Open threads at the Open Thread tab every Sunday and Wednesday

Raikoth: Economics, Relationships

Someone in one of the comments asked whether I am actually pushing Raikoth as a perfect society, and the answer is sorta. My con-culturing philosophy is trying to optimize for goodness while erring on the side of weirdness. This stands in contrast to a reasonable philosophy of the real world, where one might optimize for goodness while erring on the side of safety.

The reason I mention this may become clear further down in this post.

I promise no more than one or two more posts about this before I get back to regularly scheduled programming.


The economic system of Raikoth takes the form of a basic income guarantee and very little else.

There’s not much government intervention in health care – but people are strongly advised to spend some of their income guarantee on health insurance.

There’s no minimum wage, labor unions, or workplace safety laws – but people are encouraged to remain unemployed and live off their income guarantee if they’re not entirely satisfied with the job options available to them.

There’s not even a budget, per se. There’s a basket of taxes and the ability to slightly raise or slightly lower all taxes in the basket in order to fund individual programs. For example, a policy proposal would not take the form “Let’s build a new bridge, it would only cost $100 million and I’m sure there’s room in the budget for that”, but rather “Raise all taxes 0.5% to get $100 million, then use that to buy a new bridge”. Spending proposals not linked to revenue proposals would be considered incomplete; thus decision-making takes the form of real cost-benefit calculations instead of just saying yes to all pleasant-sounding spending and no to all unpleasant-sounding taxes.

It’s not that there’s no deficit – a small one is maintained for Keynesian reasons – but there’s complete control over its size and it’s never going to increase unexpectedly or against everyone’s wishes.

The taxes in the basket are primarily land taxes, very high estate taxes, and a tax on large corporations proportional to their size. This latter is meant to approximately balance economies of scale and give small mom-and-pop stores and start-ups ability to compete on an equal footing. It is recognized that this makes things a bit less efficient, but the Angels have decided it is still a net good for hard-to-measure reasons.

The government also taxes externalities – carbon taxes, pollution taxes, noise pollution taxes – as well as some less obvious cases. One of the weirder ones is destroying-the-social-fabric taxes. In these latter, Priests of Truth predict whether some piece of media will cause increases in crime or racism or whatever, calculate the costs of these effects, and then levy them as a tax on the producer. You can still get violent videogames that demean women or whatever, but they’re going to cost much more than the other videogames and that money is going to go to law enforcement, educational campaigns, and generally cleaning up after them.

Income taxes supposedly don’t exist, but symbolic beads place wealthy people under extreme social pressure to donate, and given the government’s (widely believed) pretension of being a perfect utilitarian system, most of these donations take the form of “bonus taxes”, providing an extreme boost to the national coffers and subsidizing most of the basic income guarantee program.


One of my goals in con-culturing was to subvert dystopian tropes in unexpected ways, and one of my least favorite dystopian tropes is Bureaucrats Or Computers Decide Who You Are Allowed To Marry.

So fine. Let’s give these people their frickin’ Bureaucrats And Computers Deciding Who They Are Allowed To Marry.

Temion Mirun is a festival held on May 1st (approximate, based on solilunar calendar), a sort of combination celebration of springtime / Valentine’s Day / group marriage day. Each year in the runup to May 1st, everyone who has undergone the coming of age ritual but is not currently married writes down the names of the people they’re attracted to and want to date, in order of how much they like them. Then they send their lists to – say it with me – a centralized database.

Each year, on April 30th, the computer with the database runs a special variant of the Gale-Shapley stable marriage algorithm on everyone’s ranked lists. On May 1st if it is sunny, or the first sunny day after if not, everyone gets together with their chosen partners and are handfasted in a group ceremony – where handfasting means that they are considered a couple until the next May 1st.

Springtime in Raikoth

Handfasted couples are polyamorous by default, and people still ask each other out on dates in the traditional manner. Sometimes dates people get one year become their handfasts the next; other times people end up with partners they have only admired from afar and didn’t have the slightest idea returned their affection.

It is considered somewhat uncouth, but certainly allowed, to break up with your partner before the next May, unless there’s something really horrible going on like domestic violence (which is always provable – good old Third Eyes). On the other hand, it is not considered unusual or a sign of dislike to change partners the next year, even if the relationship was mostly happy. It’s not considered uncommon for A to handfast B one year, rank C higher the next year and handfast zir while maintaining a secondary relationship with B, and then decide B is better after all and go back to zir the year after.

Most people start considering marriage around the third year of being handfasted to the same person. Marriages are a bit like the handfasting relationship except that they are in theory permanent. There are a variety of different marriage contracts available, but most people go for neither no-fault divorce nor for full-on covenant marriage. The most common contract says that the marriage cannot be annulled except by the High Priest of Tala, and that the petitioner must go to Tala without using any form of motorized transportation. That is, they must walk through a few hundred miles of freezing tundra and then up the slopes of a towering ice volcano until they reach the holy city, where the High Priest will free them from their oath. The idea is to have a committment mechanism: it’s not impossible to end a marriage, but it’s not something you do without a lot of thought and an absolute certainty that it’s the only remaining option.

Marriage contracts may also include other stipulations. Although handfastings are polyamorous by default, marriages may or may not be; a couple that chooses not to be polyamorous may include this in their marriage vows, at which point people who cheat become liable for civil penalties relating to contract violation. Although some couples will happily marry with nothing but standard marriage vows, others will include a whole host of rules that then become inviolable unless both sides agree to edit the contract together.

Some people never take these stronger marriage contracts, preferring to move from handfasting to handfasting for their whole life; this is quite acceptable. But those who do usually do because they want children. Not only are most people who ask for the removal of contraception married, but the Priests of Truth use statistics about differential outcomes in married and unmarried couples when deciding whom to offer opportunities for subsidized child-raising.

This entry was posted in Uncategorized. Bookmark the permalink.

44 Responses to Raikoth: Economics, Relationships

  1. Michael Vassar says:

    I’m very impressed by the overall quality of this con-culture. Props!
    Wishing it was real.

  2. Anonymous says:

    Sorry, what about people who don’t want to be in a relationship? Do they just not write anything down, or what?

    • Scott Alexander says:

      Yeah, you can write down as many or as few people as you want, including zero.

  3. Deiseach says:

    I like your society (at least, I like the way you’ve structured it, though I’m not sure I’d like to live there) more and more the more you talk about it, and the little flaws and cracks and fudges that make it like a real society.

    I’m trying to decide if the “destroying-the-social-fabric taxes” incline the Raikoth to hypocrisy, or to more hypocrisy than our version of doing things, or what. Because we’ve all been young and stupid and done things simply because we were told “Don’t do that!” So if the thrill of “forbidden fruit” hangs over ultra-violent/demeaning video games, for instance, does that make them more attractive, which would seem to defeat the purpose? At least the Raikoth seem genuine about wanting to penalise the undesirable elements, in contrast to our version of things, where in order to get a higher rating on a movie, they’ll throw in gratuitous swearing/a bit of nudity/whatever, because the idea is that the teenage market – who are the ones who generate the income with repeat viewings – will want to go to see a film with a higher rating, so to appeal to them stuff that is not necessary to the plot or does anything other than bump up the ratings is edited in.

    And what of those who are prepared to sit through the finger-wagging lectures afterwards, because they want to play the games/see the movies/read the book, but have no intention of changing their minds? Do they rack up a certain level of “Okay, you’ve had six sets of educational sessions about how this stuff rots your brain and you’re still consuming it; this indicates a psychological problem so here’s your hospital appointment for assessment, citizen”? Again, this reminds me of the compulsory workplace training sessions where most people show up because they have to; they let the lecture from HR pass in through one ear and out the other; they fill in the forms ticking all the ‘correct’ answers because they know how the game is played, but it really doesn’t affect their attitudes re: race, gender, workplace safety and the likes, nor do they intend to permit it to change those attitudes?

    And your marriage customs gives me the perfect excuse to mention the Cáin Lánamna, a mediaeval Irish legal text detailing the kinds and associated rights of the various marriages under Brehon law:

    “Question: how many couples of cohabitation and procreation are there in Irish law? Answer: ten – (1) union of common contribution; (2) union of a woman on a man’s contribution; (3) union of a man on a woman’s contribution with service; (4) union of a woman who accepts a man’s solicitation; (5) union of a man who visits the woman, without work, without solicitation, without provision, without material contribution; (6) union by abduction; (7) union of wandering mercenaries; (8) union by criminal seduction; (9) union by rape; (10) union of mockery.”

    • nemryn says:

      Well, the extra tax levied on Maximum Blood ‘n’ Boobs means that it’s going to be correspondingly more expensive for the consumer to buy; and this is an actual cost that helps balance out the ‘forbidden fruit’ effect, and is more likely to dissuade people than the threat of a lecture. And if you go to the lectures and don’t learn anything, that’s fine, because you’re still subsidizing lectures aimed at people who are going to learn something from them.

      I imagine that Raikolin PSA campaigns are general and probabilistic: the goal is to make sure that at least X% of the population is educated, rather than attempting to identify and target each individual person who doesn’t ‘get it’.

    • Scott Alexander says:

      I think taxes are a little better than bans/ratings. Like if the nonviolent videogame cost $20 and the violent videogame cost $25, then not only does that extra $5 fail to make it seem “edgy” or “prohibited” but it might not even be noticeable.

      I would think all public service campaigns would have to be vetted by Priests of Truth to see if they had a good chance of working. Even our own society does that sometime – one commenter, can’t remember who, pointed out a study saying that rape decreases significantly after public service campaigns against rape, which hugely surprises me but seems to be born out.

  4. Fnord says:

    There’s not much government intervention in health care – but people are strongly advised to spend some of their income guarantee on health insurance.

    Unfortunately, the pre-existing condition thing, particularly when it leads to post-claims underwriting, means that an unregulated market in health insurance has problems that take more solving than a basic income guarantee.

    • Scott Alexander says:

      Good point. I’m not sure how to best solve that.

      • naath says:

        Have the state provide the healthcare?

        • Mary says:

          Or not provide it.

          The reason why things have prices is to provide incentive to minimize use. Things without prices tend to be exploited without measure.

      • Fnord says:

        I’m leaning towards “determining the best solution with high confidence is a problem that requires access to Kadhemic, Risurian-Silk, and the Angels”.

  5. Blake R says:

    As someone doing research in microeconomic theory / market design (both matching markets and information aggregation), I’ve found this series inspiring. Finally a society that listens to my tribe 😉

    Scott, how many people tend to participate in the matching ceremony yearly? What’s the gender ratio of participants? How often to people include others of the same gender in their rankings (either exclusively or partially)? Same-sex attraction can make it impossible to find a set of stable pairs, so polyamory might need to built into the system, with some handfasted into triads.

    Re anonymous above: These algorithms are fairly flexible about who you find acceptable. Everyone includes the option of remaining unmatched in the rankings they submit. For instance, I could say Alice > Betty > Being single > Carol, and any reasonable algorithm would leave me unmatched rather than pair me with Carol. If you preferred being single to being any relationship, that would just go at the top of your ranking.

    For those who don’t know, matching theory has led to practical improvements in everything from in the market for hospital residents (which I think Scott had some first-hand experience in) to public school choice to exchanging kidneys for transplant. Research was kicked off by Gale and Shapley (1962), has seen contributions from luminaries like Conway and Knuth, and was the subject of the Nobel memorial prize in economics to Shapley and Roth last year. The primary focus is on finding stable matchings, where no one would prefer being single over their assigned partner and no one can find someone who they prefer over their current partner who also prefers them.

    • Deiseach says:

      “no one can find someone who they prefer over their current partner who also prefers them”

      That seems to me to be the tricky part; Bob may like Jane and prefer her to Lisa, but Jane doesn’t like Bob and Lisa does. So trying to match up Bob with Sally who he would like better than Jane seems to me to necessitate some kind of movement between population centres; after all, if Sally never sees Bob, how will she know she might like him, which then means he doesn’t go down on her card as a selection.

    • Scott Alexander says:

      Glad there’s someone who understands this so I can ask you my questions.

      The hospital algorithm has to privilege one side (the residents), in that they can write down their true preferences but the hospitals have to write their strategic preferences (am I right?) Would a marriage algorithm have to be the same way (maybe alternating male-favorite and female-favorite with alternate years) or is there a way to avoid that entirely?

      • Blake R says:

        The man-proposing version of Gale-Shapley deferred acceptance is strategy-proof for the men (vice versa for the woman-proposing version), but there is no possible mechanism that is strategy-proof for both sides. However, as the size of the market gets large enough, the potential for manipulation vanishes, so it’s not all that concerning.

        The man-proposing version also returns the stable matching that is unanimously preferred by all men (and unanimously the worst stable matching by the women’s preferences). This can be resolved in a few ways. First, there is a well-defined “median” stable match where each person gets their median-ranked partner of all stable matches. However, the median stable match is computationally hard to compute, so it’s unlikely to be feasible despite its appeal. Second, the problem can be solved through linear programming techniques to maximize average ranks or a cardinal measure of how each person likes each partner, but this probably introduces unreasonable opportunities for manipulation. Finally, randomized algorithms can help, but it’s hard to sell randomized institutions.

        Luckily, the difference between the man-optimal and woman-optimal stable matches tends to vanish as the market gets large. I also suspect that most people have preferences that will lead to a small number of fairly similar stable matches (analogously to how it’s really hard to find examples of Condorcet’s paradox in practice). So, unless you really want the most elegant, optimal solution, you could just alternate year-by-year and it won’t matter all that much.

      • Douglas Knight says:

        Perhaps only trivia, but if you force everyone to get married, GS is strategy-proof for both sides.

      • Tab Atkins says:

        There is a variant for roommate selection, where there’s no natural way to divide the set into two sides.

        Even if that didn’t exist, though, alternating m/f would be horrible. You’d be able to game on predictable years, trans* people run into classification awkwardness, and the relationship graph isn’t bipartite, so the division isn’t even natural. Just split the population randomly each year into suitors and acceptors.

  6. naath says:

    Is it permitted to write “no-one, I like being single” on your matching-algorithm form? or are people actually forced to list people?

    • Blake R says:

      Gale-Shapley allows participants to say any subset of potential partners is acceptable, including no one. This can be done by allowing short lists or by explicitly including an unmatched option in the ranking. If you never want to be in a relationship, the outcome always leaves you unmatched. If you only found one person acceptable, the algorithm would either pair you with that one person or leave you single.

    • Scott Alexander says:

      Yes. You can put as many ranks on your list as you want, including zero.

  7. Von Kalifornen says:

    It seems evident that Raikoth would not match people against their will. Why does everybody seem to anticipate this?

    • Deiseach says:

      Because you can’t always get what you want? If six people put down Bill as their top preference, okay polyamory, but if Bill only reciprocates to three of those six, that means three people are going to get their second preference. Or maybe have a choice between “This person or nobody”.

      Also, eugenics: the selection committee may refuse your application to have a child with or by Tom and match you up with Joe instead, and if you don’t like it, then no baby for you.

    • nemryn says:

      Because in Raikoth, The Government Dictates Who You Can Marry; and The Government Dictating Who You Can Marry is a Bad Thing; and Bad Things inevitably lead to Unpleasant Outcomes.

    • Mary says:

      Because we suspect human nature?

    • Eric Rall says:

      1. When applying a novel algorithmic solution to a complex social problem, we expect failure modes. This is because most algorithmic solutions to complex problems have failure modes (at least until a great deal of trial and error leads to them being discovered and corrected), and social problems are particularly messy in this respect.

      2. Endowment effects and risk aversion: we are accustomed to the known failure modes of the mate-selection methods we’re used to, and are inclined to prefer them to the uncertain failure modes of the proposed new mate-selection method.

      3. The readership and especially the commentariat of this blog are self-selected to be inclined to enjoy finding problems with things.

    • Mary says:

      And why would it not match people against their will? Do we have any reason to think that would not maximize utility?

  8. Rachael says:

    That’s not really “computers decide who you marry”, though, any more than Google Maps is “computers decide where you go on holiday”. It’s just like what we do now, except ritualised and formalised, and optimised.

    • Deiseach says:

      It does seem to be more like “computers decide who will be your future co-parent” which then translates, societally, into “computers decide who you will marry” if the trend is for only married couples (triads? more?) to be the ones raising children.

      The rest of it sounds like a souped-up dating agency, and I suppose there are still people who do things the old-fashioned way (finding someone on their own); it’s just that if most of the rest of them are filling in the forms and waiting to see the computer matches, the odd ones out may feel very much ostracised if they don’t play along (something like the in-practice-it’s-compulsory wearing the social status beads to give information about yourself; if you say to your potential love interest that you didn’t enter the Love Lottery, this may be seen as ‘you are a weirdo and I want nothing to do with you, you unsociable nutcase’).

  9. Nestor says:

    It feels much like any utopian construct in that it’s a place where the creator would be comfortable living and everyone else gets coerced into being like him/her.

    I guess I’d appreciate the “head off to the wilderness” option being available after a few months of little brother omnisurveillance.

    • nemryn says:

      Or you could just emigrate to one of the other nations on Micras.

    • Von Kalifornen says:

      I’d say that there is a certain sense of that to every utopia. Although I wouldn’t say coerced. it’s more like acculturated or something.

      • Randy M says:

        Or, er, drugged into enjoying it. 😉

        • Mary says:

          reminds me of John C Wright’s The Hermetic Millennia where there’s one culture that heavily relies on drugs to keep everyone all happy. Including drugs that cause people to forget that they had to go to war.

        • Von Kalifornen says:

          Whaaaaaat? This makes no sense. It seems like there’s a confusion between ‘wouldn’t it be cool if this country existed and had always existed’ and ‘I wish I could force a countries worth of people to have this culture’/

  10. Alex Theisen says:

    I loved, loved, loved the description of Raikoth economic policy. It’s very close to my own ideal. The only things left out that would have been very interesting to me were the monetary institutions: what does Raikoth’s central bank (if it exists) look like? What about the financial system? Inquiring minds (or okay, just me) want to know!

    • Scott Alexander says:

      I don’t know. One of the reasons I don’t know is that in-game considerations require a near-zero-growth economy (Raikoth has been at near-modern tech level for about 5000 years, but is stuck in a telluric field that prevents technological advance), and the idea of central banks and stuff seems so linked to a growing economy that I don’t have the economic background to imagine how I would have to modify it to make sense.

      • Von Kalifornen says:

        Does this mean that an expansionist empire might have some kind of huge economic power behind it?

  11. Michael Vassar says:

    So Rakioth’s technology has a low peak. OK. How about their science? What’s the strangest thing a Raikothian could tell me?

    • Deiseach says:

      I can’t tell you, man, you have to experience it for yourself, but those particular psychedelics I did for my Fifth Year meditation class – the colours…


  12. a says:

    thanks, nice post.

  13. @johnwbh says:

    The taxes attached to costs idea is nice in theory, but assumes you can perfectly predict the cost of a project and how much a tax will raise.

    There can be genuine unexpected problems, e.g. part of the bridge unexpectedly collapses when its 90% finished do they simply abandon the project when funds are used up, a massive cost, or if not who pays for the additional costs? If overspends are paid for then you get structural incentives to underestimate the cost of a policy you want passed.

    • Scott Alexander says:

      Prediction markets! If they’re well-calibrated, then while some projects may end up cheaper or more expensive, averaged over all projects started in a year the amount of money raised should be predictably close to correct.