Making Voting Pay

This is the presentation I'm giving later this afternoon:

Some countries, such as Belgium and Australia, already have compulsory voting. Such proposals have been suggested in Britain by the IPPR. In this paper, which has a ‘practical’ rather than ‘philosophical’ focus, I do not dispute the legitimacy of action to increase turnout as such – though some have argued the freedom not to vote is important and turnout provides an important indicator of voter engagement. Rather, I assume that some form of encouraging voting is legitimate, but dispute the claim that the best way to do this is to make it compulsory – instead, I argue, we should incentivize voting.

1) The Alternatives Described
Some countries have compulsory voting laws, but do nothing to enforce them – they can be seen as nothing more than collective endorsement of a citizen’s obligation to vote, although whether they themselves ground a moral duty depends on one’s views on political obligation. My concern is with countries where compulsory voting is enforced. I hold that this coercion is, roughly, a denial of freedom. I will argue that instead of fining non-voters, we should pay people that do vote. Aristotle argued that democracies should pay people for participation, but to my knowledge the suggestion has not been made with regard to contemporary voting.

I do not here specify a figure, but I have in mind a relatively nominal sum – as compensation for the time and effort involved – probably less than fines imposed for non-voting where it is compulsory. While RCT is notoriously bad at predicting voting turnout, it does seem that Probability, Benefit and Cost do play a part in voting decisions, and this compensation should offset the cost, thereby increasing voting without coercion. If the figure seems too low, we could offer people lottery tickets, so each voter stands a chance of winning a more sizeable sum.

2) The Freedom Argument
The defender of compulsory voting can claim that, in fact, all laws interfere with our actions, but in fact we are still as free as before – because we can choose whether or not to vote, we will simply be fined if we do not (this is the claim that threats merely change payoffs, but do not diminish liberty any more than offers). I accept that there is still in fact a free choice – it would, after all, be odd to say someone does votes unfreely given that they could actually have not voted. However, I do not think it is legitimate to impose certain choices on people. The difference between a threat and an offer is that the former wrongs you by making your option set worse.

Suppose I face a choice between X and Y, and I prefer X – that is because X makes me, in some sense, better off. If you threaten to fine me £50 for choosing X, then I may prefer Y, because I prefer Y to X-£50, but I am worse off than I would have been given an uninterfered with choice. On the other hand, if you pay me £20, that may make Y better for me than X. It is not, strictly, that my choice is any more free, but I have been made better off by being presented with a more attractive option.

3) The Equivalence Objection
It will be objected that there is not really any difference between threatening a negative fine and offering a positive inducement, since money cannot come from nowhere – fines could be used to swell the public coffers, reducing the general tax burden, whereas incentive payments would have to be funded by taxing all more heavily. At this point, I should admit I do not find taxation so normatively problematic as some – I think much of our wealth consists in natural resources or inheritance that is relevantly ‘gift-like’ that it can be taxed without threatening people’s rights, and even Nozick objects only to labour taxes, not the poll tax that would presumably be necessary to fund even his minimal state.

Still, it may be held that fines and inducements will have the same net result: e.g. if we have ten people, six of whom vote, then we could pay each of those six £12 for doing so, but would have to tax all ten £7.20 each to pay for this – when the alternative would be to fine the four non-voters £12 and use that money to reduce everyone’s tax burden by £4.80 each. In this case, what’s the difference? Well, firstly, I would argue that framing effects matter. People generally feel their freedom is diminished by a threat or legal requirement to act in a certain way, whereas they are not so threatened by an incentive payment and are unlikely to notice a marginal increase in their taxes. Secondly, I would argue that the incentive system would likely be easier and cheaper to administer – there would be no need to track down non-voters, but it would be easy to give people a cheque as they turned up at the polling station.

4) Is it Wrong to Pay People to Vote?
Most people feel it is obviously wrong to pay people to vote a certain way, because this gives money undue influence in politics. Some would go as far as to say it is wrong to pay people to vote at all. It seems there are two possible reasons for this – one that people should not being paid for doing their duty, the other that for the state to do it violates some sort of principle of neutrality.

4a) The Duty to Vote
Turnout may be seen as an example of the ‘Titmuss trilemma’. We want:
1) No payment for voting.
2) A sufficiently high turnout.
3) Freedom of choice.
These three may be consistent if we postulate that people freely accept an act on a moral duty to vote. However, this doesn’t seem to be the current practice. At the moment, arguably condition 2 is violated – people are left free and the result is low turnout. Those who advocate compulsory voting address turnout by denying freedom (condition 3); I call into question condition 1, suggesting that – just as the US encourages blood donation by paying donors – we could encourage people to give up their time to vote by compensating them for doing so.

The claim that it is wrong to pay people to vote because it is their duty presupposes that people do in fact have a duty to vote, but this is not obvious – certainly an act-utilitarian is unlikely to accept that their voting is likely to maximize expected benefit, while a Kantian need not be able to universalize such a duty – and even if there is a moral reason to vote it need not outweigh competing considerations. In any case, I think it is wrong to say that people should not be paid for doing their duty. Maybe they should not profit from it, but to clarify what I am advocating is mere compensation. Now, perhaps there are some duties that are not really done if one doesn’t pay the cost – e.g. charity – but there are plenty of others where it does not seem the individual has to bear the burden – so, for instance, people are paid while on jury duty, for military service, etc.

It may be argued that if people only vote because they are paid to do so then the benefits, such as personal moral development, will be undermined. Psychological studies have shown that even if people are paid for things they would do anyway, they may come to regard that activity as ‘work’, and refuse to do it if not paid. This suggests people may come to take the ‘wrong’ attitude to voting. However, this is not an objection that can be voiced by those who would otherwise advocate compulsory voting, for surely forcing people to vote does not make them do it for the right reason. Moreover, I am assuming we are concerned merely with increasing turnout to support democracy; judgements as to reasons why people should vote seem to impose non-neutral conceptions of the good on them.

4b) Non-Neutrality
It may be argued there is something particularly wrong about the state paying people to influence their behaviour. Parents may make pocket money conditional on their children’s behaviour, but paternalism is justified when it comes to children – we do not need to respect someone’s autonomy until they have it. Many liberals hold that the state should be neutral about the good life, and not try to promote some particular forms of behaviour at the expense of others (e.g. not promote church-going).

However, this is again not an objection that can be put by those who favour compulsory voting, since if incentivizing people to vote is non-neutral then forcing them to is even more objectionable. In any case, it misunderstands our motivation. I do not claim that people should be encouraged to vote because it is good for them, but to uphold democracy, which I take to be a neutral value. State encouragement is therefore more like enforcing the rules of the road than promoting religiosity. Even if the state’s role is to be restricted to matters of justice, insofar as democracy (equal involvement of all in making the laws all will be bound by) is a requirement of justice, then actions directed at preserving and promoting democracy will be legitimate. Currently the British government spends considerable money on advertising to inform and encourage voters, but maybe some of that would be more effective if simply spent on paying people to vote.

5) Effects of Paying People to Vote
Those who do vote are compensated for their time, and this can be important if, e.g., they must take an hour off work or away from caring for (elderly or young) relatives. Even if voting is a duty, I argued above there is no reason it must make people worse off, and all the less if it is something optional but that we wish to encourage. There will be a redistribution from non-voting taxpayers to voters, but this is justified if we assume a certain amount of participation is necessary for meaningful democracy and – this being a presumptive good – that the non-voters are actually free-riding on those that do vote and thereby support democracy. As such, it does not seem they have much objection – particularly because the tax burden is one they could themselves offset by choosing to vote themselves (and if they have something they’d rather do, so be it). The scheme can be seen as the kind of ‘selective incentive’ for participation that Mancur Olson argued was necessary when free riding is possible. It can also be seen as their fulfilling their duty of ‘fair play’ – in cash, rather than kind – to support the just institution of democracy, from which they benefit.

6) Conclusion: What I Have and Have Not Done
I have not argued that we should do anything to encourage voting. Perhaps our current practice – uninterfered with voter choice – really is best, despite low turnout. Many democrats are, however, concerned – not only that turnout is low generally, but that it is particularly low among low income groups, whom one might think will be more responsive to even relatively modest financial inducements. It is such concerns that have motivated some calls for compulsory voting, which is one way out (as illustrated by the above ‘Titmuss trilemma’) – I call such proposals into question, by pointing to another neglected solution.

Nor have I addressed the forms voting should take. I think one considerable problem is the electoral system: even ‘marginal’ seats are rarely decided by a single vote, while votes in many safe seats are effectively meaningless. It is no wonder that, in such conditions, turnout is low, and maybe more could be done to address the problem by electoral reform or other ways of making politicians more responsive to the concerns of ordinary voters.

What I have offered is an ad hominem argument against proponents of compulsory voting: I am therefore entitled to assume it is legitimate for the state to encourage voting. I have merely argued that the carrot is better than the stick – it is both less objectionable and probably more efficient to offer incentives than to threaten penalties. Maybe my empirical claims can be disputed, but we should at least consider the possibility of paying people to vote.