a blog about philosophy in public affairs

Voters or residents: how should we draw our electoral map?

In recent months, the Right in both Britain and the US have been accused of trying to manipulate electoral rules to increase the influence of their supporters, and diminish the power of left-leaning voters. Both cases raise important questions about the objectives and principles underpinning electoral democracy, and specifically who elected representatives are supposed to represent.

At the heart of the dilemma is the question of how electoral districts should be drawn.

It is generally accepted that electoral districts should be roughly the same size. The British constituency boundaries prior to 1832 – when Old Sarum had consisted of three houses, Manchester consisted of 142,000 people, yet both elected two Members of Parliament (MP) – would be rightly considered farcical today.

So we want equal districts – but equality of what? One approach is to seek to make districts equal in terms of their population. This is the process favoured in the US. The most common alternative is to seek equal districts in terms of the number of registered voters, as has historically been the case in the UK.

The American approach is currently being challenged in the US Supreme Court, which is hearing the case of Evenwel v Abbott. As a result of this method, Sue Evenwel’s rural State Senate district in Texas has 584,000 eligible voters, while a neighbouring urban district has only 372,000. This disparity occurs because the urban district has many more non-voters (children,  non-citizens, felons and the mentally ill).

The plaintiffs want the state of Texas to redraw its electoral districts to contain equal numbers of voters, even if this means that some districts end up considerably larger in terms of total population. In practical terms, this would mean fewer urban districts with ethnic minorities, and more rural white districts, which would damage the prospects of the Democrats.

In the UK, the Conservative government has made registering to vote a little harder and more complicated, with the result that an estimated 800,000 people (according to the Labour Party) have dropped off the register entirely. The UK has always had more-or-less the system favoured by Evenwel et al*, but the increase in the proportion of the population that is unregistered to vote has made this much more of an issue – there will now be more rural seats with older electorates, and so a bleaker electoral landscape for the Labour Party.

Electoral rules should not be designed according to who they benefit. We should decide which way of doing things is best in principle, and design our democratic processes accordingly. I think there are two separate principles at stake in this debate. The first is the idea that individuals in constituencies with more voters are less likely to get their way – that their vote is ‘diluted’. For example, imagine district A has 500,000 voters who want to raise taxes, while district B and C have 250,000 voters each who want to cut them. The representatives of B and and C will outvote A, even though there are equal numbers of voters on each side. The underlying principle this contravenes is that every voter ought to have an equal chance of influencing the ultimate outcome of the election.

But such a principle is inconsistent with the very logic of having an electoral system based on districts. Such an arrangement almost inevitably leads to ‘safe’ seats, where voters have minimal chance of getting their way, and ‘swing’ seats, where their vote seems to count for more. To put it another way, Britons concerned about diluted votes might want to focus their attention on the 3.9 million UKIP voters who have just one MP .  The way to avoid such outcomes is to have a proportional electoral system – but neither the UK nor the US do, or indeed show much sign of wanting one.

The second key issue is who elected representatives represent. Those in favour of drawing boundaries according to population believe that they should stand for everybody in their district. They fear a different type of dilution: dilution of consideration. On this view, those living in large constituencies will receive less consideration by the legislature because they have to share their representative with more people. Practically, this means they will be less likely to be able to meet their representative, to receive timely correspondence or to have their casework take up.

Those in favour of equalising voters might respond that representatives should only have duties to those who elect them. This might have some plausibility in the case of non-citizens, though there are many rights, such as health and safety protections or labour regulations, that most people would agree ought to be extended to all residents in a jurisdiction, regardless of nationality. But this position also implies withdrawing the services and protection of the state from anybody who does not vote. Worse still, it implies representatives have no responsibilities to children. Such a position cannot be tenable.

The population approach to drawing electoral districts seems to me the only plausible one that is consistent with the internal logic of such a voting system. Therefore districts ought to be equal in terms of population, not voters.

* It’s a little ambiguous whether they want boundaries to be based on the number of registered voters (as they are in the UK) or eligible voters (i.e. including people who are entitled to vote but who have not registered)

Aveek Bhattacharya is a PhD student in Social Policy at the London School of Economics. He has an MPhil in Political Theory from the University of Oxford. His research interests include cosmopolitanism, migration and political economy. He blogs at https://socialproblemsarelikemaths.wordpress.com/



Good books for a ‘first read’ in political philosophy / ethics


Is the £35,000 rule unjust?


  1. Lisa Herzog

    Hi Aveek, thanks for your post. Coming from a country with a (mostly) proportional system, this debate is new to me (although there are related debates here). I was quite struck by the disparity in numbers you quote: districts with similar population sizes having 584,000 and 372,000 eligible voters respectively. It made me wonder what exactly is going on in this case, and whether things are “as they are”, or whether there are other normative issues hidden underneath these numbers? For example, if it is inmates that make up a large chunk of the difference: should they really not be allowed to vote? It would be great if you could expand a bit on this case! Thanks!

    • Jesper L Pedersen

      I may be mistaken, but I believe proponents of allowing inmates to vote generally assume that they would be voting in the area they are connected to (probably where they lived before their sentence) rather than where their prison happens to be. Presumably they would have to vote by mail anyway.

    • Thanks for the comment Lisa. I’m struggling to find a specific source for those data points, so can’t be certain what is driving such a large disparity. My guess would be that immigration is the main factor, though. In total, 11% of Texas’ population are non-citizens, so it’s conceivable that one area could have almost no non-citizens, and others 30%, which would explain such differences. As to whether the fact there are so many non-citizens affects the question of whether they should have voting rights – I would say that is a question of principle that shouldn’t depend on numbers

  2. Jesper L Pedersen

    Hi Aveek, thanks for a great read. One thing you don’t touch upon here, but which is extremely important when it comes to fair voting in the US, is gerrymandering. Looking at the Texas congressional map (https://en.wikipedia.org/wiki/United_States_congressional_delegations_from_Texas), some of these districts are outright crimes against geography. The article you mention doesn’t say which district the plaintiff lives in, but it seems it’s highly unlikely that their vote would make any real difference to who gets elected either way. Do you think it’s possible to decouple the problem you’re mentioning from the wider issues of gerrymandering and election fraud, or do they have to be considered as one?

    • I think it is separable, for the reason that the basic question of voters vs residents applies to the UK as well, where gerrymandering is much less of an issue.

Leave a Reply

Your email address will not be published. Required fields are marked *

Powered by WordPress & Theme by Anders Norén