Eric A. Posner

Overall, one of the best books I've ever read. Basically describes ways markets could be used to increase economic inefficiency, decrease inequality (and in particular, increase the proportion of rewards going to productive labor vs. accumulated capital), and fix various systems (political and social). Largely an extension of the ideas of William S. Vickrey.

Things I particularly liked about the book: a credible argument against "conventional" Georgist land-value tax (difficulties in valuation), and an interesting alternative. A superior form of voting (accumulated/bankable votes). "Quadratic" increases in cost for certain policy preferences (such as reducing pollution, or regulations. An interesting immigration scheme where individuals could sponsor foreign workers, gaining a portion of their income, to more broadly distribute the benefits of immigration along with costs.

The chapter on data (data sovereignty, data markets, etc.) seemed pretty weak in comparison to the rest of the book (ironic given the background of the authors), and detracted from the whole.

One interesting element was prefacing each chapter with a fictional story of what life would be like under their proposed rules -- more abstract policy arguments should include these.

There were a lot of flaws with the specific proposals they make, and overall I think for most private property, taxation is theft, and their taxing schemes were in some ways even more immoral than the status quo (personal/portable property being taxed in such a way that third parties could forcibly purchase it for the declared value at any time seems rife for malicious exploitation by trolls, or effective censorship of unpopular people), but they propose testing in much less challenging environments (such as as an alternative way to distribute public assets like radio spectrum or resource exploitation on public property), which I'd support.

I strongly recommend this book.… (mehr)

This is an excellent book full of well-designed mechanisms for finding good Nash equilibria in society. From taxing possessions based on how valuable their owners consider them (and being required to sell them at that price if someone asks,) to a quadratic voting scheme for balancing the weight of the many against the passion of the few, this book nails lots of weird ideas for what life could be like. But of course people are too risk-adverse to try anything of the nature, and so we'll be stuck with the status quo until the collapse of society.… (mehr)

I'm not entirely convinced by any of the proposals, the writing sometimes bogs down, and despite containing more detail than the average popular book, some of the explanations feel quite handwavy. But this is tasty food for thought.

The proposals include:

* A 'common ownership self-assessed tax' (COST), which in its most radical form would be applied to all property. Owners would have to self-assess the value of everything they own, and would be taxed accordingly, at a rate that would vary according to the turnover rate of each type of asset. (Items that we expect the owner to hold for a long time would be taxed at relatively low rates; this is done to incentivise accurate assessments, but also has the side-benefit of making it relatively inexpensive to protect one's ownership of items with primarily sentimental value.) The catch is that you would have no choice but to sell your possessions to anyone willing to pay the values you assigned.

* Quadratic Voting (QV), in which voters would be granted 'voice credits' which they could distribute across elections (or referenda -- QV is supposed to be combined with direct democracy) based on the relative strength of their preferences. Voice credits are used to buy votes, and the price of a number of votes on a single decision is the square of that number. This non-linear structure encourages voters to express their opinions broadly rather than banking all of their votes for maximum influence on one or two issues, but still enables people to express the strength of their preferences and to have disproportionate influence over the issues that matter most to them.

* A more open international labour market, in which regular citizens of rich countries could sponsor migrant workers, taking some responsibility for their living conditions and conduct in exchange for a share of their income. (The authors call this a Visas Between Individuals Program, or VIP.) The idea is that this would improve economic opportunity for citizens of poor countries, spread the benefits of migrant labour across average people in host countries rather than concentrating them with business owners, and soften the ethnic/cultural/racial tensions that are most likely to arise when host country workers feel that they are being displaced. The idea is that host country citizens know the conditions on the ground, so they would be able to accurately assess the demand for certain types of workers, as well as making informed judgments about cultural fit. The authors predict that, as well as economically benefiting both the global poor and the poorer members of rich countries, their system would soften xenophobic attitudes in these countries, replacing hostility with benign condescenscion. This is not exactly a utopian vision, but would (if it worked) do more actual good than paying lip service to ideals of global equality while taking an out-of-sight out-of-mind approach in practice.

* Financial reforms, the details of which I can't remember. Part of the idea was to weaken institutional investors' control over entire industries, which has predictably anticompetitive effects. Seemed sensible but politically difficult.

* Creating an explicit market for the personal data that we currently give away for 'free' (i.e. in exchange for 'free' services) to the big tech companies. This chapter seemed kind of handwavy to me, and I don't really see a plausible path forward. The authors are big fans of Jaron Lanier, whose book I gave up on because it was frustratingly vague and contained at least one glaring error. I guess they must all have good ideas that I'm failing to grasp, but I wish these were communicated more precisely, along with answers to the obvious sceptical questions.


I'd say (on the basis of no relevant expertise) that the VIP is probably the most likely proposal to be useful in practice, though the details would be tricky and there are all sorts of ways it could go wrong. The COST and QV are the most interesting to me intellectually, but I have a lot of concerns about their likely pathologies. I made some notes on these, which I'll copy here with a few edits, because I don't have the energy to try to shape them into proper arguments right now. (Spoiler-tagged for being long and half-baked.)

COST
* poor people lose the freedom to retain their property -- it's all very well to say people can set an above-market price if the asset is that important to them, but that's meaningless if they literally can't afford the associated tax (rent) rate
** yes, governments can already seize property, but this is relatively rare and not entirely unpredictable
* poor people would not benefit much from appreciation, e.g. of a once-cheap house? (if they buy it for $100k, and can only afford the corresponding level of rent (tax), what can they do once the value rises? It seems like they have to sell at a price low enough that they can still afford the rent.)
** maybe they can borrow against the (new, higher) value, and use that money to pay the rent? This is pretty risky though -- if the value drops, they're left with a debt they can't afford
** It seems that the economic logic here is that the turnover of this house is a good thing -- the asset is being shifted to a 'higher-value use'. But this is obvious bullshit -- the person who buys the house doesn't necessarily value it more than the current owner in any meaningful sense, they're just richer, so a dollar is worth less to them on the margin.

Quadratic Voting

* would people vote more selfishly, primarily by saving up their votes for issues that affect them directly? I think currently we are able to vote on moral grounds partly because we know we don't have much influence. If we felt we were capable of saving up enough to have a real chance of affecting the result, I suspect self-interest might take over. Suppose you are a member of a vulnerable minority -- are you generous enough to spend many of your votes on issues affecting other minorities, when you know one day your own minority might desperately need as many votes as it can get? Currently self-interest might even push toward helping the other minority (or at least claiming to), in the hope of reciprocation. But if voting were something like a zero-sum game (votes spent on one issue could not be spent on another) would you really be willing to trust that your generosity would be returned? And even if you could, why bother, unless the other minority is larger than yours (and can therefore reciprocate with interest)?
** I suppose this is the system working as intended -- what we want is people honestly and accurately voting their own interests. I'll have to think more about this, but currently it leaves me uneasy.
** By likening QV to a 'zero-sum game' I was kind of missing the point -- you have a fixed supply of voice credits, but because of the quadratic pricing, the more evenly you distribute credits between issues, the more votes you cast in total. Still, the way to maximise your influence over issues that affect you personally is not to vote on anything else.
* strategic voting would be rife (unless things were so unpredictable that it came to be seen as futile). If you save up credits by underplaying your preferences on issues that seem unlikely to be close, you can have a greater chance of affecting close votes in the future.
* some old people would have no saved votes, not necessarily because they care any less than when they're young, just because they didn't think things through. Others would have many saved votes, not because they care a lot now, just because they couldn't be bothered voting before.
** maybe in order to bank votes you have to vote? e.g. if you turn up and cast one vote, or even no vote (i.e. register your indifference), you get to bank the votes granted for that issue, whereas if you don't even turn up you forfeit them?
* people do not all care the same amount overall, nor are they equally affected overall. In the ideal case QV registers the relative strength of my preferences (i.e. relative to each other) but not the strength of my preferences relative to those of someone else. There might be a lot of apathetic people who accumulate a lot of votes, then blow them all on one issue on a whim, or because they're whipped up by someone who does care.
** would political campaigning become even nastier and more fear-based? it would no longer be enough that people favour your side over the other, and (in non-compulsory systems) care enough to vote -- you would benefit greatly if you could make them care intensely.
* future politics are unpredictable. If issues A and B are extremely important to me, how do I decide how many votes to spend on issue A today, when I can't predict whether issue B will come to a vote in my lifetime? (Maybe I'm not even consciously aware that B could come up at all.)
** maybe it would work better to allocate votes per-election, have many issues decided at a single election, and not allow banking of votes across elections.


Overall this book isn't everything I wanted, but I recommend it. It would be well suited to a book club or university tutorial, or anywhere else you'd have a chance to discuss it with other readers and perhaps drill down into the technical details.… (mehr)