Desktop version

Home arrow Political science arrow Evaluating Democratic Innovations: Curing the Democratic Malaise?

Pros and cons of direct democracy

As already noted, direct democracy, in the sense of the people voting directly on the questions parliaments now vote on, has a driving appeal in the sense of forming the most obvious institutionalization of democracy itself. If the object is to reinforce the ‘necessary link’ between popular preferences and public policy, how better than to have the latter directly decided by the citizens? Opponents generally accept this argument, but argue against direct voting on three broad grounds: difficulty/impossibility of achievement (especially since we already have policy voting on overall government programmes); incapacity of citizens to make detailed policy decisions; and instability of decisions (as one popular majority succeeds another). Various forms that these objections take, together with counter-responses, are summarized in Table 1.1.

1 The most general objection to direct democracy, and the killer argument for most of the nineteenth and twentieth centuries, is the impossibility of gathering all the citizens of any modern state together for discussion and voting on policy proposals. This only works as a criticism in the contemporary context if one considers face-to-face discussion the only legitimate form of debate, and discounts the referendums and initiatives traditionally carried on in Switzerland through the press and postal ballots. With the development of the electronic media, capacities for interactive discussion are obviously increased - indeed as the introduction to this book makes clear, the possibilities are almost unbounded. Many countries, perhaps most, now have referendums of one form or another, with postal ballots and discussion in the media, accompanied by private blogs and chat rooms, with two- or multiway discussion between individuals. Deliberative theorists (for example, Dryzek 2000) or advocates of ‘deep’ participation (Barber 1984) may object that much of this debate is superficial and does not rest on the deep personal engagement that face-to-face discussion produces. The latter, however, has other disadvantages (dominance by powerful or forceful individuals, fudged compromises and so on). More impersonal forms of discussion avoid this and are perfectly capable of stimulating the ‘deliberations we have with ourselves’. On balance, it seems that modern developments offer citizens more than adequate opportunities for developing and stabilizing their issue preferences, and also easy facilities for registering their votes. The argument against direct democracy, in terms of the feasibility of its preconditions, now looks the most outdated of these in the table and really can be totally dismissed for the twenty-first century.

Table 1.1 Criticisms of direct democracy with responses to them

Criticisms

Responses

1. It is impossible to have direct debate and voting in modern democracies

Even postal ballots and the print media, let alone two-way communication devices, allow interactive debate and voting among physically separated citizens

2. General elections already let

citizens choose between alternative governments and programmes, hence there is no need for direct policy voting

Many issues are not discussed at general elections so if the people are to decide they need to vote on them directly. Moreover such elections primarily choose governments, which voters may do on non-policy grounds

3. Ordinary citizens do not have the education, interest, time, expertise and other qualities required to make good political decisions

Politicians do not necessarily show expertise and interest either. Participation expands citizen capacities. Citizens currently spend a lot of time informing themselves about politics through TV and radio

4. Good decisions are most likely to be produced where popular participation is balanced by expert judgement. This is representative democracy where citizens can indicate the general direction policy should take, but leave it to be carried out by professionals

Expertise is important but not infallible. In any case it can inform popular decisions. Modern representative (party) democracies are heavily imbalanced against popular participation

5. Those who vote against a particular decision do not give their consent to it, particularly if the same people are always in the minority

The problem is general and not confined to direct democracy. Voting on issues one by one gives minorities more voice

6. No procedure for democratic collective decision making can be guaranteed not to produce arbitrary outcomes. What seems like a strong majority may be destabilized by setting out the alternatives another way

Such problems are generic to democratic voting procedures. Voting on dichotomous questions one by one (the usual procedure in popular policy consultations) does, however, eliminate cyclical voting and guarantees a stable majority for one side or the other

7. Without intermediary institutions (parties, legislatures, governments) no coherent, stable or informed policies will be made. Direct democracy undermines intermediary institutions including parties

Direct democracy does not have to be unmediated. Parties and governments can and do play the same role as in representative (party) democracies today

2 Even though direct policy voting, with its associated discussion and campaigns on one side or the other, may be technically quite easy, there are still costs involved, which could be quite heavy in terms of money and time - both public and individual. A more subtle variant of the feasibility argument is that moves towards direct democracy are costly and unnecessary, because citizens are already empowered to vote on policy in general elections. That is, they can already choose between alternative party governments in terms of their medium-term policy programme. In some ways, voting on a general programme gives voters an opportunity for control over a range of policies rather than discrete single ones. Thus, voting in representative elections (which in actual fact are now party elections) might even empower electors more than direct policy voting in referendums and initiatives.There is something in this latter argument, as we recognise below in our comments about synthesizing direct and representative democracy. However, it is also specious in suggesting that representative elections ascertain citizen preferences as well as direct votes on policy. This is for two reasons:

a General elections mix all sorts of other considerations with policy ones in choosing a government - competence and attractiveness of candidates, for example. True, the parties also present a policy programme for government as one basis for choice. If elected, they will even claim a ‘mandate’ for effecting any policy contained in the programme. The mass of voters, however, may well have chosen the party on non-policy grounds. Thus, general elections are far from providing a ‘necessary connection’ between popular preferences and public policy. And even when voters do choose on general policy grounds, this does not mean they have endorsed the party stand on every last issue contained in the programme. b This point is strengthened if we consider the ‘paradox of the platform’ (Brams 1976) illustrated in Table 1.2.This shows electors’ preference orderings over alternatives ‘a’ and ‘b’ (the positions put forward on each issue [x, y and z] by the political parties A and B). Party B wins in a programmatic election with pure policy voting, because three out of five voters (voters 1, 2 and 3) prefer its policy on two out of the three issues covered by the programmes. Had referendums been held on each issue separately, however, three out of five voters on each issue prefer policy ‘a’ and would have voted for it. A direct policy vote thus records citizen preferences more sensitively than an overall, programmatic vote - an additional fact that tells against the argument that general elections sufficiently record citizens’ policy preferences.

All this is not to say that general elections should be totally abolished. There is indeed room for a synthesis of general and policy-specific elections (see below). But the former urgently need to be supplemented by the latter, if citizen preferences are to be transmuted accurately into public policy.

3 A key point in debates about direct democracy is whether citizens really have the capacity to formulate clear public preferences in the first place, and to apply them intelligently to decide between, and vote on, policies. That is

Table 1.2 Party B wins on its overall programme even though a majority opposes its position on each specific issue

Issues

Voter

y

z

1

a

b

b

2

b

a

b

3

b

b

a

4

a

a

a

5

a

a

a

Note: a and b are different alternatives on each issue x, y and z, which are the positions endorsed by political parties A and B respectively.

why, the argument goes, citizens need intermediaries or representatives to do it for them. This argument is often buttressed by pointing to the allegedly bad decisions to which referendums and popular initiatives have led, from the launch of the ill-fated Syracuse expedition by Athens in 413 BC, to Proposition 13 limiting taxes in California in 1978. Kriesi’s chapter takes care of the assertion that direct democracy produces bad policy, with its summary of research that indicates that it produces better results in comparisons of American states and Swiss cantons with more and less initiatives and referendums (see also Geissel in this volume).

Doubts about the ability of citizens to generate informed or even stable preferences about policy have been touched on by Converse’s (1964) findings about citizen ‘non-attitudes’ on public matters. These started from the observation that while aggregate issue preferences remained stable over time (see also Marcus and Hanson 1993) individuals’ opinions did not - and indeed seemed to vary almost randomly around their over-time mean. Forty years of debate on this point have, however, led to a consensus that the variation follows from the different circumstances under which repeat issue questions are asked and which are uppermost in the individual’s mind at the time (Zaller 1992: 54). The effect of political debate and party campaigning is to produce a more informed consensus on which aspects of the issues are salient, so that citizens can make good judgements of what their preference actually is and what their vote ought to be.

In point of fact, democracy itself - not just direct democracy - would be in a bad way if we could not rely on citizens making informed judgements and good decisions, as a result of free political debate. If we cannot rely on their good sense in judging their own interests, even in general elections, we must turn to philosopher kings, experts or unelected politicians - who individually, however, may be no wiser than many ordinary electors. Democracy puts its trust in the aggregate and considered judgements of its citizens.1 Criticisms of the electorate’s ability to do so rapidly lead us to doubt not just direct, but representative democracy (cf. Budge. 1996: 59-83).

  • 4 A better argument is the one about balance. Citizens can take very broad decisions but are not qualified to decide technical ones. Counter-arguments range from ones that ask why non-expert politicians are better qualified to decide than non-expert citizens to the observation that many important policies have technical aspects but these can be ventilated in popular as well as parliamentary debate and a general decision then made on their merits.
  • 5 Balance and compromise are also considerations in dealing with minorities. A traditional fear expressed with regard to direct voting is majority tyranny. Without safeguards or intermediaries, the majority may well steamroller minorities. Of course, this fear has also been expressed with regard to democracy generally and is the reason for entrenched constitutional provisions or requirements for super-majorities on certain issues. These can co-exist with direct voting just as they do with parliamentary voting. One point to note, however, is that voting issue by issue is less likely to lead to one clearly defined minority being consistently defeated than is package voting, where the defeated minority has to wait for the next general election to overturn the previous decision.
  • 6 The effects of a tyrannical majority may be compounded by certain features of majority voting that could lead to arbitrary decisions being taken that are not even desired by the real majority. The argument takes its start from the well-known voting cycle phenomenon (Condorcet 1785; Arrow 1951) illustrated in Table 1.3.

Succinctly put, the theorem states:

a rational individual who prefers A to B to C must prefer A to C ... it is always possible that majority rule is intransitive (i.e. irrational). In the simplest case, if voter 1 prefers A to B and B to C, voter 2 prefers C to A and A to B, and voter 3 prefers B to C and C to A, there is a majority for A over B, a majority for B over C, and a majority for C over A. Transitive individual preferences lead to an intransitive social ordering, otherwise known as a cycle.

(McLean 1991: 506)

This cycle implies first that the majority decision would be overturned on another vote (so it is not a true majority decision) and secondly that it is not a real reflection of majority wishes, but dependent on the order in which alternatives were voted on.

It is easy to see how this pattern of voting - where rational individual voting leads to non-rational and unstable aggregate outcomes - might generalize over large populations, and how it could occur often enough to cast doubt on the pretension of any popular vote to reflect true majority opinion (McLean 1989: 123). It would be equally likely, on the basis of these arguments, to reflect an arbitrary placement of topics on the agenda, or even deliberate manipulation of it, rather than a true majority opinion.

Table 1.3 Voting circles: consistent individual preference orders giving rise to cyclical and unstable majority choices

Preference ordering of policies or candidates A, B and C

Classic case of the paradox (% of electors)

Less extreme case of the paradox (% of electors)

A ^ B ^ C

33.3

22.2

A ^ C ^ B

0

11.1

B ^ C ^ A

33.3

22.2

B ^ A ^ C

0

11.1

C ^ A ^ B

33.3

22.2

C ^ B ^ A

0

11.1

% choosing A over B

66.6

56.5

% choosing B over C

66.6

56.5

% choosing C over A

66.6

56.5

Note: A, B and C represent three policy alternatives or candidates. The arrow ^ represents preferences between alternatives. Thus A^ B ^ C stands for A is preferred to B is preferred to C.

Riker (1982) generalizes this argument into a claim that we can never know whether a true majority exists among the population, as the majority that emerges is always ‘constructed’ by procedures. Hence, liberalism - a series of checks, balances and entrenched rights - is better than majority voting, even for representative elections. It may be observed that this argument, if correct, again tells against democracy in any of its forms. The argument - and the very possibility of voting cycles - only holds, however, insofar as the decision space is neither unidimensional nor separable (that is, each dimension is discussed and voted on separately). Insofar as decisions are made on unidimensional left-right priorities, or on issues voted on individually, one by one, a true majority is guaranteed. This is because there is always a middle position mathematically on a single line, which has to be included in a majority to make it the majority. Referendums of course do generally take the form of issues being voted on one by one, as we saw from the ‘paradox of the platform’ (Brams 1976) discussed above.

7 Parties impose additional constraints on the dimensionality of decision space and thus enhance the probability (already high) that stable, ‘real’ majorities will emerge (Niemi 1969). A telling criticism of direct democracy is, therefore, that it necessarily dispenses with intermediary institutions like parties, legislatures and governments. The shifting majorities that emerge under such circumstances then produce ill-considered policies, which are subject to sudden reversals as the majority collapses or comes under the influence of different demagogues.

This seems a valid criticism of unmediated direct democracy and that is certainly the kind of set-up that many radicals yearn for: a direct and undiluted expression of the popular will, uncontaminated by wheeling and dealing and party fixes. However, we have already made the point that the kind of informative debates necessary for fixing and defining individual preferences themselves depend on protagonists like political parties taking up the issues, and focusing and defining them. In order to assess the force of criticisms of direct democracy as essentially unstable, therefore, we have to ask if an unmediated form is the only one that direct policy voting can take. We have already suggested that, in practice, parties often intervene in referendums or sponsor initiatives for their own ideological or office-seeking purposes. In the next section, we ask whether this is a valid expression of direct democracy or a perversion of it, and whether, therefore, the criticism of shifting preferences and unstable majority tyranny applies to direct democracy as such, or simply to particular manifestations of it.

 
Source
< Prev   CONTENTS   Source   Next >

Related topics