Why You Should Invest In Upgrading Democracy And Give To The Center For Election Science

This post is mak­ing the (fol­low-up) case for why giv­ing to The Cen­ter for Elec­tion Science (CES) is a good idea. Here’s a sum­mary:

  1. What we do

  2. Why we do what we do

  3. What we’ve done and what we in­tend to do

  4. Why we need funding

  5. Why there is urgency

  6. Our ask

  7. FAQ

What we do

We study and ad­vance bet­ter vot­ing meth­ods.

On the study end, we de­ci­pher vot­ing the­ory com­plex­ities for the gen­eral pub­lic. We do that through in­ter­nal anal­y­sis, gath­er­ing ex­ist­ing re­search, and through pri­mary data col­lec­tion. All this shines light on the is­sue of vot­ing meth­ods and helps in­form our ac­tion.

On the ad­vanc­ing end, we run ed­u­ca­tional cam­paigns alongside bal­lot ini­ti­a­tives. We work with lo­cal groups who in­volve their com­mu­ni­ties so that they can im­ple­ment bet­ter vot­ing meth­ods. We fo­cus heav­ily on ap­proval vot­ing right now due to its sim­plic­ity and strong perfor­mance as a vot­ing method.

Why we do what we do

Vir­tu­ally across the globe, we all use the worst vot­ing method there is—a choose-one vot­ing method—to elect peo­ple to ex­ec­u­tive and other offices. We trust those same peo­ple we elect us­ing that ter­rible vot­ing method to (1) spend vast sums of tax­payer money and (2) ex­e­cute the poli­cies that con­trol our daily lives.

Our cur­rent choose-one vot­ing method causes vote split­ting be­tween can­di­dates. This vote split­ting causes spoilers and can squeeze out mod­er­ate can­di­dates. Good can­di­dates some­times don’t even run for fear of be­ing la­beled a spoiler and not be­ing per­ceived as vi­able. This re­sults in bad gov­ern­ment and a poor en­vi­ron­ment for good ideas to be dis­cov­ered and im­ple­mented.

There are much bet­ter ways of elect­ing these peo­ple to office, one of which is very easy—ap­proval vot­ing.

Ap­proval vot­ing is clas­si­cally a sin­gle-win­ner method that lets vot­ers choose as many can­di­dates as they want. The can­di­date with the most votes wins. This vot­ing method tends to elect more con­sen­sus-style can­di­dates and give an ac­cu­rate re­flec­tion of sup­port for third-party and in­de­pen­dent can­di­dates. Vot­ers can always vote for their fa­vorite no mat­ter what.

Th­ese fea­tures per­mit new ideas to de­velop that oth­er­wise couldn’t un­der our cur­rent choose-one method. It also pushes for a more sta­ble gov­ern­ment over time so that the win­ner doesn’t wildly shift in ide­ol­ogy from one elec­tion to the next.

What we’ve done and what we in­tend to do

We brought ap­proval vot­ing to its first US city and mod­ern use. We did that by col­lab­o­rat­ing with a lo­cal or­ga­ni­za­tion, Re­form Fargo, who ran the ad­vo­cacy cam­paign while we did the ed­u­ca­tion cam­paign. While this was per­ceived as a long shot among me­dia who both­ered to re­port on it, the bal­lot ini­ti­a­tive passed with 64% of the vote.

We in­tend to strate­gi­cally repli­cate this in cities neigh­bor­ing Fargo, ND. Fol­low­ing that repli­ca­tion, we will move to larger cities. We can then keep scal­ing to fo­cus pre­dom­i­nantly on large cities and states. Even­tu­ally, we can go out­side the US, but that strat­egy is cur­rently un­clear.

We in­tend to achieve these out­comes by scal­ing up our pre­vi­ously suc­cess­ful efforts. That means col­lab­o­rat­ing with more lo­cal 501(c)4 groups and hiring a Direc­tor of Cam­paigns to co­or­di­nate cur­rent cam­paigns and set up fu­ture ones We’ll also be hiring a Direc­tor of Re­search to high­light is­sues in cur­rent elec­tions that are cur­rently in­visi­ble and to bring aware­ness to our other work.

Why we need funding

We have been ex­tremely grate­ful for the fund­ing we re­ceived from Open Philan­thropy Pro­ject. And we are grate­ful to Will MacAskill for be­liev­ing in our abil­ity and prospec­tive im­pact enough to recom­mend us to Open Phil in the first place. We wasted no time in trans­lat­ing that fund­ing to a his­toric win within less than a year. We’re proud of that turnaround.

To suc­ceed in our mis­sion, we need to scale up our fund­ing to han­dle large cities and statewide efforts. Early on, we’ve been lucky and our part­ners have been able to avoid spend­ing funds on sig­na­ture gath­er­ing for the ini­ti­a­tives. We also haven’t faced or­ga­nized op­po­si­tion.

In the fu­ture, how­ever, that will likely no longer be the case. Thus, the cost per per­son to use ap­proval vot­ing will in­crease. Count­ing our in­fras­truc­ture, the spend per our part­ners, and run­ning ed­u­ca­tion cam­paigns, I’d es­ti­mate the cost per per­son us­ing ap­proval vot­ing will likely in­crease to about $3. I don’t fore­see it sur­pass­ing $5/​per­son. Some of that spend will start the cal­en­dar year be­fore the ac­tual ini­ti­a­tive.

[Note that we re­quire economies of scale to get at the $3-$4/​per­son be­cause of the ini­tial in­fras­truc­ture re­quired. Tar­get­ing pop­u­la­tions of 500K+ gets us in the economies of scale range. You ba­si­cally get a dis­count on your re­turn by in­vest­ing on the mis­sion in bulk.]

So, if we’re tar­get­ing cities and states within a par­tic­u­lar year so that the to­tal pop­u­la­tions add up to 2M peo­ple, then we’ll need a bud­get of $6M. If 10M peo­ple, then $30M, and so on. In ad­di­tion to the ini­ti­a­tive work, this helps us to sup­port an­cillary or­ga­ni­za­tional pro­grams that help our over­all efforts—ac­tivi­ties like re­search, cre­at­ing elec­tion tools, and perform­ing out­reach. We are largely a funds-capped or­ga­ni­za­tion in terms of the im­pact we can make.

We’ll also have to set up a 501(c)4 our­selves soon to ac­cept fund­ing to di­rect out to other col­lab­o­rat­ing 501(c)4s. This is be­cause we’re limited by the IRS on the amount we can di­rectly spend or give sup­port­ing bal­lot ini­ti­a­tive ad­vo­cacy.

Why there is urgency

Push­ing vot­ing method re­form through leg­is­la­tion is a non­starter due to the con­flict of in­ter­est from leg­is­la­tors. It’s par­tic­u­larly a non­starter given the rel­a­tively short mod­ern timeline for ap­proval vot­ing. In­stead, bal­lot ini­ti­a­tives are the main tool for push­ing ap­proval vot­ing. Only about half the US states per­mit bal­lot ini­ti­a­tives. Even fewer have the le­gal frame­work to do ini­ti­a­tives to change vot­ing meth­ods at the city or county level.

That may still sound like a lot of places, but in­stant runoff vot­ing (IRV) (also called ranked choice vot­ing or the al­ter­na­tive vote) re­form­ers are start­ing to take up more space. One lead­ing or­ga­ni­za­tion claimed they would try to run as many as over a dozen statewide efforts by 2022. That’s not count­ing other in­de­pen­dent or­ga­ni­za­tions who have started on their own statewide cam­paigns.

As more peo­ple know about in­stant runoff vot­ing and are less aware of its sub­stan­tial in­fe­ri­or­ity to ap­proval vot­ing, it has more per­ceived trac­tion. Con­se­quently, cam­paigns for IRV are now fol­low­ing Maine’s statewide im­ple­men­ta­tion. As IRV is im­ple­mented, it can re­move op­por­tu­nity for ap­proval vot­ing re­form within a state. It can even nul­lify our wins within a state. On the other side, if the in­stant runoff vot­ing cam­paign fails or is re­pealed, it could sour vot­ers to the idea of al­ter­na­tive vot­ing meth­ods al­to­gether.

This means we have to act quickly. There are other 501(c)4 or­ga­ni­za­tions who are open to ap­proval vot­ing, but they will go with the in­fe­rior IRV if they per­ceive that we can­not provide the fund­ing. This means our fund­ing and fran­tic pace have to con­tinue.

In ad­di­tion to de­clin­ing op­por­tu­nity, there are fu­ture peo­ple to con­sider. When con­sid­er­ing fu­ture peo­ple, there’s fur­ther ur­gency to act now. If ap­proval vot­ing can im­prove poli­cies and gov­ern­ment over time, then we want those pos­i­tive effects to build as quickly as pos­si­ble to those in the fu­ture.

Our ask

Many of you are already fa­mil­iar with our work. Maybe you heard about the Open Phil grant. You may have heard me speak at EA Global or REACH Berkeley. Or per­haps you listened to the 80,000 Hours epi­sode. Re­gard­less, our team and I are grate­ful that you take this is­sue se­ri­ously enough to di­rect your at­ten­tion to it.

And to those of you in the EA com­mu­nity who have already donated, thank you. It re­ally takes a spe­cial kind of donor for this think-heavy cause. Many of you have reached out to us fol­low­ing your gift to let us know about your in­ter­est in the cause fol­low­ing the 80,000 Hours pod­cast. EA mem­bers have already donated at lev­els of $50, to $500, to $20K.

Please con­sider a gift that matches both your ca­pac­ity and your com­mit­ment to fun­da­men­tally im­prov­ing gov­ern­ment. You can give on­line through our web­site. Dona­tions up to $30K given by De­cem­ber 31 will be matched. To give by other means or at a sig­nifi­cant level, please reach out to our Direc­tor of Philan­thropy. She will hap­pily get back to you within 24 hours, even over the holi­days. We are a so­phis­ti­cated team so we can also han­dle com­plex as­sets if you let us know your situ­a­tion.

Thank you again to those in the com­mu­nity who have already given.


Q: I heard there was this thing about ap­proval vot­ing that wasn’t so good or that an­other vot­ing method was bet­ter. Also, don’t for­get about Ar­row’s The­o­rem.

A: All vot­ing meth­ods have quirks, but we main­tain that the quirks of ap­proval vot­ing are com­par­a­tively mild com­pared to the al­ter­na­tives. You can see this ar­ti­cle where we go into all the de­tails. Also, I talked with Ken­neth Ar­row per­son­ally for an hour and he said that our choose-one vot­ing method was bad. Really.

Q: How does IRV match up to ap­proval vot­ing?

A: Not very well. From en­coun­ter­ing avoid­able anoma­lies to be­ing need­lessly com­plex, IRV falls well short of what ap­proval vot­ing can offer. Here’s an ar­ti­cle on that topic.

Q: How do you de­cide what makes a vot­ing method good?

A: We look at the type of win­ner it tends to elect as well as prac­ti­cal is­sues from sim­plic­ity to im­ple­men­ta­tion. Here’s an ar­ti­cle on that topic.

Q: Will ap­proval vot­ing in­crease the num­ber of par­ties?

A: Prob­a­bly, but not by much. Those par­ties can, how­ever, get their voice heard (and ig­nored if they have bad ideas). Here’s an ar­ti­cle on Du­verger’s Law. (Fun video here). Also, third par­ties and in­de­pen­dents clearly benefit from ap­proval vot­ing. Note that the multi-win­ner pro­por­tional ver­sion of ap­proval vot­ing would en­courage more par­ties. But it’s more com­pli­cated on the calcu­la­tion end. We’ll write a post on that pro­por­tional method be­fore too long since it’s just been up­dated.

Q: Why don’t you go af­ter or­ga­ni­za­tions that do achieve­ment awards?

A: We do, though we limit our re­sources to high-im­pact op­por­tu­ni­ties. Here’s an ar­ti­cle about how we worked with The Webby Awards. We’ve also done an ar­ti­cle on giv­ing games. I’ve per­son­ally en­coun­tered some re­sis­tance when talk­ing with some large awards or­ga­ni­za­tions. They don’t col­lect the data to know whether their cur­rent vot­ing method is bad. Plus they likely per­ceive that chang­ing their vot­ing method may re­veal that their pre­vi­ously given awards have less value.

Q: The Elec­toral Col­lege is awful. Why aren’t you work­ing to get rid of it?

A: The cur­rent ac­tions to make the elec­toral col­lege moot would still leave us with that awful choose-one vot­ing method. Ap­proval vot­ing would work with this cur­rent ap­proach though (IRV wouldn’t). We wrote a whole ar­ti­cle about it.

Q: Why don’t you go af­ter pri­maries? You should be go­ing af­ter pri­maries.

A: In ar­eas where we run ini­ti­a­tives and there are pri­maries, we will be hav­ing them use ap­proval vot­ing. We’ve writ­ten lots about pri­maries. Here’s an ar­ti­cle. Here’s one, too. Here’s one more. We’ll likely write an­other one be­fore too long as well.

Q: Why don’t you tar­get third par­ties to get their sup­port?

A: We tar­get third par­ties to get their sup­port. Green and Liber­tar­ian chap­ters in mul­ti­ple states sup­port and use ap­proval vot­ing. The Liber­tar­ian Party even uses it for na­tional in­ter­nal po­si­tions. Other third par­ties use it, too. Many of those folks have already bought that IRV will help them, so we have to ex­plain how ap­proval vot­ing would be bet­ter.

Q: I listened to the 80,000 Hours Epi­sode, but I felt that you didn’t go into enough de­tail in cer­tain ar­eas.

A: Here are some quick fol­low-up de­tails into ar­eas like voter turnout where I could have given a more com­plete an­swer.

Q: Let’s talk about that web­site of yours.

A: We’ll be launch­ing our new web­site by the end of De­cem­ber to early Jan­uary. In the mean­time, we have a land­ing page for our dona­tions.

Q: How can I help again?

A: Let other peo­ple know about our work and in­vest in a bet­ter bal­lot to im­prove gov­ern­ment.