There is a war.

Households vs markets

The first symptom was the clutter on the kitchen counter. One cutting board, two pans, one knife. My colleagues had arrived at the rented house the day before, so they'd had plenty of time to arrange things to their liking. I was sure the clutter was not to their liking, if they noticed it at all.

A teachable moment. Instead of tidying the counter myself, and accreting a small amount of resentment, I suggested to one of them that she think of the things on the counter as things that were in her power to arrange however she liked, to suit her taste, selfishly. (“Just as you might optimize your text editor to suit your workflow,” my other colleague chimed in.) She took this suggestion, and spent a few minutes arranging and rearranging the items on the counter. She put away the knife in the knife block.

A puzzle. She noted that she doesn’t usually think of the items in her home this way. Instead, household chores feel as though they are impositions from an abstract, outside authority. She was capable of accessing this other, more pleasant way of working on the things around her. Why wasn’t it natural for her to feel that way in her own home?

An hypothesis. In our usual mode of life, there is a separation between a job - which is done for someone else, to satisfy someone else’s standards, outside the home - and consumption, which is at least ostensibly done to suit one’s own taste. One of the goods you can buy with an income from a job is a nice place to live, and you can also buy services to keep the place clean and tidy. For the most part, you maintain the place you live by leaving it, and entering the domain of an outside authority. Household chores are the remainder that cannot efficiently be outsourced, or an echo of a previous era in which such outsourcing was less common.

A household. I pointed out that in the past, people mostly did not work for a salary. Instead, they worked on things in their local environment. For the most part, this was not in order to receive money, but in order to have the thing, even if it was part of a process directed towards producing goods for trade. Home improvement used to be, not a hobby or special interest, but what one did if one wanted an improved home. Even those without a substantial amount of arable land to work might keep a garden. A minority engaged in trades or professions. Loops were mostly closed, either at the level of the household or at the level of the village.

A leak. My colleague pointed out that taxation means you can’t actually just have a closed loop anymore, even if you were willing to do without modern inconveniences. To occupy a space, even a space you legally own, unencumbered by any debt or lien, one must pay rent in the state’s currency. If you don’t pay, you have to leave. This means that at least some of a household’s activity has to be, in effect, defense spending: providing the state with mercenary services, directly or indirectly.

External standards. It’s worth noting that the phrase “good enough for government work” was originally a compliment; it referred to the exacting procurement standards of the US Government, more precise than the standards people were otherwise accustomed to. Things need to be machined with more legible precision if the customer can’t just go consult with the craftsman.

Tribute. Likewise, it’s a commonplace that a village can’t improve its prospects by taking in each other’s washing. But, actually, a village can’t improve its prospects by engaging in any sort of closed-loop commerce; something must be procured as tribute. In a country with an income tax, exchanging services - if legibly enough to be detected - actually causes an outflow of wealth.

Why are we doing this to each other?

Taxation is mobilization

Imagine a peaceful village with a closed economy. Much of the villagers’ productive activity is not transactional, but simply working to make the village a better place. The reward for producing food is that the village has more food. Perhaps some transactions are market-based, though few will be arms-length. If so, it is easy to imagine that a precious metal like gold might be, if not an actual medium of exchange, at least an unit of account. In addition, people often maintain their own households, and improve them. Again, the reward is an improved household, not money.

This way of life affords a freedom to which the people of my generation and the generation before are not accustomed: if you don’t like the activities by which one might get gold, you can just not do those things. Accordingly, social policing has to be comparatively direct. If you want to exclude someone from village life, you have to actually coordinate to run them out of the village. Especially if they own land to deprive them of their livelihood you have to physically expel them from their own house and land. Anything less is merely an inconvenience.

Suppose the Golden Benevolent Empire decides that this village, like many others along its border, needs defending against the Barbarian Horde. The Golden Benevolent Empire has limited resources, and insists that villages contribute to their own defense, in the form of professional soldiers.

One way to do this would be to demand specific in-kind contributions from each village. But centralized resource allocation of this sort is tricky, especially without reliable record collection and a very large computer, so it settles on a simpler expedient. Require a certain, small quantity of gold from each villager (a head tax, or maybe a land tax), on pain of death.

Now, everyone in the village needs to come up with at least a little gold. Not only will some people be happy to accept pay as professional soldiers, but the other villagers will be eager to provide whatever goods and services your army needs - so long as you are willing to pay in the same coin you demand from the villagers. Like magic, your army is fed, clothed, and equipped.

But something else happens. In the village, gold - which was perhaps used occasionally, either for transactions of ritual significance, the occasional foreign trade, or largely optional market transactions - is now just as necessary to sustain life, as food and water and shelter are.

Different villages may solve this problem differently. Some may coordinate to produce trade goods that can be sold elsewhere, otherwise retaining their accustomed mode of life. Others, already integrated into larger markets, may simply shift their production person-by-person slightly more towards trade. On the margin, closed-loop households are destroyed; no one can be fully self-sufficient. Villages nearest the military encampments will be in a good position to serve as intermediaries. Villagers from poor or remote villages, that can’t afford their tax, may move closer to these trade centers, or accept employment as mercenaries themselves.

In any case, all across the Golden Benevolent Empire, production shifts towards things that can be traded for gold, and people are more eager to keep accounts. Commerce, in other words, is booming.

In some ways, this points towards our present-day situation. We pay taxes, and this would be sufficient to explain why we can’t have totally closed loops. Small towns are dying, and people are moving towards participation in the global economy when they can. What it doesn’t explain is our apparent reliance on this process, such that when the demands made on us by the economy decline, our society doesn’t revert to more home production; instead, we anxiously seek out employment, and hope the state will stimulate additional demand.

MobsterBucks: an interlude

The need to concentrate resources to administer the Golden Benevolent Empire has led to the emergence of a few very large cities, where its administrators and soldiers have gold to spend. This drives up wages, which draws peasants from the poorer villages, the villages which are not productive enough to bear the burden of their land taxes. There are also great merchant ports, enabling gains from trade that could not have been realized before so many lands were unified and pacified under Imperial rule. Both of these cities also draw immigrants from other lands, because gold has no national allegiance.

The growth of these cities has outpaced the Empire’s ability to enforce its own laws, and in one of them, something strange and new is happening.

It is not at all uncommon that immigrant communities sharing a nation of origin have their own customs, norms, and ways of enforcing these. This can take the refined form of the Ottoman millet system, or the crude form of street gangs.

What is uncommon is how one of these gangs is administering its territory.

It is not at all uncommon for a gang to demand payment from the residents of its territory, in exchange for “protection” - both from outsiders, and from the gang itself. But in many neighborhoods, the amount of hard currency available is nearly nil - the people live on credit. Demanding in-kind payment is administratively difficult, for the same reason the Golden Benevolent Empire ultimately decided against it. One enterprising gang has a solution: MobsterBucks.

The proposition is simple, perhaps inspired by a folk tale about a village oppressed by archers. Each person in this gang’s territory has to pay one MobsterBuck per year, lest they meet with some sort of mishap like a broken kneecap or worse. How does one acquire MobsterBucks? Well, the gang is the sole issuer of said currency, and is happy to exchange it for goods and services it needs. Soon, the internal economy of this community is denominated primarily in MobsterBucks.

Having implemented this scheme, this enterprising gang now needs a monetary policy. If it spends too many MobsterBucks, then people will be less afraid of injury for want of MobsterBucks, and the exchange value of the currency will decline. This can of course be corrected by simply raising the rate of taxation - demanding more MobsterBucks from each person.

On the other hand, with an imbalance in the other direction, an inefficiently high number of people will be kneecapped, permanently reducing the productive capacity of the gang’s territory. Another problem with deflation is that inside the gang’s territory, debts are now denominated in the convenient unit of MobsterBucks. Anything that increases demand for MobsterBucks may unsustainably immiserate a large number of debtors, forcing them into less-productive debt slavery, and substantially eroding any good will towards the gang.

What to do if the gang is spending too few MobsterBucks? Giving money directly to the people lacking money would eliminate the incentive to work, which would reduce the productivity of the gang’s territory, but the gang can correct the imbalance indirectly by increasing spending or reducing taxes.

One might imagine that to avoid either an excess or shortfall of MobsterBucks, the correct policy would be a balanced budget: demanding and spending the same number of MobsterBucks each year (or other budgetary period). But in practice, some wealthy people in their territory may hoard MobsterBucks against a future need, leaving less than nothing for others, who may go into debt - leading ultimately either to debt slavery to the richer ones, or to kneecapping. So the optimal MobsterBucks spending and taxation levels are not obvious, and require considerable sensitivity to economic conditions.

A second question the gang may face is the optimal level of resource extraction. If the neighborhood is transient, or they face an emergency situation that requires all the resources they can bring to bear, they may as well extract all they can, as quickly as they can. On the other hand, if the neighborhood is comparatively stable, they may want to extract as little as they can, in order to allow the reinvestment of productive resources. The exception is the occasional public good; where the gang is well-positioned to make productivity-enhancing infrastructure improvements to the neighborhood, or subsidize otherwise undercompensated activity.

To some extent the Golden Benevolent Empire must account for the same things, but these are somewhat obscured by the fact that gold also has foreign exchange value, limiting the Empire’s freedom of action. I waited for MobsterBucks to introduce these complications, as a simpler example.

Wartime inflation

The Golden Benevolent Empire is at war! Its very survival is at stake! More soldiers are needed! More equipment is needed! The treasury is spent down, and taxes are raised. Whole villages are reorganized, to meet the increased demand for arms, supplies, and soldiers. People swarm towards the cities and fortresses and arms factories, where the money is flowing. Areas that are not so useful to the state suffer under the new taxes, and sink into debt.

The value of money changes during this period. The Empire’s demand for additional resources is greater than its territory’s ability to supply them, and more money chases increasingly scarce goods. Prices rise, and common citizens must make do on less. In war time, one needs national economy.

Interest rates also change. As the decisive battle approaches, the Empire would much rather have money now, than a year from now. It is willing to borrow, even at high rates of interest, since if it loses this battle, it loses everything. The towns and villages of the empire acquire a correspondingly high time preference. If you can get a better return on your money by lending it, than by reinvesting it in productive assets, then it is more profitable to do the former than the latter. This, too, is appropriate behavior for wartime; spend down resources, and recover once you are at peace.

But a side effect of this policy is that the poor sink into debt to pay their taxes, while those with money are further enriched by a high return on investment.

Then the decisive battle is won, and the Golden Benevolent Empire is at peace. Soldiers are released from duty, and sent back to their villages. Conscientious administrators and common citizens alike breathe a sigh of relief, and expect that after the austerity of war, they will reap the rewards of peace.

But things are not so easy. Enterprises that sprung up in the cities to serve the war effort find themselves suddenly out of business. The populations brought into the cities during wartime now find themselves with little to do. Many of them go back to their villages, but the villages themselves have less of their former character, and are oriented towards serving the national economy.

What’s more, while spending is reduced, this is not so much true of taxation. The Empire is honestly administered, and if it borrowed, then now it must pay back its debts. But this deflationary policy forces indebted farmers off their lands, reduces businesses that borrowed to support once-profitable enterprises to bankruptcy, and immiserates whole villages.

The Imperial government is not pleased. Imperial administrators have hearts, like anyone else, and they did not fight a war to immiserate their citizens. In the heart of the Imperial Treasury, a clever bureaucrat from humble origins in a poor area of the capital city comes up with a plan, based on what he saw his neighborhood’s street gang do.

In his plan, the Empire will confiscate all gold, and replace it with a scrip issued by the Empire: GoldenBucks. Creditors must now accept payment for all private debts in this currency, and it can also be used to pay public debts - taxes. The gold will be enough to pay off foreign creditors, but the Empire is no longer constrained by honor to tax as much as it spends. It can simply issue more GoldenBucks. As the people need employment, the Empire will spend or lend its newly minted GoldenBucks, until demand matches supply.

A massive, disruptive deurbanization is thus averted, and the Imperial economy continues its operation. However, the resulting taxation means that resources and attention still flow towards the places the Empire spends its money.

Of course, the government will still be constrained by a need to preserve the value of GoldenBucks - it would not do to disrupt the national economy too much, too quickly - so taxes will still be collected.

The world wars

In the first half of the 20th Century, the world order was shattered by two successive, cataclysmic wars. Great empires were brought down. The country that ultimately emerged victorious - and became the world’s sole superpower - won with a strategy of complete economic mobilization. In between these world wars, this country experienced a massive economic disruption in which a full quarter of people who wanted a job could not find one. The government responded by confiscating all hard currency, and substituting pieces of paper that its citizens were required to accept for debts previously denominated in gold. It then engaged in massive public works projects, paid for by this new, fiat currency.

The symbolism of this new currency is a bit too on the nose for me to include in a fictionalized narrative. On the obverse, a portrait of the country’s highest-ranking general, and a statement that this note is legal tender for all debts, public and private. On the reverse, a pyramid, perhaps the most famous of monuments built by centrally planned, conscripted labor, topped with an eye, a symbol of surveillance. And a bird of prey, holding an olive branch and an arrow; production and violence bound together by the agency of an apex predator.

These world wars coincided with unprecedented levels of urbanization. At the beginning of the century, 40% of the country’s population lived in cities. Halfway through, after the wars, 64% did. These wars were not isolated events. The transition to a state of total mobilization was happening before them. (In 1860, at the beginning of the US Civil War, only 20% of the US population was urbanized.)

The timeframe of this massive urbanization has roughly coincided with the timeframe in which recessions begin to be a thing. In earlier societies, we hear about immiseration, about famines, about disruptions in trade, about heavy taxation and debt serfdom, but not recessions. That’s a modern thing - we start hearing about them and related financial panics after the US Civil War, which was perhaps the first modern war with total economic mobilization.

What’s perhaps more surprising is that the trend has continued. After the conclusion of the second world war, the victorious US elected its foremost general as president. In his two successive four-year terms as leader of the free world, he presided over a transition from a system with the capacity to mobilize on demand, to a system of permanent readiness, the military-industrial complex he described in his valedictory speech:

Our military organization today bears little relation to that known by any of my predecessors in peacetime, or indeed by the fighting men of World War II or Korea.

Until the latest of our world conflicts, the United States had no armaments industry. American makers of plowshares could, with time and as required, make swords as well. But now we can no longer risk emergency improvisation of national defense; we have been compelled to create a permanent armaments industry of vast proportions. Added to this, three and a half million men and women are directly engaged in the defense establishment. We annually spend on military security more than the net income of all United States corporations.

This conjunction of an immense military establishment and a large arms industry is new in the American experience. The total influence -- economic, political, even spiritual -- is felt in every city, every State house, every office of the Federal government. We recognize the imperative need for this development. Yet we must not fail to comprehend its grave implications. Our toil, resources and livelihood are all involved; so is the very structure of our society.

In the councils of government, we must guard against the acquisition of unwarranted influence, whether sought or unsought, by the military-industrial complex. The potential for the disastrous rise of misplaced power exists and will persist.

[]

Another factor in maintaining balance involves the element of time. As we peer into society's future, we -- you and I, and our government -- must avoid the impulse to live only for today, plundering, for our own ease and convenience, the precious resources of tomorrow. We cannot mortgage the material assets of our grandchildren without risking the loss also of their political and spiritual heritage.

We do not seem to be in the middle of a cataclysmic war. Even the “cold war” against Russia appears to have been decisively won by the liberal Anglo-American system. But we behave as though we are still at war. I don’t just mean the routine bombing of foreigners, or maintenance of a huge permanent military establishment. I mean that we still have an elevated time preference, as though we were in a state of emergency; 10% is not an unusual internal rate of return for major corporations. Many businesses’ effective time preference is even higher. Perhaps we have become so accustomed to wartime mobilization that we don’t remember any other way of life.

The scarcity factory

Recall that between the world wars, the US (and much of the rest of the world) experienced, not peacetime prosperity, but a massive economic contraction leading to immiserated laborers, driven out of their homes and forced to wander as vagrants.

The orthodox policy solution is to create demand for labor. To manufacture scarcity. To create a pressure differential between money sources and money sinks, such that almost everyone in the country is required to do things, to alleviate that pressure.

But in the process we have become accustomed to accumulating wealth in the form of financial instruments and rising prices, rather than improved homesteads. These are the sort of wealth one can accumulate during wartime, but are only valuable as claims on the work of others. We can’t become richer by all going into debt to each other. So someone has to become poorer.

The composition of major businesses reflects this. The growth of the financial services industry is often cited in this context, but a classic business with a real physical product, like Coca-Cola, is a marketing company dedicated to persuading people to drink flavored sugar water. Facebook is mostly trying to maximize the attention it uses up. More generally, business as we know - especially weighed by profitability - is largely marketing, in the sense of creating needs. When it isn’t, it’s often about bottleneck capture. (There are of course major exceptions.) This is a battle for control of a fixed resource (cash flows), not production. Business books and news articles routinely use the framing of war to describe the running of a business. One is reminded of stories about how groups of chimps that encounter a stable food source begin to fight over it.

Our rulers didn’t create this system out of perversity; they did it to win two successive world wars. We can think of countries as engaged in an adversarial game, making tradeoffs between creating resources, and using resources in adversarial contests. The character of modern war has been such that while rapid militarization typical of the German strategy has failed, persistent economic mobilization of the kind employed by the Anglo-American alliance has dominated. Countries that succeeded in relaxing back to peacetime standard were perhaps simply selected out of the pool by means of conquest long before.

That doesn’t make the consequences of permanent mobilization any less unfortunate. Lifespans are declining at the center of the empire, though not yet at the periphery. This suggests that near the center time preference has increased to the point where we’re creating scarcity faster than we’re alleviating it, while at the periphery scarcity is still actually being alleviated because there’s enough scarcity to go around, or perhaps marginal areas do not suffer so much from total mobilization.

The friends I grew up with still live in a world where they can’t help with the family homestead because there is no homestead, even though in most cases they grew up not in a city, but in a house in the woods. If they can’t find a job with wages, they’re in serious trouble, even if they’re willing to work, even though there’s no war on, and no conscription. A high cost of housing, and tax burden, means you can have a situation, but not really property. The state will be happy to kick you off your land if you don’t cough up tribute, and this is the new normal.

https://imgflip.com/i/2aw543

https://imgflip.com/i/2aw543


Thanks to Jessica Taylor for helping me think through the households example, Jack Gallagher and Ben Pace for independently suggesting the idea of MobsterBucks, Wei Dai for pointing out the possibility of strong selection for political orders that can engage in total mobilization, Michael Vassar for independently noting that WWII in a sense never ended, and I owe a debt to David Graeber for the basic framework of taxation as mobilization.

45 thoughts on “There is a war.

  1. Paul Christiano

    I agree with some but not all of this story.

    The time preferences thing doesn't really add up, the current real risk-free rate is ~0% (and investors can borrow several times their net worth at that rate if secured by equities), even equity yields are 3-4%.

    It seems like there is actually a premium on risk, especially on risky illegible projects (and essentially all illegible projects are implicitly considered risky by the market). This seems important but not really connected to wartime time preferences, more connected to long-term shifts towards decision-making systems that require legibility.

    In general, I don't think you can give a rational-society explanation for the observed behavior, being ready for war on a moment's notice is expensive but not this expensive.

    It seems to me like there is quite a lot wrong with our current economic situation, but I'm not convinced that it is related to war. That doesn't even seem like the right place to lay the blame in your hypothetical of the Golden Benevolent Empire's war; beyond the total burden of increased taxes (i.e. the real cost of diverting production to wartime activities) there is no particular reason that the mobilization should have further costs, something else has gone wrong. Obviously the problem isn't actually diverting production to wartime activities, since you can't manufacture food by by printing money. It doesn't seem to me like the actual problem here is connected to war or mobilization in particular.

    I might be misunderstanding your example though, I don't really get why they solved the problem by printing money. It seems to me like the only relevant effects are (a) moving real wealth from creditors to debtors, (b) making people more willing to accept a real pay cut (by raising nominal wages).

    Reply
    1. Benquo Post author

      the current real risk-free rate is ~0% (and investors can borrow several times their net worth at that rate if secured by equities), even equity yields are 3-4%.

      The behavior of businesses when considering concrete investments doesn't seem to reflect this, e.g. in that old Overcoming Bias post I linked to.

      Reply
      1. Paul Christiano

        I think this is largely the risk premium rather than time preferences.

        Also: in 2012 stable corporations really were borrowing at negative real interest (and in rare cases negative nominal interest). If their time preferences were really so high, despite the availability of cheap credit, they would have kept borrowing until interest rates rose. Real equity yields were 5-6%, suggesting a 10-15 year time horizon for projects with "normal" risk.

        Another aspect is that corporations tend to grossly overestimate the rate of return on internal projects (analysts are also preposterously optimistic in general), just like they overestimate the return on acquisitions and are generally too optimistic about themselves. Strong folklore is that business reinvestment on average manages to underperform the stock market, it could be a self-serving bias of the finance sector but I don't think that's the case.

        Reply
      2. Zvi Mowshowitz

        There's lots to unpack, but on this particular point, 5 years ago I would have said this is a clear contradiction but now I'm not convinced. This could easily be consistent.

        Suppose that the real risk-free rate is 3%. But also suppose that it turns out that historically, profitability this quarter has correlated highly with future profitability, such that algorithmic and short-orientation traders have learned they can make excellent profits by selling companies that do not do well in the short term (unless they are explicitly 'growth' companies like Amazon or Netflix, which we'll ignore here), and buying those that make higher profits than expected, unless the price has already adjusted to reflect the new information. Let us also presume that you can bark about how you're making investments in your factory all you like, but such improvements are hard to verify, thus easy to lie about, and also somewhat illegible to such algorithmic traders even if you are trustworthy. Now suppose that your job security and pay as the CEO or company board depends mostly on the short-to-medium-term share price.

        And of course, failure to placate those who determine the share price shows you are bad at doing what's best for the share price, and bad at playing such games in general, further depressing your career along with the share price.

        None of those presumptions seems especially unrealistic to me.

        The result is that it might be that the company acts as if illegible investments need to have a 50% return rate, despite the company being able to issue 7% bonds and the risk-free rate being 3%, because we're caught in an inadequate equilibrium. The solution to this problem, as Robin suggests, is for someone or some group with a lot of wealth, who notices such investment opportunities, to pay a premium (often quite large) to take the company private, and make such investments, and enjoy the arbitrage gains.

        Reply
        1. Benquo Post author

          This framework is helpful to me, but a more compact way to put it might be:

          1. Nominal interest rates are not as informative about time preference when you have a fiat currency.
          2. The need for a short-run cash flow profit implies something like a high time preference.
          3. The need to sustain numbers justifying a quarter-over-quarter growth narrative implies an extremely high time preference along a very different vector.

          I'm still a bit puzzled as to how and why our nominal interest rates can be so low sustainably in such an obviously high-time-preference society if you just look at things, but it seems like it might just be a persistent wealth transfer to banks and highly leveraged corporations - individuals mostly can't borrow at that rate, and high corporate rates of return on investment relative to their cost of capital imply high barriers to entry.

          Reply
          1. Zvi Mowshowitz

            It's a good point that saying that the CEO is under a different form of time preference still implies that a different vector ends up with a very short time preference. I think we currently have a huge problem that, because people use current measurement - and much worse, they actually use the rate of change of the current measurement and extrapolate it and use it as a measure of hidden variables - as a predictor of future measurement, no matter what the financial time preference is, we are forced to have an extremely short time preference on all metrics that are used for such prediction. An investor can have a true time preference of 0% and still force very short-time-preference behavior from places he might invest or short.

            As an individual, I can't loan money out safely and earn much; that's what the risk-free rate means. That's true because the banks (which are safe only because the US backs them, but that still counts) and other safe places to loan money don't need more money. This is symmetric if I can actually prove I'm risk free e.g. if I wanted to leverage up using my stock portfolio, I'd only pay a few percent. That has to be true, because otherwise someone could do an easy arbitrage (even if it had hidden risks, they wouldn't be taking them). Similarly, if I own a house I can borrow at 3.8% (as of when I checked a week ago), and the reason that rate is even that high is that I'm buying an option - I can repay that loan if rates go down, but keep the loan if they go up, so the bank has to charge me for that.

            What I can't do, is borrow money at a reasonable rate without giving the bank some assurance that they'll get paid back. And that's because often banks don't get paid back, so you're paying to compensate for that risk.

            If you could issue a 7% interest credit card and make money, someone would. It's not a profitable thing to offer most people even when you can borrow at 2%. The banks are not ideal citizens but they do very much compete on price.

            So we have a in many ways very high-time-preference society, where a lot of people decline to accumulate wealth, but because they can't credibly go into debt that they'll pay back, debt remains expensive, is I think the story.

            That implies that where one can create a corporation that can credibly go into debt, and thus can borrow money, they can often make great investments on that basis, and this is a great benefit to being able to create a business. I think this is just correct? One thing I don't understand is, the people who buy that debt from companies seem to just get a worse deal than equity holders, yet they do it anyway, I think because they're doing two different and not entirely compatible calculations.

            A lot of the barrier to entry for corporations is that there are not so many identified good opportunities for investment, and even less that can be demonstrated as good and safe in a legible way that allows the borrowing of money.

          1. Benquo Post author

            Cf. Dan Davies' "don't believe fibbers' estimates".

            One explanation is that the effect of high internal rates of return and incidence of "optimism" is not uniform - some people are more willing to fudge the numbers than others - and the IRR has been calibrated to transfer resources from the honest to the dishonest, at the cost of forgoing some genuinely good honest projects, but not the very most valuable honest projects.

          2. John Salvatier

            I'm not sure what you mean by "why can't they just become more optimistic?".

            There are two sets of decision makers: the people who set the IRR cutoff and the people who evaluate capital projects (typically engineers). The reason optimism in the people who evaluate capital projects is bad is that they will do projects that they think have positive return but in fact have negative return.

            You might ask why you can't just have a multiplier for the costs, instead of changing the IRR, and the reason was that engineers will learn this and reduce their cost estimates. Also they do in fact do this, its just not enough to keep engineers from being overconfident.

          3. John Salvatier

            Re: fudging the numbers,

            Yes, I think this is probably part of the story. Though I'm not sure how setting the IRR high makes this worse, if anything I think it probably makes things better. If you had a low IRR cutoff it would be much easier to fudge the numbers and there would be more liars getting their bad projects through.

          4. Benquo Post author

            If you had a low IRR cutoff it would be much easier to fudge the numbers and there would be more liars getting their bad projects through.

            This seems like it favors liars and narcissists over mildly wishful thinkers or actually well-calibrated forecasters.

          5. John Salvatier

            This being the high IRR?

            High IRR does seem like it will proportionally favor liars over mild overconfidence, even as it reduces the ratio of bad projects to good projects.

            I have the sense that you think having a high IRR is 'rigged' which suggests an intuition that there's some better system, but I'm not sure what it is. Unless you're imagining getting rid of the liars and narcissists, which is obviously superior, but depends on having the social tech for that.

          6. Benquo Post author

            You could tie compensation very closely to the accuracy of projections, as one example. This could be done with some sort of escrow and third-party-review arrangement to vastly reduce counterparty risk. That's one example of what a company might do if it were trying to have consistently accurate projections.

            But even if there's no viable alternative within the corporate context, it's still important to track the extent to which your model is compatible with the existence of two separate types of project:

            (1) Fudgers collude with fudgers of a similar type, with shared norms about how to produce correct "optimistic" estimates (and scapegoat people outside their coalitions as fraudsters for fudging estimates).
            (2) Honest well-calibrated estimators are funded when their project achieves a very high rate of return.

            The IRR threshold is part of what determines the share of resources available to (1) vs (2). Of course, projected mean IRR will be a lot higher than the measured average true IRR, because the true IRR threshold for fudgers is actually much lower and may even be negative.

            This implies that some projects pitched by honest estimators, that would be funded in an unbiased rank-ordering of projects, will end up unfunded. This can be modeled as a sort of deadweight loss due to taxation as a means of transferring credit from honest estimators to a coalition of fudgers.

            However, it's not obvious that the situation I'm describing is stable.

          7. John Salvatier

            Yes, good point. You want to keep track of how much you're cutting out unbiased proposers and favoring adversarial proposers.

          8. Benquo Post author

            I feel like the thing I was trying to say here - the apparently uniform standard serving as a means for a credit transfer from honest players to dishonest ones - should have a short message length, since things like it seem pretty common. But it apparently has a long message length - we had to engage in a substantial back-and-forth for you to get what I meant to say.

            Any ideas how I could have said this more clearly and distinctly up front? How could I have made this conversation easier?

          9. John Salvatier

            Not that this fixes your problem, but one thing you could do is give the phenomena "apparently uniform standard serving as a means for a credit transfer from honest players to dishonest ones" a name. What's the prototypical example of this? Maybe name it after that?

            I think its a bit hard for me to wrap my brain around the mechanisms at work here, so its hard for me to say when this is plausible and when it isn't.

            1. There's a difference between a relative benefit and an absolute benefit (the dishonest need more absolute resources than otherwise). It seems to me that in order for coordination of a bad rule to be worth it you need an absolute favoring.
            [NOTE: this was my original thinking; but upon writing this down, I think this is probably wrong because the resource you're talking about is /credit/ which seems pretty fixed and thus relative-benefit==absolute-benefit. But I didn't see that clearly at the time].

            2. The mechanism by which this operates is confusing to me and probably involves no one ever knowing that they're doing it. Like maybe it involves savvy people liking rules that make savviness rewarding? Its a bit harder for me to jump to this kind of hypothesis because I don't really have the parts to /build it/ from scratch.

            3. I get a small bit of existential panic whenever I think about the world primarily operating on a competition for a fixed resource (e.g. credit) because there's no place for me in such a world. (I think this makes me emotionally shy away from this kind of thing a bit).

          10. John Salvatier

            But I failed to actually answer your question.
            Things that might help are
            1. Naming it
            2. Emphasizing that it's about social credit rather than wealth
            3. Talking about mechanisms? maybe

    2. Benquo Post author

      I don't really get why they solved the problem by printing money.

      I honestly don't fully understand why this happened in real life either, but it seems like there are a bunch of people in the main cities who have zero productive capacity there, and the administrative work involved in resettling them elsewhere is likely pretty expensive.

      Reply
      1. Paul Christiano

        OK, but resettling is almost certainly worth it, you aren't going to reduce real misery by procrastinating. If you expect to mobilize again you might need to keep them in the city, but you would do that by clearly signaling future demand (e.g. by signing contracts, by continuing to procure things-like-weapons in peacetime, etc.).

        (If resettling isn't worth it in real terms, and you actually needed the productive capacity of those citizens, then you are just in for misery. Either you underestimated the real cost of the war, or the moves were malinvestment. If you don't need the productive capacity of those citizens and this is just a distributional issue, that seems different from the story you are telling.)

        Reply
        1. Benquo Post author

          resettling is almost certainly worth it

          I'm all for figuring out a humane way to do that. But by this point it would involve a radical change to many people's way of life, and centrally planned resettlements have a very poor track record and tend to involve lots of violence, even when they're undoing the effects of a prior resettlement, so we should be pretty cautious here.

          Likewise, the Golden Benevolent Empire may be pretty worried about facing either an immediate rebellion if a large share of its urbanized laborers suddenly find themselves without an income, the creation of a useless and miserable lumpenproletariat mob if it just puts them on the dole, and horrific coordination problems leading to large numbers of deaths in the short run if it just directly reallocates people to existing or new villages via a centrally planned resettlement.

          Reply
      2. Zvi Mowshowitz

        I recommend reading Scott Sumner's The Midas Paradox if you want to explore the role of gold in the great depression in detail. I tried to write a sentence-to-paragraph version of why this helps with the problem without assuming the person basically already knows the answer, and mostly failed. In the context of this post's metaphors, once people are using contracts and norms denominated in gold (or MobsterBucks, or GoldenBucks, etc) the money is a commodity like any other, and screwing up what it is worth causes a lot of bad effects that can both wipe out a lot of people (which happened, a lot) and crowd out investment in producing things (which also happened, a lot). Getting the number of MonsterBucks wrong is, as Benquo explored somewhat, pretty bad.

        I hope to have enough time to create a good short explanation, because I can't think of one, and it would be valuable, plus until I can create one I should worry I have it wrong.

        Reply
  2. Alyssa Vance

    I think this explores a lot of interesting ideas, but the claims and data points aren't precise enough to reasonably establish cause and effect. Eg.:

    "That doesn’t make the consequences of permanent mobilization any less unfortunate. Lifespans are declining at the center of the empire, though not yet at the periphery. This suggests that near the center time preference has increased to the point where we’re creating scarcity faster than we’re alleviating it, while at the periphery scarcity is still actually being alleviated because there’s enough scarcity to go around, or perhaps marginal areas do not suffer so much from total mobilization."

    US life expectancy grew the most during the period of greatest military mobilization (1900 - 1950). The (small) decreases seen recently are at a time when military spending, as a percentage of GDP, is close to pre-WWII lows:

    https://www.defense.gov/Portals/1/features/2015/0215_budget/info04.jpg

    http://mjperry.blogspot.com/2011/03/us-life-expectancy-rose-to-new-record.html

    Reply
    1. Benquo Post author

      I'm not saying the problem is increased military spending per se. The problem is a control system designed to mobilize labor for legible purposes, left running for several successive decades, even though there's not a war on that would require this. The point of the Coca-Cola and Facebook examples, taken jointly with recently declining lifespans, is that mobilizing people to work on harmful nonsense has begun to crowd out or overpower helpful work.

      Reply
      1. Alyssa Vance

        I'm not really sure what kind of control system is being described here; in general, the US federal government (and to some extent governments elsewhere) have become less controlling of markets since at least the 1970s. Currently, around 5% of the American labor force is composed of people who are not even legally allowed to live in the US, so whatever form of control is being exerted doesn't seem very controlling. Tax evasion in labor markets is common enough that practically everyone is familiar with it, although it's not quite as ubiquitous as in eg. Italy or Greece.

        Coca-Cola and Facebook seem to me like the exact opposite of some kind of top-down control system; they exist because people are hyperbolic discounters, and will gladly buy things that hurt them in the long-term for short-term gain, and no one will stop them from doing this or even create trivial inconveniences. (And in Facebook's case, it gets to continue doing this partly because no outside corporation or state has the power to dent its network-effect monopoly.) A top-down control system doesn't look like Coca-Cola, it looks like the War on Drugs, which has been slackening off since the Nixon and Reagan administrations. Indeed, the War on Drugs is the only reason that Coca-Cola doesn't literally contain cocaine, which was one of its original ingredients in the 1890s.

        Reply
        1. Benquo Post author

          they exist because people are hyperbolic discounters, and will gladly buy things that hurt them in the long-term for short-term gain, and no one will stop them from doing this or even create trivial inconveniences

          How come Coca-Cola is so aggressively marketed, then? Sometimes a sugar packet is right there, and getting a Coca-Cola would require some sort of minor or even substantial inconvenience, and people pretty regularly choose the Coca-Cola over the sugar packet.

          Reply
        2. Benquo Post author

          The War on Drugs doesn't look like a control system to me at all and I'm unsure why you'd use that example. Contemporary monetary policy, on the other hand, is talked about almost exactly the way people talk about thermostats, even using the metaphor of the economy getting "overheated" when monetary policy is too loose.

          Reply
  3. Alyssa Vance

    "This way of life affords a freedom to which the people of my generation and the generation before are not accustomed: if you don’t like the activities by which one might get gold, you can just not do those things."

    I'm not sure what historical period this is intended to refer to, but it seems like it must have been so long ago that there wouldn't be any written records. That would make it difficult to say very much beyond pure speculation, or to meaningfully compare that time with ours. A village which doesn't belong to any larger polity anywhere probably doesn't have writing, since writing only emerged with complex cities and their long-term administrative requirements. (If it's intended to describe, eg., 12th century Europe, that period very much did have landlords and rents and taxes, and also various kinds of forced or unfree labor on top of that.)

    Reply
    1. Benquo Post author

      I didn't mean to claim that this hypothetical literally happened. I agree that few if any places spend nothing on defense or tribute, but it's still important to understand what parts of our economic life that sort of pressure is a response to.

      My sense is that there have been times and places that were much, much more similar to that description than ours is.

      Reply
      1. Alyssa Vance

        I think the government-replacing-local-authority model explains a lot more of the data than talking about war or defense; I've most recently read about it in Baliocene, although I'm sure other authors have covered it before him. Basically, in early societies without a strong administrative state, local authorities of various sorts (families, tribes, clans, priests, etc.) would have a sort of social contract with their populations. They would take care of those too young to work, too old to work, or who had some sort of sickness or disability. In exchange, they demanded not only "tax" (labor of various kinds), but also rigid obedience to whatever the high-status members wanted. It's easy to see the remnants of this in eg. India or China, which were pre-modern much more recently than western Europe. How many adult Indian and Chinese children still live under the thumb of parents, grandparents, and other older relatives? A modern state reduces dependence on this local infrastructure by providing these services itself, and the majority of its budget is spent on elder care, child care, caring for the sick, and other types of social spending. Hence, at least in some ways, the unfreedom of an elaborate system of laws and tax codes is just replacing a previous, even worse unfreedom, where your powerful uncle tries to control who you're friends with and who you date and where you live and how you spend your time. The United States wants its cut of my paycheck, but it couldn't care less what I'm doing at 9 PM on a Friday, as long as it's not weapons smuggling or otherwise blatantly starting trouble.

        Granted, it is much easier to opt-out of a pre-modern local authority than a modern state. In the former case, one can simply walk away, and some people always did (especially unattached young men), while it's harder to give up citizenship. However, walking away usually meant both the permanent loss of one's entire social sphere, and also a significant chance of serious injury or death. So there are a lot of advantages to the modern system.

        Reply
        1. Jessica Taylor

          A shift from local to federal authority is often caused by war. A major reason the Articles of Confederation failed was that they made it hard for the federal government to raise an army. The US Civil War caused an expansion of federal power, as did World War II.

          I agree that there are advantages to a shift to federal power other than war. I am not actually sure how to check whether overall social pressure (including economic pressure resulting from artificial scarcity) is higher or lower when there is more federal power. Social pressure in the US is currently sufficient to force most people to spend 40+ hours per week doing something they generally don't like doing (yes some of this is about actual scarcity, but not most of it), and there are also complex behavioral standards people must follow to be accepted in social groups (think about what is involved in being "chill" or "social" or "not cringeworthy").

          Would also be interesting to look at _why_ local authorities would demand obedience to rigid rules. I don't currently have a model for this. I don't currently have a model for this but the forager vs farmer distinction (http://www.overcomingbias.com/2012/05/forager-vs-farmer-morality.html) seems relevant.

          Reply
  4. Alyssa Vance

    "But something else happens. In the village, gold - which was perhaps used occasionally, either for transactions of ritual significance, the occasional foreign trade, or largely optional market transactions - is now just as necessary to sustain life, as food and water and shelter are."

    This seems like an overly complex story regarding monetization, especially since there's a long history of rents and taxes of various sorts being paid in kind (through sharecropping, labor, etc.) in highly agricultural societies. From my understanding of economic history, monetization of the economy simply goes hand in hand with the local urbanization rate. The most important good in a pre-modern economy is food, and since a subsistence farmer can produce that themselves, they can largely get by without money. An urban dweller doesn't, of course, so they must trade on a regular basis for the food they need to eat, hence money becoming more important. That seems like a simpler explanation than tying it to forms of governance or effective local tax rates or political organization, unless there's solid evidence that the simple theory doesn't fit the data well enough.

    Reply
    1. Benquo Post author

      From my understanding of economic history, monetization of the economy simply goes hand in hand with the local urbanization rate.

      This seems to me like it treats urbanization as an unexplained exogenous factor.

      Reply
      1. Alyssa Vance

        That's true, but I'm not sure urbanization can be explained by any kind of universal model, as it's closely tied to general specialization and economic prosperity, and that can have any number of causes. Eg., in the year 300, what is now England was much less urbanized than Europe as a whole, and what is now London had a population of ~10,000-20,000. In the year 1700, London was arguably the largest city in Europe, at a population of ~700,000, and England was arguably the most urbanized society in the world. Why? At an extremely rough guess, advances in shipbuilding techniques allowed the Atlantic rather than the Med to carry the main European trade routes; but that's hardly a general theory, and I'm sure the full story is much much more complicated anyway. (One can explain the huge population of Rome by it being the center of a giant empire; but what about Antioch, a huge Greco-Roman metropolis that had never ruled anything outside itself? And in the 15th century, when Italy was again the most urbanized part of Europe, why did it have no central government, being divided into dozens of squabbling city-states? Questions, questions... )

        Reply
  5. Alyssa Vance

    "I mean that we still have an elevated time preference, as though we were in a state of emergency; 10% is not an unusual internal rate of return for major corporations. Many businesses’ effective time preference is even higher. Perhaps we have become so accustomed to wartime mobilization that we don’t remember any other way of life."

    Not sure how much of this is standard academic history, and how much is me just drawing dots through my own observations, but what I see is that humans naturally have very short-term time preferences by civilizational standards, and most humans will only have longer preferences if they are bludgeoned into it through strong social pressure and a rigid system of rules. This does seem to match what we know about the ancestral environment and pre-agricultural societies; there's little meaningful property and no fixed capital investments, so there's no reason to sacrifice now for benefits reaped five years from now (never mind decades or centuries). If we're seeing a return to shorter time preferences in the 21st century West, to me the most obvious explanation is an extension of Hanson's farmer/forager thesis, ie. when most humans don't directly participate in agriculture, the pressure to follow farmer norms gets weaker, and people will some extent revert back to forager norms.

    Reply
    1. Benquo Post author

      These communal deliberative, planning, and resource-allocation mechanisms have implied time preferences, which inform the ways in which they shape individuals' environment and incentives, and our society's central resource-allocation mechanism seems to in practice have a higher one than it used to.

      Reply
  6. Alyssa Vance

    "But in the process we have become accustomed to accumulating wealth in the form of financial instruments and rising prices, rather than improved homesteads. These are the sort of wealth one can accumulate during wartime, but are only valuable as claims on the work of others. We can’t become richer by all going into debt to each other."

    Sure we can - this is possible because the other side of the balance sheet is a corporation, not an individual. Suppose that a thousand people are currently working hard to grow their own food, by tilling and plowing fields themselves. Rather than continue doing this, they decide to pool their labor, and build a magical automatic food-producing-factory. (Real corporations have both capital and labor inputs rather than purely capital, but this is deliberately simplified for clarity.) Each person gets an 0.1% share of the food the factory produces, which is more food than they were previously growing on their own, and they no longer have to work to get this food. So everyone is richer - they have both more food and more leisure time. Technically, all of their "assets" (their 0.1% shares in the factory) are balanced by "debt", in that on the other side of the balance sheet, there is a claim against the corporate entity that owns the factory for a percentage of its assets and future output. However, there are no humans in debt anywhere in the system; there is no person anywhere who has to eat less so that others can eat more. It all comes down to the accounting equation, Assets = Liabilities + Shareholder's Equity:

    https://www.principlesofaccounting.com/chapter-1/accounting-equation/

    and more generally the IS-LM model of macroeconomics, where interest rates act as a pricing mechanism to balance individual consumption (eg. growing food to eat now) vs. savings and investment (eg. building a robot factory to make food automatically):

    https://en.wikipedia.org/wiki/IS%E2%80%93LM_model

    (as Peter Thiel says, there is currently an issue in first-world countries with how the capital is being allocated among investment opportunities, but this is a time-and-place-specific problem that is separate from the ideas of capital and investment generally)

    (and there are currently issues in wealthy Anglosphere cities with regulatory capture on local governments, allowing landlords to grab monopoly profits from real estate ownership, but this is again a much narrower issue than what the quote describes, and has nothing to do with the ideas of investment or financial instruments per se)

    Reply
    1. Benquo Post author

      This seems theoretically possible, but for the most part businesses don't really work like that as far as I can tell.

      Reply
      1. Zvi Mowshowitz

        Not quite this cleanly, no. But the basic concept is, that while it's in an important sense correct that we can't make everyone wealthy by borrowing money from each other, it's also importantly wrong - if I lend money to you, and this allows you to make valuable things, borrowing money led to wealth. If a group pools its money in some form for a group project, that too can lead to wealth, even if it's not as clean as Alyssa's example. The active action is of course when the humans build the food-making machine, but if I invest in a food-machine-making company, which leads in expectancy to food-making, that too created wealth, even if you'd prefer to use a measure where it doesn't count until the machine starts pumping out food, and ignore the expected value of the future food production from the as-yet unbuilt factory.

        And in the general case of just doing business, the question to ask is, is the business itself a form of wealth? If you think that the business's wealth is rent extraction and thus its value is the impoverishment of others, than its existence doesn't create wealth. If it does make actually valuable things, it does. That's then distinct from the question of, how do we distribute those gains? If I make a food-making factory, I create wealth. If I then sell shares in it to others, that doesn't generate wealth (although doing this so I can build the factory in the first place, perhaps did), but it also doesn't destroy wealth unless it leads to bad management of the factory (which it might). It reallocates wealth. And if it diversifies people, satisfies their time preferences and risk profiles, it can be a win versus everyone owning their particular homestead.

        To focus back on the original claim, if I have a generic claim to 'work of others' that claim existing doesn't make the world wealthier, but I hopefully got it in a wealth-creating way. Even when the currency is shields, as you noted last time, my having shields doesn't tell you whether I've produced bread or arrows to get them. It just means we're doing a better job of allocating the wealth that does exist in an efficient manner, at some cost in alienation, and at the cost of making it harder to tell which thing I was making.

        Reply
  7. Pingback: Weekend Readings for May 25, 2018 – Verywhen

  8. Pingback: Rational Feed – deluks917

  9. Pingback: Rational Newsletter | Issue #9: Full List

  10. Quioxte

    This post glosses over a few things of very high importance. “To occupy a space, even a space you legally own, unencumbered by any debt or lien, one must pay rent in the state’s currency.” This sentence refers to ‘, even a space you legally own’ as if ownership is a physical condition or an ontological natural primary.
    Land doesn’t have meta data. It doesn’t have tags or labels. If I dig up a piece of dirt it’s just dirt. “Ownership” isn’t something that’s ‘in’ the land, or something that’s in you. What it means to “own” land is multifaceted, but basically it boils down to two main things. One, in a legal property register, which is maintained, continuity assured, and quality verified in various ways, you are listed as owning the land. Two, if someone is on the land who you don’t want to be there or the land is being put to a use you didn’t approve, you are able to contact a group of armed people that are capable of evicting the adverse users with overwhelming force that the adverse users are not able to contest.
    The post users “ownership” as a bit of light magic, where a mental category is treated as a “real” or physical thing. But ownership is more of a shorthand or compression algorithm, when you say, “I own this land” what it means is “there is a durable recording through [some long process] which entitles me to contact [some large and complex institution] to request the use of force in [some semi predictable way in response to a properly made request]”.
    There never was a time when someone could just “own” land in piece and live off it without interaction with a registry mechanism and an enforcement mechanism. Even when the registry was just the memory of the village elder and the enforcement mechanism was the chief and 3 guys with large builds, those mechanisms still existed. If you pissed of the village elder or alienated folks in the village you might exiled. Good luck accessing the land you “own” then.

    Reply
    1. Benquo Post author

      Possibly you skimmed past this part:

      Accordingly, social policing has to be comparatively direct. If you want to exclude someone from village life, you have to actually coordinate to run them out of the village. Especially if they own land to deprive them of their livelihood you have to physically expel them from their own house and land. Anything less is merely an inconvenience.

      Obviously there's never been "absolute" ownership. Maybe you can say more, in more detail, about exactly what about the post seems to be operating under the implied assumption that there has been?

      Reply

Leave a Reply

Your email address will not be published.