September 25, 2016

Can a Different Voting System Improve our Democracy?

After watching the Presidential nominating process over the last year, I find it puzzling that this system ever worked. Going in, I had some vague notion that primaries and caucuses were a mechanism to make the nominating process more democratic. They were to give the party’s rank and file more control over who ends up on the ballot in November.

Given the outcome of the 2016 nomination process, this perspective is in tatters. More than anything, media attention seemed to drive the results. The Democratic Party leadership actively colluded with the media to disparage Sanders. Quite the opposite in the Republican Party. The Party had no mechanism to counter Trump and his showmanship, even though many of his positions ran counter to many positions held by most Republicans. And other Parties were rarely covered by the media at all.

So, if not media attention, what are the values that should drive the design of the nominating process? Off the top of my head I would say -
  • make the process more democratic and more inclusive
  • find out exactly where candidates stand on the issues (and ditch all the negative campaigning)
  • craft a ballot with candidates who each represent different points of view, and 
  • (based on my personal pet peeve) spend a month or less making the nomination decisions

It seems to me that the first item is achievable, but the rest are more pie in the sky. So how might we improve the current voting system so as to foster more inclusivity and democracy? In the not too distant past, I had naively thought that the way I vote was how all voting takes place. Each citizen gets one vote and the candidate with the most votes wins. It turns out that there are many ways to run an election. They all have their pitfalls and their strengths. Thankfully, I didn't have to wade through volumes of scholarly political science journals to get to the bottom of it. I found some entertaining YouTube videos that explain some of these systems.

Call me a spoon-fed learner, but it was fun to think about the first election in the Animal Kingdom, in which the race was between a turtle, a gorilla, a snake, a tiger, a monkey, an owl, and a leopard. 

Our current voting method in most elections in the United States is known as winner-take-all or first-past-the-post (FPTP). In the video, CGP Grey demonstrates why FPTP voting always trends toward a two-party system with gerrymandering. After watching this video, I understood how our winner-take-all voting method results in two major parties controlling our elections. There are a couple of troubling side effects with a two-party system. One is that voters must worry about how everybody else is voting, rather than voting for whomever they like the most. If they don't, they might end up with someone they feel is truly atrocious. Another downside of a two-party system is that there is rarely much of a choice on the ballot and new parties don't stand a chance of gaining a foothold. A last consequence of a FPTP voting is that it is quite susceptible to gerrymandering, where those in power draw the voting districts to advantage their party and incumbents in their party. Such gerrymandering tends to result in disproportionate representation and/or elected officials who aren't responsive to their constituents.

Taking the catchphrase  Think Globally; Act Locally  to heart, I decided to attend a local Election Reform Working Group. They started meeting just subsequent to the National Conventions, so I thought they might share my frustration with the current system. I also hoped they were working up some useful strategies for improvement. When the group asked for someone to study the "Top Two Open Primary" ballot amendment, I raised my hand. Perhaps structurally changing the primary election mechanisms might help solve the problem of making the process more democratic and more inclusive. 

The full name of this (potential) Florida ballot initiative is All Voters Vote in Top Two Primary Elections for Congress, State, Legislature, Governor, and Cabinet. Currently, the fastest growing segment of Florida voters are those who have chosen not to affiliate with any party. With Florida's closed primaries and its history of gerrymandering, many races are actually decided in one party's primary, which is then followed by an uncontested general election. This means that in those uncontested elections, ALL voters who are not registered with the majority party are effectively disenfranchised. All-voters-vote would allow voters of all persuasions to choose from a single list of candidates, no matter the party. The two who receive the most votes would move on to the general election in November - whether they are from the major party, the same party, or the minor party.

Other states have implemented similar amendments, most notably California, Louisiana, and Washington State. The results are not encouraging. Certainly, independents have the opportunity to fully participate in primaries, but unfortunately it hasn't boosted overall voter turnout in any of these states. The 2000 presidential election recount crisis in Florida resulted in Florida legislation to change the mechanics of voting technology, including voting machines, automatic recounts, a voter registration database, poll worker training, and more attention to ballot design. But the Florida recount crisis spurred the state of California into high gear. Voters enacted a series of state and local election reforms. After the enactment of these reforms, Eric McGhee of the Public Policy Institute of California concluded that "Independents appear to be fickle primary voters, inclined to participate only when a ballot includes a close race. Recent efforts to increase turnout by making registration easier — such as online and same-day registration — also appear to have no meaningful impact on primary turnout." I came across a fair amount of election research, while trying to learn about voting systems. From my limited reading, the research shows that the consequences of changes to primary election rules are hard to predict and unlikely to have consistent and uniform effects. 

The All-Voters-Vote ballot amendment tinkers with the voting method a bit, but leaves the main elements in place. The results have been fairly unimpressive. So, if we want to do something to rein in two party dominance and its associated downsides, we might need to consider a more dramatic change to our voting methods. I turned back to CGP Grey. He outlines instant-runoff-voting or ranked-ballot voting a.k.a the alternative-vote in this video

The main advantage to this system is the elimination of the spoiler effect, where a "spoiler" candidate's presence causes vote splitting between similar candidates and causes an opponent of both to win. Voters don't have to worry about how everyone else is voting, and the winners tend to be those that a larger majority liked. This presentation of instant-runoff-voting did not delve into all possible complexities, so I will add that it can still be advantageous to be less than honest in your vote. However, such strategic voting, is neither as obvious nor as clear-cut as it is with a winner-take-all system. Instant-runoff-voting is also vulnerable to gerrymandering. But, all in all, it seems a far superior system to winner-take-all.

How easy would it be to make this more dramatic change to our voting system? As it turns out, voters where I live have already weighed in and 77% approved instant runoff voting (IRV) for the Sarasota City Commission races. In Sarasota's case, the ballot initiative was aimed at reducing the overall cost of elections by eliminating runoff elections. As a side note, these costly runoff races aren't particularly inclusive, because they happen at the onset of summer, when many residents have fled to cooler climes. So, IRV was an easy sell to voters.  But it's one thing for voters to decide something, but something entirely different for that decision to be implemented. The first hurdle stopped this change dead in its tracks. All voting machines must be certified by the State of Florida, and the State of Florida had not certified any IRV machines. Nine years later, I haven't heard another peep about this.

To be sure, there are some locations that have successfully approved AND implemented instant runoff elections. I am curious to know about their experience with instant runoff voting. How are the ballots laid out and how difficult is it for voters to understand such ballots? What kind of up-front and operating costs are associated with this form of voting? Has it affected voter turnout, ballot errors, the range of candidates on the ballot, or the tenor of elections? From my initial look, voters don't seem to have a problem understanding the ballot. There are greater costs associated with voter education. The answer to the overall cost question seems to depend on who is answering; those in support of IRV find it less expensive and those against IRV find it more expensive. It does appear that where runoffs were already mandated, most seem to think IRV is less expensive. Some research shows that IRV reduces negative campaigning. Several Leagues of Women Voters (LWVs) have studied IRV extensively.  Since an instant runoff voting initiative will appear on the Maine ballot in the upcoming 2016 general election, LWV of Maine put together this Q&A.

But then there is a relatively untried voting system called range voting (or sometimes score voting), which claims to solve many of the problems associated with winner-take-all and instant-runoff-voting. In range voting, voters give each candidate a score, the scores are added or averaged, and the candidate with the highest total is elected. I say it is relatively untried, but if you are an avid online shopper, you may have had a lot of experience with it. Do you check out all those 1 to 5 star reviews that others have given products, before making your final buy? Then you have utilized something akin to range voting results. Alas, CGP Grey has not made a video about range voting, so I was left to scour the arsenal of political science literature.

Here are the highlights of what I found in my somewhat random, cursory look at the voting system literature. In 1950, Kenneth Arrow won a Nobel Prize, when he showed that there was no perfect ranked-order voting system. That's because some important criteria for a fair voting system are mutually exclusive. Opponents of any particular system could always come up with a worst-case scenario that showed how unfair it was. Consequently, there's a lot of discussion about which criteria a given system does or doesn't fulfill, which are the most important, and which are the most likely to occur. Of note in the early discussions is that they generally dismissed non-ranked-order systems. And range voting does not ask voters to rank one candidate against another, so it is not a ranked-order system. More recently, computer scientists and game theorists jumped in with simulations. Range voting looks pretty good in many of these simulations. But there's not much real-world evidence as to how voters would actually behave in a range voting election. Nor could I find anybody working on range voting ballot initiatives.

Based on my preliminary investigation, both range voting and instant-runoff-voting look far more promising than top-two-open-primaries. But note that all of these voting systems result in the selection of a single candidate.  In other parts of the world, voters don't directly elect candidates. Rather they elect parties, and the parties select those who will fill elected government positions, e.g. members of the legislature. Electing candidates directly sounds far more democratic than electing parties who select the winners on my behalf. But having never paid any attention to these popular voting systems, I thought it only fair to see if electing parties might result in more representative outcomes. Out of curiosity, I checked in with CGP Grey and viewed his video about one such system called mixed-member-proportional (MMP) representation.

On Kiwi Island, the Animal Kingdom is selecting between kea, tuatara, and kakapo - a ballot of unfamiliar animals. MMP is a mixed approach. The advantages to this system are that it reins in gerrymandering, prevents minority rule, and allows for political diversity. This sounds more inclusive and democratic to me, but so very different from our current candidate-focused elections.

Question: What will I do about this new trove of voting system related info? 
Answer: Not sure yet.  To be continued...

No comments:

Post a Comment