Rethinking Voting
When designing an electoral system, the most fundamental objective should always be to come up with a mechanism that closely translates the will of the people into political outcomes. Of course, there is much more to consider beyond merely majoritarian sentiment, as the Framers of the U.S. Constitution observed. That being said, across the Western world faith in democratic practices and institutions is dwindling and I believe that our electoral system deserves a fair portion of the blame. With some frank conversations about the state of our electoral system, and a few key reforms, there lies the exciting promise of progress.
A fairly rudimentary way to diagnose the current problems with our electoral systems would be to look at how closely votes map on to outcomes. For starters, two of the last five U.S. presidential elections have gone to the candidate who lost a majority of the popular vote. In the House of Representatives, where all candidates are up for election every two years, the Republicans won a plurality 49.1% of the vote in 2016 but currently make up 54.8% of the seats. In 2012, even more surprisingly, they won fewer votes than Democrats but still made up a majority of the seats. The predictions for this year’s midterm elections are similarly grim, wherein if Democrats win every district won by Clinton and those narrowly won by Trump they would still fall short of a majority.
This poor translation of votes is largely the result of two factors: the Electoral College and gerrymandering. The Electoral College is what favors smaller, more rural states in Presidential elections, and gave Republicans victory in both 2000 and 2016. This fosters disenchantment amongst voters, especially for the partisan minorities in a respective state. If you are a Republican in California or a Democrat in Alabama, there is little motivating you to turn up and vote come Election Day, since all votes from that state are given to the candidate who wins a bare majority there. This means your state will undoubtedly vote against your preference, so your personal ability to impact change or voice your opinion through voting seems lower.
Gerrymandering is the practice of drawing district lines to favor a particular candidate or to isolate specific groups of voters. This is prevalent at elections below the state-level, such as for the House of Representatives. Similar to the Electoral College, this increases disenfranchisement by literally moving the goal posts to ensure incumbents have an edge in their own reelection. Gerrymandering practices such as “packing” or “cracking” can dilute the influence of selected voter groups. When combined with the amount of money that has entered politics in recent years, we see why despite Congress’ abysmal approval rating over 97% of incumbents are still reelected in the House. Thus, it too increases the divide between the will of voters and the outcome of elections.
Both gerrymandering and the Electoral College contribute heavily to low voter turnout. Voter turnout is more a symptom of a poor electoral system than a cause. Obviously, it is a serious problem for democracies, as we should want as many people as possible playing a role in deciding key decisions such as who will make our laws or what those laws ought to be. Time and again, it has been shown that groups perform better than individuals at intellectual tasks. Low voter turnout is also driven in part by designs to make voting harder.
In recent years, numerous states have enacted stricter photo ID laws to be able to vote, cut mail-in ballots and early voting, purged voter rolls, closed polling locations, and changed requirements for voter registration. Increasing the minimum time between registering and voting, as well as cuts to early voting and absentee voting, make it tougher for people to be able to actually vote. Currently, Election Day is not a holiday in the United States. So if early voting is restricted — and polling locations are being closed — it can be impossible for some voters to actually make it to the polls. This has a disproportionate impact on minorities and working class people, who are less likely to have a photo ID and more likely to work minimum wage jobs that do not allow flexibility of hours on Election Day.
In the name of expanding the electorate, we should also consider lowering the legal voting age to 16. Research has shown that 16-year-olds possess the same level of civic knowledge as young people in their 20’s, and that so-called “cool” executive function skills are almost fully developed by age 15. 16 has been shown to be a better age at which to form habits, and encouraging young people to participate in their democracy while they are still actively learning civics in school could be a boon to overall voter turnout and participation rates.
Our electoral system today is largely based around the concept of first-past-the-post (FPTP) voting. In this system, whoever receives the most votes wins outright. For that reason, this system is also commonly referred to as winner-take-all voting. Canada, the UK, Brazil, and numerous other democracies also use FPTP voting. This kind of system encourages a great deal of tactical voting, as any vote not given to the top two finishers can be seen as wasted. Take the 2000 election, where if all the votes for Ralph Nader had broken for Bush or Gore then the Gore-lean amongst Nader voters would have pushed Al Gore over the top to win the election. Such a system gives a tremendous amount of power to the media, as who is projected and publicized to finish in the top two spots will likely be the beneficiary of tactical voting. This makes it even more difficult for lesser-known candidates to be elected and can incentivize voting for a candidate who is not your most preferred.
FPTP voting lends itself more easily to gerrymandering since it allows for a large number of wasted votes. More broadly, wasted votes constitute any for the winning candidate beyond the threshold needed to win or any for losing candidates. In the UK’s 2005 General Election, some 70% of votes were ‘wasted’ in this sense as 52% were cast for losing candidates and another 18% were over the threshold for victory, meaning this 70% played no part in determining the outcome. In Britain’s 2015 General Election, the SNP received 37% of the votes that UKIP did, but they took 56 times the number of seats in Parliament. For this reason, FPTP systems tend to have lower rates of voter turnout than other democracies.
To briefly get a little more into the weeds on electoral systems, FPTP voting fails all mathematical criteria scholars measure voting systems on save for one. If more than 50% of voters prefer one candidate, they will win in a FPTP system. This is referred to as the Majority Criterion. However, FPTP voting fails many other very important metrics; notably the two Condorcet Criteria. The Condorcet Criteria state that if a candidate would win or lose in head-to-head elections against all other candidates, then they must either win or lose overall. FPTP voting fails to meet this. Consider an election wherein the field is so fragmented that someone who is least preferred by a plurality of voters may, in fact, win if the opposition to them is split enough.
In order to rectify some of these problems with FPTP voting, I propose we move to a system of ranked voting. Ranked voting entails ordinal ranking of candidates based on preferences, so instead of voting merely for your favorite candidate, you would have the opportunity to rank your preferences. To bring back the 2000 Election example, Nader voters could have ranked Gore or Bush as their second choice to broadcast their preferences between those two candidates, and not just their preference for Nader over both Gore and Bush.
There are many methods of ranked voting, most commonly used in practice are Instant-Runoff (IRV) and Single Transferable Vote (STV) systems. IRV runs as many iterations as it takes to produce a majority winner through reassigning votes. If in the first round no candidate wins more than half of all #1 ranked votes, all votes from the candidate who received the fewest #1 ranked votes are reassigned to each ballot’s respective #2 candidate. This is automatically repeated until a candidate receives a majority of votes. For single-seat elections, STV is actually exactly the same as IRV. STV differs in how it can be used for multi-member constituencies. Votes are distributed to elect the required quota number of candidates, and then all excess votes are reassigned to the voters next preferred candidate. If the quota is not hit, the candidate with the least votes is eliminated and their votes are redistributed.
Ranked voting allows voters to transmit more information regarding their preferences. It also makes manipulative voting harder as one would have to highly rank candidates they want to lose in order for manipulation to work. IRV meets the Condorcet loser criterion described earlier, wherein if a candidate loses in all head-to-head matches they must lose overall. Imagine if in the 2016 Republican Primary elections there had been ranked voting similar to an IRV system. Perhaps Donald Trump would have been the first choice of a minority of voters but one of the last choices of a majority of voters. It seems highly plausible that Rubio voters would prefer Jeb Bush or John Kasich to Donald Trump. Thus, perhaps as these other candidates were eliminated in instant runoffs their votes may have been reassigned to another non-Trump candidate. This would mean the anti-Trump vote would have eventually coalesced around one alternative instead of being fractured between half a dozen candidates. Of course, this example is purely speculative, but it does highlight the way ranked voting can help elect a consensus or compromise candidate.
Having IRV for the presidency would replace the Electoral College. Doing away with this antiquated system would mean voters no longer have such disparate relative powers of influence based solely on their state of residence. Currently, voters in Wyoming have disproportionately more say on who the President will be than voters in California. This goes directly against the democratic ideal of each person having an equal vote.
If we were to implement STV voting with newly formed multi-member constituencies, this would mean single seat elections like for the presidency would be carried out as instant runoffs as outlined above, but lower level elections would have multiple candidates winning in a given area. Ten states currently use multimember districts to elect state legislative seats. This would reduce gerrymandering and importantly is a necessary condition for proportional representation. There is evidence that multimember districts more regularly elect women and minorities. A concern often cited is that it dilutes the relationship between elected officials and their constituents. I would counter that the relationship is already imperfect, and as I will outline below there are other ways to get voters more closely engaged in politics.
Ranked voting does require a significant degree more information about candidates than merely voting for one person, as voters need to know enough about all candidates to effectively rank their preferences. This could be disadvantageous to voters who have less time and access to such information, such as lower-income voters. Luckily, IRV systems do meet what is called the Later-no-harm Criterion, which states that switching the relative ranking of lower than first choices does not influence the probability of your preferred candidate being elected. Additionally, any of the electoral reform discussed here must be accompanied with an expansive effort to increase accessibility to voting.
Ease of voter access should be a top priority in a democracy. This would begin with making Election Day a holiday, or at the very least mandating that employers allow time for workers to vote. If we expand early voting and mail-in options we will also reduce the barriers to voting. Making registering to vote easier would go a long way as well, and I support Automatic Voter Registration laws that make being registered to vote opt-out instead of opt-in. States as varied as California, Georgia, West Virginia, Illinois, and most recently Massachusetts already have Automatic Voter Registration laws on the books.
If we make voting easier for individuals we can encourage greater participation. If those people also know that their vote is less likely to be wasted, they should have more incentive to partake in the democratic process. Congress should take the lead on these changes. Pass laws that ease the voting process all the way through from registration to Election Day. Work on innovative solutions to campaign finance that challenges the current dynamic of dark money influences. Congress has the power to do a lot of a good when it comes to reforming elections in America, but greater participation from voters should also be encouraged.
The theory of citizen participation has its roots in ancient Greece and is most notably present today in the process of Jury Duty. Jury service requires us to take a government-sponsored break from our lives to actively participate in the carrying out of constitutional government. It provides us an opportunity to listen to opposing views and carefully weigh considerations of evidence, as well as deliberating with fellow jurors. Expanding citizen participation would help shrink the disconnect between the body politic and policy.
Examples of this have been experimented with around the world. In Canada, British Columbia and Ontario have both referred to policy juries made up of citizens from each district to consult on the matter of electoral reform. Citizens were chosen through stages of self-selection and sortition. In 2013, Ireland created a temporary body to draft new articles for the constitution where two-thirds were comprised of citizens who were chosen by lot. Similar councils have been used in the Netherlands and Iceland for advising on specific topics of electoral or constitutional reform.
During the founding of the American political system, the Framers took influence both from Lockean classical liberalism as well as from republicanism. The former pushes democracy as the ultimate tool of freedom since individuals are best positioned to determine their own self-interest and should have the right to act on that. The latter argued for democracy as an end in and of itself — that there was a sort of “public happiness” in allowing everyone to actively participate in political life. Ultimately, democracy is all about self-governance. Government of the people, by the people, and for the people. In instituting the reforms I have laid out here, I believe we will increase enthusiasm for and participation in democratic institutions and practices. If we pass the correct laws and restructure the system to incentive easy access to the voting booth, greater citizen involvement, and better translation of voting preferences, we can reduce the deficit between policy preferences and policy outcomes.
Image Source: https://pxhere.com/en/photo/1444335
