Too Much Dark Money In Almonds

Everyone always talks about how much money there is in politics. This is the wrong framing. The right framing is Ansolabehere et al’s: why is there so little money in politics? But Ansolabehere focuses on elections, and the mystery is wider than that.

Sure, during the 2018 election, candidates, parties, PACs, and outsiders combined spent about $5 billion – $2.5 billion on Democrats, $2 billion on Republicans, and $0.5 billion on third parties. And although that sounds like a lot of money to you or me, on the national scale, it’s puny. The US almond industry earns $12 billion per year. Americans spent about 2.5x as much on almonds as on candidates last year.

But also, what about lobbying? Open Secrets reports $3.5 billion in lobbying spending in 2018. Again, sounds like a lot. But when we add $3.5 billion in lobbying to the $5 billion in election spending, we only get $8.5 billion – still less than almonds.

What about think tanks? Based on numbers discussed in this post, I estimate that the budget for all US think tanks, liberal and conservative combined, is probably around $500 million per year. Again, an amount of money that I wish I had. But add it to the total, and we’re only at $9 billion. Still less than almonds!

What about political activist organizations? The National Rifle Association, the two-ton gorilla of advocacy groups, has a yearly budget of $400 million. The ACLU is a little smaller, at $234 million. AIPAC is $80 million. The NAACP is $24 million. None of them are anywhere close to the first-person shooter video game “Overwatch”, which made $1 billion last year. And when we add them all to the total, we’re still less than almonds.

Add up all US spending on candidates, PACs, lobbying, think tanks, and advocacy organizations – liberal and conservative combined – and we’re still $2 billion short of what we spend on almonds each year. In fact, we’re still less than Elon Musk’s personal fortune; Musk could personally fund the entire US political ecosystem on both sides for a whole two-year election cycle.

But let’s go further.

According to this article, sold for less than $5 million. Mashable sold for less than $50 million. The whole Gawker network (plus some other stuff including the Onion) sold for $50 million. There are some hints that Vox is worth a high-eight-digit to low-nine-digit amount of money. The Washington Post was sold for $250 million in 2013 (though it’s probably worth more now). These properties seem to be priced entirely as cash cows – based on their ability to make money through subscriptions or ads. The extra value of using them for political influence seems to be priced around zero, and this price seems to be correct based on how little money is spent on political causes.

Or: Jacobin spends a lot of time advocating socialism. The Economist spends a lot of time advocating liberalism. First Things spends a lot of time advocating conservatism. They all have one thing in common: paywalls. How could this be efficient? There are millions of people who follow all of these philosophies and really want to spread them. And there are other people who have dedicated their lives to producing great stories and essays advocating and explaining these philosophies – but people have to pay $29.99 for a subscription to read their work? Why do ideologies make people pay to read their propaganda?

Maybe the most extreme example here is, which recently sold for $3 million, ie the cost of a medium-sized house in San Francisco. Tumblr has 400 million monthly visitors, and at least tens of millions of active users. These people talk politics all the time, usually of a far-left variety. Nobody thinks that one of the central political discussion platforms of the far-left is worth more than $3 million? Nobody on the right wants to shut it down? Nobody on the left wants to prevent that from happening? Nobody with a weird idiosyncratic agenda thinks being able to promote, censor, or advertise different topics on a site with tens of millions of politically engaged people is at all interesting?

(in case you’re keeping track: all donations to all candidates, all lobbying, all think tanks, all advocacy organizations, the Washington Post, Vox, Mic, Mashable, Gawker, and Tumblr, combined, are still worth a little bit less than the almond industry. And Musk could buy them all.)

The low level of money in politics should be really surprising for three reasons.

First, we should expect ordinary people to donate more to politics. A lot of the ordinary people I know care a lot about politics. In many of the events they care about most, like the presidential primaries, small donations matter a lot – just witness Tom Steyer begging for small donations despite being a billionaire. If every American donated $25 to some candidate they supported, election spending would surpass the almond industry. But this isn’t even close to happening. Bernie Sanders is rightly famous for getting unusually many small donations from ordinary people. It’s not clear exactly how much he’s received, but it looks like about $50 million total. This sounds like a lot of money, but if you use polls to estimate how many supporters he has, it looks like each supporter has on average given him $2. This is a nice token gesture, but surely less than these people’s yearly almond budget.

Second, we should expect the rich to donate more to politics. Many politicians want to tax billionaires; billionaires presumably want to prevent that from happening. Or wealthy people might just have honestly-held political opinions of their own. As rich as Elon Musk is, he’s only one of five hundred billionaires, and some of the others are even richer. So how come the amount of money in politics is so much less than many individual billionaires’ personal fortunes?

Third, we should expect big corporations to donate more to politics. Post Citizens United, corporations can supposedly put as much money into politics as they want. And they should want a lot. The government regulates corporations, so having friendly politicians in power can mean life or death for entire industries. Suppose hostile government regulation could decrease Exxon Mobil’s revenues 5% – you would think Exxon Mobil would be willing to spend 4% of its revenue to prevent this. But Exxon makes $280 billion per year. 4% of its revenue would already be larger than the whole US political ecosystem! In fact, according to Exxon’s own records, they only spend about $1 million per cycle. While they’re probably hiding something, they couldn’t hide donations the size of the whole rest of the political ecosystem, so it’s still pretty mysterious.

I think there are individual factors affecting all of these. As mentioned before, elections have spending limits (however inconsistently enforced) and may not be tractable to money. Think tanks may be more talent-limited than funding-limited. Media properties may be limited by the opinions of their journalists and subscribers (the Washington Post couldn’t pivot to being a conservative outlet without getting completely different employees and customers). Tumblr has already proven unable to censor its users without sparking a mass exodus. These issues are probably responsible for part of the underpricing. But it still seems surprising.

In his paper on elections, Ansolabehere focuses on the corporate perspective. He argues that money neither makes a candidate much more likely to win, nor buys much influence with a candidate who does win. Corporations know this, which is why they don’t bother spending more. Most research (plus the 2016 results) confirms that money has little effect on victory, so maybe this is true. But it would also have to be true that lobbying, the NRA, the media, etc don’t affect politics very much, which seems like a harder sell.

That leaves the Bernie Sanders supporters. Even if money doesn’t affect politics, Sanders supporters seem like about the least likely people to believe that. I think here we have to go back to the same explanation I give in Does Class Warfare Have A Free Rider Problem? People just can’t coordinate. If everyone who cared about homelessness donated $100 to the problem, homelessness would be solved. Nobody does this, because they know that nobody else is going to do it, and their $100 is just going to feel like a tiny drop in the ocean that doesn’t change anything. People know that a single person can’t make a difference, so they don’t want to spend any money, so no money gets spent. This is true for ordinary people, but it’s also true for billionaires and greedy corporations. No single greedy corporation wants to pony up the money to change the laws to favor greedy corporations all on its own, while its competitors lie back and free-ride on its hard work. So they basically donate token amounts and do nothing. By all accounts the Koch brothers actually believed in everything they were doing, and they had to, because you couldn’t make billionaires spend Koch-brothers-like levels of time and money out of self-interest.

In this model, the difference between politics and almonds is that if you spend $2 on almonds, you get $2 worth of almonds. In politics, if you spend $2 on Bernie Sanders, you get nothing, unless millions of other people also spend their $2 on him. People are great at spending money on direct consumption goods, and terrible at spending money on coordination problems.

I don’t want more money in politics. But the same factors that keep money out of politics keep it out of charity too.

The politics case is interesting because it’s so obvious. Nobody’s going to cynically declare “Oh, people don’t really care who wins the election, they just pretend to.” It’s coordination problems! It has to be!

So when I hear stories like that Americans could end homelessness by redirecting the money they spend on Christmas decorations, I don’t think that’s because they’re evil or hypocritical or don’t really care about the issue. I think they would if they could but the coordination problem gets in the way.

This is one reason I’m so gung ho about people pledging to donate 10% of their income to charity. It mows through these kinds of problems. I may not be a great person. But I spend more each year on the things I consider most important than I do on almonds, and this is the kind of thing that doesn’t happen naturally. It’s the kind of thing where I have to force myself to ignore the feeling of “just a drop in the ocean”, ignore whether I feel like other people are free-riding on me, and just do it. Pledging to donate money (and then figuring out what to do with it later) ensures I will take that effort, and not end up with revealed preferences that seem ridiculous in light of my values.

This entry was posted in Uncategorized and tagged , . Bookmark the permalink.

263 Responses to Too Much Dark Money In Almonds

  1. “What surprises me is how little is spent on political campaigns, considering the stakes.” (Price Theory, Chapter 19, 1986)

    • Frederic Mari says:

      Yep. I’ve always said that, for people on the take, US politicians were dirt cheap…

      • onyomi says:

        Couldn’t a surprisingly small amount of money in politics equally be interpreted as evidence that politicians are too expensive to be worth buying and/or that what they can reasonably offer within the bounds of realpolitik isn’t worth buying?

        • Frederic Mari says:

          Both you and Garrett makes good points but I don’t really agree.

          See below my other response to see my own thesis. But, as an answer to donations supposed uselessness, can I introduce you to the latest book on two brothers I really despise?

          In the late 80s, Bush the 1st was for doing something about global warming. The Koch brothers swung into action and made sure Congress Republicans would forever act as a bottleneck/veto point on anything meaningful.

          Seems like their political spending was pretty effective…

          • salvorhardin says:

            There were lots of other forces pushing against climate change action at that time, including many that were not just political spending. How would we disaggregate the effect of Koch brothers spending from those?

          • sourcreamus says:

            Given that very few countries in the world have done anything substantive about global warming it seems unlikely that the Koch brothers money has made a huge difference in this country.
            Much more likely that fixing global warming will be very expensive for little immediate gain, and that is a hard sell to the voters.

          • Loris says:

            Sourcreamus, that’s not particularly good evidence.

            Suppose you lived in a village where people would graze their livestock on the commons. It becomes pretty clear that the commons is being overgrazed, and will be degraded (over a few years) if nothing is done. A village meeting is held, at which the head of the largest family – which has more than double the number of cattle per person than the average – makes it clear that they’re going to carry on regardless.
            Would you really expect the other families to do much substantive?

          • JayT says:

            The EU is of a comparable size/value to the US though, so it’s not like the biggest family overshadowed all the other families.

          • Loris says:

            The EU isn’t a country, it’s lots of countries. Nevertheless, it was represented at the appropriate G7 conference (although not actually one of the eponymous 7), and I don’t want to argue it, because it’s not particularly important for the argument.
            The analogy description was optimised for conciseness and instructive power rather than for an exact mapping. The take-home message is that if any of the most significant large contributors won’t take part in managing a shared resource, the other players can’t solve the problem. Anything they do just means more short-term gains for the defector; effectively it’s just a subsidy for them.

        • John Schilling says:

          and/or that what they can reasonably offer within the bounds of realpolitik isn’t worth buying?

          It seems quite plausible that, once someone has been elected as Representative X (Party Y) for District Z, the scope of action consistent with maintaining that position is too narrow to be worth a whole lot. If the issue is controversial, e.g. cap-and-trade carbon restrictions, you can’t pay the Congressman from North Dakota enough to vote for it, or the one from Portland to vote against it. If it is uncontroversial, e.g. occupational licensing for truck drivers, then either zero or one group will have the institutional focus to offer a bribe, and if it’s one then there’s no competition so it doesn’t need to be a very big bribe.

          Meanwhile, most of the real work in lobbying is done by getting Congressional staffers on your side. And they can generally be had for the cost of a three-star meal plus hiring an expert to do their homework for them, which multiplied by thousands of staffers and thousands of issues adds up to almond money but not terabuck money.

      • Garrett says:

        Part of the issue, I suspect, is that bribing politicians directly is generally illegal. That is, if someone gives a million dollars directly to a politician in exchange for a particular cause, they’re both going to prison.

        So the money needs to be spent indirectly (as campaign contributions, political ads, think tanks, or whatever) where the only benefit to the politician is that they have a slightly better chance of being reelected. But “keeping a job that pays less than they could make in the private sector” isn’t quite the greatest selling point. And I doubt that scales linearly, either.

        • Richard says:

          The interesting thing about that story is that it cuts in the opposite direction from the normal story about Congressional pay. We generally tell a story in which Congress’s low wages make them more susceptible to corruption, because they need the marginal dollar more. This story implies the opposite, that the low wages reduce the incentive for corruption because keeping their current job isn’t that valuable, especially from the self-interested lens that presumably motivates the most corrupt actors.

          • Anthony says:

            The big corruption happens after the politician leaves elected office when they get jobs (and book deals) where they get millions of dollars to lobby their colleagues or lend their endorsement to something.

            The Clintons went from being broke in 2001 to being hectamillionaires in 2016. I’m not sure what all they did for that money, but I’m sure that they were paid that much for it because Bill had been President and Hillary was a potential (and actual) political power player, and not because of the intrinsic value of those services.

    • nagydani says:

      There is, I believe, another aspect of politics that also keeps money from being poured into it (most of the time): its game theory is very similar to that of the dollar auction. In a simplified model, spending on opposing political objectives are mutually annihilating efforts (i.e. those who spend more win, irrespecitve of the absolute amount) and the losers won’t get their spending back.

      Dollar auctions have two stable outcomes: either all parties spend very little or a runaway contest when the two biggest spenders spend way more than the prize is worth. When politics flips into this second equilibrium, we call it war. The trick to keep it in the first one is to make a credible committment to spend a lot in case it gets into the second phase.

      Usually, all stable political systems tend to have such a committment, and a sufficiently rich/powerful outside player can push them into a civil war by credibly demonstrating that they are willing to help the current underdogs in excess of the credible committment of the current incumbents. Works pretty reliably around the world.

  2. Hochreiter says:

    It’s an odd sort of coordination problem given that the totality of the political organism is small enough to be personally funded by several interested agents. This explains the absence of small donations in politics, but Exxon could buy out essential institutions of Washington with pocket change and reorder the language of politics to their liking. I cannot see how the return on that investment wouldn’t be large enough to be a rational pursuit.

    (I also expected this to be a deep dive into the financial sector’s interest in water-intensive agriculture through almond farms)

    • melolontha says:

      I would have liked to read a few paragraphs on this point — and I was a bit surprised that Scott didn’t address it, given the emphasis in the first section on entities whose potential spending would be far more than a ‘drop in the ocean’.

      But I think there are limits to the degree of control you can (openly) exert over an institution without undermining its legitimacy so badly that you seriously weaken it. This definitely applies to media outlets, where competition exists and barriers to entry are becoming relatively low. And I think it even applies to governments, though inevitably in a messier and more complicated and uncertain way.

      edit: Also, I don’t think the ROI of massive political spending would be quite so obviously and greatly positive as you suggest, even if I’m wrong about the importance of preserving legitimacy — especially when we think about it from the perspective of the individuals who would actually make and execute the decisions, and in terms of utility rather than linear $. But I don’t have a clear model of this in my mind.

      • Fluffy Buffalo says:

        But I think there are limits to the degree of control you can (openly) exert over an institution without undermining its legitimacy so badly that you seriously weaken it.

        That would essentially be the point of the exercise. I mean, why bother to buy an outfit that is already doing God’s work (from your perspective)? It would be much more efficient to pick the most popular propaganda machines of the opposing side, and then either shut them down, or divert them to harmless pursuits, or just launch enough (conveniently true) rumours that they’re now working for the enemy, grab some popcorn and watch the infighting begin.

      • Aapje says:


        Successful revolutions often involve the opposition working together to topple the regime, even if they hate each other, because they hate the regime more than each other.

        An attempt to control politics by a single actor may bring about unity against the looming threat.

        • Lambert says:

          ‘One entity becomes powerful enough to threaten all the others so everybody else forms a coalition to stop them’ is an important dynamic.

          It’s one of the biggest drivers of European history from the 30 Years’ War to WWI.
          If Exxon tried to corner DC, all their competitors and anyone else this might harm would unite against them.

    • Nornagest says:

      Exxon isn’t stupid, and it isn’t shy about trying to manipulate politics. If it isn’t buying up think tanks and newspapers, that’s probably because it doesn’t think buying up think tanks and newspapers is worth doing.

      There are probably a few reasons for this, but the first two that come to mind are, first, that authenticity has value in politics; and, second, that in an open society propaganda that’s too blatant tends to spur its own counter-propaganda. What’s the last time an industry made a really big push to try and change the political landscape in favor of itself? Maybe Philip Morris et al.? How’d that work out for them?

      • Aapje says:

        Also, think tanks might be mostly preaching to the choir, in which case, they are at most useful to Exxon to make their concerns more prominent among those who already favor them.

    • Wency says:

      Scott misleadingly uses XOM’s revenues, which are massive, but the oil & gas business is relatively low-margin due to the enormous multi-year capital expenditures involved. Its profits were on the order of $40 billion at their peak (with $100 oil), closer to $20 billion in recent years. Much less profitable than, say, AAPL. So it wouldn’t be pocket change to allocate, say, $5 or $10 billion to US politics — it would be 25-50% of their earnings, and deeply upsetting to investors. If this move backfired in any way, the CEO would definitely be fired for such a bizarre decision. Whereas he can deliver mediocre performance in the oil biz and still expect to keep his job and be well-compensated until he’s ready to retire (or briefly become Secretary of State, as the case may be).

      I’m also not sure what they could hope to get from the U.S. government that would be worth that kind of money, unless you imagine the Treasury just forking over cash hand over fist to subsidize drilling. It’s not that heavily regulated an industry, and the U.S. regulations when it comes to drilling are already pretty favorable. Plus XOM doesn’t even need to drill in the U.S. and wasn’t too focused on drilling here for many years until recently. XOM, like most megacorps, is global. Buying the U.S. government can only do so much for you when you’re all over the world and would need to buy lots of governments.

      If I could name a single company that would most benefit from owning the U.S. government, it’s probably UNH. It operates exclusively in the U.S., the biggest player in the most heavily-regulated industry, and the business always runs the existential risk of being snuffed out by single-payer. But it only generates about $10 billion in earnings, and so far the healthcare industry’s level of political spending seems to have worked out OK.

      Also, if a health insurance company owned the government for a period of time, then if they ever lost their grip on it, I imagine single-payer would become just about inevitable in reaction. Their current level of spending can plausibly say “We just want to make sure our voice is heard.”

      • Hackworth says:

        I’m also not sure what they could hope to get from the U.S. government that would be worth that kind of money, unless you imagine the Treasury just forking over cash hand over fist to subsidize drilling. It’s not that heavily regulated an industry, and the U.S. regulations when it comes to drilling are already pretty favorable. Plus XOM doesn’t even need to drill in the U.S. and wasn’t too focused on drilling here for many years until recently. XOM, like most megacorps, is global. Buying the U.S. government can only do so much for you when you’re all over the world and would need to buy lots of governments.

        Exxon being a US company with multinational interests is exactly why they would want the US government having their back. The US has never had many qualms about throwing around its weight (no matter how soft-sopken the respective president may have been) when it comes to advancing or protecting its economic interests, up to and including toppling democratically elected governments with military power. The very term “banana republic” was coined to describe countries exploited by the United Fruit Company, which had extensive government entanglement.

        Even with the existential risks involved in exploring, drilling, and burning ever more fossil fuel, the reality is that the melting of the polar caps is not going to horrify everyone equally – there will be an unprecedented resource rush, oil among those resources, because the logic of capitalism demands it. Many global powers have been positioning themselves for decades when it comes to the resources of the arctic/antarctic, such as symbolically planting flags, and performing military and resource-exploring expeditions. It will be a land grab with all available methods, including military if international political negotiations don’t bring the desired agreements. As a company, you will want that kind of backing.

    • gbankmanfried says:

      I’d also claim that it’s not really a coordination problem, it’s more of a scope-insensitivity problem. The effectiveness of political donations is fairly linear. If $10 billion buys the whole political ecosystem, $10 million probably buys .1% of it. There are reasons why it is superlinear — you can fund research into how to spend your money better, lobbying has some fixed costs, etc. But there are also reasons why it is sublinear — money has diminishing returns in political campaigns, you run out of the close elections or swayable politicians, etc.

      So if people really thought that the political system was a “steal” for $10 billion, they should probably still be donating more than $2.

      (credentials: former election analyst and political donor advisor, current Congressional staffer)

  3. IvanFyodorovich says:

    Not the main gist, but could we really end homelessness on $20 billion a year? Los Angeles passed initiatives to spend hundreds of millions a year on the problem, so far to no effect. I don’t know if this is because Los Angeles is just terrible at accomplishing stuff, or because this is a hard problem to solve even on a big budget.

    • gbear605 says:

      With that $20 billion, you could give each homeless person in the US 35k per year (550k homeless population in 2018, per Wikipedia). That’s a gross simplification, but here’s a few more details on how to make them not homeless:

      Theoretically, you could house the entire US homeless population for less than $20 billion (550k homeless, three of them in a house, build each house for ~$90,000, per Habitat for Humanity, 550k / 3 * $90,000 = $16.5 billion). Many of them face substance abuse or mental illness (about half, per Sunrise House), but the $20 billion per year could also pay the salary ($80,000/year, which is fairly high) for one helper per current homeless person who faces substance abuse or mental illness. So build the houses in the first year and from then on pay people to take care of the homeless population who faces those issues.

      Obviously this is a simplification to the greatest degree, but I’d say that ending homelessness on $20 billion a year is definitely possible.

      • rbwabd says:

        35k a year is not enough to pay for housing + 275k helpers at 80k a year even under your simplified scenario.

        program administration costs will be huge.

        550k isn’t a constant number. improving relative attractiveness of homelessness with free housing will increase figure quite sharply (“i’m 30 and jobless and living in parents basement and dad doesn’t allow me to smoke weed, let me get into that government program instead…”)

        worse, that 550 figure you cite is estimate for homeless people on any given night. but homelessness is episodic. this link estimates that 3.5m people will be homeless at any point during a single year. if you take multiple years it will probaby careen towards 8 digits pretty quickly. if you offer a nice free lodging option (better than the struggle to pay rent many many low income people face on daily basis) then all these people will become permanent customers. your mission is no longer solving homelessness for 550k people but basically ending poverty in the US.

        you really think having a personal assistant solves substance abuse issues? they are going to be stripping sinks, toilets and copper wiring from your nice houses within the first week to sell for dope.

        solving homelessness doesn’t literally just mean put a roof over their heads. They also need food, healthcare, other necessities, jobs, help reintegrating society. If you wanted to provide a comprehensive program that could actually make a real and permanent difference i bet you are looking at a mid-6 digits figure per head, minimum (although maybe in that case a portion of it won’t need to be permanent if we take the optimistic outcome where we actually save many of these people for good with a one-off spend)

        i also wonder where you’d go to build couple hundred thousand houses for former homeless people. not in white suburbia for sure. basically bound to be housing projects which turn quickly into lawless ghettos.

        would 20bn a year create an improvement over current situation? probably (i hesitate to be more certain than that). will it actually solve the problem? very unlikely.

        • Ttar says:

          As someone who works in property preservation I can absolutely back up your claim that these houses will be stripped bare of every fixture, wire, and pipe by day 30, rat and roach infested by day 60, and completely unlivable and condemned in whatever timeframe it takes a city code officer to drive by. We already see this in a shocking proportion of FHA and VA loans, which should be a much more responsible pool than the homeless population, and in which case the people generally had to pay to get in.

          • Elementaldex says:

            Could you expand a little on how often you see these behaviors in VA/FHA loans. As a recent user of an FHA loan its hard to imagine putting time and money into a property and then just destroying its value.

          • Antistotle says:

            When people are getting foreclosed on they trash the place.

      • eric23 says:

        If you give all homeless people $35k/year, there will suddenly be many more homeless people.

        • Exactly. The suggestion to just calculate how much it would cost to give all these people places to live, and claim that is how much it would cost to end homelessness” is utterly ignorant. In reality eliminating homelessness would have a marginal cost which would increase with each case. Consequently completely eliminating it has an effectively infinite cost.

        • Viliam says:

          Also there will suddenly be many people selling alcohol in large packages costing $35k per package.

      • keaswaran says:

        You don’t solve homelessness by building a bunch of cheap houses in areas with conveniently open land and relocating the homeless people there. People who don’t have a home usually have geographic needs of some sort, and moving them to cheap land isn’t going to work for very long (they’ll either lose access to their entire social support network, or they’ll abandon that cheap house in the suburbs and come back to homelessness in the city center).

        Building housing in the places where homeless people can benefit from it is a lot more expensive than just building Habitat for Humanity houses for people that already have cars and suburban social support networks.

        • IvanFyodorovich says:

          Good point, and perhaps I am being a little unfair to LA. They may not be able to solve this without better housing policy. In LA, or especially SF, giving homeless people easier access to cheap housing is a little like giving one kid an advantage in a horrible game of musical chairs.

      • DragonMilk says:

        NYC spends what amounts to nearly 50k/year per homeless person.

        Given the excrement and public defecation/urination I see from time to time when walking to work, I am in doubt that the money is efficiently flowing down to goods and services for the homeless.

      • John Schilling says:

        build each house for ~$90,000

        You can’t build a house in the Bay Area for $90,000. And Plumber will gladly explain to you why you can’t entice people to move from the Bay Area to e.g. Appalachia even with the promise of a free room in a $90,000 house. They’ll crash on their friends’ couches, live out of their beater cars, sleep in the rough if need be, but they’ll be home even as you count them “homeless”.

      • eqdw says:

        The last time I ran the numbers, San Francisco spent something like $60,000 per homeless person per year (~$700M budget divided by ~11k homeless).

        If double your budget was incapable of solving anything (the homeless problem, both by the numbers and by subjective reports, is getting worse), why should I accept your numbers here?

        • Le Maistre Chat says:

          A $700 million budget for the bureau of homeless people?!
          Hrm, city budget of $12 billion. I suppose that checks out.

          • DragonMilk says:

            Maybe more per capita, but NYC is spending $3.2bn on 61,415 homeless people

            I’d be happy to solve homelessness in this city for that kind of money, but it’s probably better to spend most of it on policy wonks and bureaucracy. After all, how could homelessness be solved when you’re not even spending the median US household income per homeless person?

    • mobile says:

      What if you spent $20 billion on homelessness and wound up with $20 billion worth of more homelessness?

    • Robert Jones says:

      I also find this claim highly surprising. The linked article is very short, but provides a further link for that claim, which just says, “Mark Johnston, the acting assistant housing secretary for community planning and development, estimated that homelessness could be effectively eradicated in the United States at an annual cost of about $20 billion.” It’s not clear in what context he provided that estimate or how he arrived at it. An acting assistant housing secretary seems like the sort of person who might be inclined to overestimate the ability of government to solve the problem.

    • It’s a coordination problem. It’s Moloch. SF spends $300M a year on homelessness. There are 678 homelessness-focused groups and 1,800+ services offered by those groups.

      I went to a seminar on ending hunger, and my line of questioning ended up with this conclusion, “there is no shortage of excess food, and no shortage of people willing to provide that abundance to others.” Hunger is not a supply problem, it’s a distribution one.

      The Head of the SF Marin Food bank said the no. 1 thing that would do the most good on ending hunger would be to make the application for Cal Fresh easier.

      So, we have on one end a bunch of private organizations with lots of random donors, and then on the other end we have the government that makes a half-ass attempt at solving these problems. Government is supposed to be how we solve coordination problems. Or maybe it’s not. Perhaps this is why I’m a fan of UBI because it solves the political problem of government-based giving; nobody can be against it if we all get it.

      • Ttar says:

        If some people with UBI (maybe 1%, who might be marginally metnally I’ll or challenged) spend it all on drugs or other non-housing services every month and then end upon the street, how has UBI helped? Bear in mind that would mean the same 3.5m occasionally homeless people we have today…

      • matthewravery says:

        The complexity of “solving hunger” or “solving homelessness” are so overwhelming that I find it exasperating when I read posts on reddit or twitter bemoaning, e.g., Bezos’s decision to not “end world hunger” with his wealth, as if it were as easy as filling up a novelty donation thermometer.

        Truth is, I imagine Bezos could put a hefty dent in world hunger if he made it his life goal and devoted his efforts to that instead of running Amazon and the other things he does with his time. But that’s more of a “proven entrepreneur shifts focus to altruistic causes” rather than a “he has lots of money so ezpz” argument, which is not what I think the twitter/reddit posts are talking about.

        I think this is one of the reasons why charities, where possible, try to offer comoditized results from donating. “Giving $8 will feed a needy child in Africa for 30 days” is a very concrete transaction. You understand what your money is “purchasing” in terms of reduced suffering. Or at least that’s how its advertised; it’s likely far more complex than the 1990s-era TV ads made it seem.

      • BD says:

        Perhaps this is why I’m a fan of UBI because it solves the political problem of government-based giving; nobody can be against it if we all get it.

        My view of “the political problem of government-based giving” is that the receivers are often more organized and effective than the payers, leading to distortions and inefficiency in whatever problem you’re trying to solve.

        I suspect you believe that this is solved for the UBI because “everybody” is both receiver and payer, but that is not true. Every CURRENT adult becomes a receiver, but it’s the FUTURE taxpayers footing the bill. So you still have the same distortions and inefficiencies…in this case which will be manifested as crushing national debt (beyond the not-quite-crushing debt we have now).

        • Why can’t we start UBI @ $1/mo., and put no cap on how much lawmakers can scale it. That way the benefit at least becomes in play, providing another lever for the 99% for whom UBI would be a net benefit. Maybe one day it’ll be Yang’s $1,000/mo., but it’ll go up and down based on the general political tug-of-war, which to some degree involves a consciousness of our deficit?

          Total honesty, though, I’d like to see a small, but developed, nation try this first, before we slap it on the U.S.

          • Viliam says:

            it’ll go up and down based on the general political tug-of-war

            You would likely get a positive feedback loop here: higher UBI -> more people quit their jobs and live on UBI -> more people who want to increase UBI -> politicians further increase UBI -> etc.

            To achieve balance, you need some kind of negative feedback loop. Technically, the easiest way to achieve it would be if only net payers could vote on increasing or decreasing UBI. But that could cause other kinds of problems.

          • Anthony says:

            I have a proposal for a UBI test site:

            Nevada, plus the California side of the Tahoe Basin.

            The only place with jobs that’s a reasonable commute from Nevada is the Tahoe Basin, and vice versa. The Tahoe Basin is not a reasonable commute to any other part of California.

            To solve the migration-for-UBI problem, we could limit eligibility to people who living in the zone on the day of the 2020 Census, or who can show they’ve been there for at least one year, which will make moving unattractive for people who don’t have or get jobs.

      • John Schilling says:

        Perhaps this is why I’m a fan of UBI because it solves the political problem of government-based giving; nobody can be against it if we all get it.

        Scaling problems and status-quo bias are going to keep this from being truly universal for quite some time, but we can start small as a proof of principle. How about the philosophistry-Schilling special supplementary income? Way it works is, you max out your credit cards for a cash advance and put $20,000 into the pSssi fund. Then we each get $10,000. You should be a great fan of this; neither of us will be against it because we’re both getting it.

      • Antistotle says:

        Government is supposed to be how we solve coordination problems.

        Where do you get that from?

    • DeservingPorcupine says:

      Agreed, mostly for reasons already mentioned. There is absolutely no chance whatsoever that 20 yards a year would cut homelessness by large amount.

    • baconbits9 says:

      General reply:

      If you are attempting to solve homelessness and are talking about average/median spending then you have the wrong approach. Imagine a school system that has decided to commit to every student getting an 800 on the SATs (honestly). How much money/time/effort would it take to get a severely developmentally disabled child, with Downs perhaps, to that score level? On the extreme end it isn’t even a question of resources, there is no dollar amount that will get that child above a certain threshold, with the homeless there is no dollar amount that will get the most severely mentally ill off the street indefinitely (barring forced incarceration). Likewise there is no known treatment that approaches 100% cure rate, even with multiple trips, for addiction, and an addict can consume indefinite amounts of resources.

  4. rho says:

    Imagine paying $3 million just to erase tumblr.

    Honestly though, i wish there was a note-taking app with tumblr’s posting interface. Add a way to rearrange stuff and it would be primo.

  5. EricN says:

    Great post — the analysis makes sense to me. I agree it’s confusing that there’s so little money in politics, and that this is because of a failure to coordinate. Interestingly, there’s no problem with voting — even though naively voting isn’t in your own interest (it takes up time and only very slightly influences the probability that the person you like will win), people do it — perhaps because it’s something that’s encouraged by society and easy to be proud of!

    I also agree that more money in politics wouldn’t be a good thing, because it seems like a big waste: if people donate $1 billion more to both the Republican and Democratic nominee for president, that’s $2 billion lost right there without a substantial change in the outcome.

    I think there’s a pretty neat solution to this problem, though: say two people want to donate $100 to opposing political campaigns. Realizing that they would cancel each other out and their money would go to waste (read: dueling TV ads no one wants to see), they could agree to give their money to charity instead! More concretely, I think it would be great if a platform could be set up to facilitate such coordination. If this idea interests you, here’s some further discussion. (Thoughts on this idea?)

    (See here for some follow-up thoughts I’ve had concerning incentives in such a platform.)

  6. slartibartfast says:

    I’ve always assumed that the reason there was such miniscule amounts of money in politics was because it was all being channelled illegally, and consequently not reported.

    • Viliam says:

      You don’t give money to a politician; you offer a good business deal to a company they own, or you employ their cousin at a company you own.

      (That’s what I could think of in 30 seconds. I am sure that people who do this for living know many other options.)

      • EchoChaos says:

        Or you donate to their foundation or you give them a book deal or any one of a huge number of legal ways to directly give them money.

        • Ttar says:

          Exactly. Paid speeches. Sweetheart business deals and nepotism. Lucrative consulting careers post political retirement. Invitations to swanky parties. Epstein sex stuff. None of this is going on the books in the above statistics as lobbying or campaign spending.

        • tayfie says:

          There is massive amounts of money that go to politicians and former politicians that is probably accounted as Cost Disease.

  7. theodidactus says:

    Hiya: former campaign finance researcher and current law student with an interest in campaign finance here.

    Here’s some work I did in the 2012 election:

    My personal take on the matter is really cynical: When you look at campaign finance as a whole, what you’re looking at is a system that’s carefully calculated to pull money from certain true believers, and put it to absolutely no use. Most of the money, I’d argue, is “wasted” on pass-throughs and media consultancies, who pocket a big chunk on every pass.

    As evidence, I’d offer up the fact that most election spending is objectively badly spent, especially by people who should in theory be really good at spending the money well. Intuitively, this makes sense as well. If I gave literally anyone here 15 million dollars and charged you to meaningfully influence a senate or congressional race with that money, I believe you could do a better job than stuff like this.

    • Enkidum says:

      If I gave literally anyone here 15 million dollars and charged you to meaningfully influence a senate or congressional race with that money, I believe you could do a better job than stuff like this.

      You can’t just have faith without evidence. I volunteer as a test subject.

    • dogiv says:

      Candidates definitely care about winning. They’re certainly not always good at spending money effectively: they hire the wrong people (often because of who those people know, owing somebody a favor, etc), they don’t make optimal trades between staff and TV ads, etc. But they’re trying, and while extra money might not make a huge difference, having enough to keep your campaign up and running (and get your name out there) is absolutely essential.

      Super PACs are different–it’s an intentionally indirect way of giving money, so it’s probably easier to divert some. But regular people (even true believers) mostly don’t give to PACs, they give to candidates. Super PACs only make sense for mega-donors who will quickly hit their per-candidate donation caps.

      Also, that ad didn’t cost anything remotely like $15 million, it was $700,000. Maybe they should have spent a bit less on airtime and a little more on writing, though…

      • theodidactus says:

        I mixed that ad up with a different one (and it’s an entirely new experience being fact-checked by an article that I wrote). My broader point still stands. $700,000 is a lot of money from a “get **** done” perspective. I think that ad is the sort of thing you’d make if you wanted to convince a donor you were doing a good job, rather than an ad you’d make if you actually wanted to change anyone’s mind.

        While I characterized this as a nefarious swindle, I think I might have been overly cynical for rhetorical effect. Naturally what the ad-placer and the ad-maker are going to say is their job is to do what their donors want them to do.

        …but I think my larger point is that a lot of this money ends up as basically waste heat, and you get really diminishing returns on any election spending of consequence. I don’t know as much about 2016 but I can speak on 2012 given my recollection: with the exception of a few wacky buys in minnesota toward the end of the race, and two campaigns from then-relative unknowns Cruz and Flake, I don’t think big money actually did anything.

    • sty_silver says:

      I’ve long felt that TV ads, in general, can’t possibly be cost-effecitve, based on a) rumors I’ve heard that they’re extremely expensive, b) the fact that there aren’t even that many people who watch TV anymore and c) that watching TV ads is super unconvincing.

      I mean, just think about that last point. How much more likely are you to be convinced by someone whom you are talking to personally than by watching an ad? 100 times?

      Using the money to get more and better people to phone bank / knock on doors has to be better. As an authority on this subject, would you say that’s accurate?

      • theodidactus says:

        They are extremely expensive, and a big part of the cost is figuring out where to air them. As some of the articles I linked to above mention, some of the biggest gurus in this area seem to be exceptionally bad at figuring out where to place these ads.

      • dogiv says:

        In terms of getting name recognition, TV is really hard to beat. You can phone bank for turnout or even to persuade undecided voters, but you can’t call every voter who’s never heard of you.

        Phone banking and knocking doors are far more cost-effective when they’re done by volunteers, which is a finite pool. Once you’ve got all the willing volunteers doing stuff for you (not easy, but possible with good organization), the only way to increase your call volume would be to hire a lot more staff, and that quickly gets way too expensive.

        One reason getting on TV is so expensive is that media markets don’t correlate very well with political divisions, so you end up airing your ad to a bunch of people who can’t even vote in the election.

        • sty_silver says:

          Phone banking and knocking doors are far more cost-effective when they’re done by volunteers, which is a finite pool. Once you’ve got all the willing volunteers doing stuff for you (not easy, but possible with good organization),

          Don’t you think you could almost always improve organization with money, though? I would be pretty surprised if the Sanders campaign was within, say, 50% of what volunteers could accomplish if perfectly organized. Admittedly that’s total speculation.

          • Alsadius says:

            The stories I’ve heard about the Sanders campaign say he was impressively well-organized.

            As a really rough guess, I’d say the first million or so that you spend on organization does a lot of good. That gets you offices, a core of full-time staff to do research and media and logistics, a professional website, good campaign materials, pizza for the volunteers, a few polls to gauge your messaging, and so on. After that, the diminishing returns start to cut into you pretty quickly.

            The ceiling might be a bit higher for a Presidential campaign, which needs more staff, but I’d bet you could run a full organization for a full campaign on $10M and not feel too pinched. Everything else goes to ads and hiring paid foot soldiers.

      • Logan says:

        I’ve found political TV ads convincing. Not in the “What? Your opponent wants to raise taxes? Guess I’ll vote for you” sense, but rather in the “Huh, forgot that guy was running. How seriously should I take him?” sense.

        When I’m actually undecided, every little thing that reminds me of a candidate (a conversation at work, a thinkpiece or even just a headline, a tv ad, news of a rally in town) makes me re-evaluate each candidate, slightly changing my conception of them. If I see two in a week, I’m likely to consciously think over whether my opinion of them is outdated. If I talk to someone personally and they parrot an idea that I’d already seen (in an article or on an ad) I’ll probably get the sense that there’s a zeitgeist I’m missing and become eager to learn more.

        Ads don’t single-handedly change everything (you can put anything in an ad, but ultimately the voter can see the candidate) but they can be one component of a convincing campaign.

    • sourcreamus says:

      There is a saying about advertising that half of it is wasted but no one know which half.
      In politics it is probably more like 90% is wasted but no one knows which 10% works.

      Since getting less votes than your opponent is the same as getting 0 votes, it makes candidates willing to spend everything they are given and try for as much as possible just in case it works.

      When stakes are high and information is low, fraud and whoo abound.

    • Douglas Knight says:

      What other of your posts do you recommend?

      What do the competent people do? Proposing that the flood of money attracts parasites is sensible. But does the portion that isn’t wasted shine through and let us understand things?

      If campaign spending were completely worthless, then competent interest groups would get out entirely and focus on lobbying. But you think that advertising is possible. Can you point to advertising you think is effective?

      If money is so ineffective, why do politicians spend so much time raising funds? Aren’t some politicians competent to understand that? Are they buying something other than votes, via patronage?

      • Alsadius says:

        A lot of politicians are dumb, and a lot of others want to be the leader in the “fundraising race” stories the media likes to run (because it’s one of two objective data sources they have, and you can’t easily mess with poll results).

  8. Said Achmiz says:

    Nobody’s going to cynically declare “Oh, people don’t really care who wins the election, they just pretend to.”

    Huh? Why not?!

    I thought that this was the obvious answer when I started reading this post, and I still think this is the obvious answer. Why in the world are you dismissing this so peremptorily, Scott?

    • melolontha says:

      Why in the world are you dismissing this so peremptorily, Scott?

      I think because of the financial rewards that seem to be on the line. (Especially if we translate ‘who wins the election’ to ‘the policies we actually end up with’, because winning this game might involve getting both parties on your side, and/or influencing the non-partisan machinery of government enough to constrain whoever is in power.)

      (Admittedly Scott didn’t specify whether this argument was supposed to apply to the case of average people too. Obviously scepticism is much easier to maintain there, though the failure of the ‘nobody actually cares’ explanation in the case of big business might imply we don’t need to rely on it in the case of individuals either, because whatever actually explains the behaviour of big business might also explain the behaviour of individuals who do care.)

      • Said Achmiz says:

        Especially if we translate ‘who wins the election’ to ‘the policies we actually end up with’,

        But we shouldn’t do this translation. That, actually, is basically the whole point.

        • melolontha says:

          Okay, but political spending is about influencing what the winners do, as well as who wins, and Scott talks about both of those purposes in the article. So even if it were true that nobody cares who wins, there’d still be the question of why more money isn’t spent buying influence.

          edit: if you believe that nobody cares who wins because that doesn’t really affect the policies we end up with, presumably that’s because you think that whoever wins will face the same incentives and constraints. So the question is, why is more money not spent trying to change those incentives and constraints. Maybe some of them are too deeply rooted to be bought, but some of them are things like ‘what the public wants/will tolerate’ and ‘who will give me money for my reelection campaign, and under what conditions they will do so’.

          • Said Achmiz says:

            So the question is, why is more money not spent trying to change those incentives and constraints.

            Who says it isn’t? (Certainly nothing like this conclusion is warranted given what Scott writes in this post; in fact, nothing in the post even touches on this.)

    • Dacyn says:

      Yeah, my response to that remark was “when did Robin Hanson get renamed ‘nobody’?” But I think the reason Scott is dismissing it is because it’s the kind of thing you can’t really argue with, like Howard Holmes’ beliefs about how no one should care about anything because no outcome is better or worse than any other outcome. And most people do think that people care about elections.

      • Said Achmiz says:

        To the contrary—this is a very different case.

        “No one should care about anything because no outcome is better or worse than any other outcome” is obviously absurd, because the first clause is a “should” claim (and thus depends on one’s own values, and is unsupportable across gaps in values), and the second clause is manifestly false if taken literally, while being totally underspecified otherwise.

        The claim I’m making (and that Scott dismissed without consideration), on the other hand, has no normative component whatever. I don’t say that people should, or should not, care about election results—only that in fact, they (to a first approximation) do not.

        And most people do think that people care about elections.

        Yes, your phrasing is apt: most people do, indeed, think that people (including themselves) care about elections. But actually, they don’t care. It’s a sort of belief in belief.

        • Dacyn says:

          What I meant was that they are both a sort of arguing about definitions. Holmes has a different definition of what “should” means, and in this instance you seem to have a different definition of what “caring” means — does it mean “acts in a way so as to effectively achieve a certain goal”? or does it mean “has an internal experience of the sort that most people would describe to others as ‘caring’, if they wanted to convey what their internal experience was like”?

          • Said Achmiz says:

            Yes, this is a good time to taboo the word ‘care’.

            I think that people have the internal experience of feeling in a way that they describe as ‘caring’ about politics.

            I also think that people are not motivated to, and indeed do not, act in a way consistent with an assumption that they have ‘affecting politics’ as a goal, or that ‘politics’ (in the specific sense of “who wins the election, at the federal level”) affects their lives in any meaningful way.

            (The latter is, obviously, much more relevant than the former is, when it comes to predicting people’s behavior.)

          • Dacyn says:

            OK. I agree with all that.

            Regarding what Scott said. Of course external caring is more useful for prediction because it’s defined that way. But internal caring can also be useful for prediction. For example, we can predict that if someone internally cares about a goal and there are literally no obstacles to them acheiving it, then they will just do it. After all, why wouldn’t they? So if they don’t, it makes sense to talk about what the obstacles are.

            In this case, I read Scott as saying the obstacles are things like: you don’t have direct control over the outcome, you won’t even know whether you made a difference, versus buying almonds where you know exactly what the results of your actions will be. In Hansonian terms we could say these obstacles move people’s thinking into far mode, where signalling is more relevant. I don’t know if “coordination failure” is a good way of describing these obstacles, but I think I can at least see where Scott is coming from.

            Note also that Scott’s statement about caring about elections may have been intended to contrast with the behavior of corporations, for whom there’s often no reason to believe anyone at the corporation internally cares about anything the corporation claims to care about.

            Anyway I’m not sure whether my initial comment in this thread was a good pointer to what I wanted to say, but at least we seem to have gotten somewhere useful with it.

      • aristides says:

        My thought as well. If you add Hanson’s work with Bryan Caplan’s The Myth if the Rational Voter, there is not much of a mystery left. Of course many disagree with their conclusions, but that is still my prior based on the objective evidence that Scott presents here. Even if I look introspectively, I realize I care more about who wins the next super bowl than who wins the next presidential election, and I’m the kind of person that reads Scott, Caplan, and Hanson for fun. I also work for the government, so that person will technically be my boss. How much do you think the average person truly cares about politics?

        Edit note: I corrected grammar and accidentally reported myself. Please ignore report Scott

        • Dacyn says:

          I am a little confused that in your first sentence you seem to be agreeing with me, but in the rest of your comment you seem to be disagreeing. Anyway, I agree you may be more emotionally invested in the Super Bowl outcome but if you had the choice of which one you could control the outcome of, is that really the one you would pick?

    • Ursus Arctos says:

      Anecdotally, If you placed in front of me a button that would let me change the results of the next US election, but in the process I would die, I would push it in a heart beat.

      • Glen Raphael says:

        Anecdotally, If you placed in front of me a button that would let me change the results of the next US election, but in the process I would die, I would push it in a heart beat.

        Wouldn’t you want to wait and see what the result is first? You might not need to push the button!

      • Said Achmiz says:

        Talk is cheap. Anecdotes aren’t data, but what you’ve just given us isn’t even an anecdote. The fact that your described scenario cannot possibly ever happen makes your talk even cheaper.

        What we actually see, in reality, is people not spending much money on elections. That is our data. “In this totally impossible hypothetical scenario, I would…” is worth less than nothing.

        • niohiki says:

          Yeah, and a bunch of data is not a model, either.

          Saying that you must be right because your model happens to be compatible with specifically the data you present… I don’t think it even qualifies as a fallacy.

          • Said Achmiz says:

            I didn’t say anything even remotely resembling that, so you’re attacking a strawman.

            EDIT: And, more importantly, what you said is not even slightly pertinent to the point I was making in the grandparent.

      • Tarpitz says:

        I, on the other hand, would not sacrifice my friend’s dog to control the outcome of the next UK election, and I think I’m far more invested in politics than the average person.

        In fairness, it is a very cute dog.

    • niohiki says:

      Ok, let’s do an overly simplistic calculation, because numbers. How much do people care about politics, donations aside? “Care” is very ill-defined, and the worth people place on their time equally so, but we can make do with a very crude quantification by using opportunity cost, which at least gives us an order-of-magnitude estimate. I read in several places that average income is around 25$/hour.

      The first thing to think about is, obviously, how much time people devote to voting. Voter turnout in the US over the last years has hovered slightly over 50%say the average voter spends half an hour between commuting, queuing and whatnot. With 300+ million people of voting age, that’s 1.8 billion dollars of opportunity cost per election. One should multiply this by some factor to account for local elections and where applicable, referendums.

      But wait, do people not think about politics prior to that vote? “No”, you will say, by virtue of the obviousness of people’s lack of interest in politics. Let’s see. Forbes tells me that peak ratings for Fox News, MSNBC, and CNN sum together an audience of about 5 million. Assuming very conservatively that this core audience watches about 3 hours of cable news per week, and (even more conservatively) that only half of its content is political, this gives me about 1.4 billion dollars per year. This is without counting the cost of cable TV itself, which may arguably be mostly paid for other reasons.

      What about activism? In 2017, about 4 million people participated in the Women’s March in the US. Assuming they spent about 5 hours in the whole endeavor, that’s 0.5 billion extra dollars. Of course, there are plenty of other demonstrations and strikes (arguably, these are pretty direct forms of opportunity cost) going on all the time.

      And, well, not everyone watches Fox News or attends strikes. Can we quantify the amount of time people spend arguing politics on the internet or with their relatives and friends, even if working in echo-chamber mode? I’m not sure how to, to be honest, but even if the average person devoted only 0.01% of their average 16 daily waking hours to discussing politics (and this is for sure a large understatement, because it amounts to 35 minutes per year), that 300+ million population accumulates 4.3 billion dollars per year.

      One could go on finding examples like this. All the numbers are pretty wild, and the assumptions correct only at order of magnitude – if anything. The point is not what the exact number should be, but rather that even if I am off by a factor of 10 in the estimation of the value people place on their time, I could get a general personal investment in politics of at least the order of magnitude of the almond industry. People care about politics. In at least a comparable amount to what they care about almonds. It’s just that they don’t seem to place a lot of value on donations or other ways of directly fueling the political machinery.

      For sure, people could care more. They could be demonstrating all the time or writing angry letters to their representatives using all their spare time. But it makes no sense, quantitatively, to say people care “a lot” or “a little” about politics. You need a reference. People seem to care more about politics than about almonds, and more about almonds than about donations.

      I mean, personally, I would have said that was the completely obvious thing. But an internal feeling of obviousness should never be an argument.

      PS: In case it is not clear enough. The average smoker is more than happy to spend absurd amounts of money on tobacco every year. Yet whenever it gets banned on this or that public space, they don’t spend anywhere a similar amount of money in funding campaigns to get their previous right back. Why? Because maybe deep down people are not so stupid, and they understand, even if only intuitively, this:

      In this model, the difference between politics and almonds is that if you spend $2 on almonds, you get $2 worth of almonds. In politics, if you spend $2 on Bernie Sanders, you get nothing, unless millions of other people also spend their $2 on him. People are great at spending money on direct consumption goods, and terrible at spending money on coordination problems.

      • dogiv says:

        The things you list here are essentially for signaling and/or amusement, and tell us very little about how much people care for the election result.

        Voting is very visible, and in many cases, if you skip it people will be upset with you. Also, while the average wage may be $25, in practice people generally value their time much lower than that.

        Watching the news is basically entertainment and has little to do with actual motivation to influence the election; same goes for “discussing politics”. Marching in a protest is even more visible and social than voting. Lots of people will go join a march, but almost none of those people will volunteer to make phone calls to strangers on behalf of a candidate.

        Only anonymous donations would really work as a measurement, and I wouldn’t know how to interpret the numbers if we had them–you’d want to adjust for the personal consumption effect somehow. Hypothetically, say a labor union could coordinate to spend $10 million in a state race by raising union dues ($100 per member), and that spending would guarantee their candidate wins an election they would otherwise have lost. What percentage of union members would consider it worthwhile? It seems plausible that people might care $100 about who’s governor, in conditions like that. But I don’t know of any data that would tell us.

        • discountdoublecheck says:

          The things above *may* be about signaling or amusement.

          If you want to defend the claim ‘politics uses few resources’, conservative estimates for resources will include the opportunity costs of those activities. The burden of proof is on you to show that people aren’t focused solely on the resource argument.

          As point by point disagreements — where I live, I vote by mail. This reduces ‘commute’ costs, but it also makes it easy to spend more effort on down-ballot candidates. It also means that my voting behavior is totally invisible to anybody external. I could easily gain the ‘signaling/social’ benefits without incurring the time costs associated with my voting. While not everywhere allows for voting by mail, between early voting, and absentee voting, it would be trivial to lie about your actions to people around you. “Oh yeah, I went last week.”

          As for “discussing politics” the sheer number of people who feel a need to convince family of their opinions despite large social pressure of the form “JUST SHUP UP UNCLE JOE ITS THANKSGIVING”, is solid evidence for an actual caring perspective.

          Another important category niohiki failed to include (and scott as well) is the opportunity cost to individuals within this system. If (as often is the case) a think tank employee could double their pay by going and working for a bank, they are valuing their work extraordinarily highly. That opportunity cost is potentially tremendous — if there are 10k such people earning $100k/year less than they could, that is an annual contribution of $1 billion from that small group.

          • dogiv says:

            Minor nitpick: in the US, whether you voted is public record, so it’s theoretically visible even if you vote absentee. In practice, yeah, you could probably lie about it without anybody noticing and I’m sure some people do.

            The “need to convince” is in my experience a very different impulse from the desire for large-scale changes to actually occur.

            If total spending on think tanks is $500 million, then the in-kind donations of labor by think tank employees are presumably not much more than that (probably less).

          • niohiki says:


            The “need to convince” is in my experience a very different impulse from the desire for large-scale changes to actually occur.

            I agree. Why is it so different, when at first approximation convincing your uncle doesn’t change much, and large-scale changes would actually change your life a lot? Well, precisely because one is aware that it is unlikely that large scale changes will actually happen. That’s basically the point of the post.

            About signaling (and replying to other similar complaints). As @discountdoublecheck mentioned, it seems a bit of a stretch to assume that all angry dinner table, web forum debates and demonstrations happen only due to signaling. Again, the numbers I made up are… made up, and you can multiply it by some factor to account for people only actually caring 10% of their effort vs 90% signaling, and for valuing their time only 50% of their opportunity cost, or whatever. The point was: the claim of people caring so extremely little about politics that it justifies in itself low donations is not sound. This is the typical very-low effect over very-large group calculation where assuming that the effect is just zero is not at all the same as assuming it is very small, but not quite zero. People care at least a little. People make fools of themselves and say things they later regret in this kind of situations, because they are driven by emotion and care more about their “cause” than about their in-group status (ie being dismissed by the rest of the family as being a bigot/commie/whatever). But that “a little” integrated over everyone is definitely much more than what is spent on donations (and, sure, that is easily explained by people seeing donations as pointless because of coordination, which again is the point of the post).

            On the other hand, I know enough people that really like to remind you of the causes they donate to. Since one is as free to signal that as to signal buying diamonds or fancy cars, it works pretty well as a way of conspicuous spending, if signaling was really all one cared about. I’d even say more – it’s quite efficient, because since not a lot of people do it, saying “yeah, you went to some demonstrations as entertainment, but I actually put my money in it” gives way more status points per unit effort.

            Finally, @Said may consider that only countries in a borderline crisis situation where I’d never want to find myself in qualify for “people caring about politics”, but going again to the contents of Scott’s post, it’s not about people caring at levels of life-or-death situations, it’s that people care less about donations than about almonds, over which no time whatsoever is spent arguing on the dinner table. By @Said’s argument, no one who, eg, has not given birth or been amputated without anesthesia knows what pain really is, so the pain from twisting your ankle while running should be rounded down to exactly zero, and the discussion of whether it merits analgesics should not even be had to begin with.

          • Said Achmiz says:


            By @Said’s argument, no one who, eg, has not given birth or been amputated without anesthesia knows what pain really is, so the pain from twisting your ankle while running should be rounded down to exactly zero, and the discussion of whether it merits analgesics should not even be had to begin with.

            I’ve had surgery (not amputation, but tonsillectomy) without anesthesia. I endorse your analogy—no one who hasn’t experienced at least that level of pain, knows what pain really is.

            One man’s modus ponens is another’s modus tollens.

        • Randy M says:

          The things you list here are essentially for signaling and/or amusement, and tell us very little about how much people care for the election result.

          This. Going to a protest is the form of politics our monkey brains actually care about–forming coalitions with specific other people as a demonstration of power.
          Giving money to marginally increase the chance that an election will change, based on the hope that the other candidate will act in a way that noticeably advances our interests… pretty much not.

      • Said Achmiz says:

        Others have already responded re: caring about election outcomes vs. merely caring about signaling/entertainment.

        I’ll just add that I’ve lived in a country, and at a time, where people actually, really cared about election outcomes.

        It looks very, very different from this. (And I most fervently wish never to experience it again, and may you, also, always remain ignorant of what that’s like.)

      • zzzzort says:

        Signaling/amusement is overly abstract, so let’s go with an analogy. It’s September, I’m a baseball fan, and I will spend several hours a week watching who wins baseball games, arguing with people about how baseball teams should win baseball games, and explaining how my baseball team is in fact superior to their baseball team. And while I do care about my preferred team winning, if you offered to let me “buy out” the time spent watching/talking about baseball with a donation to the team, I would of course say no, even though it would obviously be a more efficient way of affecting the outcome. I contend that most people care about politics the same way that I care about baseball, which is consistent with spending a lot of time consuming it and with not wanting to spend much money affecting the outcome.

  9. salvorhardin says:

    “But it would also have to be true that lobbying, the NRA, the media, etc don’t affect politics very much, which seems like a harder sell.”

    Why is this a harder sell? Suppose the world works like this:

    (a) To make an activist (or lobbyist) cause successful, you need “enough” money, but this is not remotely sufficient, only necessary

    (b) Once you have “enough” money, pouring additional money into an activist cause doesn’t buy you much additional chance of success

    (c) There’s already sufficient money sloshing around in the system that any plausibly successful activist cause has a reasonable chance of capturing “enough” of it

    (d) Therefore additional money spent on activism has approximately no effect, and potential funders know this and don’t spend more

    Why is it implausible that this is the way the world actually works?

    • gbear605 says:

      (b) seems unlikely, since any given cause (that is controlled by, say, Congress) can be influenced by every single congress person out there. So if I’m interested in some cause that Congress could either vote in favor or against, then I want to put large amounts of money into every congressional race, of which there are a couple hundred every two years.

      • salvorhardin says:

        But putting those amounts of money into those races doesn’t actually buy you anything, as Scott’s links on campaign spending say.

      • keaswaran says:

        Look at the list of most expensive congressional races in the last political cycle:

        It seems clear that without a million or two, the candidates wouldn’t have been competitive. But Beto’s $75 million just prompted Republicans to raise another $40 million for Cruz, and moved the needle by a few points, and didn’t end up with a change in the outcome. In New Jersey, the Republican outspent the convicted corrupt Menendez by a factor of 3 to 1 and still barely cleared 40% of the vote. In Florida the sitting governor outspent the Democrat by 2.5 to 1 and managed to beat him, but it’s really unclear that the tens of millions spent campaigning made all that much difference. Would another $10 million by the Democrats have affected turnout in the right way, or just had some unknown butterfly effect?

        There’s a lot of diminishing marginal returns to campaign spending, especially because the general public becomes very aware of one candidate hugely outspending the other and has a negative reaction to that.

  10. Ursus Arctos says:

    X-Posted from Reddit. A quibble re: Jacobin. Obviously there’s additional content in the printed magazine, but playing around with the Jacobin website I can’t find anything that qualifies as a paywall per se. I don’t recall ever running into one, despite checking it out fairly regularly. Even if I am missing something, Jacobin seems to make an effort to offer quite a bit on its website freely, which is exactly what you’d expect them to do if they truly want to spread the ideals they espouse. I’ve also noticed that their more immediate and urgent stuff is on the website for free, while more theoretical stuff tends to be in the magazine, which again, makes sense if their motives are sincere (not that Scott is saying their motives are insincere per se, but an important clarification nonetheless).

  11. arbitraryvalue says:

    >revealed preferences that seem ridiculous in light of my values

    Perhaps you’re looking at this backwards? I don’t want to make assumptions about your personal motivations, but my own experience has been that modifying my (ultimately arbitrary) mental model of morality to better match my own revealed preferences has resulted in significant personal growth. I rid myself of useless guilt (the sort that doesn’t change my behavior but only makes me feel bad) and focus more on the things that really do matter to me (according to those revealed preferences).

    • melolontha says:

      But Scott’s refusal to give up on his moral values has directly caused him to pledge 10% of his income to charity, i.e. it has changed his behaviour and moved it more in line with those values.

      (More pedantically, what work is ‘revealed preferences’ doing here? If they’re the marker of your true values, doesn’t every action you take become self-justifying?)

      • arbitraryvalue says:

        it has changed his behaviour and moved it more in line with those values.

        My argument is that there’s a distinction to be made between one’s values and one’s mental model of those values. Aligning your behavior with your values is great, but trying to align your behavior with an inaccurate model which doesn’t accurately represent what you actually value isn’t so great. Maybe Scott is truly a utilitarian at heart. But I’m not, as it turns out.

        doesn’t every action you take become self-justifying?

        I don’t find that to be the case. Your revealed preferences are what you feel is right, not necessarily what you do. For example, I truly feel that it is right to be a good friend. Sometimes I let my friends down, but I don’t think that indicates a revealed preference for letting friends down sometimes.

        Your consciously-held moral system, meanwhile, is what you think is right. The difference between the two is like the difference between wanting something and wanting to want something. If you don’t want to give to charity, but you want to want to give to charity, I think you’ll be better off (in the deep-personal-fulfillment sense, not just the financial sense) by letting go of what you want to want than by forcing yourself to give to charity when you don’t want to.

    • Enkidum says:

      Nominative determinism.

    • Ttar says:

      Some people with high levels of scrupulosity struggle to join us übermenschen.

  12. Glen Raphael says:

    Nobody’s going to cynically declare “Oh, people don’t really care who wins the election, they just pretend to.”

    I might be willing to bite that bullet! The fact that I don’t really care who wins the election makes it hard to empathize with the people who claim to care, so…yes, people certainly sound like they care, but sounding like you care is a good way of signaling group affiliation. So how much of this is just signaling? Do they really care who wins the election more than they care about, say global warming or killer bees or whatever?

    I guess the question isn’t really if people care so much as how they care. For instance, I’m pretty sure a lot of my liberal friends are secretly glad Trump won because it gives them a new excuse to loudly complain about how bad the world is and they really enjoy getting angry about that.

    • MawBTS says:

      I guess the question isn’t really if people care so much as how they care. For instance, I’m pretty sure a lot of my liberal friends are secretly glad Trump won because it gives them a new excuse to loudly complain about how bad the world is and they really enjoy getting angry about that.

      I agree to a small extent. I know people like that too.

      On Facebook, one of them just shared a link to her (young, female, progressive) friends with the message “just in case you didn’t feel sufficiently stabby this morning” [it was clickbait story about some minor CW related thing].

      It was surreal: this was a story guaranteed to upset her friends…and she shared it like a drug dealer giving out speedballs. And they ate it up. They seemed to take pleasure from being upset. I’m 90% sure they were crisis actors, hired by Scott to demonstrate his Toxoplasma of Rage article.

      I don’t believe this is central to opposition to Trump: I think most people who oppose him are genuinely anxious, miserable, etc about his presidency.

      But some seem to love it.

      • eric23 says:

        Presumably she meant her message “ironically”, as the kids do this day.

        And even a story that makes people angry can be good for the world if those angry people are motivated to go and fix the problem (which probably means voting the “right way”, or else convincing other people to vote that way, but could also mean charity or volunteer work). They don’t enjoy being angry, but they do want to motivate people to solve the problem.

      • keaswaran says:

        Moral outrage porn sells:

        It’s a lot like tragedy and horror and spicy food and skydiving – people enjoy feeling intense emotions and sensations, especially when they can do so without the commitment to action normally involved.

      • eqdw says:

        In the wake of the 2016 election I realized that our entire media industry is in the business of emotionally abusing their readers. I’m still not sure what to do with this information, but it is distressing

    • Ttar says:

      I think a lot of people care about election outcomes in similar way that sportsfans do. You might REALLY want the Sox to beat the Yankees, but you wouldn’t personally donate $50 to have some tiny imperceptible impact on the outcome, and besides — it’s not just about winning, it’s supposed to be about winning because your team is better, not just because they have more money.

    • sty_silver says:

      I just want to give my first-person report here. I care a lot about who wins. I don’t donate any money or time (except a 1$ donation to increase the number of individual donors) because I don’t believe it’s effective altruism to invest into politics. But that feels like it’s keeping me from doing something I want to, not like it’s giving me an excuse. I have a desire to donate to my candidate’s campaign.

      I also claim that I would have accepted a lot of personal damage if that could have prevented a Trump election, even if no-one ever found out about it. I’m pretty sure I would even have accepted such a deal (high personal damage for Trump loss) if it involved me forgetting about it afterward.

  13. Eric Crampton says:

    This is a long-standing puzzle in the public choice literature. I’m not sure that coordination problems solve it though. As you point out, there are a lot of wealthy people who could presumably be made substantially wealthier by appropriate changes in regulation or legislation, and there’s nothing out there that looks like efficient rent-seeking (in which people pay in lobbying and contributions, in aggregate, about the value of the rent). The closest thing I’ve seen to efficient rent-seeking is, strangely enough, in work looking at aggregate effort expended by academics seeking research grants – and I think that’s reasonably well explained by that universities provide a lot of encouragement for academics to pursue grants despite low odds of success.

    Anyway – the Ansolabehene et al paper is a good summary. But if you haven’t seen it before, Denis Mueller’s Public Choice textbook gives a very good summary of the kinds of explanations that have been given for the relatively low amounts of money in politics. Danegeld is a potential one: people may be reluctant to reveal how much they value particular policies if allocated rents are not durable. But it’s hard to come up with good explanations for why there’s so little money in politics – or ones that seem to stand up.

  14. Alethios says:

    My wife and I keep talking about giving more to charity, but never seem to quite get around to it. Part of the reason for this is that, managing our finances, I don’t feel comfortable donating without her input into how her money is allocated. So then we have discussions about how I’d like to donate to such and such a cause, which takes time to explain, and there’s always the question of how much is appropriate. Then after all that, I have to remember to go back and actually take care of the transaction.

    Having read your post, I think a quasi-automatic mechanism is probably the way to go. I think we’ll start with diverting a percentage of our fortnightly savings (i.e. the quantum left in our cheque account at the end of each pay cycle) into a separate donation account, to be allocated 50/50 by each of us to causes we care about at some reasonable interval. Using a percentage of gross income is probably too blunt of an instrument to my mind, as our pay varies reasonably substantially while our outgoings stay relatively stable – so the tendency would be to make changes to the percentages each cycle in order to balance the checkbook, which sort of defeats half the purpose. Savings meanwhile are surplus by definition, so its easy to manage and justify their partial reallocation.

    • Elementaldex says:

      My wife and I automatically spend 2% of our income every month buying an ETF that invests in developing nations. In theory boosting demand for businesses there and maybe doing some good. Then every two or three years we cash that ETF out for a couple thousand dollars and donate it all to something (clean water something something last time). It gives us all the tax advantage in one year. It feels good to do a big thing occasionally. We get to talk about and decide what to donate it to as the month approaches but the amount in question is automatically built up. And it might actually be helpful to have it sitting in developing country stocks in between donations. I like it as a system to donate.

  15. glorkvorn says:

    It’s an issue of diminishing returns. Sharply diminishing returns, because there’s a fixed number of high-level politicians and they’re all very busy people.

    You can’t just “buy a vote”. Even the most corrupt politician isn’t *that* corrupt. The cliche is that money will buy you *access*. A rich guy or PAC donates a million to a congressmen (no strings attached!). Then some time later, he calls up their office and asks, won’t you pretty please take an hour of your time to talk to me after I was so generous to your campaign? Again, no strings attached, he just wants to talk. The congressman says sure, I guess it won’t hurt just to listen.

    The PAC/rich guy brings in a lobbyist. This is a very important meeting so you don’t want just anyone talking to the congressman, you bring in the smoothest operator you can possibly find. As an example, check out the last episode of “The Dream” podcast, where you can hear the lobbyist for the MLM industry in action. His powers of persuasion are amazing. And I assume he’s getting paid top dollar, but he’s still just one guy, maybe a million or two per year, relatively little in the grand scheme of things.

    So what else are you supposed to spend political money on? You can’t bring in an army of lobbyists to talk to a congressman, it’s a one on one meeting. You could donate more money to political campaigns, but it won’t get you much except a few more one-hour meetings. If you go too far you’re just annoying the congressmen by wasting their time. You’ve already got all the slickest political operators in DC working as lobbyists for one side or another.

    The individual political donations are kind of the same thing. It’s one thing to *say* you’re going to vote for Sanders, but donating money, even just $2, is a pretty strong signal that you really are going to vote for him. But donating more money doesn’t change much. I’ve still got just one vote, and he’s not going to take a phone call from me just because I donate $200 or whatever amount I could reasonably afford.

  16. Clutzy says:

    If the collection action problem was really a problem, a candidate could solve it with a kickstarter style pitch of contingent funding.

    You could set it up with any sized goal and then people would know that they only pay if you end up getting payments from a million other people. Or something like that.

    Also, with regards to the larger media outlets, I think they were sold around a time that a lot of people thought they were going to end up as ideological money sinks, and IMO were purchased with self protection as a goal. Carlos Slim bailed out the Times and has essentially bought their silence on his rampant corruption south of the border. Bezos has re-oriented the Post to a more West Coast style of liberalism that has a much greater emphasis on identity politics and a lite socialism instead of the older lunchpail liberalism that was expressed by WaPo stalwarts like EJ Dionne and Eugene Robinson.

    • Sortale says:

      my exact thought when I read the article, can we find come candidate that supports the “less wrong” cause and set up a kick-starter for him?

      It kinda clashes with my prior that politic is mired with corruption, considering how little money is spent here. Maybe they are able to hide more money than we thought [one order of magnitude more]?

      • dogiv says:

        What would such a candidate even do? I think you’ll find that rationalists don’t agree that closely on hot-button political topics, other than avoiding some of the really dumb stuff. And long-term views don’t have that much impact on current government policy.

  17. mtl1882 says:

    It seems like the modern situation isn’t really straightforward enough for just giving money to seem worth it. It disappears into consultants, etc., and you can’t reliably control someone’s vote. Congress hasn’t been known for getting a lot of major changes done in recent times–it may not make a vital difference to have a certain number of seats. Of course, it can, and I’m sure people donate based on this, but I can see that for many people the reward doesn’t seem at all certain, and perhaps other ways of handling it are more reliable or satisfying, such as media pressure or loopholes/workarounds. A lot of campaign promises are vague or impossible. Bills completely change form by the time they become law, to the point of being sometimes pointless or counterproductive or just confusing. Securing someone’s vote once may do little if amendments are made. I just don’t see the concrete interests that call out for money. In the 1800s, it may have paid off to spend a ton of money on organizing rural voters in some states and getting them to the polls or other events. While getting to the polls etc. can still be an issue for some voters today, transportation options and populations are such that this may not be the payoff it once was. Money for ads is probably not as straightforward either, now that there is social media and 24/7 news. The bureaucracy is so big that few people have any real power, and there are a lot more restrictions in hiring, so bribing lower level people connected to government also seems less attractive. Businesses have ways of getting what they want without buying votes, and politicians have ways of getting what they want without taking bribes–they can go work for the industry afterward. It just seems like there are other ways of going about it that seem more preferable to many people.

  18. len says:

    I’m surprised that nobody has pointed this out yet, but it’s not how “loud” or influential you are that matters, but how loud and influential you are relative to your competitors. There’s a limited pool of attention you can capture, and raising your bid may only cause your opponents to raise their bids accordingly.

    If everyone spending $2 for Bernie spends $3 instead, plausibly supporters of other candidates sees this (e.g. through campaign finance records, or through the other candidates being more aggressive in asking for donations) and donates 50% more as well, and you return to the status quo.

    • Juanita del Valle says:

      This doesn’t answer the question: what you’re positing is that the coordination problem has been solved, with everyone settling on a Nash equilibrium of spending little money. What stops the amount of money being spent increasing?

      • len says:

        Even assuming no coordination problems, there is a the finite amount of attention you can get, and the diminishing returns on how much loudness helps. It’s obvious if you think about advertising. Doubling your advertising budget does not double your rate of return, and the marginal return on advertising quickly go to 0 or negative if you advertise too much. Once all potential customers (or swing voters, or legislators, or senators) knows about your product (or platform, or recommended industrial policy), spending more doesn’t help.

        Tl;dr: Equilibrum is at spending little money because marginal return on outreach spending drops really quickly.

  19. OxytocinLove says:

    “Americans could end homelessness by redirecting the money they spend on Christmas decorations” also means the US government could end homelessness by raising taxes by that amount. I definitely think charitable donations should be a part of how we solve large-scale problems but “homelessness in the governed country” seems like a reasonable problem to expect the government to solve, and it sort of exists to solve the kinds of coordination problems you’re talking about. What I’m saying is that all the arguments in this post seem to me like better arguments for socialism and welfare than for more charitable giving.

  20. stopandgo says:

    1. $8.5 billion to influence maybe a couple thousand people is several million dollars per decision-maker. Not an insignificant amount of money.

    2. The relevant comparison point is advertising spending, not sales of a physical product (and that was sales of almonds you quoted, not “earnings”). Total US ad spending is ~$200 billion per year. Meaning that influencing decision-makers at just the federal level makes up multiple percentage points of all advertising.

    (The actual figure for almond advertising is around $100 million per year, putting political advertising around two orders of magnitude higher than the spend on influencing almond purchasing decisions.)

    3. “These properties seem to be priced entirely as cash cows – based on their ability to make money through subscriptions or ads. The extra value of using them for political influence seems to be priced around zero, and this price seems to be correct based on how little money is spent on political causes.” — You really believe people like Jeff Bezos or Carlos Slim take major stakes in newspapers for their cash flows and that their ownership does not impact what is printed about them?

    4. “They all have one thing in common: paywalls. How could this be efficient? There are millions of people who follow all of these philosophies and really want to spread them.” — because people assign more credibility to things they (more) pay for. Extraordinarily well-documented in behavioral psychology and behavioral economics.

    5. “By all accounts the Koch brothers actually believed in everything they were doing, and they had to, because you couldn’t make billionaires spend Koch-brothers-like levels of time and money out of self-interest.” — They spent only about 1% of their combined family net worth on politics. It doesn’t take having much effect in highly regulated industries to make that very smart spending.

    6. Notice the trend here: most of the spending is by owners of closely-held businesses who are often in highly regulated industries (including avoiding antitrust enforcement). Clearly being a public company with diffuse ownership, barely aligned managers, and shareholders and board members with cross-cutting views tends to half political spending. If you’re looking for a “why”–this is it. Only a few business are both privately held by aligned managers and in industries or situations where political influence would produce much of a return.

    • len says:

      4. Also because it’s a lot easier to collect money from True Believers who can be convinced that by buying a subscription, they are Contributing to the Cause.

    • keaswaran says:

      Are you sure advertising spending is the relevant comparison? A very large amount of campaign spending is on turnout and organization efforts. I’m not sure exactly what industry to compare that to, but it’s not obviously advertising.

      • dogiv says:

        I think the parts that aren’t exactly advertising still correspond to other parts of “sales” expenditures, like door-to-door sales or telemarketing. Maybe actually driving people to the polls wouldn’t, but that’s only ever needed by a few voters.

    • peterispaikens says:

      The disagreement about #2 is about a fundamentally different perspective of that money – you seem to be targeting the consumption side i.e. the budget of influencing almond buyers vs the budget of influencing voters; however, this article is not about the perspective of a candidate or party paying money to get elected, it’s about the perspective from the *buyer* of the whole package which includes not only buying votes but also “buying” candidates who are going to be elected anyway.

      People in USA are willing to pay $36 annually per capita to the almond industry to get an occasional tasty snack.

      People in USA, including all the corporations and billionaires, are not willing to pay as much (i.e. $36 annually per capita) to influence who gets elected and what laws get passed. So, whatever they say, the influence that can be bought apparently is less valuable than a single small sub-group of snacks. How come?

  21. Steve Sailer says:

    The last time I was in Mexico I was struck by the wall-to-wall political advertising. Rich people take politics seriously in Mexico.

    Here’s Andres Oppenheimer’s account of the 1993 Billionaire’s Banquet in which Mexican president Carlos Salinas asked 30 zillionaires to whom he had sold government businesses to donate $25 million each to the PRI election fund for 1994.

    Then he made clear he expected not 25 million pesos each but 25 million U.S. dollars each (or $750 million from one fundraising event):

    “”Mexican pesos or dollars?” one of the billionaire guests asked. “Dollars,” responded Hernandez and Borrego, almost in chorus. Twenty-five million dollars each? ! There were hmms and ahhs around the table. Don Garza Sada, of Monterrey’s Visa soft drinks empire, said he agreed — it was the business community’s responsibility to support the party Telecommunications magnate [Carlos] Slim, who had won the government bid to privatize the national telephone monopoly, supported the motion, adding only that he wished the funds had been collected privately, rather than at a dinner, because publicity over the banquet could “turn into a political scandal.””

  22. Thomas says:

    People may not much care who wins, but they sure do care who loses. I sure felt a thrill when HC lost and even now have moments of pleasure at the thought.

  23. sentientbeings says:

    This one is definitely going into my pile of go-to posts to share with the SSC-uninitiated.

  24. Steve Sailer says:

    My impression is that more money is given by rich guys to college sports than to politics. E.g., the LA Times just reported that USC had gotten $400 million over the years from B. Wayne Hughes, founder of Public Storage.

    O.J. Simpson’s purported 1994 suicide note during the White Bronco Chase thanked Hughes, who may or may not have paid for the subsequent Dream Team defense lawyers.

    Other 9 figure donors to college football programs include Phil Knight (U. of Oregon), T. Boone Pickens (Oklahoma St.), and John Arrillaga (Stanford).

  25. Juanita del Valle says:

    The classic coordination problem is one where you only get the end product if everyone can force themselves to commit to contributing.

    This is not true of homelessness: if Alice pays $50 to buy one night of shelter for one homeless person, that benefit is still achieved regardless of Bob decides to pay $50 to sponsor an additional night.

    It might be that there are economies of scale – i.e. the setup costs of establishing a charity for homeless people will eat up the first $X millions donated, so the marginal benefit is zero until you reach a certain threshold – but in the real world such charities (as well as various other mechanisms) already exist, so that is an unlikely explanation.

    • Bugmaster says:

      if Alice pays $50 to buy one night of shelter for one homeless person

      Is that actually possible ? As I said below, when I buy Christmas lights for $50, I verifiably gain $50 worth of Christmas lights. I’m not sure this is the case with homeless shelters.

      • Juanita del Valle says:

        Agreed, the example is abstract. But again, if the answer is that there’s simply no effective way to make a difference to homelessness, that is still not a coordination problem.

    • keaswaran says:

      Presumably a lot of what Alice wants out of homelessness spending is the benefit of having a city free of homeless people. If it takes $5 million to get that, and 100,000 people are all willing to pay $50 to have the benefits of a city free of homeless people, they could do it if they could coordinate. But if each has to make their choices independently, then this is a classic Prisoners Dilemma, like any public good. There’s no way to ensure that only the people who chipped in get the benefits of a city free of homelessness.

      Even if you think people are motivated entirely by the altruistic goal of helping the homeless people, there’s still the problem of free-ridership. The homeless people mostly get helped even if I don’t chip in, so why should I bother if everyone else is already taking care of it? And if the others aren’t, then the homeless people basically don’t get helped even if I chip in, so why should I bother?

      • Juanita del Valle says:

        I don’t think many people believe that homelessness is only worth fixing if you can fix it completely. Even if you only care about homelessness because you think the homeless are criminals, a 50% reduction in homeless people implies (speaking very roughly) a 50% reduction in criminal activity.

        The situation put forward is one where no-one else is contributing, not one where everybody except Alice already is.

        So suppose Bob is not contributing because he just doesn’t care. Alice however is altruistic, and initially believes it would be possible to make a marginal reduction in homelessness below the cost of what such a reduction is worth to her. As you say however, it’s conceivable that she might feel that there’s no need to bother if Bob doesn’t care. There are several possible interpretations here, but one is not that Alice still wants to contribute (she could, after all, still make a difference), but that her preferences themselves have changed: Bob’s indifference is social information that affects what she cares about. This is not a classic coordination problem: if you offered a binding contract to Alice and Bob that each will donate $50 if the other does, Bob will not accept, and indeed now neither will Alice.

    • Jan_Rzymkowski says:

      This is exactly, what I wanted to comment. The problems Scott’s describes are not coordination problems. In those, the agent’s reward depends on the choice of the other agents. But in most of described cases there either is no reward to the agent anyway or even if there is, it’s still better to cop out EVEN if there is a capability to communicate and make commitments.
      Giving money to solve homelessness, to politics or just simply voting are all examples of charitable action, where the agent gets practically nothing but does it for the greater total utility.
      The actual motive that connects these cases is that the cost leads to achieving *a very tiny improvement for a very large number of people*. People don’t redirect money to homeless charities not because of lack of coordination but because it’s _probably helping in some very small way some large number of homeless people_ which is really ethereal compared to tangibly enjoying Christmas decoration.

      • Uncorrelated says:

        This is also what I was thinking as I read. The coordination problem is a good fit for elections where if you try to get a particular politician elected and you fail, you get no benefit at all. Charity, at least if it is well thought out, can get proportional results. Spend a little and help one person, spend more and help two, …

  26. Scratch says:

    I’m guessing the bottomless well of lucrative and prestigious sinecures for retired politicians of all stripes might be the fundamental (and still dirt cheap) method of securing compliance.

    Even the £60m someone put in the purportedly “socialist” Tony Blair’s pocket for instance is remarkably cheap for a decade’s worth of governmental not scaring the horses.

    • Steve Sailer says:

      Right, it would be good idea to try to quantify other streams of money into politics besides campaign donations, such as how much Bill and Hill made from speaking fees, book advances, and donations to the Clinton Foundation.

      • Steve Sailer says:

        I’m struck by how much members of the public want to shower money on their party’s ex-Presidents by buying their dull books. For example, Bill Clinton made $29.6 million from his post-Presidential memoir, but I only made it through his first sentence:

        “Early on the morning of August 19, 1946, I was born under a clear sky after a violent summer storm to a widowed mother in the Julia Chester Hospital in Hope, a town of about six thousand in southwest Arkansas, thirty-three miles east of the Texas border at Texarkana.”

        Marco Rubio made something like $750k off his book. Without it, he’d be quite poor because he has a lot of kids. With it, he can afford, barely, to have a fishing boat (although it’s nothing special by Florida standards).

        I really think federal politicians should be paid more lavishly.

        • EchoChaos says:

          Is there any way to see how much of those book deals are actually recouped by the publisher and how much them are lavish advances that don’t pan out?

          I suppose that’s proprietary within the publishers, but such books seem to always hit the clearance shelves.

          • Steve Sailer says:

            The fictional thriller “The Ghost Writer’ by Robert Harris implied that Tony Blair was being paid off by Rupert Murdoch via a $10 million advance for the Blair character’s memoirs from a publishing company controlled by the Murdochish character.

            On the other hand, a lot of regular people really do buy these memoirs. Michelle Obama’s recent book is so popular that she’s doing a tour so readers can pay to worship her in person in hockey arenas.

        • Jaskologist says:

          How much of this is really just a way to launder campaign money around to the candidate? Sanders spent $440 of his campaign’s money buying his own book. In light of that kind of deal, the big advances from the publisher make more sense.

          • Douglas Knight says:

            How does this laundry work? Sanders spent $444k on his 1997 book and then a different publisher gave him a $500k advance. That new book was the best seller, so that was a good investment and with the book published 11/2016, I don’t think he spent campaign money on it.

            As for Echo’s question, Hillary’s 2008 tax returns show that Bill and Hillary both made royalties, not just advances, so the books sold well and the advances were well calibrated. Bill made $23m, including $15m advanced and Hillary $10m, including $8m advanced.

            Most of their money is from Bill giving speeches. He gave a lot of speeches and they seem to be market rate. Before the returns were released, I had heard that he gave 200 speeches a year and that $50k was a typical price, but I hadn’t multiplied them together to get $10m/year. People worth bribing don’t seem to charge more than ordinary celebrities.

  27. Krisztian says:

    It’s not a coordination problem.

    Elon Musk doesn’t have to coordinate with anyone — as you mentioned, he alone could outspend the system. More likely: political spending has no (or such a tiny) effect that it makes it a bad investment (both privately and socially). Musk gets more done spending a billion on space exploration than spending a billion to lobby for space exploration.

    How about the Bernie supporters? I think they are loud for tribal signaling reasons and ‘feeling-righteous-anger-feels-good’ reasons, but otherwise they don’t care. The lack of donations is not the only evidence. Another one: as soon as you offer them a bet on their (supposedly strong) beliefs, the loud support disappears. Also: while not solving the free-rider problem, a kickstarter campaign would solve the coordination problem.

    Otherwise good piece. It bothers me that people don’t talk about this enough, and think money controls politics.

    • Freddie deBoer says:

      “How about the Bernie supporters? I think they are loud for tribal signaling reasons and ‘feeling-righteous-anger-feels-good’ reasons, but otherwise they don’t care. The lack of donations is not the only evidence”

      Wow, what an objective take, backed up with compelling evidence.

      • axiomsofdominion says:

        Well Slate Star Codex commentators trashing Bernie and his supporters isn’t exactly news Freddie.

        • Krisztian says:

          Sorry, didn’t mean to offend your tribe. Only used Elon Musk and Bernie supporters as those were the examples in Scott’s post.

          Just to clear things up, the same reasons apply to other billionaires, and other grassroots movements. (Heavy rhetorics, near zero material support)

  28. Robert Jones says:

    But it would also have to be true that lobbying, the NRA, the media, etc don’t affect politics very much, which seems like a harder sell.

    Doesn’t seem like a very hard sell to me. I’m pretty certain I’ve seen things on this very blog showing that lobbying has no effect. Since lobbying is basically just asking politicians nicely to do what you want, and since politicans usually have their own incentives to act in one way or another, it doesn’t seem likely to have much of an effect, save perhaps on non-contentious issues. In my model, people spend money on lobbying for much the same reason they spend money on acupuncture: they don’t really have much expectation it will work, but they think it worth punting a few dollars on the off chance.

    Similarly I’m pretty sure that partisan media are following rather than leading their audience.

    I’m not sure why the NRA merits a special mention. I don’t know much about it, but I doubt that it has a unique ability to influence politics. It may be particularly good at lobbying, but that’s like being particularly good at acupuncture.

    I don’t think the Sanders supporters really pose a great problem for the model. Like everyone else, there’s a divergence between their stated beliefs and their actions. They may say that they believe money affects politics, but at some level they know this is wrong, and they act in accordance with this knowledge.

    • gkai says:

      But does lobbying really works like that (i.e. asking more or less nicely politicians to do something)?
      I think it does not, at least that’s not the part of lobbying that has the most likely effectiveness…
      It’s more like a coordination effect: an efficient lobby is just a group of people coordinated enough that politicians know the group official position will control the votes of the members. Basically the vote of many of the members will be determined by a very specific subset of political program/promises. That’s imho is the real power of NRA (or other NRA-like things, left or right leaning), a coordinated mass-vote on a focussed subject. The power does not come from any lobbying or message relayed by the media, so the lobby do not really have to spend money of lobbying activities, except make his opinion on a policy public…
      But having many paying member is good for the lobby: a paying member is more likly to align vote than a non-invested sympathisant…So paying members gives weight, and you don’t really have to spend. Seems like beeing part of a lobby organisation with many paying members is a good position, especially if earning a living from a pretend-to-work job is not bothering you 😉

      • keaswaran says:

        The NRA doesn’t need very much money to demonstrate that there’s a big group of organized voters behind it, and doesn’t need to spend very much to organize those voters. A couple million or a couple hundred million will do nearly equally, in terms of mobilizing voters. Marginal votes matter. But a few marginal million dollars don’t.

        • gkai says:

          Yes, that’s my point. I don’t know how much (if any) you need to pay to get affiliated to the NRA, i’m european so not really aware of how it works in practice.
          But the point is that just having (a lot of) paying members makes you efficient at lobbying, without having to spend this money on any lobbying action.
          That’s free money to burn, which lead to an inflated organisation with a lot of pretend-jobs politically obtained…I mean you get your job by internal political manoeuvers, not by doing anything measurable…

        • John Schilling says:

          NRA membership is $45/year for basically a magazine subscription and web site access. The magazine/web site will, among other things, tell you who to vote for if you’re a single-issue voter on the subject of gun control. And encourage you to keep being a single-issue voter, if there’s any waffling on that point, but mostly people are committed by the time they sign up.

          But that’s basically it. The NRA’s value proposition to its members is, they get to hang out at least virtually with other members of “gun culture”, and they don’t have to do the legwork to figure out whether that Democratic candidate who is saying the right things about the union can be trusted not to take away their AR-15. Their value proposition to politicians is, they can skip the money and directly deliver a couple million votes on demand. Coordinating this requires roughly the cost of a magazine subscription per vote.

          It’s not quite unique, but it is a very different model than the usual money->legislation model of political lobbying.

    • Enkidum says:

      I’m not sure why the NRA merits a special mention. I don’t know much about it, but I doubt that it has a unique ability to influence politics. It may be particularly good at lobbying, but that’s like being particularly good at acupuncture.

      So, my beliefs about the NRA are definitely formed by the bubble I happen to have spent most of the past twenty years in (left-wing, highly-educated, non gun owner, folk music, etc), and I don’t trust them that much for that reason. But the dominant narrative is that the NRA is one of the most successful lobbying organizations that has ever existed, that has hugely influenced politicians’ interpretations of the second amendment and directly affected their willingness to propose and/or vote for legislation concerning firearms. And that over the past, say, thirty years, these interpretations have become more extreme as a direct result of this lobbying.

      Is this narrative completely false? I’m trying to avoid making any claims about the morality of the positions the NRA takes.

      • EchoChaos says:

        Not completely, but somewhat. The NRA’s power comes from the fact that gun owners are disproportionately likely to be single-issue voters. For example, Democrats controlled the House and the Senate for 40 years until they passed an Assault Weapons Ban and then got slaughtered in 1994. The NRA using its megaphone to mobilize all those single issue voters was a major reason for that.

        But the NRA is downstream from the voters, not the other way around.

        • Enkidum says:

          But they weren’t single issue voters before the NRA mobilized them, were they?

          • EchoChaos says:

            Yes, they were. That’s why the NRA was so strong. Getting single issue voters out to vote does need a bit of a megaphone and mobilization because single issue doesn’t mean the same thing as very engaged.

            The recent move of gun rights to being Republican is just that, recent. Pro-gun Democrats also had the NRA’s endorsement quite often, and the NRA’s rules are that if they have two equally pro-gun candidates they will always endorse the incumbent, so most of their pre-1994 endorsements were of Democrats.

          • Enkidum says:

            Thanks for the clarification.

        • gkai says:

          Yep, I think we fully agree: efficient lobbying comes from demonstrating vote control. It has nothing to do with spending money on lobbying action. But having members willing to pay is a good signal that the organisation indeed control the vote of their members.
          Grouping of single-issue voters (who have demonstrated to really be single issue) also have great power, on this issue and anything they can link to this issue.
          A single voice demonstratingly representing many votes is strongly heard by any politician, seems logical. And probbaly one of the only two way to do efficient lobbying. The other is bribe/threat but it’s illegal almost everywhere.

        • Anthony says:

          I think NARAL (the abortion lobbying group) is a similar case. Abortion is one of the few issues other than guns which make single-issue voters. My in-laws in $MidwesternState mostly vote Republican, even though their overall policy preferences lean more Democrat, because the Dems are almost always strongly pro-choice.

      • Garrett says:

        As an aside to what others have said, the NRA has the advantage of simplicity around the subject they are dealing with. For pretty much all gun related things, there’s a straight-forward answer of what the “pro-gun” position is. At most, there is a different prioritization.

        With other subjects, it’s more complicated. For example, is the “pro-environment” position for or against nuclear power in the US? Is the “pro-labor” position for or against more workplace safety regulation?

        • gkai says:

          Here in Europe the green NGO may have the same type of power, I suspect a majority of their members (and probaly a good part of the sympathisants) would vote as greenpeace/gaia/peta/… would tell.
          But they are not as united under a single head, and they went another way: the different green parties. I suspect a single green association/ONG would have had bigger political power than the green parties…
          I think driver associations, senior associations and such may reach NRA-like success. Trick is not to split into competing associations, avoid the trap of a direct political party, and keep the single message for multiple election cycles.

  29. Frederic Mari says:

    “In his paper on elections, Ansolabehere focuses on the corporate perspective. He argues that money neither makes a candidate much more likely to win, nor buys much influence with a candidate who does win. Corporations know this, which is why they don’t bother spending more. Most research (plus the 2016 results) confirms that money has little effect on victory, so maybe this is true. But it would also have to be true that lobbying, the NRA, the media, etc don’t affect politics very much, which seems like a harder sell”.

    I accept the results of the studies showing money does not make the winner (but flows to the likely winner). Okay.

    I don’t quite believe that lobbying has no impact.

    1- the US seems to have pretty terrible regulations, favorable to businesses compared to the rest of the world (fracking being an easy example). Some of it might be a genuine reflection of the American voters’ character but all of it?

    2- Given that the big corporations the US has are, by definition, the winners in this specific set up of the competitive environment we have (we could have other set ups with other winners), their spending is justified even if it only helps maintain the status quo. They’re the status quo winners. Not wanting the environment to change up on them and screw them up is motivation enough for lobbying/corruption.

  30. Kuiperdolin says:

    Gotta point out that some of that sweet almond cash comes from exports (Wikipedia says 4,5 billions, so about a third). Foreigners are barred/discouraged from pouring that much into the election (I think I read a few articles about some Eastern European country weighing in on the last election, but it’s hard to put a monetary value on what they did). Foreign lobbying is not illegal, but there are significant hurdles. As far as I know (little) there is no restriction on foreign thin-tankery except optics.

    So overall, not a fair comparison.

    How much the rest of the world would pour into US elections if the doors were wide open is an interesting question. My guess is a lot.

  31. Robert Jones says:

    As others have said, this isn’t a co-ordination problem. Homelessness isn’t an all or nothing problem. I can’t individually solve homelessness, but I could help a homeless person. If I give $100 to a homeless shelter, the impact of my donation isn’t lessened because others fail to do the same. If anything, it’s increased.

    Similarly for politics: if money did affect politics (which it doesn’t) then the failure of other people to donate would only increase the impact of my own donation.

    • keaswaran says:

      It depends on what you think the value of decreased homelessness is. If it’s the public good of living in a society with less homelessness (which is a good both because there will be more productive members of society doing productive things, and because you won’t personally have to encounter certain types of mess or disturbance while walking around the city) then this is a benefit that individuals get regardless of whether or not they pay, and from which it’s impossible to exclude them. This is the standard setting for coordination problems. No one voluntarily chips in for a military or a public fire department, because you get the benefits whether you participate or not.

  32. Long Disc says:

    Lack of coordination might explain a large under-investment only if there is a large return to scale.

    In politics, it is a plausible explanation. In a simplified form, politics is a zero-sum game between two adversaries, so if money could have a large impact on the likelihood of winning, coordination would be important. The voting mechanism of translating a 51:49 vote advantage to a 1:0 representation advantage is a pretty powerful return to scale amplifier. However, zero-sum nature of the game means that a large-scale durable effort from one side to coordinate could be counter-balanced by a similar counter-effort, so both sides would end up with little changed outcome but both spending much more money. That would be obviously detrimental to both sets of partisans, so what looks like too little coordination within one party might be actually a manifestation of a society-wide coordination of not starting a costly propaganda arms race.

    With homelessness, it is hard to see the source of return to scale. If anything, I would expect a negative return to scale, as it should be much easier to help some homeless people than other. So I do not see why somebody who cares about actual homeless people and not about existence of homelessness would be deterred from contributing by a large scale coordination problem. Even a small number of contributors with $100 each might help a homeless person or two.

    • Rana Dexsin says:

      As someone who has not studied the problem in any detail (and might be partially representative of the unfocused masses?), and also as someone who’s currently in a potential downward spiral myself, my intuition says the opposite, that donating too-small amounts of money to homelessness-as-a-cause won’t do very much for the homeless people in question. The short explanation is “heavy dilution enforced by fairness politics, combined with sigmoid-shaped effect curves due to stability phase transition activation energy”; I can try to expand on that if people want, but I didn’t want to get too OT.

    • peterispaikens says:

      The coordination problem regarding homelessness comes from the differing intepretation about what “the problem of homelessness” means. I most certainly agree to your point regarding someone who cares about actual homeless people, but I’d argue that it’s not really relevant because such people are a small minority, and are very much outnumbered by the people who really do care mostly about the existence of homelessness as much as it affects the city they live in.

      So for any actual discussion of policy change and funding that we can probably ignore the fraction of citizens who actually care about homeless people as individuals (as you say, they can and will contribute anyway), and any policy proposal should (must!) target the people who don’t, as recruting their cooperation is required to make a major difference.

      If voters have the option to voluntarily, individually pay $50 and get some housing for a particular homeless person, then altruistic Alice is willing to choose that, but non-altruistic Bob isn’t – because that $50 isn’t really giving him anything as valuable as $50. Now, if we had some policy where most people would be paying $50 and that would result in a substantial reduction in homelessness, large enough to make a noticeable different in quality of life for Bob (as opposed to particular homeless people), then that is something for which Bob (and many like him – IMHO far outnumbering altruistic Alices) would be willing to pay $50 – so solving the coordination problem can enable such policies; as “Bobs” will fund such policies if and only if they achieve (or show credible potential that they will achieve) a critical mass of homelessness reduction that makes a difference for the non-homeless population.

      However, I see no reason why this should be substantially different to the related coordination problem for other public improvements such as building and maintaining public parks, providing emergency services, etc.

  33. janrandom says:

    If it is a coordination problem it would be relatively easy to solve e.g. the homeless crisis with a Kickstarter project. Only pay if the project gets funded. Where is the catch?

    PS. Now I worry that people will get powerful political campaigns funded on Kickstarter…

  34. sarth says:

    When I read this essay the first thing that jumped into my mind was “revealed preference.”

    Maybe it’s because for most people their political stance is identity and signaling, and it is unnecessary to spend money to achieve either of those.

    Which means … maybe if a political movement could create a situation where you needed to spend money to get social status as a supporter, they would be drowning in individual contributions.

  35. Anon. says:

    The Tullock Paradox.

  36. Ninety-Three says:

    So when I hear stories like that Americans could end homelessness by redirecting the money they spend on Christmas decorations, I don’t think that’s because they’re evil or hypocritical or don’t really care about the issue. I think they would if they could but the coordination problem gets in the way.

    But there’s no coordination problem unless donating to the homeless has increasing returns, where 100 million people giving $100 produces significantly more than 100 million times the value of one person giving $100. If there are no increasing returns, then people really are selfish hypocrites because they’re still choosing almonds and Christmas decorations over helping the homelessness. Giving $100 without everyone else doing so doesn’t feel like a drop in the ocean because everyone else isn’t donating, it feels that way because it is a drop in the ocean, solving one hundred-millionth of the problem.

  37. Corey says:

    Crap. I’d figured that if automation ever ate up all other avenues for work, we could have an economy where we all worked for political campaigns. Sounds like that doesn’t scale enough.

  38. Bugmaster says:

    In this model, the difference between politics and almonds is that if you spend $2 on almonds, you get $2 worth of almonds. In politics, if you spend $2 on Bernie Sanders, you get nothing, unless millions of other people also spend their $2 on him.

    That sounds plausible; but another possibility is uncertainty, as well as the lack of immediate feedback. You say:

    …Americans could end homelessness by redirecting the money they spend on Christmas decorations…

    But, when I buy a string of Christmas lights for $50 (or however much they cost), I know exactly what I’m getting: Christmas lights. Sure, it’s possible that my package would be lost in transit, or that the lights would be defective, etc., but the chances of that happening are small. Furthermore, when I do buy those lights, I get to hold them in my hand, immediately.

    When I donate to some charity to help the homeless, there’s no way for me to know that the $50 I’d donated would actually help any homeless. At best, it would help maybe 0.01% of a single homeless person, in some vague and ineffable way (unless I picked the wrong charity) . Will it pay for an apartment ? Food ? Mental treatment ? Drug rehab ? Blankets ? What’s the most effective way to allocate this spending ? No one knows, and that’s not even counting the potential knock-on effects, such as inadvertently creating more homeless in the future. None of these issues are immediately apparent with Christmas lights. What’s worse, there’s no way for me to discover any of these effects until years, if not decades, in the future. With Christmas lights, at least, what you see is what you get.

  39. Purplehermann says:

    There’s a voter turnout around 60% percent for presidential elections- more than a third of the US can’t be bothered to vote. That’s despite it being expected socially, voting being seen by a lot of people as the important part that civilians take in politics (“so you should care!!”) and as the normal thing to do.

    It seems fair to me to say that politics is like sports. People care about their team winning, but they don’t care about it like they care about their money.

  40. jonmarcus says:

    Is this just a coordination problem? It feels like it’s also something of an…well, for lack of a better term, an anti-coordination problem. If I donate a million dollars to push lefty principals, that’s not especially likely to get me the results I seek (because of the coordination problem) but it’s ALSO quite likely to convince my friend on the right to make an equal donation, pushing back against my attempts.

    If my contribution is unlikely to spur political change and is likely to start a bidding war, it’s not unreasonable to conclude I’d be better off filling my swimming pool with almond milk.

  41. Andrew Priest says:

    ” Most research (plus the 2016 results) confirms that money has little effect on victory, so maybe this is true. But it would also have to be true that lobbying, the NRA, the media, etc don’t affect politics very much, which seems like a harder sell. ”

    I think this implied proof by contradiction is faulty. The NRA is particularly a poor example of this proof by contradiction.

    The pro-2A movement doesn’t believe its power comes from money influence. Looking at its lobbying and donations, in terms of the political ecosystem, they are a 2-bit player.
    Politically speaking, king of the hill by a long shot is the Union and Corporate interests. Small arms manufacturers are small economically speaking, new gun revenue is estimated to be ~$4B (1/3 the size of the almond industry 😀 ). And that revenue is going to include gov’t police and military sales.

    Instead, the NRA and local organizations derive power from their ability to move a large portion of voters with their propaganda, their very outspoken defenders in the public space, and the ability to mobilize large groups.

    The “Assault Weapon Ban” backlash leading to the 1994 Republican victory in Senate and House is still a relevant cautionary tale.

    I’ve seen first hand the ISRA [Illinois State Rifle Assoc] able to mobilize 8000+ to march at their own cost, in a late winter snow storm (March 6th, 2013) for their political position on Springfield’s streets and to invade their representatives office in the capital building.
    A few weeks and warmer weather later (April 11th), a gun control group had funding from the Governor to bus people from Chicago for free, and were only able to get 200 people to show up to their Springfield rally.

    Agreement with this assessment from the Democrat and gun control allied NYT:

    • Etoile says:

      While one could also probably find a left-wing counterpart to the above example, I think this is a very good one of what’s missing in the estimate.

  42. Freddie deBoer says:

    This post is a remarkably efficient argument for confiscatory tax rates that doesn’t seem to realize it is one.

    • Sniffnoy says:

      Er, how? I’m not really seeing it.

      …OK, best I can think of is that, well, government can solve coordination problems given the money! Is that what you intended? That seems a fairly weak argument; yes, the government can solve such coordination problems, but is it likely to do that, rather, than, say, buying more weapons?

      Perhaps the bigger problem with the argument is that it’s hardly the only solution; I’m surprised the post made no mention of assurance contracts. Then again, assurance contracts might just fail because people aren’t motivated enough. Would dominant assurance contracts help here? Dunno, hard to say when it comes to attentional/motivational issues.

      In any case would you mind being more explicit about your argument? Because it’s not very clear.

      • JASSCC says:

        “Dominant assurance contracts” was practically my first thought on reading this. I think the trouble with this solution is that it requires an infrastructure — experienced underwriters, escrow agencies, marketers and markets, etc. — that simply doesn’t exist yet. Who coordinates the coordinators? … is sort of the question. But I think such a thing can and (if I had to bet I’d wager that it) probably will emerge and make a huge difference for these collective action problems.

      • Freddie deBoer says:

        Sure. So I reject the “it would only cost $X to solve Problem!” canard pretty much always, because we know that we spend a lot to solve a lot of problems and they consistently go unsolved. They are almost always generated by taking a small scale solution and assuming it will scale up at a linear cost. But the variable of scale is the only variable that matters. The fact that you can cure 10 people of a disease with a new medication does not mean that you can necessarily scale up that solution to cure 100,000. Anyhow.

        If the problem is a coordination problem, then the government can use its power to tax to take the money that people might have spent solving Problem and ensure that that money does in fact go to solving Problem. Government eliminates the free rider problem by taxing everybody.

        Of course we try this, and must go on trying it, and it usually doesn’t work. Because, again, “we could cure scurvy for the cost of one city bus” estimates are bunkum.

        • Sniffnoy says:

          OK, so your argument does indeed seem to be the one I already responded to. It’s not clear why you say “we must go on trying this” when you say it usually doesn’t work and there may be better solutions — I mean, OK, sure, continue trying it, absolutely, but maybe it doesn’t deserve as much emphasis as newer, less-tried mechanisms that could potentially work better?

          You also haven’t responded to the problem of, having raised taxes, how do you get the government to actually spend it on things like these? I mean, I guess it’s still an argument in favor of high taxes, but it’s not an argument for high taxes by themselves, since having raised taxes the government is unlikely to actually undertake much of this. It could be an argument for a multi-part plan of which raised taxes are a part, however. But it’s worth keeping in mind the distinction between “we should do X” and “we should do Y (which would include X)” and etc.

        • Douglas Knight says:

          By “the post” you seem to mean the conclusion of the post. Yes, if charity is a collective action problem, that is an argument for government intervention. But the post is a poor argument for its conclusion. It really has nothing to do with coordination until it gets to populist candidates. The switch to charity is an even bigger jump.

        • Reasoner says:

          They are almost always generated by taking a small scale solution and assuming it will scale up at a linear cost. But the variable of scale is the only variable that matters. The fact that you can cure 10 people of a disease with a new medication does not mean that you can necessarily scale up that solution to cure 100,000.

          Usually in economics the assumption is that the cost is sublinear due to economies of scale. It’s cheaper to buy a pencil from the store than it is to go out in the woods and assemble one yourself from scratch. Thoughts?

          • Freddie deBoer says:

            I guess philanthropy just isn’t the same as a pencil? In large measure because you’re multiplying human/personnel costs and that’s a taller order than production costs.

          • zzzzort says:

            Dunno, supply curves tend to slope upwards in most models.

        • Anthony says:

          You’re assuming away one part of the coordination problem: the government’s action.

          Increasing taxes, even if a sufficiently large number of people agree, does not mean that the people in charge of implementing government policy will actually implement the desired policy. And it’s very hard to force things to happen that way.

  43. Nobody’s going to cynically declare “Oh, people don’t really care who wins the election, they just pretend to.”

    Oh? I am fairly sure this is the mostly correct explanation. Of course it depends on what you mean by “care”. People do actually feel bad about their party not winning. But a couple of bad feelings are not worth spending millions on. People’s lives are not, in fact, changed that much one way or another by who wins, and therefore they do not, in fact, care very much who wins.

  44. Etoile says:

    My gut feeling is, something not being accounted for – namely, the close relationships between government and industry/interest groups that arise organically, possibly AIDED by someone’s evil will, but not engineered by it. Examples:

    -The revolving door of industry and regulators – e.g. an FDA bureaucrat going on to become Director of Regulatory Relations at EvilPharm, Inc., or insurance companies suggesting insurance regulations. These aren’t 100% bad things – you don’t want to write regulation completely in the dark without hearing what the people it affects have to say. (E.g., the state of NJ in the early 2000s invited auto insurance companies to suggest improvements to its auto insurance laws after decades of no major company willing to do business there, and brought it back to sanity…. maybe someone thinks it wasn’t for the best, but there’s an argument to be made that it was.)
    Maybe a rough way to estimate this is to find out how many executives are former regulators at the top 100 US corporations, and the value of their compensation?

    -The relationships of agencies and their long-time contractors. The incumbents do a lot of not-illegal “opportunity shaping” – networking with the government agency, filing briefs why the agency should be SUPER-RISK-AVERSE in changing vendors, etc. Maybe as a proxy, you could add Booz Allen, Lockheed, etc.’s business development budgets to your estimate.

    -This one is the hardest/iffiest, but: when the “money” isn’t money so much as ideological commitment. Let’s say, for example, that all or most lawyers, public policy graduates, journalists, and other graduates of schools bound for government or media jobs tend to have the same ideological outlook, then perhaps the cost of their educations could be added to “money in politics”?

    All of these are huge amounts of resources that you don’t see with the straightforward estimate above.

    • teageegeepea says:

      The “revolving door” actually results in regulators working hard in office to demonstrate their value to prospective employers. I can’t find the exact NBER paper I read that in now, but others do indicate that the results don’t match up with a “quid pro quo” explanation:

      • Douglas Knight says:

        That’s the right paper. Matt Levine quotes it.

      • Etoile says:

        Right – perhaps I should have made that clearer, but I don’t think it’s an explicit quid pro quo situation; nonetheless, it can definitely be considered an avenue by which moneyed interests influence politics, and which you can’t count by just adding up lobbying and campaign contribution dollars.

      • Anthony says:

        But the revolving door pays off even better for elected officials, and unless the elected official is a complete bozo, there’s going to be someone willing to reward him with a job.

  45. belvarine says:

    You’re absolutely right, we do spend way too much money on almonds.

  46. The Nybbler says:

    Homelessness can’t be solved by money alone. People differ on how to solve it, and no one wants to donate money to a solution they feel is useless, counterproductive or has bad side effects. It’s even worse with more adversarial problems. If the NCTU is out collecting donations for Prohibition II, the National Alliance of Drunks and Alcohol Manufacturers is going to be collecting to stop it. Possibly at some point, getting more for your side will spur the other side to get more for the other, for little gain.

  47. Erusian says:

    Simple, really: Money can’t buy votes, either of politicians or voters.

    It’s often glibly put money buys access but what it really buys is prioritization. For example, no amount of money is going to get Trump to support more regulations. He is, in a genuine and principled way, anti-regulation. But if you’re a supporter then your regulations get higher priority than those of (for example) an anti-Trump industry like Silicon Valley. To give an (uncontroversial, imo) example: Silicon Valley was widely blue and so Obama’s administration rewrote laws that made crowdfunding legally a grey area to make it fully legal. Now, was this Obama cravenly bowing to business? I doubt it. Obama probably really did support the change. But why did he expend effort on that and not on (eg) the status of orphans? Well, because Silicon Valley gave him more money than the child welfare people.

    But that’s small potatoes, honestly. The best way to get political influence is political organization because then you control votes. And that’s what politicians really care about. The NRA claims it has about five million members most of whom vote single issue in every election. That means getting the NRA’s support is roughly equivalent to all the party voters in a major state. It means the NRA controls about 4% of actual voters. That is why the NRA has so much influence, not anything to do with money. Likewise with the Planned Parenthood crowd (feminists), the AFL-CIO (union types), the Chamber of Commerce (businesspeople), etc etc.

    The thing is, this doesn’t cost money. In fact, you’ll make money because these members will often be dues-paying. But you have to be representing something genuinely popular with some set of the population and manage to organize them.

    Basically, lobbyists don’t say, “I have such and such amount of money. Do what I want and it could be yours.” They say, “The pro-business crowd voted for you in the last election. They trust the Chamber of Commerce to tell them whether you’re pro-business and this policy looks preeeetty anti-business. I’d suggest you change it or we’ll tell our members and you’re going to lose part of your support base.” And it really doesn’t take that many people to get to this point: there are about twenty thousand cotton farms in the US. Let’s presume each of them has ten people who care about any cotton farm’s existence. That’s 200,000 votes. That could swing a lot of races, especially close ones, and that’s before you get into other agricultural interests.

    That’s the sort of thing that gets election pledges or official positions changed.

  48. YIMBY Californian says:

    This is one of those Slate Star Codex posts where Alexander writes truthfully and rationally to make a profound point, yet his implicit conclusion is flatly contradicted by the facts on the ground.

    Argument: surprisingly little money is spent on US national political campaigns, advocacy, and think tanks.

    Implicit conclusion: money in politics is not that big of a problem.

    Facts on the ground: Members of US Congress spend over half of their time fundraising, making calls to donors and attending fundraisers.

    Surely we can agree that spending over half your work hours focused on one thing will influence your way of thinking. That is, the politicians aren’t robots that can act totally fake with their donors and bamboozle them out of their money while the donors get nothing in return.

    Are these politicians wrong that spending their time in this way is in their self-interest? Or is Scott wrong?

    • Sigivald says:

      Or is this frame inapt?

      The implicit conclusion is more like “outside money influencing politics is not a big problem” (because that is counter the commonplace narrative that “Big/Dark money controls our politics!!!”).

      That politicians are desperate to win power is not questioned, nor is that they need money to do so, to compete with their opponents and to win the support of randos.

      (I don’t think that “people care about things politicians have power over and thus influence them as they can” is a problem in itself. It’s … exactly how one expects and wants a democratic system to work.

      Money isn’t the only thing, of course – a serious and reliably commitment to vote or threat to not-vote has huge power.

      Which is why e.g. the NRA’s budget (and I think Scott might want to check NRA-ILA vs. NRA; NRA doesn’t lobby, and can’t lobby, legally) is irrelevant.

      They’re a 10 ton gorilla because a huge bloc of voters takes that issue very seriously as a primary consideration and the NRA has long been a reliable source of vote guidance for that bloc.)

    • John Schilling says:

      Facts on the ground: Members of US Congress spend over half of their time fundraising, making calls to donors and attending fundraisers.

      That just means it is a buyer’s market for Congressmen, which is entirely consistent with nobody much wants to buy congressmen and so congressmen being bought isn’t a big problem.

      I’ve argued above that the value of a “bought” congressman is plausibly very small because of the constraints on their behavior and the lack of competition. That being the case, buying congressmen (and their staffers) plausibly isn’t a problem, at least for the big-ticket controversies, because nobody is that interested in doing it and it doesn’t change much when they do. But it also means that a congressman who wants to stay employed as such and thus needs to finance at least a baseline campaign, has to scramble to sell bits and pieces of themselves to a thousand different interests, offering each only the tiny bit of value he can offer in their area of interest.

      The proposed theory and the facts on the ground seem consistent enough to me,

      • salvorhardin says:

        Also, let’s not assume the candidates are acting rationally. The continuation of fundraising obsession even after one has enough money to reach diminishing returns may be an irrational “the other guys are doing it, can’t let them win” sort of arms race akin to the way some parents push their children to be hypercompetitive, rack up extracurriculars etc in order to maximize their chances of getting into an Ivy, even though if you’ve already got the means to do this and your kid has the brains to have a shot at getting into an Ivy, they’re going to be fine anyway.

      • Anthony says:

        Decades ago, I read an article that argued that unlimited campaign contributions allowed Congressmen and Senators have more independence. The example was that the Senator from Oklahoma would be owned by the oil interests, but they only cared about oil, and the Senator could vote any way he liked on any issue which wasn’t directly affecting oil company interests. These days, with more diffuse campaign money sources, that same Senator now has to keep *lots* of different interests happy, and has much less freedom of action.

  49. Sigivald says:

    Post Citizens United, corporations can supposedly put as much money into politics as they want

    Supposedly according to people who know only what dumb memes told them about it? (Not strictly a dig at you, Scott, but … that’s not how any of this works.)

    CU struck down a “reform” [sarcasm quotes very deliberate] that blacked out “electioneering” within 30/60 days of a primary/election, or advocating the election or defeat of a candidate at any time, by any “corporation”; I use quotes because while the term is clear in law, people forget that nonprofits and unions and other associations are also “corporations” in the law.

    This is ironic when Progressives complain about CU, blissfully unaware that the decision un-muzzled Progressive nonprofits and labor unions.

    The “money” the above can use without limit is money in the form of speech – ads and movies and print and the like, because CU was entirely a First Amendment case.

    Cash donations to candidates and campaigns and parties have their own regulations that are not affected by the Citizens United decision.

  50. eqdw says:

    If everyone who cared about homelessness donated $100 to the problem, homelessness would be solved.

    This is obviously wrong by inspection. It has proven difficult to find sources for these numbers, so take with grain of salt, but I recall that in 2017 a ballot initiative dramatically raised San Francisco’s budget for homelessness to $700,000,000. There are about 900,000 residents of SF. This works out to about $800 per resident per year spent on homelessness. Not only is this 8x your hypothetical budget, but “literally everyone in the city” is always, necessarily, going to be a superset of “everyone on the city who cares about homelessness“.

    As we can clearly observe that homelessness in San Francisco is not permanently solved forever by this expenditure, your comment must be found to be incorrect

    • Freddie deBoer says:

      Yeah, as I say above, the problem with this is actually the entire genre of “we could solve this problem with $X.” It almost always involves assuming that small scale interventions can be scaled with linearly scaling cost. But the entire history of social spending demonstrates that’s not true. It can’t be that EVERYONE who’s attempted to solve major social problems with money has been incompetent or corrupt. It must be that solving major problems is really hard, and that simple-looking solutions are deceptive thanks to scale.

  51. Aftagley says:

    Thought experiment:What % of Exxon Mobil would I need to own before I would feel confident that the company would respond to my directions? 51% would give me outright control, around 10% would likely make me the largest single holder of their stock, and anything over 1% probably would give me a seat at the table. Assuming the market cap for Exxon Mobil I’m seeing online is correct (between $300-$320 billion) those values correlate to roughly $160 billion, $30 billion and $3 billion.

    Now imagine the USG as a black box that we can apply these same criteria to. The most recent net worth assessment for the USG (that I’ve seen) placed it’s value at around $125 trillion back in 2014. Assuming these numbers are stable, this would mean getting “majority control” of the US would cost roughly $62 trillion, a 10% stake would set you back $12.5 trillion, and a 1% seat at the table would still require over a trillion dollars. These are quantities of money that no person or organization has access to.

    Sure, the USG isn’t a black box and the above numbers likely aren’t accurate, but how far of are they? How different is the USG from other large organizations? Assuming money is 10000% more effective at buying control over the government than it is buying regular companies and you’d still need 100 Jeff Bezoses (or 1 hectoBezsos) to even get that seat at the table.

  52. heron61 says:

    “This is one reason I’m so gung ho about people pledging to donate 10% of their income to charity. It mows through these kinds of problems. “

    That’s pretty much the exact opposite of my take-away of your post. What the coordination problem says to me is that to solve problems that require large or even moderate amounts of money, what we need is more taxes, since they easily solve the coordination problem – that’s most of the reason they exist. Ideally (from my PoV at least) this would also apply to political candidates, where it seems likely that the state of US politics would improve if there was some way to replace most money donated to political candidates with state-supplied funds.

  53. User_Riottt says:

    The upper limit on political spending by corporations is drawing attention to themselves.

  54. Jiro says:

    When you say “this isn’t even as much as what is spent on almonds”, you’ll need to first convince me that the amount which is spent on almonds is small. Agriculture is a huge industry, and this could just mean “there’s so much money in agriculture that even one arbitrary corner of it is still a huge amount of money”.

    Almonds are obviously being used as a comparison because people think of them as insignificant. But that makes the argument into “almonds are insignificant things to you personally, so whatever is spent on almonds must be an insignificant amount of money”, and that doesn’t actually follow.

  55. ECD says:

    Sorry, I’ve only skimmed the comments, so someone else may have pointed this out. Isn’t at least points 2 & 3 addressed by the fact that we’ve successfully made it rather difficult and dangerous (for the politician, at least) to bribe politicians?

    A lot is made of how you could pay off every congressperson, but all it takes is one person calling the FBI and then you’ve got to try to buy off the entire FBI. Even assuming everyone were corrupt, which I reject, I think the risk of prison for anyone with the means to buy off congress is going to be, shall we say, heavily weighted.

  56. financialwhatever says:

    “But Exxon makes $280 billion per year” – this is off by about an order of magnitude. Revenue is very different from profit.

  57. foor says:

    Ferguson concludes that money does in fact determine congressional and senatorial elections.

  58. gbankmanfried says:

    tl;dr: buying the House costs $200 million, which is surprisingly low

    Most research (plus the 2016 results) confirms that money has little effect on victory

    I’d argue that this has the same problem as “there’s so much money in politics”: considering how much elections matter, it’s surprising how easy it is to buy elections. Take Levitt’s ’94 paper, which is referenced in one of your links and remains one of the most prominent papers in the field. I think there’s a lot of reasons that Levitt underestimates the effect of money in elections. But taking his numbers at face value, we have:

    !) Giving $100,000 to a candidate increases their vote share by .1% [Levitt]
    2) Toss-up races have a vote share standard deviation of ~4% [Too lazy to post links, but look up polling errors]
    3) Giving $100,000 to a toss-up candidate increases their win percentage by 1% [Assume vote share is normally distributed with mean of .5, standard deviation of 4%]
    4) Therefore, buying a seat in Congress costs $10 million in expectation
    5) The party in control of the House usually has a ~20 seat majority []
    6) Therefore, swinging the House costs ~$200 million

    The numbers above are super rough. There are plenty of caveats, but if you spend a while thinking carefully you end up with the same order of magnitude.

    So buying the House costs $200 million. That’s fucking cheap.

    • Bugmaster says:

      Do elections matter as much as people think ? All of the stats you’d listed deal with getting a specific candidate elected; but what people really want are specific policies, not specific candidates. It might be more efficient to just nudge the election results along in a perfunctory manner, and then bribe or otherwise influence the winners once they are in office — with kickbacks, services rendered, and other off-the-book deals.

      • gbankmanfried says:

        I agree that non-electoral influence is important, perhaps more important than electoral influence. However, policies are almost entirely governed by the majority party. My math above showed that you could buy, say, a democratic majority in the house for O($200 million). I think everyone would agree that flipping the House in 2018 changed Federal policies enormously. On January 2, 2019, Republicans could pass anything they wanted (okay, not ACA repeal). On January 4, 2019, they could pass nothing they wanted.

        Note that this has an enormous effect on appropriations. The Federal government spends $4 trillion a year, or 20% of GDP. In 2017, Rs decided how to spend it. In 2019, Rs and Ds have to compromise. A lot of this spending is nonpartisan. But a lot isn’t. $200 billion would be a conservative estimate for the shift in spending from “R priorities” to “D priorities” after Dems won the House. So $200 million in election donations bought you $200 billion in Federal spending.

        1000x leverage ain’t bad.

        • Bugmaster says:

          $200 billion would be a conservative estimate for the shift in spending from “R priorities” to “D priorities” after Dems won the House.

          That’s an impressive figure indeed, but what does it mean ? Do you have any specific examples ?

          That said though, 5% of the entire appropriation budget is no small potatoes, but it also seems like it might be a small enough percentage to allow other issues to overshadow the spending. For example, if I knew that donating to a politician publicly would put me at risk of exposure, and thus jeopardize my private bribery, I might be willing to forego those 5%.

          • gbankmanfried says:

            Do you have any specific examples ?

            Great question! Here’s one (Democratic) House appropriations bill for the coming fiscal year. It would spend $1 trillion in 2020, including:

            $189.9 billion [for the Departments of Labor, Health and Human Services, and Education], an increase of $11.8 billion over the 2019 enacted level and $47.9 billion over the President’s 2020 budget request.

            So among other differences, the Democratic bill would spend $47.9 billion more than the White House wants on (mostly) health care and public education. You can look into the link for more granular information, it’s actually pretty readable.

  59. This is going to sound a bit flippant, but what if people more or less know how much to spend to get specific results and don’t spend any more than that? Politicians have a price, and if it’s a low price as long as you meet that amount you’re good. Propaganda has limits because there are only so many people to reach to begin with. These sort of marginal utility limits should exist in political markets just as they do in the economy; companies don’t just go “workers good” and then paperclip maximize workers until they run out of money and have factories filled with useless people standing around doing nothing.

    If you know how much to spend and where you only need to spend that much. Some fringe belief systems might want to spend a lot more, but they’ve probably already reached the low hanging fruit in terms of converting people to socialism or libertarianism or whatever, and don’t know how to increase their efficiency. They could spend more on market research, but that’s going to have diminishing returns of its own.

  60. wavedash says:

    “Maybe the most extreme example here is, which recently sold for $3 million”

    This is kind of apocryphal, as far as I know. The transaction also involved the buyer taking over payroll for 200+ people. It’s hard to isolate the value of the website from the cost of all the people who keep it alive; but when you do, you get a good headline.

  61. Hoopdawg says:

    As was pointed out, by Freddie DeBoer in particular, if a particular sum of money could solve some problem, then the obvious solution to the coordination problem is for the government to tax citizens and then spend the money. It’s also obvious that most people would not agree to it, because they simply do not trust the government to spend the money wisely and competently solve problems. It’s very likely that many problems are just really fucking hard to solve in the first place, but many hard problems have been solved by governments, so why doesn’t it happen more often?

    Well, one hypothesis I believe nobody mentioned so far is that the corporate money spent on electoral politics is not being spent on a difficult task of pursuing particular actions or policy positions (if something like this works out once in a while, it’s a nice bonus). It’s spent on a much easier and cheaper task of making sure that the government is devoted to maintaining status quo and incapable of implementing far-reaching policies. It’s clearly preferable to having to deal with a functional, competent government (that could unpredictably make sweeping changes to your market environment). And clearly the strategy is working.

  62. mkt42 says:

    “First, we should expect ordinary people to donate more to politics.”

    This is a fundamental mis-read of the situation facing voters and donors. It’s a classic free rider problem. Pretty much any introductory microeconomics textbook will mention that it’s in sense surprising that anyone votes at all, because one person’s vote has virtually no effect on the election. So why bother voting?

    And the argument is even stronger for small donors (“small” meaning their donation is too small to affect the election). Just as I can’t solve world hunger by donating $50 to some humanitarian organization, I can’t affect the election by donating $50 to a candidate or party. So why bother donating at all? And indeed most people don’t (and most people in the US don’t bother to vote either, unless it’s a presidential election).

    For those of you who do vote and who do donate (and I do both), good on ya, you’re doing your civic duty. But one voter or small donor is not going to change the election results, and contrary to Scott Alexander’s quote, there is no rational reason for ordinary voters to donate, quite the contrary. They’re throwing money away. It’s only due to a sense of civic engagement or wanting to do the right thing that impels them to vote or donate.

  63. nameless1 says:

    Vegetarian Indian wrestlers eat lots of almonds, so while studies are contradicting whether it is good for testosterone or bad, it is likely good plus generally healthy if you define health as strength and vigor.

    Knowing the studies of the relation between upper body strength and political views, I wonder if it is smarter for conservatives to spend on almonds than on political candidates. Just gifting big sacks of almonds to all their liberal friends. What could go wrong?

  64. ejmoncrieff says:

    If you think Tumblr is the “far left,” you’re living in centrist or a right-wing bubble.

    • EchoChaos says:

      Isn’t Tumblr pretty far left on social/gender issues? I would believe it isn’t far left when it comes to fiscal issues.

    • Hoopdawg says:

      If you think an internet site the size of Tumblr built on user-generated content can be assigned a clear, singular political leaning, you’re living in [some kind of] bubble, that’s for certain. For that to happen, you’d pretty much have to specifically moderate everything else out of the site. That’s really hard.

      But Scott’s claim was markedly different, mainly that Tumblr is especially important to the “far left”, to the point where taking it over would be a meaningful blow to it. Which, I believe, is also hugely mistaken, for pretty much the same reason. It would mostly be a blow for people who like Tumblr.

      • nameless1 says:

        I don’t get it. It’s vectors. Add ten million vectors pointing north, ten million south, ten million right and eleven million left and some in really random and weird directions, and the whole thing will end up pointing left and moving left.

        Counter-check. Look at the conservatives. As this book points out: they have many very different ideas. Still, can’t you identify some tendencies? Sure, paleos were antiwar. But didn’t the prowar neos win and had more influence?

        (EDIT: a better review: ““It would take a comic genius, an American Aristophanes, to capture the strangeness of this little world,” Mark Lilla wrote recently of neoconservative Washington, where a typical social gathering mixes “older New York intellectuals, professors in exile from politically correct universities, economic visionaries, Teddy Roosevelt enthusiasts, home-schooling advocates, evangelical Protestants, Latin-mass Catholics, Likudniks, and personalities from shock radio.” And these are the neocon gatherings, mind you, the closest thing the American Right has to an in-crowd.”

        It is a diverse bunch. And yet. American conservatism clearly had and has dominant vectors.)

    • Corey says:

      Centrist bubble is a cool idea, reminds me of Futurama “what makes a man turn *neutral*?”

      /if I don’t make it back, tell my wife I said “Hello”

  65. OriginalSeeing says:

    You’ve listed revenue instead of profit (net income) for many of these companies. That ignores the expenses necessary to earn that revenue.
    Revenue – Expenses = Net Income. Only net income is extra money that you can spend on new things without diverting money from areas you’re already spending in.

    Exxon had $279 billion revenue in 2018 but only $20.8 billion net income. That means they only had $20.8 billion extra to do alternative things with. Some companies and individuals have even more extreme ratios. Wallmart had $514.405 billion revenue in 2019 but only $6.67 billion net income.

    Your argument is still valid and I’ve wondered about the same point before, but the above cuts down the numbers in a lot of situations. For instance, how much wealth does Elon Musk have that he can spend without harming his stakes in all of the companies he owns or invests in? This is even more true for company owners who have a large net worth on paper but have almost all their money tied up in the one company that they own 51% of.

    Another thing worth noting is that people often donate money to political candidates purely so that they can obtain “access”. Access means that you’re on the list of people who the candidate may agree to meet with if you request a meeting with them. No access, probably no meeting. And that’s why lobbyists will donate to both sides of the aisle.
    But access is often not actually very expensive. I can’t find a number to quote right now and it’s been a while since I spoke with someone who worked directly under a congress member, but I think $5,000 or $10,000 was enough to obtain a large amount of access to a candidate. (Someone please correct my numbers if you have more info, and these numbers do change massively based on candidate involved.)

    • sa3 says:

      You can get limited access for less; the max donation is 2500 I think, and that usually gets you in the door. Can be more if you have more and the legislator needs to re-route funds to party committees and whatnot. However, you’re buying very limited access. You get the ability to have short conversations with politicians, maybe, which they forget in a span of minutes. It gives your lobbyists a chance to make a short presentation that the Congressman may or may not listen to. Relative to average Americans, that’s a lot of access. In the greater scheme of things it’s nothing, and influences offset almost completely. The cost of access is so low (because of donation caps and restrictions) that it’s fairly easy to create opposition to anything. The general logic is that donations are table-stakes. You don’t get much other than the right to play in the influence game.

      Following from the above, best research on lobbying that I’m aware of centers on information. Congressmen don’t know anything about most of what they legislate, and their staff barely know more. Lobbyists know a lot. They’re also biased. They don’t lie. Common misconception; a lying lobbyist loses his influence and access quickly. They are selective in the information they provide. So Congressional offices have to extract necessary information from biased sources (this includes the bureaucracy, which is the main source of policy information and which is also quite biased, in a number of idiosyncratic ways). It’s hard to write complicated, detailed legislation that can blow up in your face if you get it wrong.

      On Scott’s article, I think it’s all pretty much correct, and I appreciate the reasonable treatment of the Koch brothers. They (he, now) really do believe this stuff, and they really didn’t get much for their political investments. Charles realized this and pulled back on political investments, because his money wasn’t buying the kind of change he wanted. Money really just isn’t that impactful in politics, beyond the necessity to have enough to compete. As Scott notes, Clinton out-raised and out-spent Trump enormously, and lost. Once everyone knows who you are the value-add of more spending is kind of limited.

  66. ARabbiAndAFrog says:

    When I hear something like “Homelessness/Migration/Whatever can be solved with 10M dollars” I always feel need to snidely remark “Surely, bullets can’t cost this much!”.

    In other words, solutions to people’s behaviours are not always going to be appealing to population at large. Rehabilitating the homeless would likely involve detaining and controlling their lives quite a lot.

  67. thisheavenlyconjugation says:

    “In fact, according to Exxon’s own records, they only spend about $1 million per cycle.”
    I simply don’t believe this. Out of Exxon’s 70,000 employees, they don’t have 20 or so $100,00/year people doing political stuff of some sort?

  68. C.H. says:

    We should have a kickstarter for charities/political campaigns whereby you only get debited your pledge if the sum of all pledges reach some significant milestone to enact fully realizable effects.

    The democratic nominations kind of had this in that they gave all the candidates specific thresholds of donations to aim for in order to qualify for the debates. I’m surprised no one set up a kickstarter or something for a candidate to meet the donation threshold.

  69. borker says:

    You’re missing the mark because of the wrong frame: you should be looking per relevant capita, not total spend. Almond spend is over ~325M Americans.

    $12B / 325M Americans = $40 per American per Year. That’s 4 bags of almonds consumed, considering many of those are exported.

    $8.5B / 535 elected congressional officials = ~$16M per federal politician per year.

    Not many individuals have $16M spent on them per year.

    Certainly there’s a coordination challenge, but it’s hard to suggest that on a per capita basis, politics has a *lot* of money available to it.

    The challenge is that the currency of politics isn’t *MONEY*, it’s *social attention.* It turns out that this can be purchased, but purchasing attention is expensive and mostly ineffective. Measuring the amount of attention spent on politics (in terms of time / conversations), and *then* converting that to economic value would show a much, much, larger number.

    In politics, the attention and debates and changing of hearts and minds *is* the point. It’s just a different system that leverages, but does not serve, the economic model in our country.

    A couple other nits:
    * If everyone who cared about homelessness donated $100 to the problem, homelessness would be solved
    –> No it wouldn’t. Homelessness is a complex system, not merely a supply of housing issue.
    * Money doesn’t matter in elections
    –> The 538 article actually say it does, in meaningful ways. ~90% of the time the candidate who spends more wins. Not necessarily a causation

  70. Chad_Nine says:

    “If everyone who cared about homelessness donated $100 to the problem, homelessness would be solved.”

    Is this true? Is the problem of homelessness only that the homeless don’t have enough money? Would the organizations gifted the money use it effectively?

  71. Doctor Mist says:

    If everyone who cared about homelessness donated $100 to the problem, homelessness would be solved.

    Show your work? If the Huffpost article is correct and 20 billion would do it (which I doubt, for all the reasons people cite above, but whatever), you are assuming that 200 million out of the country’s 327 million (including children) “care about homelessness”. I’m not at all sure that’s true. For my part, I’m not sure I’d pony up the $100 even if you could creditably assure me that absolutely everyone else would if only I would. I might feel bad about it, but I would talk about things like perverse incentives and people making their own beds and wait, $100 a year for ever and ever? And even if it did work, what else are you going to dun me for next year, on top of that? Once you pay the Danegeld…

  72. echidna says:

    Spending an enormous amount of money trying to get the public to support your issue is unlikely to be successful. What it will do is focus the public’s attention on it, making it electorally costly for friendly politicians to do what you want. Much better to keep your issue out of the news, enabling friendly politicians to pass legislation without the public paying any attention. This is the strategy followed by Big Banks to neutralize Dodd-Frank.

    The alternative is to influence the public using underhanded tactics, as the Koch Brothers did with climate change. For this to work, it is essential that the effort remains covert, which again argues for a limited budget.

    That said, Fox News is a pretty big budget effort.

  73. hryanjones says:

    I’m don’t think it’s been terribly successful yet, but one interesting model for trying to circumvent the coordination in donating problem had been implemented by

    In their model each person can join a cause and if the cause takes off they agree to pay more. The idea then being you could join many causes that you think are important and if they don’t get a lot of support you only pay a few cents, but ones that do get a lot of support get funded through increasing size of donation as more people join.

  74. Deputy Van Halen says:

    If you use the more accurate, but more awkward terminology, “money largely for political advertising” as opposed to “campaign contribution”, it should become apparent why it’s not all cracked up to be.

  75. TDB says:

    Stumbled onto this, which I ought to have known about. The Tullock Paradox.

  76. kaleberg says:

    The small amount of money spent on politics is a testament to the high return on that spending. If you have to spend $1M to get a military contract or $500K to get a law changed in your favor, you could be billions or at least millions ahead. There really aren’t things you can spend money on that yield a higher return.

    An interesting question would be, how much money could we spend on politics. There’s television, radio, online media and what else. There is only so much of it. I suppose it would be possible to bid up prices, but this could be counterproductive. There are consultants fees. There is lobbying. Still, how many consultant hours could one buy before it was clear that an extra $1M wasn’t going to make a difference.

    I suppose there are think tanks, but those have long been known as efficient. Academic types, even academic types with no axe to grind, are even cheaper than consultants and lobbyists. Consider fake grass roots groups. You’d need extras to pack hearings. You’d need some media types. Maybe you could add some professional signature gatherers, but it would be like Brewster’s Millions.

    There used to be a joke about the old Horn and Hardart restaurants in New York City. Supposedly, there was an all you can eat certificate. If you had ever eaten at a Horn and Hardart, however, you knew that certificate wasn’t worth all that much.

  77. Ramiroap says:

    I see 3 possible competing / complementary explanations:
    1) Cooperation: this is a game of partial conflict, not total – if I’m a billionaire willing to give half of my money to a candidate, I can expect other billionaire to do the same for the opponent, so neutralizing my effort. Besides, there’s a reputation risk in such donations: it’d at least piss off some of my investors.
    2) Disparity between revealed preferences and values, as SSC concludes;
    3) Free rider + Elephant in the brain: when I say politics is super important, what I really want is that someone else fund it. Classic public goods dilemma. That follows from the hypothesis that individuals are already maximizing: if I’m a billionaire, that’s because I’ll only fund politics up to the point where I’ll get more utility/return investing somewhere else – business as usual, consumption, charity, or space rockets, etc.
    (How to test them?)

  78. Spiritkas says:

    I would think to invert the analysis or to widen it. Instead of looking for dollars going to individual politicians or campaigns, I’d look at how influence operates over the lifecycle of politicians and issues controlled by politics. Do we (a powerful person or group) use influence to get a tax exemption for a factory in a specific place, a particular law passed which affects many people, get a specific allied person appointed to a role as a judge or other key position, or do we use influence to have a particular politician/committee/caucus/party in our pocket?

    I think the overall currency is influence, not money. We might not see a lot of money in politics based on the pure economic values of what is being controlled, such as regulation of industry. Yet it is common knowledge and inarguable that regulatory capture is common and widespread. The influence and control is what matters, not pricing economic costs to specific economic outcomes or goals.

    Take a Soros or Koch brother, depending on which/both you dislike. They don’t need more money, it doesn’t matter to them if they take more or less money as a result of an ‘investment’ of cash into a campaign, politician, political party, or specific issue. They are not in it to make money, though they often do make a lot of money. The name of the game in politics and mega rich folks is power. The currency of power is influence, which can sometimes take the form of dollars. We are more dumb chimps who want to feel and be powerful over others, than we are (broken?) economic utilitarian robots seeking more money even when we already have more than we could ever use or need.

    Looking at the politician lifecycle, I wonder if we look at them 10 years after they leave office…how many become multi millionaires? Even adding in a 2 year waiting period for specific types of revolving door payments would be impossible to enforce realistically – are we to ban them from ever working again?

    Influence can be a hard to pin down, but you know it when you see it. The evangelicals in the US offer a voting block, organisation of conferences, connections and introductions, and endorsements; along with a lot of other benefits. They offer a lifecycle for a politician post politics with memberships and money/power not readily tracked in campaign contributions. They don’t even want money, their goals are ideological.

    If you go looking for cost benefit analysis in their spending and activism on anti-abortion laws and compare this to the lifetime economic value of workers who are born instead of aborted….then you are on completely the wrong track in terms of understanding anything about that scenario! They are not motivated in this way and money is only a partial reflection of their exercising power through the use of influence. They can give or withhold cash, voters, endorsements, and/or create negative cycles of news coverage. If they know you had an affair and you are a no-affair family values kind of politician, they can leak that story or they could make one up even if you didn’t do it.

    Rich and powerful people will almost always be rich no matter what happens. What they can lose is power and position. This is what drives them. Look to Wall Street where money is a measure of personal power and they buy bigger and bigger yachts they don’t need or use. There is perhaps too much looking at this class of powerful actors as they are very money driven to a point of no realistic or meaningful returns on improving their lives with more money. Outside of this highly economic worldview, there are many other forms power takes. I don’t know how rich Cornel West is, but he has a lot of power he can use and it does not involve spending money on political campaigns.

    Influence peddling and trading is the core of power and power is the core of politics. Most of us are very blind to this as we do not have any power or only a very limited amount of it. In terms of multiple literacies, most people are essentially illiterate in terms of reading power and influence.

    It is like money in a way where having a little vs a lot are different entities. Investing or using everyday amounts of money like $100 or $1,000 makes you think you understand something about $100,000,000 worth of money. The use of the interest alone is enough to fuel all sorts of concerns and drive interactions with others which a regular person dealing in $1,000 increments of money does not have. The same is true of power where you might only have influence over your children who are smaller than you or over a homeless person who is vulnerable. But the types and scope of concerns for someone with a large amount of power are so different and outside the scope of experience of the everyday person, that it can be hard to understand.

    Money is like the grease in the machine, you can study its properties all day in terms of amount, location, temperature and operational tolerances, etc. and wonder why the entire machine isn’t made out of a giant pool of grease…but you’d be no closer to understanding the machine of Power or its purpose or action in the world. You only use as much grease as you need.

    If someone tries to out-grease you by spending more money, then you can beat them by applying one of a hundred other forms of influence. This leads to a great discounting of money, rather than the rapidly increasing spiral of spending Scott was looking for and didn’t find.

  79. AshwinV says:

    Disagree on the obviousness of this. A lot of people actually don’t care about politics, it’s all just status signalling. Elephant in the brain type stuff.

    The exceptions of course are the wealthy billionaires that you mentioned, but I doubt the same coordination problems apply to them.

  80. Douglas Knight says:

    Almonds are an alien benchmark. A more natural benchmark is campaign spending in other times and places, such as 1994 Mexico. In the USA, campaign spending is rapidly rising. The 2003 paper you cited was from when campaigns spent less than the infamous Mexican campaign. But today they spend more.

    Any answer that purports to explain how much they spend should work at different times. Many commenters tried to explain it away. But if the reason is that people “don’t really care,” that poses the follow-up: do they care more now than before? Can we see this in something other than dollar figures?