This anonymous essay was submitted to Open Philanthropy's Cause Exploration Prizes contest and posted with the author's permission.

My idea is pretty straightforward if somewhat logistically complicated to enact. It also sounds kind of batshit on first reading, but bear with me. We should fund far-right third-party candidates to run for Senate in purple states. OK, let’s back up a bit.

Firstly, we should establish the fact that for any given issue or cause we can work on, it is hugely beneficial to have the funding and fiat power of the US government on side. As much as individual new technologies and private funding opportunities might help combat climate change for example, few of these technologies or opportunities would do as much good as the federal government committing several trillion dollars of spending towards decarbonizing the economy. Of course, we might not think climate change is the most pressing concern for whatever reason, but this is an example. For a past example of the power of the US government to do huge amounts of good without huge expenditure of (political, social or economic) capital, I invite you to look at George W. Bush’s PEPFAR program which has to date saved 20 million lives for just $90 billion. Of course, the added dimension is that this money didn’t have to come from voluntary sources, it came from the federal tax pool / government debt, which the US government crucially has sole access to.

Secondly, we should establish the potential tail risks faced by America’s political system in the current climate. As a professional electoral analyst and forecaster, I consider it a non-trivial possibility that America collapses into some form of civil unrest (or in the very worst cases, civil war) in the next 10-15 years. You’ve probably heard about the general problems, but the core idea is that the Republican Party has become increasingly anti-democracy, evidenced by their belief that Joe Biden stole the 2020 election and a broader value that any election a Democrat wins is an illegitimate one. This, at the same time as America’s political institutions are more biased against one party than ever (through the Electoral College, gerrymandering in the House and on a state level, and the Senate’s rural bias). The end result is that Democrats are very likely in their last trifecta for the foreseeable future, and that’s only thanks to Joe Manchin who is hardly a Democrat in many ways that count. They are very very likely to lose the House this November and even if they gain it back in 2024, there is realistically a <2% chance of them holding the Senate then. The next ten years or more are going to be some mix of Republican trifecta and divided government, with the Senate perpetually Republican. Why is this bad? Well, divided government is increasingly unproductive as less and less politicians are willing to work across the aisle – many of the Republican Senators who are still willing to do so are retiring soon or were elected pre-Trump and will be replaced by more hardline conservatives. But more problematic is the prospect of perpetual Republican controlled government thanks to institutional biases and outright election stealing at a time when the general population is more liberal than ever (Republicans are on a downwards trajectory in the long term if they stay on their current course, due to their problems with young voters). 

Scenario: Republicans lose the popular vote in 2028, 2030 and 2032, sometimes by a wide (>5%) margin. Yet despite this, they maintain full control of the federal government throughout this time period. They can certainly do this in the House through severe gerrymandering following the Supreme Court’s acceptance of the Independent State Legislature doctrine and the gutting of Section 2 of the Voting Rights Act. They can certainly do this in the Senate, because of the extreme bias it has towards small, rural states. And they do this for President despite losing even the crucial swing states of Pennsylvania and Georgia, because the Republican legislatures in those states simply vote for a slate of Republican electors who certify a Republican victory. This is really bad, not just for reasons predicated on my personal dislike of the Republican Party and all it stands for. There will be a point at which the Democrat-aligned majority will no longer tolerate the institutions that consistently deny them power. This could take many forms. Mass civil unrest and subsequently violence, attempts of blue states at secession, a further lack of coordination between diametrically opposed federal and state governments on crucial issues, an eroding of social trust (and flow-on effects, such as low COVID-19 vaccine uptake which cost hundreds of thousands of lives in 2021-22), and potentially damaging effects on the academic and research sectors, which Republicans see as opposed to their values. Not to mention the conservatism of the GOP resulting in a hollowing out of state capacity to tackle any global issues, and a perception from other countries that if America isn’t doing anything (e.g., on climate) why should we either? Ultimately, the modal outcome for what America looks like politically in ten years is pretty grim and the tail scenarios are America breaking apart and/or something akin to The Troubles in Ireland. Needless to say, Americans will probably have a harder time achieving effective solutions to all of humanity’s other problems in these scenarios. You would probably expect to see less billionaire money put towards malaria and more towards trying to flee the planet.

So, that’s the problem and I think it’s worth again stressing how much it underpins everything else an Effective Altruist or philanthropist or whatever might want to do. If the US government ceases to effectively and democratically operate as it has for the last 160 years … your various goals and funding plans just got a hell of a lot harder. The worst thing for trying to solve long-term problems is to throw a bunch of short-term problems at people.

Proposal

Now for my proposal. Basically, control of the US Senate is the lynch pin to this whole thing. It’s the Senate, stupid. Even despite gerrymandering and the Electoral College and potential right-wing power grabs, it is probable that the Democrats will control the House and the Presidency again at some point before say, 2036. The Senate however, is a worse picture. We’re probably looking at Republicans winning 56 Senate seats at a minimum by the end of the 2024 (picking up 1-2 seats this year & Ohio, Montana, West Virginia plus a few other potentials in 2024). A 60-seat supermajority is not at all out of the question. And the opportunities for Democrats to win the Senate back are pretty scarce; any Senate majority they’ve ever had has been built on conservative Democrats from red states and that simply can’t/doesn’t happen anymore due to polarization. To see this look no further than 2018 when incumbent moderate/conservative Democrats lost in North Dakota, Florida, Indiana and Missouri and had close calls in West Virginia and Montana – in a Democratic wave year! So, the Senate is where the focus needs to be. If Democrats are winning the Senate, they are probably winning the House and the Presidency too. The gameplan is basically this; if Democrats can cobble together a trifecta again at some point before irrevocable damage is done (say, in 2028), they can do a lot of things to entirely fix this situation. They can ban gerrymandering, they can add DC, Puerto Rico (and potentially Guam or the USVI) to balance the Senate some, they can ensure Presidential elections are secure and they can abolish the Electoral College if they get enough power on a state level (which comes from abolishing gerrymandering on a state level too). The only reason they are not doing these things at the moment is the presence of folks like Manchin and Sinema in their slim majority – no one like Manchin or Sinema will be left in the Dem caucus by 2028. 

Essentially Democrats have a slim window to do something to save their Republic now, but they’re not taking it because Manchin and Sinema are idiots. No other window presents itself, so we’re gonna have to squeeze one open. Unfortunately, winning Senate elections without any kind of new strategies is gonna be pretty hard because of the incentives placed on Democratic politicians to win primaries and appeal to liberal donors. Democratic nominees in purple states are gonna be more AOC and less Manchin going forward, and that hurts their ability to win. Given how much $$ floats around in these intra-Democrat politics, it’s also a pretty intractable problem. So, we have to try and game the general elections. The answer is simple – we fund third-party far-right spoiler candidates that cost the GOP winnable races.

It’s an idea that’s talked about a lot. Ever hear people say that Jill Stein cost Clinton the 2016 election? Except that’s not true, and it’s because normally third-party voters don’t lean towards one side or the other. If you’d forced Stein voters to choose between Trump or Clinton, many would’ve picked Trump or stayed home. There’s a pretty good instance where this isn’t true though: the 2020 South Carolina Senate race. Constitution Party candidate Bill Bledsoe was running to the right of incumbent GOP Sen. Lindsey Graham; also present in the race was Democrat Jaime Harrison who raised something crazy like $100M. Bledsoe clearly realized he was being a bit of a spoiler in what polling suggested might actually be a competitive race in a red state and dropped out, but not before qualifying for the ballot and being unable to be taken off. Bledsoe made it clear to voters, as did Graham, that he had dropped out and was endorsing Graham and that it was important that Harrison not be elected. But Harrison did something sneaky – he and affiliated groups ran ads that talked about how conservative Bledsoe was and how Democrats shouldn’t vote for him. “Bledsoe is 100% pro-Trump and pro-gun, too conservative for South Carolina”, that sort of thing. The clear strategy here was that Republicans would see these ads and think this Bledsoe guy was pretty good. And it worked, Bledsoe got 1.3% of the vote despite having dropped out a month earlier and made it clear he was backing Graham. For reference, third party candidates in South Carolina don’t often get more than 2-3% of the vote. What was special about this 1.3% though? It all came from Republicans. Whereas Libertarian Party voters might split 60-40 GOP and therefore every 100 votes for a Libertarian is only +20 for the Democrats, Bledsoe’s 1.3% probably would’ve cost Graham at least 1% in margin. Here’s a scatterplot (each dot represents a county x vote type) to show this relationship.

Apologies for the unlabelled chart, I lost the original spreadsheet with the data in it but the x-axis is % Harrison, the y-axis is % Bledsoe. As you can see, Bledsoe got very few votes in the most Harrison friendly areas, while often getting >2% of the vote in the most Republican areas. What are the implications of this? Well, if you had someone like Bledsoe in every Senate race but they didn’t ruin it by dropping out and endorsing the GOP candidate, it’d probably cost the GOP a couple points in margin in said races. Bledsoe didn’t end up mattering in South Carolina – the state’s partisan lean was too much to overcome and Harrison lost by 10 points – but he could’ve made the difference in a Senate race like Florida’s in 2018. And in the next few years there are going to be a lot of close Senate races. The problem for Democrats is that they need to win nearly all the races in purple states, Republicans only need to win a very small number. The following states are within R+6 or D+6 in partisan lean using 2020 Presidential numbers: Nevada, Arizona, Minnesota, Michigan, Wisconsin, Pennsylvania, Georgia, New Hampshire, Maine, North Carolina. That’s a lot of potential Senate seats. 

So, how would this actually work and what kind of funding is required? That’s the actual question, isn’t it? As I mentioned at the start, the logistics of recruiting candidates is a pretty big hurdle. You want to find highly conservative individuals who can be persuaded to run for a far-right third party, such as the Constitution Party, you want them to be reasonably articulate and hopefully have some kind of compelling story, background or community connections (veterans are always good, maybe a local gun shop owner or pro-life activist). Most of all, you need to be pretty duplicitous and deceitful about your motives. The groundwork for this could take years, such that even if you started next year, it might be 2028 before this bears fruit – right when we’d need it to. It would also be potentially hard to find campaign staffers who are experienced and would be comfortable working for a far-right candidate but also would be willing to work for a campaign that damages the GOP – most GOP staffers would laugh in your face. But how best to approach these logistical problems is a question that we can begin to try and answer, with expertise from people who have experience in this field, before committing any money. If we can’t get a candidate, no biggie, nothing gained nothing lost. If we can get a candidate, a million dollars could go a loooong way. 

Most third-party candidates have very little money and due to ad saturation/diminishing marginal returns your first million dollars is worth a lot less than your fiftieth million dollars. Also, the added benefit of doing this in the Senate not the House is that many key Senate states (New Hampshire, Maine, Nevada) are pretty small but still have the same weight in the Senate and so you can do quite a lot in these races. It might be tempting to look at the huge sums Democrats spent in 2020 races such as Kentucky and South Carolina only to lose by double digits and say this is money wasted. But it’s also important to note what money can and can’t do. Money can get you name recognition; it can’t get people to vote against their values. Democrats lost SC and KY in 2020 because they ran liberal candidates in conservative states. But a far-right candidate trying to appeal to the 20% of Wisconsin voters who are genuinely far right? Not such a hurdle. Remember, we’re not trying to get the far-right candidate to win; the exact opposite. Furthermore, when a normal campaign is spending its money, it’s trying to appeal to the less politically engaged moderate swing voter who consumes average media and can live anywhere. It’s a lot cheaper and easier to target specific areas that are super Republican and run ads in specific media spaces that are super Republican. Luckily, many of these places are very rural and so television advertising is a lot cheaper than in say, Los Angeles. 

Importance, Tractability, Neglectedness

Now, to try and apply the Importance, Tractability, Neglectedness framework that I suppose might be helpful.

Importance: As largely discussed above, the implications of control of the US federal government and the potential for mass civil unrest and social breakdown in America are enormous.

Tractability: A lot of research can be done to figure out how actually possible this idea is, without spending very much money at all. Most of the questions I have revolve around possible unforeseen flow-on effects and what the actual campaign reality would look like. In terms of the impact per dollar, I think investing in ten tossup Senate races between 2026 and 2032 at $2M per race could garner an average performance of 2-3% for our third-party candidate, at a cost of about 1.5% of vote margin for the GOP, which would in expectation result in the Democrats winning 2 more Senate seats than they otherwise would. At a rough estimate, this would increase their chance of Senate control by 2032 (which is quite equivalent to trifecta control) from maybe 5% to 15%, conservatively. Evaluating the net impact on human lives of this outcome is beyond me, but I’d say it’s not too out there to put it in the trillions of dollars in value, making the 10% increase in probability worth hundreds of billions. In terms of lives saved, when we look at American government undertakings such as the Iraq War or PEPFAR (both ironically by W. Bush), it’s easy to see how a federal government can save or end hundreds of thousands or millions of lives. So even if the difference between my possible scenario and the current path we’re on is only marginal, it could still be worth tens of thousands of lives. I don’t think that’s too bad for $20M.

Neglectedness: There are some pretty clear reasons why it’s been neglected. Firstly, this is a pretty new problem, this whole paradigm whereby Democrats are incapable of winning Senate races in red states has really only started since 2014 and the Trump era has accelerated things a lot. Other problems, like the evisceration of the VRA, increasing polarization and the GOP’s attempts to overturn democratic elections are also really new. Secondly, it turns a lot of people’s stomachs to think about bankrolling horrible far-right lunatics. Many people might also think that platforming such a person does more harm than good. I don’t think is true – political candidates don’t really change public opinion, especially not fringe third-party ones and if you’re worried about the Overton Window, most of the issues they’d run on (a complete abortion ban, abolition of all gun restrictions, huge tax cuts for the rich, no gay marriage) are pretty well placed in the GOP already. A few extra crazies won’t do anything. Thirdly, among people with the kind of knowhow and positioning to research and enact this sort of thing, there is a perspective that third party candidates don’t change the outcome of races. This whole “Stein cost Clinton the election” thing is a myth. And it is! But as I pointed out with the example of Bill Bledsoe, third party candidates can absolutely draw nearly all their votes from one side. You just need to make it clear where they stand, and market specifically towards their given group. Lastly it is of course a political cause, and so many in the EA area would probably be wary about wading so openly in politics, from my limited experience in that community. I don’t think that means it’s a bad idea, just that it requires a lot of thought. I have tried to present my idea as not being predicated on the GOP being bad, but more so on the immense damage that will be done by having a GOP government year after and year while the American people continue to vote for Democrats.

Next steps: I think where to go from here is to find existing third-party far-right candidates (the Constitution Party is a good place to look) for whatever office, not necessarily Senate and not necessarily in a swing state, and see what the impact of some money is on their vote share, and whether it meaningfully boosts the Democrat. It could be for State House or Lieutenant Governor, it doesn’t matter. There are plenty to choose from, and this should be done ASAP. From there, we can assess the value of spending a lot more money in these important Senate races. Also, obviously one of the biggest problems with this whole operation is keeping it confidential and not making it obvious what we’re doing so if we were to apply funding to it, we’d need to take immediate steps to obscure the process.

Some questions: I’ve discussed above, but basically we’d just want to know more about the actual logistics of running such a campaign and keeping the funding sources / our intentions under wraps. There’s definitely people with a lot of experience in this industry who can answer these questions.

Anyway, I’m going to leave it here because I’m a busy student in New Zealand and I just wrote 3,300 words while I should be studying, but if you are interested in this let me know @PoliticalKiwi on Twitter.