takes on a special attraction, at least as an ideal. Of course, the unanimity rule has obvious impracticalities for the large-scale nation-state. For one thing, it would require enormous decision-costs – i.e. time spent negotiating agreements. But some theorists have offered it as an ideal which might have to be compromised in practice for such reasons, but against which more familiar democratic forms can be judged – see, for example, James Buchanan and Gordon Tullock, The Calculus of Consent.

If, however, the possibility of tyranny-through-omission is granted, then the unanimity rule loses its special attraction even as an ideal. The distinctive fact about the unanimity rule is that it gives everyone a veto over new policies. If some group will lose out decisively in the event of a policy’s not being adopted, then the unanimity rule can produce tyranny-through-omission. Imagine some vulnerable minority on the verge of starvation, which might be saved through modest efforts at redistribution. Would we really wish to require unanimity here? Such a requirement would hold the policy hostage to vetoes from any who might disagree or who might lose even slightly from redistribution. Avoidable starvation, whether imposed by an isolated veto or by majority rule, seems a clear example of a tyrannous policy choice.

Consider another model of democracy. Majority rule and unanimity both take no account of intensities. They require reference only to the numbers of people on each side of an issue. Perhaps if people feel more strongly it would be more democratic to let them get their way. According to this notion, if one compares the numbers and intensities on each side of an issue, the preponderant weight of opinion should carry the day. Of course, such a notion would be difficult to institutionalise, since there are notorious difficulties about interpersonal comparisons of the intensity of preference. But in a rough way, we can and do make such comparisons and such a notion might provide a basis for a moral principle even if it might be difficult to institutionalise.

A little reflection, however, leads to the conclusion that such a version of democracy ends up equating it with utilitarianism – with the satisfaction of relevant preferences overall in the society, given the intensities and numbers of people on each side of an issue. After the many recent attacks on utilitarianism – John Rawls’s A Theory of Justice, Robert Nozick’s Anarchy, State and Utopia, for example – the distributional implications of utilitarianism are all too clear: it may justify extreme losses for some for the sake of minor gains to a sufficiently large number of others. Or, more generally, any deprivation to some persons may be justified provided that the numbers and intensities of preference on the other side outweigh this deprivation. As Rawls speculates, there are even conditions under which utilitarianism might support slavery provided that the benefits to the slaveholders are great enough. Clearly the notion of a balance of intensities does not offer a solution to the problem of majority tyranny.

Consider now a fourth major model of democracy – direct as opposed to indirect (or representative) democracy. Until some time in the late 18th century, democracy was generally thought to be a curiosity limited to city-states and small groups. It had to be dramatically reformulated, via the crucial innovation of representative institutions, in order to be applied to the large-scale nation-state. In many ways, the thin character of the Schumpeterian competitive model is a result of the many compromises in the democratic ideal that had to be made in the transition from the direct to the indirect (or representative) version.

Now, however, technological innovations seem to be making direct democracy more practical. In Columbus, Ohio, for example, the cable television system operates in both directions, permitting instantaneous polls on virtually any subject. Suppose an entire nation were to be wired up the way Columbus now is: would such instantaneous referenda be a more democratic way to determine public policy?

The answer depends in part on how we deal with the tyranny issue already mentioned. Without an answer to that problem we do not know how responsive to public opinion – expressed directly or indirectly – a democratic system ought to be. And the proposal brings another problem into focus. Would we really want the population at large settling, by direct referendum, the many policy issues facing a large-scale nation-state? Consider the placement of an MX missile system, the merits of Keynesian, monetarist and supply-side economics, federal funding for abortions and the complex legerdemain of Reagan’s ‘New Federalism’. Even the most ardent democratic reformers could not expect the population at large to become sufficiently knowledgeable about any large number of such complex issues. Yet such a list only begins to identify the endless variety and complexity of the issues that would be placed before the public in any major movement towards direct democracy. The trouble with direct democracy at the level of the large-scale nation-state is no longer the technological feasibility of everyone, in some sense, gathering together – although control over debate and air-time raises interesting issues. The problem is the deeper one about the quality of the preferences that such technical innovations would place in control of the details of national public policy.

The essence of Schumpeter’s thin competitive notion was that the electorate may, at best, intelligently choose a government that will deal with specific policy issues. To expect real public policy competence on any broad range of issues to become universal at the mass level would seem utopian. Of course, some people, if given the responsibility, would take politics more seriously than they do now. Nevertheless, increasing the degree of direct democracy opens the door to government by manipulation and public relations experts rather than to meaningful reform. The current disillusionment in the US with the proliferation of Presidential primaries tells precisely this story. By making the process much more direct, democratic reforms have so trivialised the process that influential observers have written nostalgically about the old smoke-filled rooms in which candidates used to be selected. This disillusionment recently culminated in a new series of counter-reforms designed to place the Democratic Presidential nomination more squarely in the hands of party leaders and office-holders. Unless major reforms were somehow directed at the process of public opinion formation itself, increasing the direct character of political institutions, at least at the national level, would not appear to be much of an improvement.

However, while much might be done to increase the variety of viewpoints represented in the mass media, it seems unimaginable that any institutional reforms might lead to near-universal mass competence on the details of national policy. There is a discouraging piece of public-opinion-polling folklore – the story of the pollster who interviews a housewife on decision-making in her family. She refers the pollster to her husband saying that she only deals with ‘minor’ issues such as the family’s finances, the choice of schooling for their children and the purchase of the family home. The pollster, somewhat surprised, asks the housewife what ‘more important’ decisions are left for her husband. She replies: ‘Détente with the Soviet Union, our relations with Red China and policy in the Middle East.’

Schumpeter’s minimal and uninspiring competitive model would leave these ‘more important’ decisions largely to specialised policymakers who have to survive the test of periodic election. It would not require ordinary citizens to acquire the same expertise on national issues that they naturally have for the ‘minor’ issues that touch their lives directly. I know that there is an uncomfortably élitist tone to the suggestion that more democracy than this might not be a good thing. But when the variety of senses that might be attributed to ‘more democracy’, at least at the national level, is made explicit, and when the serious issues facing each of these directions for reform are faced squarely, then Schumpeter’s thin notion looks better by comparison. At this stage in the 20th century, one can only conclude that ‘democratic theory’ is in such an unsatisfactory state that it can only serve as a fragile and inadequate guide to any more ambitious version of democratic practice.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.


Vol. 4 No. 14 · 5 August 1982

SIR: James Fishkin’s 2000-word survey of the current state of ‘democratic theory’ (LRB, 17 June) is a tour de force, but in at least one important respect it is seriously flawed by some national insularity. It is certainly the case that ‘in the 20th century… consensus on God-given rights has evaporated in a climate of secular scepticism and religious pluralism,’ and this proposition holds true not only for the USA but for many other democracies. But it is decidedly not the case that, as a result, ‘we currently lack any adequate criteria for the appropriate limits on majority rule within democratic theory’ – unless ‘we’ is limited to students of the USA’s domestic arrangements to the exclusion of all else.

Precisely because of the decline in the credibility of ‘natural rights’, and of the correlative expansion in the space available for ‘legitimate’ tyranny – ex-emplified in its most extreme form by Hitler’s Germany and Stalin’s Russia – the elaboration of new standards of ‘human’ rights has been conducted on the international plane since 1945. As a result, there is now in place a substantial structure of treaties, binding in positive international law on a large number of the world’s nations, which set coherent and precise limits to what governments – including those democratically constituted by the popular will of a majority – may do to the individuals under their jurisdiction. Between them, these treaties cover a wide spectrum of civil, political, economic, social and cultural rights; some are global in their extent, and others regional; and some even go to the previously unprecedented length of establishing international courts with the power to hand down binding judgments against national governments at the suit of their own citizens. Many of the 21 nations of non-Communist Europe, for example, have been compelled to repeal or modify their own legislation, validly enacted by their own democratic legislature, because the European Court of Human Rights at Strasbourg has subsequently held that it entailed an unjustifiable violation of one of the rights guaranteed by the European Convention on Human Rights and Fundamental Freedoms.

These new standards did not of course spring from a vacuum. Indeed, one of their more influential antecedents was the American Bill of Rights, and the USA played an important part in their elaboration at the United Nations and the Organisation of American States. Yet the USA is alone among the larger democracies in today’s world in not yet having ratified a single one of the major treaties in this field. Among the 35 participating States of the Helsinki Final Act, for instance, only Liechtenstein, Monaco, San Marino and the Holy See are in like case. Professor Fishkin will know better than I why that should be so: Americans are apt to ascribe it to their Senate’s notorious phobia about ratifying ‘foreign’ treaties. Whatever the reason, the USA’s non-adherence decisively weakens the force of its criticisms of oppression, persecution and deprivation elsewhere in the world.

But if students of democratic theory in that great country are searching for a modern consensus on the limits to majority tyranny to replace that shared by their founding fathers, they need look no further than the UN building in New York, the office of the OAU in Washington, and perhaps even some of their own university libraries – but in the section on international law rather than political science (US).

Paul Sieghart
London WC1

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN

Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences