There are words nobody likes to be associated with in public, such as racism and imperialism. There are others for which everyone is anxious to demonstrate enthusiasm, such as mothers and the environment. Democracy is one of these. In the days of "really existing socialism", even the most implausible regimes laid claim to it in their official titles, as in North Korea, Pol Pot's Cambodia, and Yemen. Today, it is impossible, outside a few Islamic theocracies and Middle Eastern hereditary kingdoms and sheikhdoms, to find any regime that does not pay tribute to the idea of competitively elected assemblies or presidents. Irrespective of history and culture, the constitutional features common to Sweden, Papua New Guinea and Sierra Leone (when elected presidents can be found there) officially put them into one class, Pakistan and Cuba in the other. This is why rational public discussion of democracy is both necessary and unusually difficult.
There is no necessary or logical connection between the various components of the conglomerate that make up what we call "liberal democracy". Non-democratic states may be built on the principle of the Rechtstaat, or rule of law, as Prussia and imperial Germany undoubtedly were. We have known, since de Tocqueville and John Stuart Mill, that freedom and toleration for minorities are often more threatened than protected by democracy. We have also known, since Napoleon III, that regimes that come to power by a coup d'etat can continue winning genuine majorities by successive appeals to universal (male) suffrage. And neither South Korea nor Chile in the 1970s and 1980s suggests an organic connection between capitalism and democracy.
However, the case for free voting is not that it guarantees rights, but that it enables the people (in theory) to get rid of unpopular governments. And here three critical observations should be made.
First, liberal democracy, like any other form of political regime, requires a political unit within which it can be exercised, normally a "nation state". It is not applicable where no such unit exists. The politics of the United Nations cannot be fitted into the framework of liberal democracy, except as a figure of speech. Whether those of the European Union as a whole can, remains to be seen.
The second throws doubt on the proposition that liberal- democratic government is always superior or at least preferable to non-democratic government. No doubt this is true, other things being equal, but other things sometimes are not. Ukraine has acquired democratic politics (more or less), but at the price of losing two-thirds of the modest gross national product it had in Soviet times. Colombia has never been under the rule of the military or of populist caudillos for more than brief moments; it has had virtually continuous constitutional, representative, democratic government, with two rival electoral parties, the Liberals and the Conservatives, in competition, as the theory requires. Yet the number killed, maimed and driven from their homes in Colombia over the past half-century runs into millions and is far larger than in any of the Latin American countries plagued with military dictatorships.
The third observation was expressed in Winston Churchill's phrase: "Democracy is the worst form of government, except all those other forms that have been tried from time to time." The case for democracy is essentially negative. Even as an alternative to other systems, it can be defended only with a sigh. This did not matter too much during most of the 20th century, as the political systems that challenged it were so patently awful. Until it faced these challenges, the built-in defects of liberal representative democracy as a system of government were evident to most serious thinkers as well as to satirists. Indeed, they were widely and frankly discussed even among politicians, until it became inadvisable to say in public what they really thought of the mass of voters on whom their election depended.
Yet "the people" is today the foundation and common point of reference of all state governments except the theocratic. This is not only unavoidable, but right: if government has any purpose, it must be to speak in the name of and care for the well-being of all citizens. In the age of the common man, all government is government of the people and for the people, though it cannot in any operational sense be government by the people. This was common ground to liberal democrats, communists, fascists and nationalists, even though their ideas differed on how to formulate, express and influence "the people's will". Mass propaganda was an essential element even of those regimes ready to apply unlimited coercion. Even dictatorships cannot long outlast the loss of their subjects' willingness to accept the regime. That is why, when it came to the point, the "totalitarian" regimes of eastern Europe, their state apparatus loyal, their machinery of repression in good working order, went quickly and quietly.
The governments of modern territorial or nation states rest on three presumptions: first, that they have more power than other units operating on their territory; second, that the inhabitants of their territories accept their authority more or less willingly; and third, that governments can provide services for them that could not otherwise be provided equally effectively or at all - services such as, in the proverbial phrase, "law and order". In the past 30 or 40 years, these presumptions have increasingly ceased to be valid.
First, as Northern Ireland shows, even the strongest, most stable and most effective states have lost the monopoly of coercive force, not least thanks to the flood of new, small, portable instruments of destruction and the extreme vulnerability of modern life to sudden disruption, however slight.
Second, people are no longer so willing either to give voluntary loyalty and service to a popularly legitimate government or to obey the overwhelming and established power of an illegitimate one.
The third presumption has been undermined not only by the weakening of state power but, since the 1970s, by a return among politicians and ideologists to an ultra-radical, laissez-faire critique of the state. It is argued, with more theological conviction than historical evidence, that any services our public authorities can provide are either undesirable or better supplied by "the market". Post offices, prisons, schools, water supplies and even welfare services have been handed to or transformed into business enterprises, while public employees have been transferred to independent agencies or replaced by commercial subcontractors. Even parts of warfare have been subcontracted. The modus operandi of the profit-maximising private firm has become the model to which even government aspires. Thus, the state tends to rely on private economic mechanisms to replace the active and passive mobilisation of its citizens.
Market sovereignty is not a complement to liberal democracy: it is an alternative to it. Indeed, it is an alternative to any kind of politics, as it denies the need for political decisions, which are precisely decisions about common or group interests as distinct from the sum of choices, rational or otherwise, of individuals pursuing private preferences. Participation in the market replaces participation in politics. The consumer takes the place of the citizen.
Two things compensate for the decline in citizen participation, and in the effectiveness of the traditional process of representative government. Headlines (or irresistible television images) are the immediate objective of all political campaigns, because they are far more effective (and much easier) than mobilising tens of thousands of people. The days are long gone when all work in a minister's office was put aside to answer an impending critical parliamentary question. It is the prospect of publication by an investigative journalist that brings even No 10 up short. And it is neither parliamentary debates nor even editorial policies that bring about the expressions of public discontent so patent that even governments with the safest majorities have to take notice of them between elections: as over the poll tax, taxes on petrol and the dislike of genetically modified foods. When they happen, it is quite pointless to dismiss them as the work of small, unelected and untypical minorities, although they usually are.
Thanks to the mass media, public opinion is more powerful than ever, which explains the uninterrupted rise of the professions that specialise in influencing it. What is less understood is the crucial link between media politics and direct action - action from below that influences the top decision-makers directly, bypassing the intermediate mechanisms of official representative governments. This is most obvious in transnational affairs, where no such intermediate mechanisms exist. We are all familiar with the so-called CNN effect - the politically powerful, but totally unstructured feeling that "something must be done" about Kurdistan, East Timor or wherever. More recently, the demonstrations in Seattle and Prague have shown the effectiveness of well-targeted direct action by camera-conscious small groups, even against organisations constructed to be immune to democratic political processes, such as the IMF and the World Bank.
All this confronts liberal democracy with perhaps its most immediate and serious problem. In an increasingly globalised, transnational world, national governments coexist with forces that have at least as much impact on the everyday lives of their citizens as they have, but are to a varying extent beyond their control. Yet they do not have the political option of abdicating before such forces outside their control. When, say, oil prices rise, it is the conviction of citizens, including business executives, that government can and should do something about it, even in countries such as Italy, where little or nothing is expected from the state, or in the US, where many people do not believe in the state.
But what can and should governments do? More than in the past, they are under unceasing pressure from a continuously monitored mass opinion. This constrains their choices. Nevertheless, governments cannot stop governing. Indeed, they are urged by their PR experts that they must constantly be seen to be governing, and this, as we know from late 20th-century British history, multiplies gestures, announcements and sometimes unnecessary legislation. And public authorities today are constantly faced with decisions about common interests which are technical as well as political. Here, democratic votes (or consumers' choices in the market) are no guide at all. The environmental consequences of the unlimited growth of motor traffic and the best ways of dealing with them cannot be discovered simply by referenda. Moreover, these ways may prove to be unpopular, and in a democracy, it is unwise to tell the electorate what it does not want to hear. How can state finances be rationally organised, if governments have convinced themselves that any proposals to raise taxes amount to electoral suicide, when election campaigns are therefore contests in fiscal perjury, and government budgets exercises in fiscal obfuscation?
In short, the "will of the people", however expressed, cannot determine the specific tasks of government. As Sidney and Beatrice Webb observed apropos trade unions, the people's will cannot judge projects, but only results. It is immeasurably better at voting against than for. When it achieves one of its major negative triumphs, such as toppling the 50 years of corrupt postwar regimes in Italy and Japan, it is unable by itself to supply an alternative. We shall see whether it can do so in Serbia.
And yet, government is for the people. Its effects are to be judged by what it does to people. However uninformed, ignorant or even stupid the "will of the people" is, however inadequate the methods for discovering it, it is indispensable. How else can we assess the way that techno-political solutions, however expert and technically satisfactory in other respects, affect the lives of real human beings? Soviet systems failed because there was no two-way traffic between those who took decisions "in the interests of the people" and those on whom these decisions were imposed. The laissez-faire globalisation of the past 20 years has made the same mistake.
The ideal solution is now hardly ever available to governments. It is one on which doctors and air pilots relied in the past, and still try to rely on in an increasingly suspicious world: the popular conviction that we and they shared the same interests. We did not tell them how to serve us - since, as non-experts, we could not - but until something had gone wrong we gave them our confidence. Few governments (as distinct from political regimes) today enjoy this fundamental a priori confidence. In liberal democracies, they rarely represent a majority of votes, let alone of the electorate. The mass parties and organisations, which once provided "their" governments with just such confidence and steady support, have crumbled. In the omnipresent media, backseat drivers, claiming a rival expertise to government, comment constantly on its performance.
So the most convenient, sometimes the only, solution for democratic governments is to keep as much decision-making as possible outside the range of publicity and politics, or at least to sidestep the process of representative government. In Britain, the centralisation of an already strong decision-making power has gone hand in hand with a demotion of the Commons and a transfer of functions to unelected institutions, public or private. A good deal of politics will be negotiated and decided behind the scenes. This will increase the citizens' distrust of government and lower the public opinion of politicians.
So what is the future of liberal democracy in this situation? Except for Islamic theocracy, no powerful political movements challenge this form of government in principle. The second half of the 20th century was the golden age of military dictatorships. The 21st century does not look so favourable to them - none of the ex- communist states has chosen to follow this road - and almost all such regimes lack the full courage of anti-democratic conviction, and claim only to be the saviours of the constitution until the (unspecified) date of a return to civilian rule.
Again, whatever it looked like before the economic earthquakes of 1997-98, it is now clear that the utopia of a stateless, global laissez-faire market will not arrive. Most of the world's population, and certainly those under liberal-democratic regimes deserving of the name, will continue to live in operationally effective states, even though in some unhappy regions state power and administration have virtually disintegrated. Politics will therefore continue. Democratic elections will go on.
In short, we shall face the problems of the 21st century with a collection of political mechanisms dramatically ill-suited to dealing with them. They are, in effect, confined within the borders of nation states, whose numbers are growing, and confront a global world that lies beyond their range of operations. It is not even clear how far they can apply within a vast and heterogeneous territory that does possess a common political framework, such as the European Union. They face and compete with a world economy, operating through quite different units to which considerations of political legitimacy and common interest do not apply - transnational firms. Above all, they face an age when the impact of human action on nature and the globe has become a force of geological proportions. Their solution, or mitigation, will require measures for which, almost certainly, no support will be found by counting votes or measuring consumer preferences. This is not encouraging for the long-term prospects of either democracy or the globe.
In short, we face the third millennium like the apocryphal Irishman who, asked for the way to Ballynahinch, pondered and said: "If I were you, I wouldn't start from here."
But here is where we are starting from.
Eric Hobsbawm is emeritus professor of economic and social history at London University. This is an edited version of his recent Athenaeum lecture; the full text is available from The Athenaeum, 107 Pall Mall, London SW1Y 5ER (£5.65, inc p&p)