I am excited by the prospect of other EAs redirecting our efforts and working together on political organizing.
Organizing in support of particular candidates is not a neglected area. I’m much more optimistic about working to fight the root causes of political polarization and bring an end to the culture war. Lending support to one side or the other risks accelerating the feedback loop that got us to this point. Lots of people are fighting for one side or the other. Very few are thinking about how to negotiate peace.
Perhaps the single most important fact about American politics is this: the people who participate are more ideological and more partisan, as well as angrier and more fearful, than those who don’t.
...
American politics is increasingly driven by a small group of highly ideological, highly partisan, highly politically engaged people.
My interpretation: If you think it’s obvious that one side is thoroughly good and the other side is thoroughly evil, don’t go in to politics. There are already lots of people like you in politics, and your marginal impact will be small. Your impact might even be negative: Since you’re convinced that the other side is evil, you’ll be willing to break the rules in order to win. The other side will use this rule-breaking as evidence that your side is evil. This is how polarization accelerates.
But if you’re someone who can get along with both sides, you should strongly consider going in to politics. Most people who are fair-minded enough to see the merits of both sides find conflict stressful, and avoid politics for that reason. But politics desperately needs people like this.
I suspect the most tractable approach is to brainstorm solutions for deeper incentive issues. From the second article:
The increasing levels of anger and fear in the electorate are, from the political system’s standpoint, a feature, not a bug. Pew finds that the angrier and more afraid you are, the likelier you are to actually donate or volunteer.
I’m gonna elaborate on why I think partisanship is super key to what’s going on here.
Most of you are probably familiar with Scott Alexander’s red tribe/blue tribe distinction. The red tribe leans right politically and is oriented towards nationalism. The blue tribe leans left politically and is oriented towards internationalism.
Donald Trump made history the way he got newspapers endorsing his rivals. And yet he won the election. The opinions of people working in the media and the opinions of the public seem out of sync to a historically unprecedented degree.
A lot of the early attacks the media made on Trump were for things he said about culture war topics like immigration. Culture war issues are hyper-partisan issues marked by unusually high levels of fear, anger, and mistrust. Most politicians will apologize when the media calls them out for being racist or sexist. Trump didn’t, and this made him really unusual.
Later attacks were based on non culture war topics like whether Trump could be trusted with the nuclear codes. You’d think every American wants a president we can trust with the nuclear codes. But Trump got elected anyway. Why?
I suspect many Trump supporters thought something like this: “The media says Trump would be a terrible president because he won’t apologize for his culture war stances. But the media is not on my side in the culture war. (Democrats outnumber Republicans in the media 4 to 1.) The media also says Trump’s rash personality is a threat to the world peace. But I don’t trust the media, and it sounds like they are saying that mainly to get their way in the culture war. I know they’ve made hysterical statements to get their way in the past. Trump has demonstrated that the best way to win the culture war is to never apologize. Therefore, we shouldn’t apologize for Trump’s rash personality.”
By running culture war attacks alongside global stability attacks, the media implied that unapologetically having the wrong opinions in the culture war and threatening global stability were equivalently bad. Take the popular Trump=Hitler analogy. Hitler is someone who refused to apologize for his politically incorrect opinions. He’s also someone who had a tremendous negative impact on global stability. But I never saw anyone making this analogy clearly specify which of these issues they were more concerned about. It’s easy for me to see how a person from the red tribe could have dismissed the Trump=Hitler analogy as unjustified hysteria, based on their perception that the media exaggerates how terrible people in the red tribe are. This could have lead to them dismissing the possibility of Trump having a tremendous negative impact on global stability. The person offering the analogy is not making this distinction—you’d have to steelman their analogy in order to make it yourself.
To put it another way, the initial culture war attacks on Trump made him antifragile. People rejected these attacks as invalid, and that made it easy to reject further attacks as invalid. The media was the boy who cried wolf.
Red tribe people and blue tribe people have different values. Because the tribes collectively determine who gets the nuclear codes, we would like them to thoughtfully negotiate peaceful compromises.
In a good compromise, each party offers to give up some things it doesn’t care about much in order to get the things it cares about a lot. What we saw in this election was the opposite of a good compromise. The largely blue tribe media cared so much about their ability to extract apologies from candidates that they did not make a principled distinction between having the wrong opinions and threatening global stability. Trump’s largely red tribe support base told the media they cared so much about removing this ability, they’d elect someone who was a threat to global order to make their point. With no credible third party to referee the negotiation, Trump got elected.
My interpretation: If you think it’s obvious that one side is thoroughly good and the other side is thoroughly evil, don’t go in to politics. There are already lots of people like you in politics, and your marginal impact will be small.
I think you’ll have to do much more to establish the futility of partisanship efforts, especially if the sources for this are an assortment of op-eds and blogs. Politics is an adversarial game, as opposed to being a tough technical or economic problem, so the traditional ‘neglectedness’ criterion might not apply so well. From what I have seen, game theory models of conflict seem to predict the absence of this kind of effect.
Since you’re convinced that the other side is evil, you’ll be willing to break the rules in order to win. The other side will use this rule-breaking as evidence that your side is evil.
I think there’s a pretty clear alternative here; namely, don’t break rules.
But if you’re someone who can get along with both sides, you should strongly consider going in to politics. Most people who are fair-minded enough to see the merits of both sides find conflict stressful, and avoid politics for that reason. But politics desperately needs people like this.
This is not clear; there are large coalitions of centrist Republicans and blue dog Dems who are around the middle of many issues. And you’ll find that many of them are grossly unpopular both with people on the opposing side and with the hard liners in their own party. Moreover, the value of getting them elected is lower than if you got a solid candidate in office.
Systems are hard. Institutions are hard. If your goal is to replace the current systems with better ones, then destroying the current system is 1% of the work, and building the better ones is 99% of it. Throughout history, dozens of movements have doomed entire civilizations by focusing on the “destroying the current system” step and expecting the “build a better one” step to happen on its own. That never works.
My view: agents in our political system are incentivized to make people think governance is an easy problem, and all the difficulties are caused by people of the opposing faction. I don’t think this is actually true.
I think there’s a pretty clear alternative here; namely, don’t break rules.
I’m less worried about partisans going in to politics if this is their philosophy. But “don’t break rules” is not a simple principle. There are lots of legal ways to behave in an underhanded way. I expect a partisan is more likely to (a) see the behavior of their opponents as underhanded even when it isn’t (b) engage in behavior their opponents will see as underhanded without realizing it (c) rationalize underhanded behavior as justified.
In general, I am much more optimistic about trying to change the rules of the game than playing the game. If you’re playing to win, your actions are often predetermined by the game’s structure. My model: someone who works to make the game better for everyone, in a way that doesn’t strongly favor one side or the other, often finds this surprisingly tractable due to lack of opposition.
No, an ad hominem would be if I said, “Scott Alexander is a loon, and he’s saying that politics has lots of toxoplasma; therefore, politics must not have lots of toxoplasma.” But that’s not what I’m saying. I am saying “you’re backing up your argument by taking a bunch of blogs and op eds as evidence, but blogs and op eds aren’t a very good authority so you haven’t supported your claims well.”
Something more effective would be to just reference, say, the Pew study looking at American political attitudes. That would be a good data point, and it would be easier than pouring through an op ed. That way we know that you’re supporting your argument with good evidence and we can immediately see how it’s relevant to the issue at hand. And we aren’t being told to accept the elaborate narratives and points of view of bloggers and editors at face value.
I agree with these sentences from Scott Alexander:
Well the 1% and 99% figures are clearly over-exaggerated. And I don’t think this is much of an issue at all with typical political cycles. SCOTUS rulings, for instance, get made and overturned with fairly equal difficulty (just have the president and senate appoint judges with the right views; it’s essentially symmetric). Major legislation gets repealed, but rarely; even the ACA might be preserved in part, because of the need for it and the reliance upon it.
On the flip side, making new institutions is not particularly difficult; it’s practiced all the time with passing new laws, establishing new agencies, etc., and this does not seem to be an overwhelmingly difficult practice to me. Most of American governmental history has been the successive construction of new and better government institutions against resistance, not a destructive war in which institutions are constantly created and destroyed.
I think it would be worthwhile to take a more detailed look at legislative history in order to pursue this line of thought.
If you’re playing to win, your actions are often predetermined by the game’s structure.
Whew, I hope you don’t expect everyone to read through that when they see your comment. I see that and I hear, “If you’ve read Scott’s winding 10,000 word essay, you’re in my ingroup and you get to know what I believe and why I believe it. If you haven’t, then come back to me after you’ve finished; otherwise, you don’t know enough to refute my argument, because I’m using it as a source.”
Scott’s essay is about game theoretic problems between actors leading to inevitable conflict. I’m not sure in what sense you use the term ‘predetermined’ but it certainly isn’t the case that pushing a side in partisan politics leaves the world no different.
My model: someone who works to make the game better for everyone, in a way that doesn’t strongly favor one side or the other, often finds this surprisingly tractable due to lack of opposition.
This, again, is something that would warrant an investigation of policymaking history, especially since you’re citing a blog by someone who, as far as I know, has not played a key role in any policy shifts, despite presenting it as their chosen strategy.
The institutions discussion seems to have gone off the rails. Our original disagreement was about whether governance constitutes a tough technical and economic problem. I think it does. You apparently think it doesn’t. I don’t think we’re going to make progress resolving this disagreement easily.
Our original disagreement was about whether governance constitutes a tough technical and economic problem.
No, our original disagreement was over whether partisan political activity is better than trying to systemically reform political arenas.
You apparently think it doesn’t.
No, I just think that partisan political activity is a good avenue of improving governance. I’m not making any broad claims about governance being easy or not.
I don’t think we’re going to make progress resolving this disagreement easily.
Of course not, if you want to pursue a discussion about policymaking without actually referring to legislative history.
A few of us have experience working in politics and could conceivably accomplish some good by being an influencer in Trump’s White House. Others of us have the ability to pitch Thiel on stuff. Since Thiel has sway in the Trump transition, this means we could conceivably get an EA or two into positions of influence in the Trump administration.
I’m not sure that it would be a good idea to actually do this, but I’m mentioning it because it doesn’t seem outside the realm of possibility to actually do it, and it could plausibly be highly effective. Here are some of the questions we’d need to answer:
If we had an EA inside the Trump administration, would it do more good than if they stayed in their current position instead? This depends partly on what they’re currently doing and partly on how likely we estimate they could actually make a difference in policy. My intuition is that if we expect Trump’s policies to be very bad, then even a small influence could translate into a large amount of good.
Who would be best suited for this, if we decided we wanted to try it? I’m not sure of what would count as experienced enough to do something like this. There are a few people at Effective Altruism Policy Analytics, and I believe there are a couple of people that have experience with lobbying in DC.
Who would make the pitch to Thiel?
What would the pitch consist of? We’d need to know exactly what parts of EA Thiel cares most about, and then we’d need to stress those aspects.
How likely would Thiel be swayed by such a pitch? If he endorsed Trump because he wanted influence in the administration, then I believe Thiel would be fully on board with this idea. But if he endorsed Trump because he actually believes Trump’s positions are good, then I can see where he would hate this idea.
Would Thiel be able to get the EA in a position high enough to actually influence policy? How high up would the position have to be in order for it to be influential? How influential are mid-level staffers?
I don’t know the answer to these questions. I don’t know if this is even a workable idea. I certainly would hate to convince an EA to drop their current work for this if it doesn’t turn out to be an influential position. But it seems possible that this could be a high-value opportunity, so I’m bringing it to everyone here.
I think this is the link for applying for a position in Trump’s administration. Trump has 4,000 positions to fill over just a few weeks, and lots of people will be reluctant to work for him. So you probably have as good a shot as anyone if you’ve been building career capital. You’ll probably have to pretend you support him—consider it a 4-year ideological turing test.
This is a good idea and it doesn’t change much from president to president. The big thing with politics like this is just getting your foot in the door.
Thiel will have the president’s ear on technical issues, he said, though he will not have a formal position and will stay in DC. It might be a good chance to get someone in the OSTP, as one opportunity.
I definitely became less interested in politics ever since identifying as an EA or utilitarian. But then Switzerland passed some ridiculous xenophobic propositions, and Brexit happened, and now Trump. And every time I had this worry in the back of my mind that we’re doing something wrong.
Carl mentioned “Misallocating a huge mass of idealists’ human capital to donation for easily measurable things and away from more effective things elsewhere, sabotages more effective do-gooding for a net worsening of the world” here. This point doesn’t just apply to money, but also very much to attention and activism. And the bias may not just be towards things that are easily measurable, but there may also be a bias away from “current” or “urgent” events. These events shape public discourse, which could have important flow through effects. What’s the effect if altruistic and driven people disproportionally stop caring about current events and the discussions that surround them?
Perhaps it’s negligible, but it’s certainly worth thinking about more. And I was glad to see how much attention the recent votes got within EA.
I was under the impression that a bias against urgent events is an inculcated response that takes active strength to develop, a resistance to the much stronger human bias to act on urgent but not-as-important things. It seems to me that most altruistic and driven people still disproportionally care about current events and the discussions that surround them, over the Important Things.
I’d be interested to hear a case for moving from animal advocacy to politics. If your comparative advantage was in animal advocacy before the election, it’s not immediately obvious to me that switching makes sense.
In the short term, animal welfare concerns dominate human concerns, and your marginal contribution to animal welfare via politics is unclear: welfare reform in the US is happening mostly through corporate reform, and it’s dubious that progressive politics is even good for wild animals due to the possible harms of environmentalism.
Looking farther into the future, it’s not clear that engaging in politics is has become more effective on the margin than spreading anti-speciesism.
Politics is still a crowded space and it’s looking like many other progressives have been galvanized by this result.
The case is for defending the conditions under which it’s even possible to have a group of privileged people sitting around worrying about animal advocacy while the world is burning. To the extent that you think 1) Trump is a threat to democratic norms (as described e.g. by Julia Galef )/ risks nuclear war etc. and isn’t just a herald of more conservative policy, and 2) most liberals galvanized by the threat of Trump are worrying more about the latter than the former, there’s room for EAs to be galvanized by the threat of Trump in a more bipartisan way, as described e.g. by Paul Christiano.
(In general, my personal position on animal advocacy is that the long-term future of animals on Earth is determined almost entirely by how much humans have their shit together in the long run, and that I find it very difficult to justify working directly to save animals now relative to working to help humans get their shit more together.)
Trump may represent an increased threat to democratic norms and x-risk, but that doesn’t mean the marginal value of working in those areas has changed. Perhaps it has. We’d need to see concrete examples of how EAs who previously had a comparative advantage in helping animals now can do better by working on these other things.
my personal position on animal advocacy is that the long-term future of animals on Earth is determined almost entirely by how much humans have their shit together in the long run
This may be true of massive systemic changes for animals like the abolition of factory farming or large-scale humanitarian intervention in nature. But the past few years have shown that we can reduce a lot of suffering through corporate reform. Animal product alternatives are also very promising.
Also, “having our shit together in the long run” surely includes anti-speciesism (or at least much higher moral consideration for animals). Since EAs are some of the only people strategically working to spread anti-speciesism, it seems that this remains highly valuable on the margin.
Edited to add: It’s possible that helping animals has become more valuable on the margin, as many people (EA and otherwise) may think similarly to you and divert resources to politics. Many animal advocates still think humans come first. Just a speculation.
Also, “having our shit together in the long run” surely includes anti-speciesism (or at least much higher moral consideration for animals). Since EAs are some of the only people strategically working to spread anti-speciesism, it seems that this remains highly valuable on the margin.
I’d like to see an analysis of exactly what the opportunity costs are there, before endorsing one. This analysis has no differential analysis, and as such it reads “There are many important things being neglected. This is an important thing. Therefore it is the most important thing to do.”
...as such it reads “There are many important things being neglected. This is an important thing. Therefore it is the most important thing to do.”
I never meant to say that spreading anti-speciesism is the most important thing, just that it’s still very important and it’s not obvious that its relative value has changed with the election.
Can I ask your rationale for P(AI created in next 20 years): Up?
Also note that the longer we delay the arrival of general AI, the more hardware and data will be available for it to immediately capitalize upon. So it’s not entirely clear that delaying AI development is good.
Can I ask your rationale for P(AI created in next 20 years): Up?
Increased US military spending will result in more resources attempting to create AI. In a climate of lower international cooperation, other countries will hasten their efforts as well.
So it’s not entirely clear that delaying AI development is good
Agreed. Many of my updates can be viewed as good, depending on your value system and other beliefs. However, forcing myself to guess, I think we have moved further away from the ideal, cooperative future—e.g. something inspired by CEV.
Well, the military itself spends very little on AI and none on cutting edge AI. What really matters is DARPA, IARPA, and whatever else might be going on in hidden parts of the intel community and DoD.
The bulk of AI development is the tech industry and academia. I think these areas are likely to be hurt by the administration.
You’re right, I meant to say “Increased US defense spending” not “Increased US military spending”. I inaccurately use “military” as a synonym of “defense” sometimes.
From Trump’s website: “Emphasize cyber warfare...and create a state-of-the-art cyber defense and offense.”
I think these areas are likely to be hurt by the administration
Maybe, but I’m not so sure the effect will be large enough.
Well cyber has little to do with AI; DARPA and IARPA cyber and AI teams are quite separate, and the service branches don’t do any AI with their cyber teams. The only exception is with the DARPA automated cyber grand challenge this year, but even it wasn’t really doing things that would accelerate AI development.
My point is more that since AI is isolated behind so many layers of budget allocation and is such a tiny slice, that changes at the top level are likely to not change anything.
Have you considered how Peter Thiel, big-time AI safety supporter, is to become one of Trump’s key policy advisers on technology, and will have an unprecedented opportunity to shape the policy on AI going forward?
Friends: I recently wrote a few thousand words on the implications that a Trump presidency will have for global risk. I’m fairly new to this discussion group, so I hope posting the link doesn’t contravene any community norms. Really, I would eagerly welcome feedback on this. My prognosis is not good.
Given Trump’s high uncertainty and variance, one might imagine that the chances of him doing something that increases x-risks significantly is at least one or two orders of magnitude higher than Clinton’s.
Given that, restructing the US economy to better deal with the shocks of automation and globalization, which might prevent future instances of this type of intentional systemic upheaval, suddenty seems like a potentially important new category of x-risk by proxy.
By that logic, does something like UBI become an important potential area of focus?
Given that, restructing the US economy to better deal with the shocks of automation and globalization, which might prevent future instances of this type of intentional systemic upheaval, suddenty seems like a potentially important new category of x-risk by proxy.
More like x-risk by proxy by proxy.
It should be easier to focus directly on building international institutions.
| It should be easier to focus directly on building international institutions.
Can you clarify this? As a new EA, I am unfamiliar with what this might be in reference to. Are you arguing for focusing on the creation of a second nation-state promoting liberal values and technological advancement (to backstop US leadership)? Or just circumventing politics and going straight at the underlying risks (AI, nuclear proliferation)? It seems without the air-cover of a liberal superpower, pursuing such ventures will be significantly harder.
Well for Trump increasing x risks, the potential downside is that there is going to be less international cooperation and less prospect for coordination between nations on issues like AI. That’s what I thought you had in mind and it implies that we should try to build more credible and effective groups like the UN and IAEA. It’s easier to work on this directly rather than doing things that make people less likely to support populists who are against international involvement. I guess there are other ways Ttump could be bad for x risks, so that might change things.
Gotcha. I was going the other way; considering what automation and globalization have done to low-skilled workers over the last 30 years (which is only going to get a lot worse), some sort of significant restructuring of our economic system (both wealth distribution and cultural self-actualization via employment) seems necessary to prevent full-scale upheaval (e.g. revolution, civil war, etc). Trump seems to be the first major indication that this problem is getting so bad that it could literally end America or at least American leadership and the promotion of liberal values.
The case is for defending the conditions under which it’s even possible to have a group of privileged people sitting around worrying about while the world is burning.
I do agree this is a very indirect effect but I also failed to see how more direct action would be more effective.
Well that is different but I don’t think it’s a likely threat.
considering what automation and globalization have done to low-skilled workers over the last 30 years
Low-skilled workers have seen general increases in standard of living over the last 30 years and no great increase in unemployment. Globalization generally increases wealth for all income brackets in the long run and automation has not historically increased long-term unemployment.
(which is only going to get a lot worse),
We may be able to have trucks that steer themselves on the road, but actually having unmanned vehicles take themselves from A to B reliably enough to be unmanned and without requiring a human in the cabin to do anything is years away.
In any case, the lack of truck drivers is a significant constraint on our economy alongside the fact that they can only drive 10 or so hours a day. Automation here might be unusually good.
some sort of significant restructuring of our economic system (both wealth distribution and cultural self-actualization via employment) seems necessary to prevent full-scale upheaval (e.g. revolution, civil war, etc).
It seems to me like wealth is a pretty straightforward question of welfare to resolve. As for cultural self-actualization… we’ve had retirees, part-time employees, and permanently unemployed people for a long time, but they never seem to be more troublesome or more dissatisfied than the rest of the electorate.
Trump seems to be the first major indication that this problem is getting so bad that it could literally end America or at least American leadership and the promotion of liberal values.
I don’t think so. We’ve had similar issues in the past with populist leaders and their movements, e.g. William Jennings Bryan (barely failed to get elected) and Huey Humphrey (assassinated). But our democracy has only improved over time.
Low-skilled workers have seen general increases in standard of living over the last 30 years and no great increase in unemployment. Globalization generally increases wealth for all income brackets in the long run and automation has not historically increased long-term unemployment.
Is this true in the US? My personal circle of concern certainly includes all of humanity (and beyond), but Trump’s election would appear to be an indication that displaced Americans don’t think that way.
We may be able to have trucks that steer themselves on the road, but actually having unmanned vehicles take themselves from A to B reliably enough to be unmanned and without requiring a human in the cabin to do anything is years away.
Do you have some info you can share on that? Everyone I’ve talked to in the autonomous vehicle area has indicated that adoption, not technology, will be the long pole here. And as you indicated, there are massive economic forces that will push for adoption.
And while I totally agree automation will be good for GDP, it will continue to displace a set of people who can vote.
we’ve had retirees, part-time employees, and permanently unemployed people for a long time, but they never seem to be more troublesome or more dissatisfied than the rest of the electorate.
Retirees already have cultural approval to not work: they’ve “earned” it. Part-timers and permanently unemployed are viewed as lazy or moochers by many folks in the US, where there’s a culture of self-improvement via working hard. That’s what I’m arguing will need to change.
I don’t think so. We’ve had similar issues in the past with populist leaders and their movements, e.g. William Jennings Bryan (barely failed to get elected) and Huey Humphrey (assassinated). But our democracy has only improved over time.
What metric are you using for improvement? On the axis of existential risk prevention, I would argue we’ve just taken a big step back.
Yes, average income when adjusted by PPP has increased (the standard narrative is that CPI-adjusted income has not increased, but the CPI is flawed—it overestimates inflation by focusing on cities and it ignores goods substitution), and nonwage compensation has increased (health benefits etc).
Do you have some info you can share on that?
Sure, class 8 trucks are a couple hundred thousand dollars each and the cargo is worth a lot too. They have mechanical problems. They have expensive insurance. They have complex routes. They have to reliably reach their destination and meet with the proper people to load and unload. With all this responsibility and risk it’s an easy decision to put a driver at the wheel, especially when even ordinary cars aren’t ready to go around without drivers.
Everyone I’ve talked to in the autonomous vehicle area has indicated that adoption, not technology, will be the long pole here.
Well there you have it. If it will take a long time to be adopted then there won’t be an unemployment problem for a long time.
And while I totally agree automation will be good for GDP, it will continue to displace a set of people who can vote.
They will vote for welfare, which they will get.
Retirees already have cultural approval to not work: they’ve “earned” it. Part-timers and permanently unemployed are viewed as lazy or moochers by many folks in the US, where there’s a culture of self-improvement via working hard. That’s what I’m arguing will need to change.
I don’t think any of this takes away from my point.
What metric are you using for improvement?
None of the democracy and freedom indices have been around for that long, so there is no metric. But we can look at the corruption and civil values of 100 years ago, and comparing it today, and note that it’s pretty clear that things are better all round now than they were back then. For instance, we don’t prevent women and blacks from voting anymore, and we don’t put people in prison for speaking out against the war. We don’t hold racist views about who is eligible to hold public office.
On the axis of existential risk prevention, I would argue we’ve just taken a big step back.
That’s a different axis, and it has taken many steps back, perhaps with the elections of GWB, Ronald Reagan, Eisenhower, and Truman.
Yes, average income when adjusted by PPP has increased
The distribution is what would matter for my argument. Do you have any links to data there? Where is this data from? Thanks.
Sure, class 8 trucks are a couple hundred thousand dollars each… If it will take a long time to be adopted then there won’t be an unemployment problem for a long time.
I have first hand information from those knowledgable on the topic that the technology will be ready in the next few years, and I haven’t heard a refutation of that. So I said the long pole was adoption. I didn’t say how long the pole was. You already acknowledged that the economic incentives for adoption are huge. I posit that adoption will be faster than you suspect. But we have to wait and see on this one, unless you have more information to share on why you disagree.
They will vote for welfare, which they will get.
They didn’t vote for welfare this time. Which is my point.
I don’t think any of this takes away from my point.
Maybe you can explain… you seemed to be saying that the fact we have part-timers and unemployed proves that we don’t have any cultural stigma against not working. I said that people do have such a stigma, and then you said that doesn’t take away from your point. Not sure what I missed.
That’s a different axis, and it has taken many steps back, perhaps with the elections of GWB, Ronald Reagan, Eisenhower, and Truman.
Great, that’s the axis I was trying to focus on, which is why I asked what metric you were using. And yes, it’s taken steps back before. That’s doesn’t preclude it getting worse.
I’m simply pointing out that if automation and globalization disrupt the US middle class enough, political instability could increase x-risks.
They didn’t vote for welfare this time. Which is my point.
You’re talking about what would happen in a world of high structural unemployment. But we don’t have high structural unemployment, so you can’t say that future elections will have the same dynamics. In this election cycle, Trump voters weren’t primarily motivated by the economy, they were motivated by immigration and culture issues.
Maybe you can explain… you seemed to be saying that the fact we have part-timers and unemployed proves that we don’t have any cultural stigma against not working.
I said that it shows that unemployed people aren’t likely to start a revolution or anything just because they’re unemployed.
People don’t start upheavals just because they don’t have cultural acceptance.
I’m simply pointing out that if automation and globalization disrupt the US middle class enough, political instability could increase x-risk.
If you take a broad view then half of all things will increase x risk in the long run and half of them will decrease it. I think that the issue here is too crowded and intractable and the effects are too complicated and uncertain for it to be a big goal for us to put effort on, though I agree that it would be good if the nation had a robust social system going forward.
In this election cycle, Trump voters weren’t primarily motivated by the economy, they were motivated by immigration and culture issues.
I think we need to see more data on this one. I’d put forth the idea that in fact the underlying issues were in fact economic, and that the cultural and immigration issues were mostly just utilized by Trump as things to blame for the underlying economic woes.
The terminology here can be confusing. When progressives talk about the working class, typically they mean the poor. But the poor, in the bottom 30% of American families, are very different from Americans who are literally in the middle: the middle 50% of families whose median income was $64,000 in 2008. That is the true “middle class,” and they call themselves either “middle class” or “working class.”
“The thing that really gets me is that Democrats try to offer policies (paid sick leave! minimum wage!) that would help the working class,” a friend just wrote me. A few days’ paid leave ain’t gonna support a family. Neither is minimum wage. WWC men aren’t interested in working at McDonald’s for $15 per hour instead of $9.50. What they want is what my father-in-law had: steady, stable, full-time jobs that deliver a solid middle-class life to the 75% of Americans who don’t have a college degree. Trump promises that. I doubt he’ll deliver, but at least he understands what they need.
I thought this analysis was interesting.
I said that it shows that unemployed people aren’t likely to start a revolution or anything just because they’re unemployed.
Ah, I understand your point now. Sure, that’s true, but I’m imagining what happens if the lower-middle class gets squeezed out, as mentioned above. I need to go learn more about economic trends of the middle class in American history before I can make this case, though.
If you take a broad view then half of all things will increase x risk in the long run and half of them will decrease it.
Wait, I’m new to x-risk analysis and TDT… but is that true? Why would the distribution of the space of all possible actions split 50⁄50 towards increasing or decreasing x-risks? That doesn’t seem right. Am I missing something?
“The thing that really gets me is that Democrats try to offer policies (paid sick leave! minimum wage!) that would help the working class,” a friend just wrote me. A few days’ paid leave ain’t gonna support a family. Neither is minimum wage. WWC men aren’t interested in working at McDonald’s for $15 per hour instead of $9.50. What they want is what my father-in-law had: steady, stable, full-time jobs that deliver a solid middle-class life to the 75% of Americans who don’t have a college degree. Trump promises that. I doubt he’ll deliver, but at least he understands what they need.
@rowborg/@kbog, this is a great thread, and I’m only going to dip my toe in, here, minimally, to +1 this point that rowborg has made. Employment isn’t just about exchanging labour for a wage, it’s a major source of meaning in people’s lives. The assertion that standards of living in the US have risen is probably correct, even for people at the bottom of the income distribution. But this standard of living increase can’t explain the incredibly sour mood of a huge chunk of people. Obviously it’s going to take a lot of time for researchers to unpack that, but the meaningless-employment story seems like a pretty plausible component: if people feel what they do doesn’t have any meaning, and particularly if they feel their children don’t have prospects for meaning-ful employment then despair and anger seem like totally to-be-expected reactions.
I think we need to see more data on this one. I’d put forth the idea that in fact the underlying issues were in fact economic, and that the cultural and immigration issues were mostly just utilized by Trump as things to blame for the underlying economic woes.
I saw it based on data from exit polls. I don’t know where I saw it. If there is any other evidence, sure that would be neat to see. Though it is objectively the case that our underlying economic woes are smaller now than they were in 2012 and 2008.
Wait, I’m new to x-risk analysis and TDT… but is that true? Why would the distribution of the space of all possible actions split 50⁄50 towards increasing or decreasing x-risks? That doesn’t seem right. Am I missing something?
Well think about it this way, if the majority of political positions increased x-risk, we would make the world safer just by doing things in the opposite way all the time. But that’s absurd.
The more general point is that you can construct a complicated chain of reasoning for lots of things to increase x risk, so it’s not particularly interested just that something increases it. What matters is how much and how tractable it is.
Though it is objectively the case that our underlying economic woes are smaller now than they were in 2012 and 2008.
On aggregate, yes. I’m talking about the economic situation for a certain group, which is the lower-middle class. I don’t know that this is true for them. I’ll go try and dive in on this.
Well think about it this way, if the majority of political positions increased x-risk, we would make the world safer just by doing things in the opposite way all the time.
If I’ve parsed your meaning correctly, you are saying every action has a counterfactual non-action, and if you include not just the space of possible actions but the impact of non-action, then risks break 50/50? That makes more sense, but doesn’t really seem to be helpful in deciding between actions to take. The space of possible actions to take is what should concern us, yes?
Regardless, I take your larger point, which is that we should put efforts where you can maximize impact and you are skeptical on the effective value of this one as opposed to others. I would certainly need to be more quantitative in my analysis of this and back up my ideas/claims for sure.
The more general point is that you can construct a complicated chain of reasoning for lots of things to increase x risk, so it’s not particularly interested just that something increases it. What matters is how much and how tractable it is.
Totally agreed. I have yet to hear an argument that 1) invalidates my chain of reasoning or 2) indicates there are other inflection points along the chain that are better. Whether or not this compares well with other options for x-risk reduction is an exercise I have yet to perform (as I mentioned above).
If we pull the camera back far enough, my guess is in a generation or two America will be on a good track again, so long as Trump doesn’t start a war or use our nuclear warheads. As the White House Press Secretary said, the institutions of the U.S. have survived a civil war, 2 world wars, and the Great Depression. This will be a bad 4 years with adverse consequences on the rest of the world. Putin will be on the offensive, both in reality and in cyberspace. And U.S. carbon emissions will increase. But I still believe AI risk is the most dangerous threat to humanity.
I’m with you, and I really, really want to be on the side of skepticism and saying that this election is no different from normal cycles we’ve had in the past. But this is the first time ever that we’ve given serious support to someone with a history of neither political office nor military leadership. And someone who really clearly has no consistent train of thought or policy acumen. If the economy happens to be strong and the country is successful, then people will appreciate his persona and values as the basis for a very entrenched brand of populist conservatism. So in this case I think it’s okay to say that the long term future of America could be seriously altered from what we consider to be normal, and it’s okay to say that this election was unusually consequential.
EAs like to focus on the long term and embrace probabilistic achievements. What about pursuing policy reforms that are currently inconsequential, but might have profound effects in some future state of the world? That sort of reform will probably face little resistance from established political players.
I can give an example of something I briefly tried when I was working in Lesotho, a small, poor African country. One of the problems in poor countries is called the “resource curse”. This is the counter-intuitive observation that the discovery of valuable natural resources (think oil) often leads to worse economic outcomes. There are a variety of reasons, but one is that abundant natural resources often cause countries with already-weak institutions to become even more corrupt, as powerful people scramble to get control of the resource wealth, methodically destroying checks and balances as they go.
In Lesotho, non-renewable natural resources—diamonds—currently account for only a small portion of Lesotho’s GDP (around 10%). I introduced the idea of earmarking such natural resource revenues received by the government as “special”, to be used only for infrastructure, education etc projects, instead of effectively just being consumed (for more info on this idea see this article or google “adjusted net savings”). Although this change would not have huge consequences right now, I thought that it might if there were a massive natural resource discovery in Lesotho in the future. Specifically, Lesotho might be able to avoid some of the additional corruption by already having a structure set up to protect the resource revenues from being squandered.
The idea I’m putting forward for a potential EA policy initiative is to pursue a variety of policy changes that seem painless, even inconsequential, to policymakers now, but have a small chance of a big impact in some hypothetical future. The idea is to get the right reforms passed before they become politically contentious. While it can be hard to get policymakers to pay attention to issues seen as small, there are plenty of examples of political capture that could have been mitigated by early action. And this kind of initiative is probably relatively neglected given humanity’s generally short-term focus. I think EAs are uniquely well placed to prioritize it.
Firstly, Trump’s agricultural advisors seem to be very hostile to animal welfare. This may mean that we need more people working on farmed animal welfare, not less.
In terms of going into politics, the prospect of having a group of EAs, and perhaps even an EA-associated organization, doing regular, everyday politics may turn some people off from the movement (depending on your view on whether EA is net-positive or negative overall, this may be bad or good.)
While Sentience Politics, the Open Philanthropy Project and some others I may have missed do take part in political activities, they focus on specific policies, and I suspect that what some people are talking about would involve a systematic attempt to engage in party-politics.
I think that even without Trump, the idea of having a very small number of individual EAs (maybe 1/1000 EAs) going inside politics and trying to influence administrations or even become politicians was a good one.
But, a systematic attempt to engage in party-politics would not be a good idea, partly because, even in the EA community, focusing on party-politics or even on controversial policies seems to lead to less willingness to consider other points of view.
And, partly because influencing administrations or becoming a politician on one’s own is more likely to make a difference than engaging in regular party-political campaigning, even though becoming a politician or influencing an administration is less easy to do.
Finally, I think that politics is very important, because you could potentially reduce existential risks as well as spread good values and ensure that humanity is on the right course in the future, and therefore there’s not a tension between reducing existential risks and values-spreading.
However, in order for any politicians or political advisors to be able to steer humanity in a positive direction, you need public and corporate support for it, which is why I believe that spreading anti-speciesism, working on farmed animal suffering, and so on, remains highly important too.
Overall, Trump’s election has not influenced my beliefs significantly.
This isn’t quite relevant, but I think the election cycle was sort of mishandled by the EA community.
It was mostly sort of ignored, except for some last minute stuff with people trying to put together efforts to fight Trump. Some people jumped quickly on bandwagons and picked a side. Other people just took a blank “politics is the mindkiller!!” stance and refused to care.
Ideally, we should have a rational and serious approach from the outset, looking at the candidates and seeing if there is anything that EA philosophy decisively says that we should do. And when the window opens, then we come together and say “hey, here’s our analysis, here’s our thoughts, this is what effective altruism says about the election.” And push that message as far as we can.
I don’t think a rationalist message/meme would have been successful at convincing hundreds of thousands of working class whites not to vote for Trump. Rationalism has it’s place in deciding what to do about an election, but I don’t think EA messaging is at all useful in influencing a mass audience.
We didn’t need to convince hundreds of thousands of working class whites. Just increase turnout among Berners or change the primary outcomes or show Thiel the error of his ways. There’s lots of avenues to do things.
Thiel had essentially nothing to do with the outcome of this election.
This was not primarily a turnout issue. Black turnout was down, but Hispanic turnout was up. White turnout appears relatively flat (both Democratic and Republican white turnout), but we’ll know more when actual person level vote history is released. Regardless, EA messaging is not the right way to appeal to Berners.
The easiest way to shift the outcome of the election would have been to change public opinion by a point or two by shifting the narrative of the race in the final week. Comey was successful at doing this.
Thiel had essentially nothing to do with the outcome of this election.
Well he gave about 0.15% of their money, and campaign funding is fairly important. The important thing is that in future campaigns, funding may influence the outcome of elections.
This was not primarily a turnout issue. Black turnout was down, but Hispanic turnout was up. White turnout appears relatively flat (both Democratic and Republican white turnout), but we’ll know more when actual person level vote history is released.
It was all about turnout, fewer people voted Democrat than in 2012 and 2008. Trump won just by holding onto similar GOP vote totals from previous elections while people dropped Clinton.
Regardless, EA messaging is not the right way to appeal to Berners.
I don’t know what the right type of messaging is and I don’t know who we will need to appeal to in 4 or 8 years. My point is that there are probably bigger things to be done besides phone banking.
The easiest way to shift the outcome of the election would have been to change public opinion by a point or two by shifting the narrative of the race in the final week. Comey was successful at doing this.
That’s a good point but I think it was an exceptional case as the email story had been unfolding for many months and people were already primed to watch for news about it. It’s also very hard to actually do it intentionally if you don’t have some big news you can release.
At most, campaign funds would have moved this a point or two. Campaign funding has little impact on presidential elections; Clinton far outspent Trump and Trump was far outspent in the primary election. If we assume an effective size of 5% for all of Trump’s money and assume no diminishing marginal return (both very generous assumptions), that 0.15% is 0.0075 percentage points in movement. The outcome was decided by 1, so that’s over two orders of magnitude lower than what was needed under generous assumptions. It was probably more orders of magnitude lower.
That’s not true at all.Trump gained substantially in rural areas with mostly white people where Obama had won or performed substantially better. I mean, if turnout had magically been higher among Democrats but not Republicans, we would have won, but you don’t get to do that. The composition of the electorate was roughly the same (minus some black people plus some Hispanics) as 2012. It’s conceivably possible that without the drop in black turnout, we would have won, but this was inevitable without the first black president running. There is overwhelming evidence that attitudes among the white working class moved against us. Hence our drop in the midwest.
I agree on the point that phone banking does not make much of a difference.
There were several instances that fall under the same pattern: the email story, the hollywood access tapes, the debates, probably the apprentice tapes if they had appeared, and potentially the wikileak emails, though it’s much harder to gauge their effect size.
It is true. Romney got 61 million votes and McCain got 60 million. Obama got 69 million and 66 million in 2008 and 2012 respectively. This year, Trump got 60 million votes and Hillary got 61 million.
There were several instances that fall under the same pattern: the email story, the hollywood access tapes, the debates, probably the apprentice tapes if they had appeared, and potentially the wikileak emails, though it’s much harder to gauge their effect size.
Well, depending on how early before the election you want to consider. The debates for instance were all more than a week before the election. Again, it’s basically impossible to put effort into making things like this happen, and the best way to do so might simply be conventional ways of building political clout and awareness.
You can’t look at aggregate turnout numbers being different and assume the composition of turnout was different. You’re making the assumption that there was 0 movement from Obama to Trump or from Romney to Clinton; both of which are definitely incorrect as evidenced by polling.
Secondly, turnout is much higher than that appears; much more will come in from California, Washington, Oregon and Colorado. It always takes these states forever to report. So the turnout numbers now are misleading.
I think the message of SlateStarCodex’s “Tuesday Shouldn’t Change The Narrative” is particularly relevant to EAs—any large updates to one’s beliefs about the world should have come before the election.
Agreed that large updates about things like the prevalence of regressive attitudes and the fragility of democracy should have been made before the election. But Trump’s election itself has changed many EA-relevant parameters—international cooperation, x-risk, probability of animal welfare legislation, environmental policy, etc. So there may be room for substantial updates on the fact that Trump and a Republican Congress will be governing.
That said, it’s not immediately obvious to me how the marginal value of any EA effort has changed, and I worry about major updates being made out of a kneejerk reaction to the horribleness of someone like Trump being elected.
One potential upside might be a higher chance of a swing towards the democrats in the next few elections. US politics seems to be cyclical and if Clinton had won this time, republicans would still control the senate, the house, the majority of state senates, houses and governors.
It is likely that not much would have been achievable in terms of advancing policy and 4 years later there could be a bigger push to the right with continuing economic stagnation.
This could motivate a much larger base of people, especially young people, to get involved in local politics and up and lead to a chance of democratic control of government and the chance to move policies forward that they haven’t had since 2009 with a population that is more agreeable to socially liberal ideas. This natural motivation might mean there is less of an incentive for EAs to get involved because it will be a much less neglected area.
From a ten year point of view, this might be a better outcome than a Clinton win.
On political reform, I’m interested in EAs’ opinions on this one.
In Australia, we have compulsory voting. If you are an eligible voter and you don’t register and show up on election day, you get a fine. Some people do submit a blank ballot paper, but very few. I know this policy is relatively uncommon among western democracies, but I strongly support it. Basically it leaves the government with less places to hide.
Compulsory voting of course reduces individual freedom. But that reduction is small, and the advantages from (probably) more inclusive government policy seem well worth it. I’ve heard it said that if this policy were implemented in the US, then the democrats would win easily. I can’t vouch for the accuracy of that, but if it’s true, then in my opinion it means that the democrats should be the ones in power.
I’m not sure there’s any evidence of it having changed election outcomes, the people who are forced to vote that wouldn’t normally are divided along similar lines as those that do vote.
Also there maybe more people voting who are easier to persuade because the only reason they’re voting is the risk of a fine. I used to be quite pro this idea but now think it is neutral in outcome.
One example might be the Brexit vote which saw the highest turnout since 1992.
Another way of affecting the voting balance would be extending the right to vote to felons. I think this already has something of a campaign in the US and maybe isn’t as controversial as compulsory voting.
I was going to add that the 2000 and 2016 both had bad candidates lose the popular vote but win the general election so we should think about replacing the electoral college with a popular vote. But looking at all of American history, only four out of sixty presidential elections have had this kind of outcome. So abolishing the electoral college probably isn’t worth the cost of pushing it, even though there’s a fairly good way to get there (NPVIC).
I’ve made a strong and public decision to orient much more toward making politics less irrational. For me, this means not orienting toward party politics, but addressing the problems in political discourse and culture. It’s a bipartisan/nonpartisan stance.
Organizing in support of particular candidates is not a neglected area. I’m much more optimistic about working to fight the root causes of political polarization and bring an end to the culture war. Lending support to one side or the other risks accelerating the feedback loop that got us to this point. Lots of people are fighting for one side or the other. Very few are thinking about how to negotiate peace.
From this article.
My interpretation: If you think it’s obvious that one side is thoroughly good and the other side is thoroughly evil, don’t go in to politics. There are already lots of people like you in politics, and your marginal impact will be small. Your impact might even be negative: Since you’re convinced that the other side is evil, you’ll be willing to break the rules in order to win. The other side will use this rule-breaking as evidence that your side is evil. This is how polarization accelerates.
But if you’re someone who can get along with both sides, you should strongly consider going in to politics. Most people who are fair-minded enough to see the merits of both sides find conflict stressful, and avoid politics for that reason. But politics desperately needs people like this.
I suspect the most tractable approach is to brainstorm solutions for deeper incentive issues. From the second article:
Incentives facing media organizations are also extremely important.
I’m gonna elaborate on why I think partisanship is super key to what’s going on here.
Most of you are probably familiar with Scott Alexander’s red tribe/blue tribe distinction. The red tribe leans right politically and is oriented towards nationalism. The blue tribe leans left politically and is oriented towards internationalism.
Donald Trump made history the way he got newspapers endorsing his rivals. And yet he won the election. The opinions of people working in the media and the opinions of the public seem out of sync to a historically unprecedented degree.
A lot of the early attacks the media made on Trump were for things he said about culture war topics like immigration. Culture war issues are hyper-partisan issues marked by unusually high levels of fear, anger, and mistrust. Most politicians will apologize when the media calls them out for being racist or sexist. Trump didn’t, and this made him really unusual.
Later attacks were based on non culture war topics like whether Trump could be trusted with the nuclear codes. You’d think every American wants a president we can trust with the nuclear codes. But Trump got elected anyway. Why?
I suspect many Trump supporters thought something like this: “The media says Trump would be a terrible president because he won’t apologize for his culture war stances. But the media is not on my side in the culture war. (Democrats outnumber Republicans in the media 4 to 1.) The media also says Trump’s rash personality is a threat to the world peace. But I don’t trust the media, and it sounds like they are saying that mainly to get their way in the culture war. I know they’ve made hysterical statements to get their way in the past. Trump has demonstrated that the best way to win the culture war is to never apologize. Therefore, we shouldn’t apologize for Trump’s rash personality.”
By running culture war attacks alongside global stability attacks, the media implied that unapologetically having the wrong opinions in the culture war and threatening global stability were equivalently bad. Take the popular Trump=Hitler analogy. Hitler is someone who refused to apologize for his politically incorrect opinions. He’s also someone who had a tremendous negative impact on global stability. But I never saw anyone making this analogy clearly specify which of these issues they were more concerned about. It’s easy for me to see how a person from the red tribe could have dismissed the Trump=Hitler analogy as unjustified hysteria, based on their perception that the media exaggerates how terrible people in the red tribe are. This could have lead to them dismissing the possibility of Trump having a tremendous negative impact on global stability. The person offering the analogy is not making this distinction—you’d have to steelman their analogy in order to make it yourself.
To put it another way, the initial culture war attacks on Trump made him antifragile. People rejected these attacks as invalid, and that made it easy to reject further attacks as invalid. The media was the boy who cried wolf.
Red tribe people and blue tribe people have different values. Because the tribes collectively determine who gets the nuclear codes, we would like them to thoughtfully negotiate peaceful compromises.
In a good compromise, each party offers to give up some things it doesn’t care about much in order to get the things it cares about a lot. What we saw in this election was the opposite of a good compromise. The largely blue tribe media cared so much about their ability to extract apologies from candidates that they did not make a principled distinction between having the wrong opinions and threatening global stability. Trump’s largely red tribe support base told the media they cared so much about removing this ability, they’d elect someone who was a threat to global order to make their point. With no credible third party to referee the negotiation, Trump got elected.
I think you’ll have to do much more to establish the futility of partisanship efforts, especially if the sources for this are an assortment of op-eds and blogs. Politics is an adversarial game, as opposed to being a tough technical or economic problem, so the traditional ‘neglectedness’ criterion might not apply so well. From what I have seen, game theory models of conflict seem to predict the absence of this kind of effect.
I think there’s a pretty clear alternative here; namely, don’t break rules.
This is not clear; there are large coalitions of centrist Republicans and blue dog Dems who are around the middle of many issues. And you’ll find that many of them are grossly unpopular both with people on the opposing side and with the hard liners in their own party. Moreover, the value of getting them elected is lower than if you got a solid candidate in office.
This is a DH1 argument.
I agree with these sentences from Scott Alexander:
My view: agents in our political system are incentivized to make people think governance is an easy problem, and all the difficulties are caused by people of the opposing faction. I don’t think this is actually true.
I’m less worried about partisans going in to politics if this is their philosophy. But “don’t break rules” is not a simple principle. There are lots of legal ways to behave in an underhanded way. I expect a partisan is more likely to (a) see the behavior of their opponents as underhanded even when it isn’t (b) engage in behavior their opponents will see as underhanded without realizing it (c) rationalize underhanded behavior as justified.
In general, I am much more optimistic about trying to change the rules of the game than playing the game. If you’re playing to win, your actions are often predetermined by the game’s structure. My model: someone who works to make the game better for everyone, in a way that doesn’t strongly favor one side or the other, often finds this surprisingly tractable due to lack of opposition.
No, an ad hominem would be if I said, “Scott Alexander is a loon, and he’s saying that politics has lots of toxoplasma; therefore, politics must not have lots of toxoplasma.” But that’s not what I’m saying. I am saying “you’re backing up your argument by taking a bunch of blogs and op eds as evidence, but blogs and op eds aren’t a very good authority so you haven’t supported your claims well.”
Something more effective would be to just reference, say, the Pew study looking at American political attitudes. That would be a good data point, and it would be easier than pouring through an op ed. That way we know that you’re supporting your argument with good evidence and we can immediately see how it’s relevant to the issue at hand. And we aren’t being told to accept the elaborate narratives and points of view of bloggers and editors at face value.
In any case I would encourage you to avoid calling out fallacies as a rhetorical strategy, as explained here: https://www.reddit.com/r/philosophy/comments/1lh4ip/what_is_the_name_of_this_type_of_logical_fallacy/cbz62ls/
Well the 1% and 99% figures are clearly over-exaggerated. And I don’t think this is much of an issue at all with typical political cycles. SCOTUS rulings, for instance, get made and overturned with fairly equal difficulty (just have the president and senate appoint judges with the right views; it’s essentially symmetric). Major legislation gets repealed, but rarely; even the ACA might be preserved in part, because of the need for it and the reliance upon it.
On the flip side, making new institutions is not particularly difficult; it’s practiced all the time with passing new laws, establishing new agencies, etc., and this does not seem to be an overwhelmingly difficult practice to me. Most of American governmental history has been the successive construction of new and better government institutions against resistance, not a destructive war in which institutions are constantly created and destroyed.
I think it would be worthwhile to take a more detailed look at legislative history in order to pursue this line of thought.
Whew, I hope you don’t expect everyone to read through that when they see your comment. I see that and I hear, “If you’ve read Scott’s winding 10,000 word essay, you’re in my ingroup and you get to know what I believe and why I believe it. If you haven’t, then come back to me after you’ve finished; otherwise, you don’t know enough to refute my argument, because I’m using it as a source.”
Scott’s essay is about game theoretic problems between actors leading to inevitable conflict. I’m not sure in what sense you use the term ‘predetermined’ but it certainly isn’t the case that pushing a side in partisan politics leaves the world no different.
This, again, is something that would warrant an investigation of policymaking history, especially since you’re citing a blog by someone who, as far as I know, has not played a key role in any policy shifts, despite presenting it as their chosen strategy.
The institutions discussion seems to have gone off the rails. Our original disagreement was about whether governance constitutes a tough technical and economic problem. I think it does. You apparently think it doesn’t. I don’t think we’re going to make progress resolving this disagreement easily.
No, our original disagreement was over whether partisan political activity is better than trying to systemically reform political arenas.
No, I just think that partisan political activity is a good avenue of improving governance. I’m not making any broad claims about governance being easy or not.
Of course not, if you want to pursue a discussion about policymaking without actually referring to legislative history.
You said: “Politics is an adversarial game, as opposed to being a tough technical or economic problem”. I was responding to that.
I’m tapping out of this discussion.
A few of us have experience working in politics and could conceivably accomplish some good by being an influencer in Trump’s White House. Others of us have the ability to pitch Thiel on stuff. Since Thiel has sway in the Trump transition, this means we could conceivably get an EA or two into positions of influence in the Trump administration.
I’m not sure that it would be a good idea to actually do this, but I’m mentioning it because it doesn’t seem outside the realm of possibility to actually do it, and it could plausibly be highly effective. Here are some of the questions we’d need to answer:
If we had an EA inside the Trump administration, would it do more good than if they stayed in their current position instead? This depends partly on what they’re currently doing and partly on how likely we estimate they could actually make a difference in policy. My intuition is that if we expect Trump’s policies to be very bad, then even a small influence could translate into a large amount of good.
Who would be best suited for this, if we decided we wanted to try it? I’m not sure of what would count as experienced enough to do something like this. There are a few people at Effective Altruism Policy Analytics, and I believe there are a couple of people that have experience with lobbying in DC.
Who would make the pitch to Thiel?
What would the pitch consist of? We’d need to know exactly what parts of EA Thiel cares most about, and then we’d need to stress those aspects.
How likely would Thiel be swayed by such a pitch? If he endorsed Trump because he wanted influence in the administration, then I believe Thiel would be fully on board with this idea. But if he endorsed Trump because he actually believes Trump’s positions are good, then I can see where he would hate this idea.
Would Thiel be able to get the EA in a position high enough to actually influence policy? How high up would the position have to be in order for it to be influential? How influential are mid-level staffers?
I don’t know the answer to these questions. I don’t know if this is even a workable idea. I certainly would hate to convince an EA to drop their current work for this if it doesn’t turn out to be an influential position. But it seems possible that this could be a high-value opportunity, so I’m bringing it to everyone here.
I think this is the link for applying for a position in Trump’s administration. Trump has 4,000 positions to fill over just a few weeks, and lots of people will be reluctant to work for him. So you probably have as good a shot as anyone if you’ve been building career capital. You’ll probably have to pretend you support him—consider it a 4-year ideological turing test.
This is a good idea and it doesn’t change much from president to president. The big thing with politics like this is just getting your foot in the door.
Thiel will have the president’s ear on technical issues, he said, though he will not have a formal position and will stay in DC. It might be a good chance to get someone in the OSTP, as one opportunity.
I definitely became less interested in politics ever since identifying as an EA or utilitarian. But then Switzerland passed some ridiculous xenophobic propositions, and Brexit happened, and now Trump. And every time I had this worry in the back of my mind that we’re doing something wrong.
Carl mentioned “Misallocating a huge mass of idealists’ human capital to donation for easily measurable things and away from more effective things elsewhere, sabotages more effective do-gooding for a net worsening of the world” here. This point doesn’t just apply to money, but also very much to attention and activism. And the bias may not just be towards things that are easily measurable, but there may also be a bias away from “current” or “urgent” events. These events shape public discourse, which could have important flow through effects. What’s the effect if altruistic and driven people disproportionally stop caring about current events and the discussions that surround them?
Perhaps it’s negligible, but it’s certainly worth thinking about more. And I was glad to see how much attention the recent votes got within EA.
I was under the impression that a bias against urgent events is an inculcated response that takes active strength to develop, a resistance to the much stronger human bias to act on urgent but not-as-important things. It seems to me that most altruistic and driven people still disproportionally care about current events and the discussions that surround them, over the Important Things.
I’d be interested to hear a case for moving from animal advocacy to politics. If your comparative advantage was in animal advocacy before the election, it’s not immediately obvious to me that switching makes sense.
In the short term, animal welfare concerns dominate human concerns, and your marginal contribution to animal welfare via politics is unclear: welfare reform in the US is happening mostly through corporate reform, and it’s dubious that progressive politics is even good for wild animals due to the possible harms of environmentalism.
Looking farther into the future, it’s not clear that engaging in politics is has become more effective on the margin than spreading anti-speciesism.
Politics is still a crowded space and it’s looking like many other progressives have been galvanized by this result.
The case is for defending the conditions under which it’s even possible to have a group of privileged people sitting around worrying about animal advocacy while the world is burning. To the extent that you think 1) Trump is a threat to democratic norms (as described e.g. by Julia Galef )/ risks nuclear war etc. and isn’t just a herald of more conservative policy, and 2) most liberals galvanized by the threat of Trump are worrying more about the latter than the former, there’s room for EAs to be galvanized by the threat of Trump in a more bipartisan way, as described e.g. by Paul Christiano.
(In general, my personal position on animal advocacy is that the long-term future of animals on Earth is determined almost entirely by how much humans have their shit together in the long run, and that I find it very difficult to justify working directly to save animals now relative to working to help humans get their shit more together.)
Trump may represent an increased threat to democratic norms and x-risk, but that doesn’t mean the marginal value of working in those areas has changed. Perhaps it has. We’d need to see concrete examples of how EAs who previously had a comparative advantage in helping animals now can do better by working on these other things.
This may be true of massive systemic changes for animals like the abolition of factory farming or large-scale humanitarian intervention in nature. But the past few years have shown that we can reduce a lot of suffering through corporate reform. Animal product alternatives are also very promising.
Also, “having our shit together in the long run” surely includes anti-speciesism (or at least much higher moral consideration for animals). Since EAs are some of the only people strategically working to spread anti-speciesism, it seems that this remains highly valuable on the margin.
Edited to add: It’s possible that helping animals has become more valuable on the margin, as many people (EA and otherwise) may think similarly to you and divert resources to politics. Many animal advocates still think humans come first. Just a speculation.
I’d like to see an analysis of exactly what the opportunity costs are there, before endorsing one. This analysis has no differential analysis, and as such it reads “There are many important things being neglected. This is an important thing. Therefore it is the most important thing to do.”
I never meant to say that spreading anti-speciesism is the most important thing, just that it’s still very important and it’s not obvious that its relative value has changed with the election.
Here’s a list of some updates I made, which were influenced by both the presidential and Congressional elections:
P(Anthropogenic global catastrophe): Up
P(Extinction before AI): Up
P(World government): Down
If AI is created:
P(AI created in next 20 years): Up
P(AI created by a military): Up
P(AI arms race): Up
P(Unfriendly AI): Up
In the near-term:
P(Poultry or fish included in Humane Slaughter Act): Down
P(Humane Slaughter Act effectively enforced): Down
P(Rising global productivity growth): Down
P(Higher US income inequality): Up
So...not great for most value systems.
Can I ask your rationale for P(AI created in next 20 years): Up?
Also note that the longer we delay the arrival of general AI, the more hardware and data will be available for it to immediately capitalize upon. So it’s not entirely clear that delaying AI development is good.
Increased US military spending will result in more resources attempting to create AI. In a climate of lower international cooperation, other countries will hasten their efforts as well.
Agreed. Many of my updates can be viewed as good, depending on your value system and other beliefs. However, forcing myself to guess, I think we have moved further away from the ideal, cooperative future—e.g. something inspired by CEV.
Well, the military itself spends very little on AI and none on cutting edge AI. What really matters is DARPA, IARPA, and whatever else might be going on in hidden parts of the intel community and DoD.
The bulk of AI development is the tech industry and academia. I think these areas are likely to be hurt by the administration.
You’re right, I meant to say “Increased US defense spending” not “Increased US military spending”. I inaccurately use “military” as a synonym of “defense” sometimes.
From Trump’s website: “Emphasize cyber warfare...and create a state-of-the-art cyber defense and offense.”
Maybe, but I’m not so sure the effect will be large enough.
Well cyber has little to do with AI; DARPA and IARPA cyber and AI teams are quite separate, and the service branches don’t do any AI with their cyber teams. The only exception is with the DARPA automated cyber grand challenge this year, but even it wasn’t really doing things that would accelerate AI development.
My point is more that since AI is isolated behind so many layers of budget allocation and is such a tiny slice, that changes at the top level are likely to not change anything.
Have you considered how Peter Thiel, big-time AI safety supporter, is to become one of Trump’s key policy advisers on technology, and will have an unprecedented opportunity to shape the policy on AI going forward?
Friends: I recently wrote a few thousand words on the implications that a Trump presidency will have for global risk. I’m fairly new to this discussion group, so I hope posting the link doesn’t contravene any community norms. Really, I would eagerly welcome feedback on this. My prognosis is not good.
https://medium.com/@philosophytorres/what-a-trump-presidency-means-for-human-survival-one-experts-take-ed26bf9f9a21
For many Americans, income taxes might go down; probably worth thinking about what to do with that “extra” money.
Given Trump’s high uncertainty and variance, one might imagine that the chances of him doing something that increases x-risks significantly is at least one or two orders of magnitude higher than Clinton’s.
Given that, restructing the US economy to better deal with the shocks of automation and globalization, which might prevent future instances of this type of intentional systemic upheaval, suddenty seems like a potentially important new category of x-risk by proxy.
By that logic, does something like UBI become an important potential area of focus?
More like x-risk by proxy by proxy.
It should be easier to focus directly on building international institutions.
| More like x-risk by proxy by proxy.
Fair enough.
| It should be easier to focus directly on building international institutions.
Can you clarify this? As a new EA, I am unfamiliar with what this might be in reference to. Are you arguing for focusing on the creation of a second nation-state promoting liberal values and technological advancement (to backstop US leadership)? Or just circumventing politics and going straight at the underlying risks (AI, nuclear proliferation)? It seems without the air-cover of a liberal superpower, pursuing such ventures will be significantly harder.
Well for Trump increasing x risks, the potential downside is that there is going to be less international cooperation and less prospect for coordination between nations on issues like AI. That’s what I thought you had in mind and it implies that we should try to build more credible and effective groups like the UN and IAEA. It’s easier to work on this directly rather than doing things that make people less likely to support populists who are against international involvement. I guess there are other ways Ttump could be bad for x risks, so that might change things.
Gotcha. I was going the other way; considering what automation and globalization have done to low-skilled workers over the last 30 years (which is only going to get a lot worse), some sort of significant restructuring of our economic system (both wealth distribution and cultural self-actualization via employment) seems necessary to prevent full-scale upheaval (e.g. revolution, civil war, etc). Trump seems to be the first major indication that this problem is getting so bad that it could literally end America or at least American leadership and the promotion of liberal values.
As Qloachu said in a different thread on this page:
I do agree this is a very indirect effect but I also failed to see how more direct action would be more effective.
Well that is different but I don’t think it’s a likely threat.
Low-skilled workers have seen general increases in standard of living over the last 30 years and no great increase in unemployment. Globalization generally increases wealth for all income brackets in the long run and automation has not historically increased long-term unemployment.
We may be able to have trucks that steer themselves on the road, but actually having unmanned vehicles take themselves from A to B reliably enough to be unmanned and without requiring a human in the cabin to do anything is years away.
In any case, the lack of truck drivers is a significant constraint on our economy alongside the fact that they can only drive 10 or so hours a day. Automation here might be unusually good.
It seems to me like wealth is a pretty straightforward question of welfare to resolve. As for cultural self-actualization… we’ve had retirees, part-time employees, and permanently unemployed people for a long time, but they never seem to be more troublesome or more dissatisfied than the rest of the electorate.
I don’t think so. We’ve had similar issues in the past with populist leaders and their movements, e.g. William Jennings Bryan (barely failed to get elected) and Huey Humphrey (assassinated). But our democracy has only improved over time.
Is this true in the US? My personal circle of concern certainly includes all of humanity (and beyond), but Trump’s election would appear to be an indication that displaced Americans don’t think that way.
Do you have some info you can share on that? Everyone I’ve talked to in the autonomous vehicle area has indicated that adoption, not technology, will be the long pole here. And as you indicated, there are massive economic forces that will push for adoption.
And while I totally agree automation will be good for GDP, it will continue to displace a set of people who can vote.
Retirees already have cultural approval to not work: they’ve “earned” it. Part-timers and permanently unemployed are viewed as lazy or moochers by many folks in the US, where there’s a culture of self-improvement via working hard. That’s what I’m arguing will need to change.
What metric are you using for improvement? On the axis of existential risk prevention, I would argue we’ve just taken a big step back.
Interested to hear your thoughts. Thanks!
Yes, average income when adjusted by PPP has increased (the standard narrative is that CPI-adjusted income has not increased, but the CPI is flawed—it overestimates inflation by focusing on cities and it ignores goods substitution), and nonwage compensation has increased (health benefits etc).
Sure, class 8 trucks are a couple hundred thousand dollars each and the cargo is worth a lot too. They have mechanical problems. They have expensive insurance. They have complex routes. They have to reliably reach their destination and meet with the proper people to load and unload. With all this responsibility and risk it’s an easy decision to put a driver at the wheel, especially when even ordinary cars aren’t ready to go around without drivers.
Well there you have it. If it will take a long time to be adopted then there won’t be an unemployment problem for a long time.
They will vote for welfare, which they will get.
I don’t think any of this takes away from my point.
None of the democracy and freedom indices have been around for that long, so there is no metric. But we can look at the corruption and civil values of 100 years ago, and comparing it today, and note that it’s pretty clear that things are better all round now than they were back then. For instance, we don’t prevent women and blacks from voting anymore, and we don’t put people in prison for speaking out against the war. We don’t hold racist views about who is eligible to hold public office.
That’s a different axis, and it has taken many steps back, perhaps with the elections of GWB, Ronald Reagan, Eisenhower, and Truman.
The distribution is what would matter for my argument. Do you have any links to data there? Where is this data from? Thanks.
I have first hand information from those knowledgable on the topic that the technology will be ready in the next few years, and I haven’t heard a refutation of that. So I said the long pole was adoption. I didn’t say how long the pole was. You already acknowledged that the economic incentives for adoption are huge. I posit that adoption will be faster than you suspect. But we have to wait and see on this one, unless you have more information to share on why you disagree.
They didn’t vote for welfare this time. Which is my point.
Maybe you can explain… you seemed to be saying that the fact we have part-timers and unemployed proves that we don’t have any cultural stigma against not working. I said that people do have such a stigma, and then you said that doesn’t take away from your point. Not sure what I missed.
Great, that’s the axis I was trying to focus on, which is why I asked what metric you were using. And yes, it’s taken steps back before. That’s doesn’t preclude it getting worse.
I’m simply pointing out that if automation and globalization disrupt the US middle class enough, political instability could increase x-risks.
Haven’t seen that data specifically but it’s inferred:
http://www.quickmba.com/econ/macro/cpi/
https://youtu.be/8sf3kt1KduY
https://fred.stlouisfed.org/series/MEFAINUSA672N (I guess family income increased even with cpi)
You’re talking about what would happen in a world of high structural unemployment. But we don’t have high structural unemployment, so you can’t say that future elections will have the same dynamics. In this election cycle, Trump voters weren’t primarily motivated by the economy, they were motivated by immigration and culture issues.
I said that it shows that unemployed people aren’t likely to start a revolution or anything just because they’re unemployed.
People don’t start upheavals just because they don’t have cultural acceptance.
If you take a broad view then half of all things will increase x risk in the long run and half of them will decrease it. I think that the issue here is too crowded and intractable and the effects are too complicated and uncertain for it to be a big goal for us to put effort on, though I agree that it would be good if the nation had a robust social system going forward.
I think we need to see more data on this one. I’d put forth the idea that in fact the underlying issues were in fact economic, and that the cultural and immigration issues were mostly just utilized by Trump as things to blame for the underlying economic woes.
Here’s an article I found instructive on the perspective and here’s a relevant passage:
I thought this analysis was interesting.
Ah, I understand your point now. Sure, that’s true, but I’m imagining what happens if the lower-middle class gets squeezed out, as mentioned above. I need to go learn more about economic trends of the middle class in American history before I can make this case, though.
Wait, I’m new to x-risk analysis and TDT… but is that true? Why would the distribution of the space of all possible actions split 50⁄50 towards increasing or decreasing x-risks? That doesn’t seem right. Am I missing something?
@rowborg/@kbog, this is a great thread, and I’m only going to dip my toe in, here, minimally, to +1 this point that rowborg has made. Employment isn’t just about exchanging labour for a wage, it’s a major source of meaning in people’s lives. The assertion that standards of living in the US have risen is probably correct, even for people at the bottom of the income distribution. But this standard of living increase can’t explain the incredibly sour mood of a huge chunk of people. Obviously it’s going to take a lot of time for researchers to unpack that, but the meaningless-employment story seems like a pretty plausible component: if people feel what they do doesn’t have any meaning, and particularly if they feel their children don’t have prospects for meaning-ful employment then despair and anger seem like totally to-be-expected reactions.
I saw it based on data from exit polls. I don’t know where I saw it. If there is any other evidence, sure that would be neat to see. Though it is objectively the case that our underlying economic woes are smaller now than they were in 2012 and 2008.
Well think about it this way, if the majority of political positions increased x-risk, we would make the world safer just by doing things in the opposite way all the time. But that’s absurd.
The more general point is that you can construct a complicated chain of reasoning for lots of things to increase x risk, so it’s not particularly interested just that something increases it. What matters is how much and how tractable it is.
On aggregate, yes. I’m talking about the economic situation for a certain group, which is the lower-middle class. I don’t know that this is true for them. I’ll go try and dive in on this.
If I’ve parsed your meaning correctly, you are saying every action has a counterfactual non-action, and if you include not just the space of possible actions but the impact of non-action, then risks break 50/50? That makes more sense, but doesn’t really seem to be helpful in deciding between actions to take. The space of possible actions to take is what should concern us, yes?
Regardless, I take your larger point, which is that we should put efforts where you can maximize impact and you are skeptical on the effective value of this one as opposed to others. I would certainly need to be more quantitative in my analysis of this and back up my ideas/claims for sure.
Totally agreed. I have yet to hear an argument that 1) invalidates my chain of reasoning or 2) indicates there are other inflection points along the chain that are better. Whether or not this compares well with other options for x-risk reduction is an exercise I have yet to perform (as I mentioned above).
Thanks for this!
If we pull the camera back far enough, my guess is in a generation or two America will be on a good track again, so long as Trump doesn’t start a war or use our nuclear warheads. As the White House Press Secretary said, the institutions of the U.S. have survived a civil war, 2 world wars, and the Great Depression. This will be a bad 4 years with adverse consequences on the rest of the world. Putin will be on the offensive, both in reality and in cyberspace. And U.S. carbon emissions will increase. But I still believe AI risk is the most dangerous threat to humanity.
I’m with you, and I really, really want to be on the side of skepticism and saying that this election is no different from normal cycles we’ve had in the past. But this is the first time ever that we’ve given serious support to someone with a history of neither political office nor military leadership. And someone who really clearly has no consistent train of thought or policy acumen. If the economy happens to be strong and the country is successful, then people will appreciate his persona and values as the basis for a very entrenched brand of populist conservatism. So in this case I think it’s okay to say that the long term future of America could be seriously altered from what we consider to be normal, and it’s okay to say that this election was unusually consequential.
EAs like to focus on the long term and embrace probabilistic achievements. What about pursuing policy reforms that are currently inconsequential, but might have profound effects in some future state of the world? That sort of reform will probably face little resistance from established political players.
I can give an example of something I briefly tried when I was working in Lesotho, a small, poor African country. One of the problems in poor countries is called the “resource curse”. This is the counter-intuitive observation that the discovery of valuable natural resources (think oil) often leads to worse economic outcomes. There are a variety of reasons, but one is that abundant natural resources often cause countries with already-weak institutions to become even more corrupt, as powerful people scramble to get control of the resource wealth, methodically destroying checks and balances as they go.
In Lesotho, non-renewable natural resources—diamonds—currently account for only a small portion of Lesotho’s GDP (around 10%). I introduced the idea of earmarking such natural resource revenues received by the government as “special”, to be used only for infrastructure, education etc projects, instead of effectively just being consumed (for more info on this idea see this article or google “adjusted net savings”). Although this change would not have huge consequences right now, I thought that it might if there were a massive natural resource discovery in Lesotho in the future. Specifically, Lesotho might be able to avoid some of the additional corruption by already having a structure set up to protect the resource revenues from being squandered.
The idea I’m putting forward for a potential EA policy initiative is to pursue a variety of policy changes that seem painless, even inconsequential, to policymakers now, but have a small chance of a big impact in some hypothetical future. The idea is to get the right reforms passed before they become politically contentious. While it can be hard to get policymakers to pay attention to issues seen as small, there are plenty of examples of political capture that could have been mitigated by early action. And this kind of initiative is probably relatively neglected given humanity’s generally short-term focus. I think EAs are uniquely well placed to prioritize it.
Check this out: http://effective-altruism.com/ea/147/cause_better_political_systems_and_policy_making/
Just a few thoughts.
Firstly, Trump’s agricultural advisors seem to be very hostile to animal welfare. This may mean that we need more people working on farmed animal welfare, not less.
In terms of going into politics, the prospect of having a group of EAs, and perhaps even an EA-associated organization, doing regular, everyday politics may turn some people off from the movement (depending on your view on whether EA is net-positive or negative overall, this may be bad or good.)
While Sentience Politics, the Open Philanthropy Project and some others I may have missed do take part in political activities, they focus on specific policies, and I suspect that what some people are talking about would involve a systematic attempt to engage in party-politics.
I think that even without Trump, the idea of having a very small number of individual EAs (maybe 1/1000 EAs) going inside politics and trying to influence administrations or even become politicians was a good one.
But, a systematic attempt to engage in party-politics would not be a good idea, partly because, even in the EA community, focusing on party-politics or even on controversial policies seems to lead to less willingness to consider other points of view.
And, partly because influencing administrations or becoming a politician on one’s own is more likely to make a difference than engaging in regular party-political campaigning, even though becoming a politician or influencing an administration is less easy to do.
Finally, I think that politics is very important, because you could potentially reduce existential risks as well as spread good values and ensure that humanity is on the right course in the future, and therefore there’s not a tension between reducing existential risks and values-spreading.
However, in order for any politicians or political advisors to be able to steer humanity in a positive direction, you need public and corporate support for it, which is why I believe that spreading anti-speciesism, working on farmed animal suffering, and so on, remains highly important too.
Overall, Trump’s election has not influenced my beliefs significantly.
This isn’t quite relevant, but I think the election cycle was sort of mishandled by the EA community.
It was mostly sort of ignored, except for some last minute stuff with people trying to put together efforts to fight Trump. Some people jumped quickly on bandwagons and picked a side. Other people just took a blank “politics is the mindkiller!!” stance and refused to care.
Ideally, we should have a rational and serious approach from the outset, looking at the candidates and seeing if there is anything that EA philosophy decisively says that we should do. And when the window opens, then we come together and say “hey, here’s our analysis, here’s our thoughts, this is what effective altruism says about the election.” And push that message as far as we can.
I don’t think a rationalist message/meme would have been successful at convincing hundreds of thousands of working class whites not to vote for Trump. Rationalism has it’s place in deciding what to do about an election, but I don’t think EA messaging is at all useful in influencing a mass audience.
We didn’t need to convince hundreds of thousands of working class whites. Just increase turnout among Berners or change the primary outcomes or show Thiel the error of his ways. There’s lots of avenues to do things.
Thiel had essentially nothing to do with the outcome of this election.
This was not primarily a turnout issue. Black turnout was down, but Hispanic turnout was up. White turnout appears relatively flat (both Democratic and Republican white turnout), but we’ll know more when actual person level vote history is released. Regardless, EA messaging is not the right way to appeal to Berners.
The easiest way to shift the outcome of the election would have been to change public opinion by a point or two by shifting the narrative of the race in the final week. Comey was successful at doing this.
Well he gave about 0.15% of their money, and campaign funding is fairly important. The important thing is that in future campaigns, funding may influence the outcome of elections.
It was all about turnout, fewer people voted Democrat than in 2012 and 2008. Trump won just by holding onto similar GOP vote totals from previous elections while people dropped Clinton.
I don’t know what the right type of messaging is and I don’t know who we will need to appeal to in 4 or 8 years. My point is that there are probably bigger things to be done besides phone banking.
That’s a good point but I think it was an exceptional case as the email story had been unfolding for many months and people were already primed to watch for news about it. It’s also very hard to actually do it intentionally if you don’t have some big news you can release.
At most, campaign funds would have moved this a point or two. Campaign funding has little impact on presidential elections; Clinton far outspent Trump and Trump was far outspent in the primary election. If we assume an effective size of 5% for all of Trump’s money and assume no diminishing marginal return (both very generous assumptions), that 0.15% is 0.0075 percentage points in movement. The outcome was decided by 1, so that’s over two orders of magnitude lower than what was needed under generous assumptions. It was probably more orders of magnitude lower.
That’s not true at all.Trump gained substantially in rural areas with mostly white people where Obama had won or performed substantially better. I mean, if turnout had magically been higher among Democrats but not Republicans, we would have won, but you don’t get to do that. The composition of the electorate was roughly the same (minus some black people plus some Hispanics) as 2012. It’s conceivably possible that without the drop in black turnout, we would have won, but this was inevitable without the first black president running. There is overwhelming evidence that attitudes among the white working class moved against us. Hence our drop in the midwest.
I agree on the point that phone banking does not make much of a difference.
There were several instances that fall under the same pattern: the email story, the hollywood access tapes, the debates, probably the apprentice tapes if they had appeared, and potentially the wikileak emails, though it’s much harder to gauge their effect size.
It is true. Romney got 61 million votes and McCain got 60 million. Obama got 69 million and 66 million in 2008 and 2012 respectively. This year, Trump got 60 million votes and Hillary got 61 million.
Well, depending on how early before the election you want to consider. The debates for instance were all more than a week before the election. Again, it’s basically impossible to put effort into making things like this happen, and the best way to do so might simply be conventional ways of building political clout and awareness.
You can’t look at aggregate turnout numbers being different and assume the composition of turnout was different. You’re making the assumption that there was 0 movement from Obama to Trump or from Romney to Clinton; both of which are definitely incorrect as evidenced by polling.
Secondly, turnout is much higher than that appears; much more will come in from California, Washington, Oregon and Colorado. It always takes these states forever to report. So the turnout numbers now are misleading.
I think the message of SlateStarCodex’s “Tuesday Shouldn’t Change The Narrative” is particularly relevant to EAs—any large updates to one’s beliefs about the world should have come before the election.
Agreed that large updates about things like the prevalence of regressive attitudes and the fragility of democracy should have been made before the election. But Trump’s election itself has changed many EA-relevant parameters—international cooperation, x-risk, probability of animal welfare legislation, environmental policy, etc. So there may be room for substantial updates on the fact that Trump and a Republican Congress will be governing.
That said, it’s not immediately obvious to me how the marginal value of any EA effort has changed, and I worry about major updates being made out of a kneejerk reaction to the horribleness of someone like Trump being elected.
Lots of distraction.
Has there been consideration of electoral reform with mind to proportionality as a worthwhile EA cause?
Thanks Jay! I’m definitely interested in hearing more thoughts about this. I’m putting some thought into it as well.
One potential upside might be a higher chance of a swing towards the democrats in the next few elections. US politics seems to be cyclical and if Clinton had won this time, republicans would still control the senate, the house, the majority of state senates, houses and governors.
It is likely that not much would have been achievable in terms of advancing policy and 4 years later there could be a bigger push to the right with continuing economic stagnation.
This could motivate a much larger base of people, especially young people, to get involved in local politics and up and lead to a chance of democratic control of government and the chance to move policies forward that they haven’t had since 2009 with a population that is more agreeable to socially liberal ideas. This natural motivation might mean there is less of an incentive for EAs to get involved because it will be a much less neglected area.
From a ten year point of view, this might be a better outcome than a Clinton win.
https://en.wikipedia.org/wiki/Political_party_strength_in_U.S._states
On political reform, I’m interested in EAs’ opinions on this one.
In Australia, we have compulsory voting. If you are an eligible voter and you don’t register and show up on election day, you get a fine. Some people do submit a blank ballot paper, but very few. I know this policy is relatively uncommon among western democracies, but I strongly support it. Basically it leaves the government with less places to hide.
Compulsory voting of course reduces individual freedom. But that reduction is small, and the advantages from (probably) more inclusive government policy seem well worth it. I’ve heard it said that if this policy were implemented in the US, then the democrats would win easily. I can’t vouch for the accuracy of that, but if it’s true, then in my opinion it means that the democrats should be the ones in power.
I’m not sure there’s any evidence of it having changed election outcomes, the people who are forced to vote that wouldn’t normally are divided along similar lines as those that do vote.
Also there maybe more people voting who are easier to persuade because the only reason they’re voting is the risk of a fine. I used to be quite pro this idea but now think it is neutral in outcome.
One example might be the Brexit vote which saw the highest turnout since 1992.
Another way of affecting the voting balance would be extending the right to vote to felons. I think this already has something of a campaign in the US and maybe isn’t as controversial as compulsory voting.
I was going to add that the 2000 and 2016 both had bad candidates lose the popular vote but win the general election so we should think about replacing the electoral college with a popular vote. But looking at all of American history, only four out of sixty presidential elections have had this kind of outcome. So abolishing the electoral college probably isn’t worth the cost of pushing it, even though there’s a fairly good way to get there (NPVIC).
I’ve made a strong and public decision to orient much more toward making politics less irrational. For me, this means not orienting toward party politics, but addressing the problems in political discourse and culture. It’s a bipartisan/nonpartisan stance.