As EA gets bigger, it is almost inevitable it will adopt more of a political focus (and it certainly seems to be moving this way already). Politics is a very high-leverage mechanism to achieve big changes, and many issues EA are concerned about (especially those in the longtermist area) cannot be effectively addressed without government action. I am writing this post to share some thoughts about how we can try to ensure that EA involvement in politics is beneficial and not wasteful or counterproductive.
In terms of my personal credibility in this area, I am the Executive Director of the Institute for Effective Policy, founded thanks to a generous grant from Astral Codex Ten (Scott Alexander). We engage with policymakers in the Australian political system to advance EA priorities. Previously I have worked as an advisor for a number of legislators at both the State and Federal level. I do not claim to know everything there is to know about politics, nor should anything I say be taken as incontestable - other political professionals may have very different views. And of course political institutions vary greatly between countries, so some of what I say may be true in Australia but inapplicable in other countries. So please weight my opinions accordingly.
RISKS OF ACTION
I rank these according to my personal judgement of how likely and/or serious these dangers are.
Political Polarisation of EA
One awful outcome that could happen to many EA cause areas is if they became seen as ideologically associated with one side of politics. This is bad for many reasons - the opposing political side may become actively obstructionist to an EA agenda, EA policies may not survive changes in government, political uncertainty may stifle even non-governmental action.
I would point to climate change as the key cautionary tale here. Elevating it into a contentious, high profile political issue has in my view led to a lot of highly negative outcomes (particularly here in Australia). There has been perpetual policy uncertainty, leading to underinvestment in the energy sector despite continually rising energy prices, and deeply entrenched ideological views have obstructed practical reforms.
There is a great temptation to see politics as a battle between good and evil where all that is needed is to vanquish the enemy, who will surely melt away as the masses flock to your righteous cause. In reality, it is rare for one side of politics to become dominant for an extended period of time in democratic countries, and therefore an approach that requires one party or one side of politics to win and maintain power for decades at a time is deeply flawed.
Making Enemies
Politicians are professional grudge-holders, and many people don't appreciate the extent to which political outcomes are governed by petty egos and factional infighting. Making a powerful or influential person into an enemy of EA is a bad outcome, and worse than them simply being unsupportive. It's much better if someone simply declines to support EA goals than if they actively work to undermine or frustrate them. And it is completely possible that they may do so for reasons entirely unrelated to their view of EA on the merits.
During Australia's recent election campaign, members of the Liberal party deliberately damaged the reputation of their own party in order to hurt their factional rivals. They did this both surreptitiously (by leaking damaging information to journalists), and overtly (by making speeches on the floor of parliament trashing their own leadership). Perhaps unsurprisingly, they lost badly. If this is what they will do to themselves over personal grudges, imagine what they would do to you!
Now, sometimes it's impossible to avoid getting on someone's bad side. Politicians are not always reasonable. And sometimes the benefits are worth it. We shouldn't be so afraid of this that we don't do anything. But we should be aware of the risks and smart about doing things that may make enemies.
In my view the easiest way to make enemies is to come for someone else's power. Political actors jealously guard their status and influence, and carry out retribution against those they perceive to have wronged them. For example, I have seen Senators deciding their votes on particular bills as payback against particular advocacy groups that ran negative campaigns against them.
Wasting Money
It's very easy to throw money away to little or no effect in politics. There are a number of relatively low-cost avenues for a certain amount of money to be useful, but in my opinion the marginal utility of political spending falls away very rapidly.
In particular, buying ads in a political campaign is normally a horrible waste of money. Think about this personally - when was the last time an ad convinced you to vote for a different candidate? Probably never - and you're not unusual!
The lack of success of cashed-up campaigns like Michael Bloomberg's $500m attempt to win the Democratic nomination in 2020 or Clive Palmer's spending sprees on behalf of the United Australia Party in 2019 and 2022 should serve as a stark warning to EAs - politics can be a money pit, where resources that could have done a lot of good are thrown away for no benefit - or worse.
COST EFFECTIVE INTERVENTIONS
These are, in my view, things that we can do relatively cheaply that have an outsized effect.
Direct Advocacy
I of course have to start with the strategy I am actively pursuing. Building relationships with policymakers and trying to convince them to support our issues of concern is something that I believe can be highly effective. Most politicians do not have a pre-existing opinion of or alignment for or against EA. Simply talking to them and convincing them of the merits of particular policies is a highly tractable and influential course of action.
I've seen how effective a good lobbying operation can be. Policy issues are varied and complex, and while policymakers will have certain issues that they have strong opinions on, there are many others where they don't really know what to think and a clear argument with solid support will carry the day. Put simply - there's a reason why every large corporation hires lobbyists. They get a return on investment!
Electorate Research for Favoured Candidates
Although getting EAs elected directly elected carries some risk (i.e. making enemies of the people you are trying to take the seat from), it also has enough upside potential that it should at least be considered where realistic. However I would caution against direct ad spending - that's highly visible (amplifying the risks) while also being mostly wasteful.
What is more useful is conducting proper research to determine the most appealing message and focus for a candidate. Politicians get criticised for being excessively poll-driven and focus group tested, but the reality is that politics is often embarrassingly unprofessional. Electoral appeal is always a consideration, but one of my great surprises upon entering professional politics was that it was far less of a consideration than I expected.
Many politicians occupy safe seats, meaning that any damage to the party brand from unpopular statements is borne by other people. Sometimes internal politics prevent them from publicly taking positions popular in their own electorates. Surprisingly often candidates don't really know who votes for them or why. Sometimes internal polling is hidden from candidates! This is what Liberal Senator Andrew Bragg had to say publicly about his party's practice at the recent Australian election (sorry, paywalled):
“The Liberal Party and Crosby Textor treat the candidates like absolute shit and don’t give them the information they need,” Bragg says. “The candidates, who are often members of parliament, all they are given is a phone briefing and if they’re lucky they might get a piece of paper. Crosby Textor omit key things like the favourability of the leader because they’re worried that will leak to the media. If you know the party leader is massively unpopular you’ll differentiate so you can hang onto the seat. But if you’re not told that how are you supposed to know? It’s conflicts galore.”
This all means that simply having good information about what voters want and being serious about appealing to them is likely to give a candidate a significant edge in many cases. This probably means not focusing on EA causes in your campaign statements!
[EDIT: I have removed some criticisms here of Carrick Flynn's campaign. On reflection, I think it was inappropriate for me to make uninformed assumptions about their messaging choices]
Building a Pipeline of Politically Involved EAs
There's a saying in politics that "the world is run by those who show up". In Australia at least, party structures tend to be small, with time commitments, membership costs and toxic political cultures turning off many people from getting actively involved.
A local branch for one of the main parties may have as few as a dozen people. These branches have the power to decide their local candidates. Obviously this is often exploited by branch-stacking operations (where someone gets friends to sign up to the party in order to support their candidacy, sometimes by paying their memberships fees), but you don't need to engage in such practices to have an influence. If significant numbers of EAs join a local branch (by "significant" I mean 4 or so) then simply being a part of the party and having a vote in the preselection means that candidates will be highly incentivised to pay attention to their concerns.
And of course, being actively involved creates opportunities for advancement. On an individual level, joining a party is unlikely to result in you (for example) getting a good opportunity to get elected or to work as an advisor. But the more EA people get involved in party structures, the more likely it is that some of them move higher up in the ranks and gain greater levels of influence.
Conclusion
I personally see great opportunities for EA to have a positive impact in the world through political involvement. I'm glad that others are embracing this view.
However, I don't want to see politics become a large area of EA expenditure, and I want us to be smart about the ways we try to shape political outcomes. In particular, I really hope Bankman-Fried thinks twice about his plan to spend $100m-$1B trying to defeat Trump in 2024. I'm no fan of Trump, but I think it's unlikely to affect the outcome and could contribute a lot to EA becoming seen as a Democratic/left wing thing that Republicans and other conservatives should be hostile to or mistrustful of.
We should seek to influence policy outcomes quietly, sensibly, and without unnecessary confrontation. We should design and frame our proposals to avoid entanglement in existing political battle lines. Making friends is easier and more useful than defeating enemies. I want to see EA-focused legislation sail through Parliament unanimously, barely even rating a media mention because there's no controversy.
Great post Nathalie,
Your insights make a lot of sense, and are well written.
However, while I am pretty convinced that this is probably the most effective way to influence policy, it does not sound like the most democratic way to me.
The way I read your post is that we should not try to get voters behind EA ideas (‘politicize them’) but instead tell voters what they want to hear and push policy behind closed doors without them having a chance to vote on it, preferably even without them knowing it afterwards (no media attention is good).
I’m not sure how to balance effectiveness and a vague notion of ‘democraticness’ and I was wondering if you, or others, have some thoughts on that.
I think it's unrealistic to expect voters to take a personal interest in all policy issues. There's too many, they're too complicated, and people aren't interested.
I've just been reviewing proposed legislation the government has introduced recently. There's a bill to replace the National Skills Commissioner with a statutory body called Jobs and Skills Australia. There's a bill to allow Medicare fraud-prevention mechanisms to apply to non-medical practitioners. There's a bill to remove one of two definitions of "export entry advice".
No normal person is paying the faintest attention to these bills. They don't have an opinion, and don't want to have an opinion. And fair enough! I wouldn't be paying attention either if it wasn't my job.
The reality is that the majority of legislation is like this - unheralded, uncontroversial, technical, boring, and quite often jointly supported by both sides of politics. The issues you hear about in the media are the exceptions, the ones that animate large numbers of people. Taxes, inflation, jobs, healthcare - things that directly affect lots of people in clear and tangible ways. So the question is, are EA issues like the majority of issues, or are they the exception?
I don't think they are exceptions, or that they should be seen that way. "Place regulatory restrictions on gain of function bio research" or "Allocate foreign aid according to objective measures of impact" fits right into that category of boring technical stuff. It's not undemocratic, people can still get upset about it if they want to. But realistically they're going to be happy to just shrug and say "Yeah, fair enough, I guess", and leave it to the experts. And that's not a bad thing.
Answer: There shouldn't be a balance. The most important thing for EA is to be effective, and other values are assessed on that criterion. It doesn't mean taking a stand for or against democratic politics.