No, Virginia, The State Does Not Create Surplus

With a small caveat

Over the last few months I have seen kind of a strange claim get made on various blogs and ‘Stacks I read: that the state creates surplus when otherwise subsistence farmers cannot. One might think it a little strange to keep running into that assertion, given what I write and thus presumably read. Fortunately,

Lorenzo Warby

was good enough to provide me with a trio of citations¹ for it.

Tax creates surplus references:

Joram Mayshar, Omer Moav, Zvika Neeman, ‘Geography, Transparency, and Institutions,’ American Political Science Review, June 2017, 111, 3. 622-636.

Joram Mayshar, Omer Moav, Luigi Pascali, ‘The Origin of the State: Land Productivity or Appropriability?,’ Journal of Political Economy, April 2022, 130, 1091-1144.

James C. Scott, Against the Grain: A Deep History of the Earliest States, Yale University Press, 2017.

I need to grab the James C. Scott book because I generally quite like his work, although time is rather a problem for me. However, that second citation there, The Origin of the State etc., that’s right up my alley: Political Economy and available in free PDF form!

So, what is really the claim here? The paper is itself not great² but it puts forward two general notions:

1: It isn’t just land productivity that is important to the generation of food to support a taxing elite, but what kind of food is being produced: long storing things like grains lead to states, quickly expiring foods like tubers do not.

2: A surplus of food does not need to exist before elites can become taxing elites, but rather those taxing elites exist first and create the surplus they need to exist by taxing the farmers.

Now, I don’t really have a problem with the first claim. Grain is definitely easier to appropriate (tax) than tubers. Grain is relatively easy to observe ripening, gets harvested all at once and needs to be stored for future consumption, all of which makes it pretty easy for local elites to target. They know where the farms are, they can observe who is doing the harvesting, so they just need to show up and get some, and once they do, it can sit in their storage till they want it. Tubers, on the other hand, are often³ grown in places with a year round season, meaning they are harvested as needed, and don’t keep so well, so at any given moment the farmers won’t have a lot sitting around to steal tax. It is worth noting that this isn’t entirely true, as in colder climes tubers are usually harvested before winter kills the plants and stored, and you can dry and preserve some sorts of tubers for preparation later⁴. Not to mention the fact that a local tax authority could simply require a levy on a shorter time cycle, or in a different trade good. Still, I don’t have a lot of problem with the general claim that regions that support grains as a staple crop are more likely to sprout hierarchies of elites that do not farm but tax farmers for their existence⁵.

As to point 2, well, some of you might see the problem already, and perhaps think I am presenting a straw man version of the claims. Well, let’s let the authors state their claim [emphasis mine]:

In challenging the claim that the emergence of hierarchy was a result of high farming productivity, we also contest that surplus is the mechanism that links productivity to taxes. We illustrate our argument with the following scenario: consider a farming society that subsists on a cereal grain that has to be harvested within a short period and then stored for year-round consumption. A tax collector could confiscate part of the stored grain and transport it for consumption by distant elite and other non-food producers, even without a food surplus. Ongoing confiscation can be expected to impact adversely the size of the farming population. However, due to diminishing average product of labor, the smaller population would lead to higher output per farmer. This would result in an equilibrium in which total output exceeds the farming population’s subsistence needs, with the surplus confiscated by the non-farming elite.

We concur with the conventional productivity theory to the extent that in hierarchical societies farmers produce a surplus. Our contention is that surplus isn’t a prerequisite for the emergence of a tax-levying elite, it is rather the elite that generates the food surplus on which it can flourish, once the opportunity to appropriate rises. Thus, this simple hypothetical scenario demonstrates that the availability of surplus need not be a precondition for taxation and hierarchy, as argued by the conventional productivity theory.

Indeed, the elite generates the food surplus it needs before the elite even emerges. The elites are themselves the creator, the first mover, requiring nothing to exist before they create what they require to exist.

The Journal of Political Economy is a peer reviewed journal. Not sure if I mentioned that.

Now, the authors do a poor job making their own case. Let me see if I can’t clad their argument in some steel.

Steel Mannin’

In Neolithic, pre-state subsistence farmer communities there are two relevant periods of the food cycle: high food periods directly following the harvest when more food is available than will be eaten, and low food periods while the primary crop grows where farmers need to rely on previously harvested food to subsist on while waiting for the next high food period. The cycles depend on a few different factors, notably the crops grown and the regional productivity.

Grain has a long cycle, with a large harvest all at once, and then some time spent waiting till the new plants grow and seed before the next harvest, during which time you are living off the previous harvest. The highs are quite high, and the low lasts a long time, meaning farmers are sitting on a lot of grain that is surplus to current, high food period needs, but might not have a surplus over the whole cycle because that food needs to last through the long low period.

Tubers, in regions where they are very productive⁶, can be, and in some ways must be, harvested more frequently and in smaller quantities. Thus the high period might be “right before lunch” while the low period might be “until tomorrow, right before lunch.” The rest of the crop stays in the ground. The cycle is very flat, and much more predictable, leaving little concern that the large surplus now in the high period might not even be sufficient surplus to see the farmer through the low period.

The regional productivity matters here a good bit as well, both in terms of climate and over all suitability to the crops. The Sahara is notably bad for farming, but throw the Nile river into the mix and farmers can produce multiple harvests of grain a year, compressing the normally long high/low food cycle of grains quite a bit. On the other hand, a long, cold winter might make harvesting tubers nearly impossible at certain times of year, necessitating a large harvest and subsequent lengthening of the cycle⁷.

Now, farmers facing a long cycle from a storable food have problems, but they also have a notable advantage: they don’t have a lot to do while their crops mature. This time can be spent hunting, raising other plants, or tending domestic animals (this is Neolithic, pre-state mind you, so this last might be minimal.) It can also be spent raiding the tribes nearby and trying to steal some of their food stores, great when yours are running a little lean, or stealing some of their farm land. This raiding opportunity raises another problem, as your neighbor tribes also have time to raid your stores.

The capacity for raiding then creates an opportunity for specialization, as some farmers find they are really good at raiding, and really good at fighting off raiders. So good, in fact, that maybe they don’t need to fight the neighboring tribe, but rather can fight their actual neighbors, or at least threaten to, living off their production in exchange for protection⁸. You might even still work your own farm, but as you and your friends get better at fighting off raiders you might decide to make that your full time business. Big Steve and the Lads are here to protect your crops, and all they ask is enough to keep their bellies, and those of their families, full while they are busy protecting your crops instead of taking care of their own. I mean, it’d be a real shame if something were to happen to your nice, well stocked granary there.

In lean times, however, it will not be Big Steve and the Lads who go hungry. It never is. They will collect their taxes, and if the farmers lose a few kids in the process, well, someone’s gotta go. On the plus side, this leaves more high quality farm land for the remaining farmers, who continue to produce above their own needs. Malthus’s prophecy is thus held at bay, as the farmers don’t over populate, and the elite keep their source of surplus going.

Plus, as Big Steve and the Lads’ operation grows they can access the food reserves of ever more farmers, growing the tax base while expanding their public goods provision to things like forts, fleets, pyramids with sacrificial alters on top, palaces, etc.⁹

I Am Running Out of Steel…

One big problem with this story is that point about marginal productivity, the diminishing average productivity of labor mentioned above. It necessarily requires a bit of high level gloss because it does a huge amount of work, poorly.

The overall notion, made famous by Malthus, is that humans will expand their population until they can only just feed it, then expansion stops as any extra people can’t generate enough food to keep themselves alive. This is the result of a hard limit on the output of food production, because of the quality of the land or carrying capacity of animal sources, etc. So while there is good farm land or lots of animals running about people can produce enough food to produce a new generation, but eventually those newer generations can’t find a good place to farm or hunt and starves.

There are a couple of problems there for our Neolithic story, however. Firstly, it assumes a very high proportion of humans to decent land area, such as one might find in Britain during the industrial revolution, where all the farm land is pretty much already taken so anything left over is just awful. If the population is low relative to the available land, such as was the case in North America during the same period, you don’t have this constraint. In a technical sense the marginal product of farm labor is pretty flat still, as there is more pretty good land one valley over. One might wonder whether the population to land situation 8-10,000 years ago was closer to industrial era Britain or Pennsylvania, but considering the estimates on total human population was somewhere around the 10 million mark for the entirety of Earth at the time, my money is on “there is plenty of farm land.”

Secondly, using average productivity is a bit misleading when thinking of surplus to support non-producers. Imagine I have two farmers working a given amount of land. The first farmer produces ten bushels of wheat a year, and the second farmer eight bushels, for a total of eighteen bushels and average of nine bushels per farmer. Each farmer, however, each only consumes five bushels a year, so ten bushels get eaten, leaving eight bushels as surplus. Good times. Now let’s add a third farmer, with lower marginal productivity, only producing six bushels. Whew, now average productivity drops down to only eight bushels per farmer, down roughly 11%! Oh, but surplus actually goes up to nine bushels. So productivity goes down and we have more surplus? Yes, because so long as the marginal worker is producing more than they consume average productivity and total surplus do not move in the same direction. In other words, so long as average productivity is above average consumption, adding more people adds total surplus.

Of course, when you toss in a few elites who consume food but do not produce it that is going to lower your average productivity quite a bit, and thus your available surplus, but our authors are only referring to labor productivity, not total societal productivity, and so are ignoring the drop in output caused by having some people with a food productivity of zero, and a net negative impact on food supply. As a result they are not looking at the total amount of people the society can support, merely the split between farmers and elites.

Why is that important?

Because it leads them to say silly things like “it is rather the elite that generates the food surplus on which it can flourish” when in fact the elite isn’t creating anything, but rather changing the end use of any surplus that is generated by farmers. Instead of farmers using their surplus to make new farmers (or invest in capital improvements on their farms) the surplus is taken and used to create new elites. That’s all. No new surplus is created, it is only spent in different ways. Notably, different ways that almost certainly results in a lower total population for the society and lower overall surplus.

Let’s address that last line there for a moment, because it is a rather important point. The elites are not creating surplus, because without the farmers they could not support themselves as elites. Instead they are moving the surplus that farmers create to another use other than supporting the farmer population. They do this by taking grain that has been harvested and stored, grain which is surplus to demands of the current, high food part of the cycle, but is not necessarily surplus to the total, full cycle demands of the farming population overall.

The elites starve some of the farmers as a result of taxation. Those farmers or their children die the long death of starvation so that the elites have enough to eat. That is what “Ongoing confiscation can be expected to impact adversely the size of the farming population,” means.

However, the last claim by the authors in that paragraph, “This would result in an equilibrium in which total output exceeds the farming population’s subsistence needs, with the surplus confiscated by the non-farming elite,” also does not hold true. The elites are confiscating what they want regardless of the farming populations’ subsistence needs. The confiscation happens at the high point of the cycle, not after the farmers have eaten all they need for the year just before the next harvest. The elites have to tax before the surplus can go to create more farmers in order to keep the farmer population down. Those excess kids are not going to starve themselves to keep average labor productivity high, after all.

So What?

So, why does Doc Hammer have such a bee in his bonnet about some stupid claim made by some economists in a dumb paper? So elites don’t actually create surplus, only redirect how it is spent at best. So what?

Well, I mean, there’s the whole “truth” thing. Saying that elites create the surplus that supports them is terrible double speak, as the elites do not create the surplus in any sense, they merely take what is surplus to current requirements at the cost of future requirements.

Even if one wants to stretch it to “elites prevent loss of surplus through raiding by neighbors”¹⁰ prevention of loss is not the same as positive production. That’s a point many economists get wrong. They want to say that avoiding a negative is the same as getting a positive, but that only holds true in some cases. Preventing loss is not the same as producing when that loss requires production to have happened in the first place. A really great granary that keeps out rats prevents loss, but it doesn’t put food on the table unless someone first produces the food which we are trying to prevent being lost. Preventing loss might be necessary, but it is not sufficient.

Further, and possibly most gallingly, the claim that elites create their own surplus is an example of the broken window fallacy. The framing that the elites create the surplus that sustains them highlights the easily seen fact of elites eating, while completely ignoring the fact the the society doesn’t have more people, is has fewer. Those starved farmers are rather harder to see, yet just as important to our understanding of the situation.

Falling into the broken window fallacy is a cardinal sin for an economist. No economist post Frédéric Bastiat should be making that error.

So, yea… let’s stop repeating that.

1. A really good essay, by the way; go read it after!

2. I am going to really try to avoid ripping into the paper itself here too much, but the fact it was published in JPE is embarrassing for the journal. I will try and stick to how bad this particular logical argument is rather than the structure of the paper it appears in.

3. But importantly not always.

4. It also doesn’t address staple crops like plantains… this is something of a bad paper.

5. Note, that is taxing elites as opposed to specialists who do not farm but trade for farm outputs, like hunters, fishermen, priests, or specialized tool makers.

6. This is going to conflate with the climate version for a bit, but there is only so much I can do with their bad argument.

7. Conveniently, that same cold weather makes storing the tubers a lot easier despite their high water content, so that’s nice.

8. I am torn on scare quotes around protection there, but ultimately whether it is real protection from outside raiders or a Sopranos style “protection” is probably irrelevant. Mancur Olsen was onto something, I think.

9. This is that small caveat: as Olsen points out, the stationary bandit does have incentive to provide investments in productive output with the surplus he collects. In this sense, it is possible for a future (not current) elite class to produce the surplus they live on, in much the same way that the blacksmith produces the surplus he lives on. The capital improvements allow for more production than their would be otherwise, for the same population.

Of course our authors don’t bring that up. Nope.

10. Which is in my steel man version, but notably not the original authors’, who only refer to “distant elites” who are not obviously providing any public goods.