In defense of the Wittfogel thesis

by on August 23, 2012 at 12:47 pm in Economics, History, Political Science | Permalink

There is a new paper (pdf), “Irrigation and Autocracy,” by Jeanet Sinding Bentzen, Nicolai Kaarsen, and Asger Moll Wingender.  Here is the abstract:

We show that societies with a history of irrigation-based agriculture have been less likely to adopt democracy than societies with a history of rainfed agriculture. Rather than actual irrigation, the empirical analysis is based on how much irrigation potentially can increase yields. Irrigation potential is derived from a range of exogenous geographic factors, and reverse causality is therefore ruled out. Our results hold both at the cross-country level, and at the subnational level in premodern societies surveyed by ethnographers.

1 Florian August 23, 2012 at 1:22 pm

“Irrigation potential is derived from a range of exogenous geographic factors, and reverse causality is therefore ruled out. ”

Here is my idea:

Irrigation is most useful in a flat plains with a large central river.
Flat plains are also easily conquered by a ruling warrior elite.
So the same exogeneous geographic factor favors irrigation and centralization of power.

(Whereas a hilly, fractured landscape favors small independent communities and makes irrigation more difficult).

In premodern times, large empires were generally not-democratic.
Whereas small independent communities sometimes were (such as the greek city-states, medieval German cities, Swiss cantons).

Of course, that’s a generalization, and it might be easy to find counter-examples (such as the highly irrigated and yet rather democratic 16th century Low Countries).

2 Jim August 23, 2012 at 2:33 pm

Flat plains aren’t enough to make an empire; you need control of water to make it last. Otherwise there would have always been an empire from the Ukraine to Mongolia, instead of the one-generation flash in the pan of the Genghisid period.

No one can control the water supply in a rainfed system, but an irrigation system is easy to control. And you need a pretty large and sophisticated political structure to build one in the fitrst place, even if local efforts are enough to build pieces here and there that get integrated later on.

3 Roy August 23, 2012 at 3:30 pm

But this is about irrigation potential.

In a system where irrigation potential exists the state that makes use of irrigation will have an advantage. In agricultural area, particularly a watershed with irrigation potential those who don’t adopt irrigation systems will have less surplus, they will support fewer people and thus be less militarily powerful. There must be some threshold here, but since irrigation systems radically discourage autonomy and require a central authority to settle disputes, this will gradually lead to the radical strengthening of central authority.

One aspect that always seems to be ignored in this is that at different levels of technology different irrigation schemes become possible, the same applies to different levels of political centralization. California’s system requires both, both in scale and in technology, technically it would not have been possible until the late nineteenth and early twentieth century. Another aspect is the nature of crops. In the New World the introduction of wheat dramatically changed agriculture and the value of irrigation. In much of the Western Plains and in the Pacific Northwest maize can’t be grown except with irrigation, but wheat grows so well with dry farming that it is almost never irrigated even when such infrastructure already existed.

4 Jim August 23, 2012 at 3:57 pm

“But this is about irrigation potential.”

” In the New World the introduction of wheat dramatically changed agriculture and the value of irrigation. ”

Wheat changed agriculture dramatically, but it had nothing to do with irrigation. Wheat was the first agricultural export in California, and it was a dry land crop. Wheat has naturalized all across the state because it has the same Mediterranean climate wheat evolved in, the same as almonds, olives and date palms have.

Where irrigation makes the difference in California is with rice – irrigation and rice make the area around and north of Sacramento with its horribly poor soil pretty productive, and of course irrigation makes the difference for tomatoes and melons, which are also very important exports. That’s in the valley. Lettuce and that kind of thing in the Salinas Valley are probably irrigated too. There it was refrigerated rail cars that made that industry pay.

5 Roy August 23, 2012 at 6:39 pm

I understand completely where you are coming from but as a field geologist who has spent a lot of time in the Southwest and Inland Northwest, I have a different perspective.

Native Americans tried growing maize as far north as central Utah in the Rockies and as far as Montana on the great plains and were intermittently successful at it too. The problem was it required irrigation and localized droughts regularly wiped them out. For example in the Texas Panhandle and Western Oklahoma there is considerable evidence of maize agriculture, and I have myself found pre-Colombian corn cobs in the Northwest. The problem is that the short growing season and the distribution of rain make maize agriculture nuts in these artificial environments without irrigation. Wheta is exactly as you said a dryland crop, though it can be grown under irrigation. In New Mexico, Arizona, and Colorado Native Americans largely abandoned maize, except for ceremonial purposes, in favor of wheat very quickly. Often existing irrigation systems were quickly dismantled or abandoned altogether as fields were relocated. You can see this all across the Colorado Plateau. If it wasn’t for the demographic collapse and the disruption of the reintroduction of the horse it would be obvious even today. The rapid adoption of wheat farming by some native groups in the Northwest is also very striking. Of course European settlers did not even bother with corn in these environments because wheat was a superior crop both environmentally and culturally, so it isn’t so obvious. But in many places (primitive) irrigation evidence is one of the most common, other than lithics, signs of pre Colombian occupation you can find.

As to wheat agriculture in California the frequency of even wheat killing drought is readily apparent in the geologic record, even trees that draw from the water table are killed in these drought episodes at any distance from the Western slope. Wheat agriculture is not sustainable without irrigation over a period of decades in the Central Valley and even in places like Salinas it is incredibly risky. The frequency and severity of California droughts is the reason the initial investment in large irrigation schemes was made in the first place. Californians like to think of their state as a natural paradise, but the states drought cycle makes only very small portions of it inhabitable with any density for any kind of civilization without very large technological and organizational hurdles. This is the reason that the material culture of most of California’s native inhabitants was so impoverished. (I am intentionally ignoring the Colorado Valley, the Santa Barbara Channel, and the far Northwest of the State here)

6 Steve Sailer August 23, 2012 at 8:22 pm

Before William Mulholland brought Owens Valley water to the San Fernando Valley in 1915 (not 1935, as in “Chinatown”), local farmers typically would plant huge fields of winter wheat. About one year in three, so much rain would fall that they’d make a killing. The other years, there’d be virtually no crop to harvest. Interestingly, the coming of irrigation allowed smaller farmers to thrive because there was less financial risk.

7 Jim August 27, 2012 at 4:57 pm

Roy,

I guess there is a diffenrence between wheat escaping cultivation and naturlaizing on the one hand and being a commercially viable crop on the other.

“Native Americans tried growing maize as far north as central Utah in the Rockies and as far as Montana on the great plains and were intermittently successful at it too. The problem was it required irrigation and localized droughts regularly wiped them out. For example in the Texas Panhandle and Western Oklahoma there is considerable evidence of maize agriculture, and I have myself found pre-Colombian corn cobs in the Northwest. ”

This is a very interesting datum. it bears on an ongoing question in Uto-Aztecan historical linguistics around corn terminology, and whether the ancestral Uto-Aztecans were foragers (in the north) or agriculturalists (in the Valley of Mexico). And now, depending on the dtaes for corn this far north, that may be a moot question.

Joan Didion made the same point as you, that in many ways the paradise people saw in California was the result of hard work, not some bounty settlers had fallen into. OTOH Tule Lake was pretty paradisical before it was drained for irrigation.

Steven, that’s interesting about the hisotry of wheat-growing so far south.

8 Marcos August 23, 2012 at 3:36 pm

“In premodern times, large empires were generally not-democratic. Whereas small independent communities sometimes were”

Now, rule out reverse causation on this, because the reverse makes much more sense than the way you stated it.

Or, in plain english. Yeah, empires like to expand, democratic communities don’t.

9 Bernard Guerrero August 23, 2012 at 5:37 pm

Err, Athens.

10 Roy August 23, 2012 at 6:40 pm

But Athen’s Empire was hardly very Democratic. Just try leaving the Delian League.

11 NAME REDACTED August 24, 2012 at 4:23 am

The united states is hardly very democratic, just try withdrawing from the union.

12 stalin August 23, 2012 at 9:29 pm

(such as the highly irrigated and yet rather democratic 16th century Low Countries).

The Netherlands is a flat, low-lying country criss-crossed with waterways. About half of the land is actually lower than the level of the sea. Walls called dikes, like the one pictured left, have been built to stop the land being flooded by the sea.
The famous Dutch windmills are wind powered pumps to help keep the land dry.

Doesn’t seem like irrigation to me

13 Jeanet Bentzen August 24, 2012 at 4:03 am

Using irrigation potential takes care of reverse causality. But not the omitted variables problem, which is what you are after. To deal with this issue, we throw in various control variables. Among these, ruggedness, which should take care of your concern to some extent.

14 j r August 23, 2012 at 1:48 pm

Without reading the paper, it sounds like a form of resource curse. Irrigation is a resource that can be controlled by some group of elites, who then turn it into economic and political power.

15 Orange14 August 23, 2012 at 1:58 pm

I guess Israel and California must be the exceptions to this rule.

16 Jim August 23, 2012 at 2:23 pm

Democracy in California has pretty much seized up at the moment, for a variety of reasons, but water politics may have something to do with it.

The water from the Califoria Water Project was supposed to go only to small family farms. We all know how long that lasted. An irrigation system represents a fundamental centralization of power in a hot dry place like the Central valley. It was only a matter of time.

It’s a huge industry – what percentage of US farm production does California represent? How much lobbying money do you think is avaliable to those people?

17 Boonton August 23, 2012 at 2:24 pm

Not so sure, modern day Israel was founded as a non-agricultural nation with its democracy imported from mostly European settlers. California grew not with an agricultural boom but a gold rush.

I wouldn’t phrase the problem as a ‘resource curse’ but more likely a ‘capital concentraiton’ problem. A country that starts out with nothing but agriculture will find economies of scale asserting that the most efficient irrigation system would be controlled by one entity. Rain based agriculture has no such incentive so rain based economies can grow with capital dispersed among multiple owners. Initially the irrigation based economy may trump the rain based one but over time the rain based one probably has more freedom and incentive to adopt innovations and take the risks of implementing disruptive ideas.

18 Jim August 23, 2012 at 2:39 pm

“California grew not with an agricultural boom but a gold rush. ”

It started with a gold rush. It grew on agriculture, specifically on irrigation agriculture in the Central Valley, and on oil exploration in the southern, barely arable end of the Central Valley. This was decades before non-irrigated products like wine began to count for anything.

“Rain based agriculture has no such incentive so rain based economies can grow with capital dispersed among multiple owners. ”

That part is true and there is another aspect to consider: rain-based farmers cannot be intimidated or manipulated with threats to their water supply, so it has direct political implications. You need rain-based agriculture for the Jeffersonian utopia.

19 Steve Sailer August 23, 2012 at 2:46 pm

Try building a big housing development in the Santa Barbara area. You can’t because you can’t get any water. Santa Barbara decided decades ago to not hook up to the California Water Project to strangle new development.

Is that democracy or not?

20 Jim August 23, 2012 at 3:43 pm

It was democratic to the extent it was a local anti-growth initiative. It was not democratic to the extent that it disregarded the wishes of people who didn’t have any standing in the decision process in the first place, so it was still democratic.

21 chuck martel August 24, 2012 at 10:23 am

It’s astonishing that supposedly educated people (educated in schools dominated by the “democratic” government, however) keep assuming that democracy, despite its many obvious failures and its inevitable capture by the elites, is somehow the most superior form of government, or societal organization.

22 Jim August 27, 2012 at 5:00 pm

Thank you.

Although “superior” is a matter of personal preference. It’s not obvious to me that excluding this or that number of humans from a landscape is “inferior” to allowing them to come in and overwhelm it.

23 mulp August 25, 2012 at 4:38 pm

How is it that Santa Barbara spent $600 million from 1997-2004 connecting to the CWP with a 144 mile pipeline?

Santa Barbara looked into other pipelines as well, but those were not reliable year round sources of water.

That was delayed during a drought that limited the water available to the CWP ended in the early 90s. But the shortage of water has only gotten worse and the latest proposal is the Brown-Obama administration for $25B in Federal spending plus billions in State spending for a vague general approach to feed more water into the CWP.

The water for the CWP was all taken from farmers back in the 20s and 30s (with some going to farmers until their farm land has been turned into developments or water rights bought.

Perhaps the thesis is water for farmers is incompatible with farming, with the CWP proving democracy takes water from farmers to build cities in water poor regions which unlike farming need water every day of the year no matter what.

24 ohwilleke August 24, 2012 at 5:26 pm

I suspect that the condition only holds so long as local food production is economically material to a jurisdiction.

25 Brian Donohue August 23, 2012 at 2:44 pm

It’s a fascinating theory. To me, it’s not about the control of resources as much as transparency and the ability to monitor effort, both of which seem like good things on grounds of economic efficiency and accountability but here are put in the service of consolidating power at the top. Very thought-provoking.

26 Steve Sailer August 23, 2012 at 2:57 pm

Rainfed farming was more or less of a prerequisite for Jeffersonian America. In contrast, when the U.S. took over California and replaced Spanish law with English common law, it proved a disaster in terms of water rights. The common law let anybody divert as much water out of a river as they wanted, because there is no shortage of water in England. So, energetic American farmers in the San Fernando Valley in the mid-19th Century diverted all the water out of the Los Angeles River, leaving the downstream city of Los Angeles high and dry. Eventually, American courts had to dream up a legal excuse to restore the old Spanish system giving Los Angeles its customary amount of water.

The movie “Chinatown” gives an exaggerated version of a later bit of the history of water in Southern California, but the basic lesson is right: irrigation means that water will inevitably be Big Politics, which it usually isn’t in rainfed places, except for flood control.

Similarly, Israel’s occupation of the Golan Heights, both sides of the Sea Galilee, and the West Bank of the Jordan involves water for irrigation.

27 mulp August 25, 2012 at 5:06 pm

How many cities exist under the constraints of “rainfed water”? Not many I can think of.

28 Jim August 23, 2012 at 3:47 pm

“Rainfed farming was more or less of a prerequisite for Jeffersonian America. In contrast, when the U.S. took over California and replaced Spanish law with English common law, it proved a disaster in terms of water rights. The common law let anybody divert as much water out of a river as they wanted, because there is no shortage of water in England. ”

….or New England or New York, more to the point, since California was a mostly Yankee project. I didn’t know they even tried to apply Comon Law to water in california, I thought they had just taken that part of Spanish>Mexican water law over, since they had none of their own, so this is a TIL for me. So it was that business down in LA that triggered it. Stands to reason.

29 Doc Merlin August 23, 2012 at 4:28 pm

Correct, they didn’t apply normal common law water rights. The western US uses spanish water law (to varying degrees).

30 TGGP August 26, 2012 at 7:36 pm

“Governing the Commons” by the late Elinor Ostrom (econ Nobelist of a few years back) discusses the development of water rights in California (I believe it was the first Common Resource Problem she studied). There’s no mention of common vs Spanish law, though there is another section on long-lasting irrigation systems in Spain.

31 Ryan August 23, 2012 at 4:10 pm

Anyone interested on this topic would be well served by the early chapters of “The Wealth and Poverty of Nations” by David Landes following Wittfogel. The hydraulic despotism thesis has had its ups and downs over the last several decades, but it’s good to see the idea be treated with serious empirical study.

32 mulp August 25, 2012 at 5:17 pm

The politics of water for a city do not result in democracy water policy.

But no city gets by on “rainfed” water. Large cities concurrent with large scale irrigation farming were no more democratic than the irrigated farming societies.

Allocating water for production in a city or on farms is driven by the same objectives and constraints, and democracy is not efficient at resolving all the conflicts with rationing water whether for cities or farming or fisheries.

Comments on this entry are closed.

Previous post:

Next post: