Larry Bradley’s Ezine #379 How Would Our Electoral College Proposal Have Worked in 2016?
Weekly Ezine Number 379
Notice: Our Ezines are not intended to either support or oppose any individual candidate for election or government official. The thoughts are, instead, always intended to demonstrate the inadequacy of our current election processes and offer solutions to make those processes better. Our hope is by implementing better processes American citizens will gain a more efficient and effective government with less conflict and more consensus.
What Did the Peru State Election Reform Survey Show?
How Would Our Electoral College Proposal Have Worked in 2016?
What Did the Peru State Election Reform Survey Show?
We were very pleased to have the opportunity to work with Dr. Sara Crook and Peru State College in Peru, Nebraska. I once did a talk on Election Reform at Peru State. Afterwards, I talked with Dr. Crook and she remarked she teaches courses where she involves her students in political projects.
I remembered her remark and so I suggested to her that she conduct a survey of the 2016 candidates for the Nebraska Unicameral about Election Reform. She liked the idea and did the survey. The survey was based on my own suggested letter to candidates I published in September. http://blog.thecenterstrikesback.org/?p=132
She and her students decided to modify my second and third proposal. For the second, they decided to retain the method used by Maine and Nebraska to split the state’s Electoral votes. That method awards one elector each to the winners of the state’s Congressional Districts and two electors (representing the state’s U.S. Senators) to whoever wins the statewide popular vote. The modification they proposed was that the ranked choice ballot be used and that the winners in both the Congressional Districts and state wide must be majority, not plurality winners.
For the third proposal, the students wanted to propose implementing a non-partisan redistricting commission. They preferred this over proportional representation.
So, the students did conduct the survey. We were all disappointed by how many responses we got, but even the League of Women Voters has difficulty getting candidates to respond for their Voters’ Guide. Here is a link to the Press Release Peru State put out on the results. http://www.peru.edu/media/peru-state-political-science-class-conducts-survey-of-legislative-candidates-on-topic-of-election-reform/ Here are some key points.
As Dr. Crook pointed out, it was the female candidates who were most open to considering alternative ways of voting. Several of the respondents noted that a commission had been proposed in the legislature and vetoed by the Governor. The remarks by Mr. Clark criticizing the third proposal actually provided justification for our original proposal for Proportional Representation. Perhaps most amazing of all was the feeling of Mr. Roland that the winning of an election carried with it the power to Gerrymander. Mr. Roland also seemed to struggle with the concept of majority, rather than plurality winners. Mr. Roland lost.
Overall, this was a worthwhile effort that we will attempt again.
How Would Our Electoral College Proposal Have Worked in 2016?
What we’re going to do today is to show you how the results of the Presidential Election might have changed if our proposed reforms to the Electoral College had been in place.
Before beginning, we want to say the debate about whether the Electoral College should go away in favor of a popular vote is not the most challenging issue we have with our election. We are missing the most challenging issue we have with our elections. The most challenging issue we have with our elections is our process is not allowing voters to validate whether a candidate is a consensus majority choice of all voters (and thereby win the election) or is merely the leader of a motivated minority (and is, therefore, defeated). More on that in a bit.
To review, our proposal is that we would use a ranked choice ballot in each state. That ballot would be used to determine the top two popular vote getters for President in each state. The Electors would then be allocated to those top two candidates based on the percentage of the vote they got. The State Elector totals would be carried forward to a final national total.
Obviously, there has been a firestorm over the fact that Clinton beat Trump in the popular vote, but not the Electoral College. Here is a key point we are not hearing.
If the situation were reversed— if Clinton had 46% of the vote but won the Electoral College and Trump had 48% of the vote, would any Trump supporters be satisfied with the result? The answer is an emphatic “No!” isn’t it? Had Trump lost, would or would not Trump supporters be using the very same arguments Clinton supporters are using now? They would, wouldn’t they?
And what about the other 5.6% of voters who voted for someone else and yet have no opportunity to weigh in and take one of those two candidates to over 50%? Are they satisfied with the results? Again, we think the answer is, again, an emphatic “No!”
So, the moral of the story is this. Neither Trump nor Clinton got a majority. No matter which candidate you supported, the evidence is clear that this system we’re using is increasingly susceptible to not providing us a definitive decision and a legitimate majority winner.
That means that all of us should be dissatisfied with the system as it is and want to find a way to provide more definitive results than the ones the system is giving us now.
We believe the Electoral College should be retained in order to properly represent the cultural and economic interests of the low population states, but believe our method would deliver a result that is more likely to logically reflect the two results (Popular Vote and Electoral College) and ultimately provide a majority winner. That alignment would not, however, be guaranteed. What our method would do is produce a result that more accurately reflects how the country voted. We found that more accurate reflection when we did this exercise in 2008 and 2012. We found that more accurate reflection this time, too.
So, what did we learn from doing this exercise in 2016?
First, there were significantly more votes for third party candidates in 2016. That forced us to make an assumption decision to account for those votes.
We debated about making some sort of assumption about how the votes for third party candidates would have split in the second or third round of voting. In other words, if someone voted for Gary Johnson as their first choice, who would have been their second choice? We decided that was impossible to know, so here is what we did in order to illustrate our method.
(By the way, you’re welcome. We had to hand jam a lot of this data into a spread sheet and then make calculations with a calculator to produce these results. Further, we did it twice. We did it once on Nov 15 with data from Politico. But when Clinton’s popular vote margin got ever bigger, we did it again using data from the Cook Report as of Nov 26. (Their spreadsheet was more user friendly.) Here is the Cook Reports’ Spread Sheet. https://docs.google.com/spreadsheets/d/133Eb4qQmOxNvtesw2hdVns073R68EZx4SfCnP4IGQf8/htmlview?usp=sharing&sle=true I couldn’t get my own spreadsheet to post properly on the Blog. It’s similar to the Cook Report’s. If you want to see my spreadsheet email me and say so. I’ll be happy to send it to you.)
First, a disclaimer. The voting results are changing every day as provisional and absentee ballots are approved. We took a snapshot of the results on a given day and did our estimates. Don’t bite our heads off two weeks from now and try to say we manipulated the data. We didn’t. We did a demonstration of our proposal given the data we had available on a given day.
Our first step was to total the votes for Clinton and Trump in each state. We divided the total votes for Clinton by the total votes for the two. (It was a coin flip. We could have also divided using the total votes for Trump and it would have led to the same result.) That gave us a decimal/percentage figure.
We multiplied the decimal/percentage figure by the total number of Electors for the state. That gave us the number of Electors Clinton should get versus how many for Trump.
Then we applied what we call the “Chin over the bar” rule for the next step. When you go to Airborne School, you have to be able to do 6 “good” pullups to be allowed to jump. For a pullup to be considered “good”, you had to get your chin over the pullup bar. What constituted the “chin over the bar” here was a rounding rule.
For example, let’s say a State has 7 Electors. Let’s presume we do the calculation described above and each candidate is entitled to 3 point something Electors. So how do we decide which candidate gets only 3 Electors and which gets 4? We decide based on the calculation. The candidate who has a calculation of 3.50 or better gets the fourth Elector. The candidate with 3.49 or less gets 3 Electors.
The other interpretation of “chin over the bar” is this. If a candidate doesn’t get enough popular votes to earn at least one elector, then they get zero electors in that state. As you will see, there were two states where Clinton got zero electors using our method. Trump got zero in DC.
This points up another limitation of our demonstration. The data we have is for behavior based on the old existing rules. People are voting based on the old rules. If the new rules were in effect and there was no penalty for making your first vote with your heart (as opposed to your strategic brain) then the numbers would probably be more votes for third parties. And what that might have meant for the 2016 election is that the final two might have been Trump and Johnson. Or, it might also have meant that with those realities to consider, Democrats might not have nominated Clinton. But, once again, we point out that this is a demonstration to show you how our method might work.
So, for each state, we calculated how many Electors both Clinton and Trump should get with our method. We should note that as things stand today, Trump has 306 electors and Clinton 232. California gave Clinton all 55 of its electors. Texas gave Trump all 38 of its electors.
When we did these calculations the first time our initial result was that Trump got 271 Electors and Clinton 267. When we did it the second time with updated popular vote numbers, Trump got 270 electors (exactly the number needed to win) and Clinton got 268. California gave Clinton 36 electors and Trump 19 with our method. Similarly, Texas gave Clinton 17 electors and Trump 21. So, what does that result show? Why didn’t Clinton win if our method was supposed to better reflect the popular vote in the Electoral College?
One explanation is that this result may illustrate a point critics often make about the Electoral College. That point is that the College is tilted towards protecting the interests of small population states. We think this is sour grapes. The key is to be someone who appeals through both personality and policy to voters throughout the U.S. We think a saying of the late Al Davis, owner of the Oakland Raiders, is pertinent. “Just win, baby.” Candidates must get votes in all states, not just the big ones, especially with our method.
We also think Clinton didn’t clearly win in our demonstration because we can’t clearly predict the choices of the 5.6% of voters who voted for third party candidates. And again, we are not saying our method will guarantee the popular vote winner will also win the Electoral College. We’re simply saying our method will be more likely to sustain the cultural and economic interests of low population states while more closely reflecting the preferences expressed by the popular vote. Also, to say again, more voters voted third party this time than in either 2008 or 2012.
Remember we told you by our calculations Clinton got zero electors from two states? Those two states were North Dakota and Wyoming. In North Dakota, she only needed another 9760 votes to earn an elector. That one elector would have put the two candidates at a 269 to 269 tie.
We doubt, however, that there would have been a tie. Why? Because our calculations found 13 states (AZ, CO, FL, ID, ME, MI, MN, NH, NM, PA, UT, VA and WI) where it was very likely that at least one Elector might be moved from one candidate’s column to the other’s. Accordingly, it might be possible with our method to have a final result as high as 283 to 255 with Trump winning or 281 to 257 with Clinton winning. It would all depend on how the other 5.6% third party voters mark their second and third choices on their ranked choice ballot. How they mark their ballots would also, of course, impact the final popular vote results, too.
All in all, our method would make for a much more competitive Election.
• Voters would feel free to vote for their hearts’ desire without worrying they were enabling their least favorite choice to win.
• There would be far fewer spoiled ballots where voters refuse to vote for a Presidential candidate. There would be more voters voting because they felt their vote would truly count, no matter where they live.
• Voters would have a majority winner the majority can be satisfied with, rather than the majority being dissatisfied with the result.
• And it’s something we can have done by 2020, unlike a Constitutional Amendment which might never be approved.
All of these are the reasons we need to start demanding these changes now.
See you next time.