Matthew Yglesias, crackhead, and his snowflake theory of identity politics

snowflake01Are Democrats currently engaged in identity politics?  Is that working out for them?  Should they double down on identity politics, or pull back?  These are important questions!  Maybe we should ask Matt Y., noted intellectual crackhead, for his opinion:

Democrats neither can nor should ditch “identity politics”

There’s no other kind of politics. They just need to do it better.

Wow, insightful.  Tell us more, Matt.

The beginning of wisdom on this subject is to recognize that human identity is more complicated than a quick glance at an exit poll table would suggest. If I describe my father to you as a “white male over 60 with no college degree,” you would say it is overwhelmingly likely that he is a Donald Trump voter. But if I describe him to you as a “Hispanic union member who didn’t graduate high school” or a “Jewish screenwriter and novelist who lives in Chelsea,” you would say it is overwhelmingly likely that he is a Hillary Clinton voter.

Now, my dad happens to be a demographically unusual person. But we’re all beautiful unique snowflakes in one way or another.

Yes, dear reader, there are more things in Greenwich Village then are dreamt of in your philosophy.


Yglesias’s ultimate conclusion is that there is no barrier to the Democratic Party representing the identities of the white voters it lost in this election, because winning the support of a group is something you do by running token candidates with thick Southern accents and stuff like that; and, as his subtitle brilliantly foreshadows, there is no reason for the Democratic Party to try to draw back from the precipice of identity politics, because there’s no way you could do that — there’s only one kind of politics, you big silly!


I’m not going to engage with Yglesias’ piece in detail.  I’m also not going to deal with the the emptiness of the snowflake theory in its general form: DR has already covered the conceptual side for us here and here (not to mention here).  Instead I am going to focus on what you miss when you pretend that demographic groups don’t have distinct identities, don’t clash over  competing interests, and don’t face any choices about how to engage in politics.

A. Group interests

albinosGroups have competing interests.  Not all members share all of these interests, but enough of them do. Shared interests are like any other shared traits.  Under unusual circumstances, a member of a kind may lack one of that kind’s characteristic traits, but his odds of lacking most of these traits become vanishingly small if the kind as a whole shares even a dozen distinct characteristics.  Appearance, they say, is only skin deep; yet African albinos do not look white!  Skin, it transpires, is only one of several dozen visually salient features of human appearance, so those rare Bantu with atypical depigmentation still have typical skull shape, typical facial structure, typical hair texture, typical body habitus, and as a result can be immediately identified as albinos rather than Swedes.  The relationship between group interests and individual interests is exactly parallel; many are outliers with respect to one or two of the characteristic interests of their group, but only a tiny number are outliers with respect to a majority of these interests.

If you, as a candidate, associate yourself with a group identity you will attract its members, who will assume you will promote their interests; you will repel any groups with opposed interests for the same reason.  Such groups will almost certainly exist!  Group interests are only politically salient to the extent that there is some cost or risk to pursuing them.  Ethnic Albanians have an interest in having an atmosphere with adequate oxygen for respiration; but because there are no groups working against an oxygen-rich atmosphere (in fact, the Eternal Albanian shares this interest with all other groups) there is no risk of the opposite policy being embraced.  So there will never be an Albanian Lobby dedicated to pushing an oxygenation agenda.  While groups strive to portray their own narrow interests as identical to the common good, and imply that members of other groups neglect them only due to lack of “awareness” (so that the oversight can be resolved by “raising awareness”), in reality there are no free lunches.  Sometimes a group has interests which are matters of indifference to some other groups, but any politically salient proposal which is in the interests of one group must be against the interests of at least one other group.  The proposal wouldn’t become an issue otherwise.

No discussion of identity politics, therefore, can be considered complete without a discussion of (a) which groups you are planning on screwing, and (b) whether you expect the groups you are planning on screwing to vote against you (but in smaller numbers than the groups you’re helping) or to ignore their own group interests and vote for you.  In fact, no discussion of identity politics can be considered informative without this discussion.  The snowflake theory of identity politics, by this standard, is non-informative.

B. Political principles and identity politics are substitutes

If you campaign on principle, you attract people (of all ethnic groups) who find the principle compelling or profitable, and repel those who find it distasteful or burdensome.  If you campaign on identity, you attract people (of all different ideological leanings) who fall under that human kind and expect you to favor their kind in your policies and decisions; not only will this favoritism probably benefit each of them personally, but it will be even more attractive to the extent that they have more friends within their own kind and identify/sympathize with strangers of their own kind.  However, people who belong to different kinds, who mainly have friends of different kinds, or who mainly identify/sympathize with different kinds, will be repelled for exactly the same reason.

If your principles are sufficiently attractive to voters to win elections, you don’t need to campaign on identity.  If your identity is sufficiently attractive, you don’t need to campaign on principles.

Politicians who campaign on principles don’t need to pledge absolute devotion to a single principle.  Generally a principled candidate will give at least some weight to a basket of multiple principles, and then run on an official platform which shows how the principles, in combination, justify a concrete list of policy stances.  In some cases different principles have an affinity for one another, and in others they simple cover each other’s blind spots.  But it is also perfectly possible to campaign on flatly inconsistent principles in order to appeal to adherents of both principles, or to principled voters who can’t decide between them, or (most frequently!) to more self-interested voters who would not benefit very much from a rigorous application of either principle by itself.  For example, consider a utilitarian principle of maximizing the total consumption of all citizens and a maximin principle of minimizing inequality in consumption.  While these principles are logically incompatible and give different answers to almost any social question you might ask, they actually combine very smoothly.  Any two principles (in fact, any number of principles) which try to maximize some outcome can be combined into a hybrid principle which maximizes a weighted index of the maximands of the individual principles.  In practice this can mean that, on every issue, a candidate running on two principles offers a proposal which is a compromise between the “pure” policies either principle would have required on its own, or it can mean that each issue is governed by the principle whose goal it affects most strongly.

The same logic applies to politicians who campaign on identity.  When you engage in identity politics, you don’t need to pledge absolute devotion to a single group.  Sometimes different groups, despite having different interests overall, have a fortuitous alignment in all of their major interests which makes a tribal alliance easy.  In other cases, two groups’ interests center on such unrelated issues that there are no difficulties promoting the interests of both groups at the same time.  And because a group member’s allegiance to group interests which are not his personal interests may be motivated by social ties and sympathies within the group, two groups which socialize and identify with each other will be more inclined to support each other’s demands.  But even beyond potential grounds for cooperation, candidates who campaign on identity can offer platforms which identify them with multiple groups with disparate interests. There is no practical obstacle to putting a certain weight on the interests of each group and, for each issue, promising to support a policy that will maximize a weighted index of the welfare of the different groups in his coalition.

Principled candidates who go into politics to champion one particular principle can easily tack on additional principles to build a broader base of support, if the original core principle animating the candidacy doesn’t seem to appeal to an absolute majority of voters.  Identity candidates who go into politics to champion one particular group can easily tack on additional groups if their own ethnic machine is not, by itself, enough for a win.  But it is difficult (even in theory) to combine appeals to principle and appeals to group identity.  Principled politics asks what features of a policy make it a good policy; identity politics asks which groups benefit most from it.  Adding even a little bit of identity politics into your principles is like adding a little bit of piss into a glass of wine: even if the glass is still mostly wine, if you take a sip you’re drinking piss.  On a logical level, a machine politician’s group identity gives him a criterion for choosing which abstract political principles he prefers (whichever one will be best for the group whose interest he represents), rather than a goal which can be weighted against the goal of distributing benefits impartially according to some general principle.

Fortunately our politicians are not strictly rational, so nothing prevents them from promising some people they will look out for the interests of certain group, other people that they will follow a principle in an even-handed manner, and governing in a way that is equally disappointing to both groups of constituents.  However, when you assure a voter that you don’t actually believe anything you say so he shouldn’t worry about inconsistent promises that you’ve made to other groups, he probably won’t feel reassured!  At best, if you run on a “compromise” platform of doing what’s best for Group A and follows principle X, rigid Xists will worry that the “compromise” will take the form of following X when X is good for A, and ignoring X when X is bad for A.  That kind of compromise would similar to (if not indistinguishable from) just doing what is best for A, and ignoring X.  Pragmatists who don’t adhere to X as an ideology but would be happy to see the electorate converge on X as a stable consensus, a fair way to resolve conflicts that avoids a see-saw struggle for total victory, will be skeptical for similar reasons.  If you, the candidate, are primarily motivated by A’s interests and are only willing to suggest following X because it seems to you that X is in A’s interests, then as soon as X no longer seems beneficial to A you will drop X.

This piss-effect further polarizes the two types of political appeals because, to the extent that principled voters expect you to know how suspicious they are of appeals to identity, if you do make appeals to identity that means you aren’t very afraid of losing suspicious principled voters.  This in turn implies that you expect to win without strong support from the voters whose principles you have appealed to, and if you aren’t too worried about losing their support before the election, then you won’t be anxious to keep their support by following through on your promises to respect their principles, either.  (The same logic applies, in attenuated form, when principled candidates make anemic attempts to identify with a voting bloc they want to add to their coalition.  Think Ted Cruz and the butter cow.)

C. Adapting to identity politics

Imagine a teacher bakes a cake for her class and has them vote on how to split it up.  Whichever proposal for splitting it up gets the most votes determines how the cake is split up.  Maybe there are some students in the class who vote to get the whole cake themselves, but so long as each naively selfish student only votes for himself, there is only one vote for his proposal, and it only takes two good-natured students voting “give every student an equal slice” to overrule the many different selfish plans.

Now, you could imagine more than one idea for a “fair” distribution of cake.  You might give every student an equal share. You might give students on the honor roll a double-share. You might make the size of the slices proportional to each student’s height.  You could have a spelling bee and split the cake between the last eight students standing. You could have some sort of auction.  These are all at least nominally fair, although they correspond to different ideas of what makes two students equally deserving of cake (how much food he needs, whether he’s a good student).  Different judgments about who deserves cake are likely to appeal differently to different students both for selfish reasons (saying “tall people deserve more cake” is worth it to get more cake, even if you don’t believe it) and due to different perspectives on the evidence (you know you’re ravenously hungry when your midget friends are pecking away at lunch and throwing a lot of it out).

You could also imagine some group of students realizing (maybe after the teacher has already brought in one or two cakes) that continuing to vote for naive selfish proposals is imbecilic, but since only a few goody two-shoes are joining together to vote for a fair division, a small clique of selfish students who voted to split the cake among themselves could get the whole thing.  Say the class has two goody two-shoes, Xander and Yvonne (maybe Zane is principled too, but he voted for a height-proportionate division of the cake); then if Alice, Bert, and Caroline vote for all the cake to go to Alice, Bert, and Caroline, the ABC proposal has a plurality of the votes.  They get to split the whole cake three ways.

“If the students were at all sensible, then the next time the teacher baked them a cake the naively-selfish kids would stop dicking around, join Xander and Yvonne, and stop the ABC-triumvirate from taking all the cake again!”  But more likely in the real world you’d get the other students scheming and conspiring to make a four-man alliance to beat ABC at their own game, and then a five-man alliance to beat that one.  Along the way, as the naively-selfish voters form sophisticated cliques in pursuit of cake and glory, they will start to learn the difficulties of coordinating cliques with arbitrary members.  Difficulty #1: people lie.  If Alice, Bert, and Caroline add Dick to their clique to form a 4-vote bloc, they have no reason not to encourage Edna to vote ABCDE, Fred to vote ABCDF, and Georgia to vote ABCDG.  These proposals would fail, because each will only receive one dupe’s vote.  But by fooling Edna, Fred, or Georgia into thinking they belong to the ABCD clique, ABCD prevents them from adding a fourth vote to some rival voting bloc while still getting a quarter of a cake each.  Difficulty #2: people are greedy.  If Caroline and Dick have agreed to the ABCDE coalition with honest intentions, but then Alice and Bert come to them secretly and say “We’re voting ABCD, take it or leave it”, what are they to do? If ABCDE only gets three votes (remember, A & B say they won’t support it), it is likely to lose; sticking with Edna and cutting out A & B with a CDE-bloc still only has three votes; so C & D have no better option than to go along with the greedy ultimatum that A & B are pushing.  Difficulty #3: people can be complicated.  Edna may be willing to join your clique, but only if she can bring her friend Fred in.  Fred will betray his current clique for yours, but he doesn’t want his ally Georgia to think he’s leaving her because he prefers Dick: so to get Fred, your clique has to add Georgia or kick out Dick.  And remember, not only does the membership of a clique need to match the demands of each of its current members, it also has to remain the same up until the vote actually takes place.  In other words, kicking out D to gain E & F may seem like a good idea… until E & F defect to a different clique at the last moment, leaving your voting bloc smaller than it was to begin with.  In turn, worries about group stability may be what motivates the members’ clique-membership demands in the first place.  They want a clique they can trust, and they also want other clique members who can trust them.

These difficulties students face in forming stable cliques are essentially coordination problems.  If a student’s vote is driven by sophisticated selfishness, he is probably happy to vote for any proposal that (a) includes him as one of the kids who gets cake and (b) would win a plurality of the votes; of possible proposals that match -a- and -b- he will lean towards those that (c) award the biggest slices to the winning clique. The problem isn’t finding a proposal that he is willing to vote for, the problem is making sure that he’s voting for the same proposal as a plurality of the other students.  There are too many possible proposals rather than too few, and too many ways for a big bloc like ABCDEFG to be torn apart into factions proposing ABCDEF and BCDEFG… or even just factions who are afraid that their (current) allies are secretly planning to defect in that way. The bigger the group, the greater the uncertainty about whether one of the marginal members will be dealt in or out when the final ballot comes.

The solution to this coordination problem is to find a easily identified stable point of agreement.  For example, the black students in the class might start suggesting that they could vote for the proposal “Only black students get cake.” This proposal has a few downsides… for example, some black students might vote against the Black Cake Matters platform, and they would get pie anyway.  But the overwhelming advantage is that “Only black students get cake” is a much more salient proposal than any of the competing versions where all of the black kids plus one more white kid get cake, and it is also much more salient than any competing version where all of the black kids except for one black kid get cake.  If all of the black kids who are considering joining the BCM clique know that there is no alternative similar to BCM which could possibly do better than BCM, then they know that everyone who says they are going to support BCM is probably going to stick with BCM (instead of changing to some similar but more complicated proposal at the last minute); this is, in itself, a reason to believe that BCM might be a winning platform, and could be worthy of support.

Group members share the traits that are typical within a group, and a fortiori they share typical interests; but even if the group has no common interests, identity politics emerges nonetheless as the rational solution to the coordination problems that selfish voters face.  If you are a sophisticated selfish voter, your tribe is your Schelling point.

If you’re a principled voter… well, most principled people are only willing to act in a principled way so long as they believe they are dealing with other principled people.  If no team ever wins without playing dirty, either you learn to play dirty or you find a different sport.  If only tribal candidates ever win the election, you stop seeing it as a referendum on how the government should be run and starting seeing it as a referendum on which tribe should be running it.  Most people are very strongly motivated by the realm of the possibility, whether in politics or in anything else; once one tribal coalition mobilizes itself with enough success to make the victory of some tribe inevitable, everyone else will end up concluding they want it to be their own tribe, sooner or later.  In fact, once tribalization is underway, members of a group start expecting every member of the group will do his part to ensure the group wins, and treating anyone who violates this norm as a traitor and a worm.

So, returning to our example: if the black students were to successfully band together and vote to give themselves the whole cake, the other students would learn from the experience and split into whatever ethnic factions were appropriate to the make-up  of the class: Hispanic Cake Matters, White Cake Matters, Jew Cake Matters, etc.  If in the next cake-vote one of these groups lost due to principled students like Xander and Yvonne ignoring their group interests to vote for a fair proposal, some combination of the futility of defeat and bullying from their friends will push them to vote on tribal lines.

To recap the Parable of the Cake:

1. Naive selfish voters learn that pure selfishness is pointless, and become sophisticated.

2. Sophisticated selfish voters discover the magnetic attraction of identity politics.

3. All selfish voters are drawn to identity politics.

4. Principled voters learn that there is no point in not voting for their identity group.

And a final recap of the whole post:

A. If you are promoting the interests of one group, you are hindering the interests of some other group; if the one has reason to support you, the other has reason to oppose you.

B. If you hoping to win the support of some groups by appealing to their group interest, that hampers your ability to appeal to other voters’ principles.

C. If you are hoping to win elections primarily on the basis of appeals to group interests, your opponents will learn to do the same thing.

Am I crazy to think these points are so simple and incontrovertible that they don’t normally need to be explained or justified?  Perhaps the main function of the Lügenpresse is simply to make the average citizen feel like he’s crazy whenever he exercises common sense.  Once in a while we must rebuild common sense from its foundations to verify that we are not.

6 thoughts on “Matthew Yglesias, crackhead, and his snowflake theory of identity politics

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s