See the Tragedy of the Commons

Warning: Stefan Geens, an expert in the use of Google Earth, has taken a close look at the Zimbabwe photos and he argues that they are the unfortunate result of misinterpreting Google post-processing and updating.  I am not an expert but at the present moment I find his analysis compelling and I have taken down the powerpoints in which I incorporated the flash animation produced by the Center for Global Development.  

A few points of importance.  First, the sharply delineated lines in the photos dividing communal and privately owned land is real and is evidence consistent with the tragedy of the commons.  Geens writes to me:

The observation that the communal farmland seems worse off than the private farmland is a plausible one. And since the Landsat imagery could well have been taken around the year 2000 (or even earlier), Richardson's image can reasonably be used to support the thesis that communal farmland is not as well managed as private land.

What cannot be supported is the "animation" aspect which appears to show but in fact does not show a change over time.  

I should note that I have no reason to think that there was any malfeasance on the part of the Center for Global Development, a great organization, or Craig Richardson who originally brought attention to the photos.  Indeed, my reading of Geens is that without some expert help it would be hard not to make some errors in interpreting Google images, which have a more complicated provenance than one would imagine.  In addition, Richardson emailed me that he showed the images to farmers and government agriculture officials in Zimbabwe who agreed that they were representative of what was going on on the ground.  (See also Geens on this point noting burning.)  Nevertheless, although the redistribution of land in Zimbabwe clearly led to a drop in productivity and output, the animation is not evidence on this point.

One interesting point about this episode is that the story and the flash animation have been on the CGD website for several years but the huge exposure of MR means that we can draw on the expertise of many people around the world. Although I put some work into putting the powerpoints together I am pleased that we were able to correct this so quickly.

Original post follows but with PPs deleted.


In 2000 Zimbabwe began to forcibly redistribute land from private but predominantly white-owned commercial farms to much poorer black farmers who toiled on communal lands.  Stunning pictures from Google Earth collected by Craig Richardson show the result.   

Take a look at the Before picture.  The communal land on the left is dry, dusty and unproductive compared to the private farmland on the right which is green and dotted with blue ponds and lakes.  Why?  There were two theories to explain this difference. 

  • The Tragedy of the Commons – the farmers on the communal lands did not have the incentives to invest in the land and thus the land eroded and turned to desert.
  • The land on the right (which was owned mostly by whites) was better quality land.

Both theories could be true.  Regarding the latter explanation, however, notice that the dry communal lands on the left are sharply delineated from the green private farms on the right–so sharply that soil quality and rainfall alone are unlikely to explain the difference.

So what happened after the land was redistributed beginning in 2000 and all of it made communal?

Click on the arrow to progress between before and after photos

After reform the land quality worsened everywhere. In particular, note that the blue lakes and ponds on the right became dry and empty as farmers no longer had an incentive to invest in maintaining these resources. The tragedy of the commons.

This excellent visual look at the tragedy of the commons was produced by
Todd Moss at The Center for Global Development based on pictures and
ideas from Craig Richardson.  Of course Zimbabwe had many problems before and after this forcible land redistribution.  You can find more pictures, background information and a lengthier discussion of this episode here.


This is great, thanks Alex!

It's possible that "tragedy of the commons" has less to do with the problem, than with putting loyalists and cronies in charge of said land. Much of the seized land has gone to (black) cronies, not to poor (black) farmers outside the cities.

Interesting, but not as demonstrative as is wants to be. The overall difference in color could as easily be pictures taken at different times of year, or in years with different precipitation levels, as a secular difference in greenness. The loss of the lakes is obvious, but the the other land-destruction outcome that might be expected, deforestation, is not very evident.

Very effective demonstration, but it does need a little more investigation:

Did the communal farmers (esp in the after case) have the capital required to maintain the land and water resources?

What were the weather conditions between the two cases?


Are you kidding me? You think Zimbabwe as a whole is better off now than 10 years ago?

I disagree with Alex's interpretation of the images.

The previously communal land on the left also looks much worse in the "after" image, and the ponds there are also gone. I suspect the changes have much more to do with the collapse of the Zimbabwean economy and the total breakdown of the rule of law. If you examined before-and-after photos of Zimbabwean workshops and businesses over the same period you'd probably also see dramatic changes. The Mugabe regime's desperate and violent tactics are not a good source of 'clean' natural experiments.

South Africa has had a number of similar problems with its businesses, which are required to have a black co-owner.

The white farmers had typically been farmers for generations and had learned to take care of their land and make it productive(In fact, Rhodesia, the precursor to Zimbabwe, was called the "breadbasket of Africa" back in the day because those farmers were so productive). The farms were stripped and given to poor blacks with no farming experience. The white farmers by and large left, and the inexperienced black farmers didn't know how to manage their land and resources.

Is it a tragedy of the commons issue? Maybe partially, but it's just as much a "when you redistribute ownership of fragile physical resources to people who have no idea what to do with the resources you give them, those resources get destroyed".

South Africa is having a similar problem - their education and infrastructure resources have been absolutely crushed and exportable manufactured goods reduced substantially because every white owner had to find a black co-owner, which led to brain drain as talented whites left to find better opportunities elsewhere. Sadly, people can't just learn to operate a factory with no substantial training.

They both also have the problem that their governments redirected investment from infrastructure to education, but those educational attempts largely failed (language differences are one major reason, AIDS is another). South Africa now has frequent rolling brownouts and blackouts. Zimbabwe has crushing food shortages, hyperinflation and no ability to get credit. The poorest Zimbabweans are going to South Africa, but still having trouble making a living there, which has increased crime and the presence of the drug trade.

South African blacks have more rights than they did under apartheid, and that has been crucial to any human decency/sense of fairness, but the blunt redistribution has, by and large, cratered South Africa as a world economy. Zimbabwe started from a better situation than South Africa and ended up worse 50-60 years later.

(and Tomasz, given the extent of the food riots that have occurred in Zimbabwe over the last decade, you'd be hard pressed to say ANYONE was better off than before redistribution, unless you're counting politically-connected members of the ANC).

The right-hand side absolutely looks worse off in the "after" case. But then again, so does the left-hand side. We can't really rule out other causes like changes in the broader economy or in weather (like a drought). If the left-hand side had looked the same in each case, you'd have a great argument to make; but that's not what the evidence appears to say. In fact, given the fact that both sides seem to deteriorate at about the same rate, that seems like evidence that it is not due to the redistribution of land.


Sidenote on Tomasz's suggestion that the benefits of redistribution could potentially outweigh the costs of productivity. Other commentors like Cliff and Trevor F ridiculed this suggestion, citing the fact of Zimbabwe's growing poverty. That is no doubt true, but that has to do with many policy changes, whereas Tomasz is only considering the costs and benefits of this single one. Sure, all of the changes were bad, when you consider them as a bundle, but that does not necessarily imply that every individual change was bad. So that evidence actually has no bearing on Tomasz's hypothesis.

+1 to david. This isn't land redistribution to the poor, it is land distribution to supporters who have no incentive to farm the land.

Rather: see the tragedy of absolute power and central planning.

This proves cimate change is a serious problem. That explains why Zimbabwe is doing so bad. They need more taxes on CO2 and then you'll see things finally start to improve.

Yep, those are the only two theories. Definitely no other possibilities. No sir.

Great images; I think it's good, illustrative evidence of what we knew already--Zimbabwe is much worse after the land redistribution.

However, I agree with the other commenters who attribute it primarily to capital flight and economies of scale. I think black landowners have the same incentive to invest, but didn't have the resources, credit, and perhaps know-how to do so. Plantations used for cash crops may also have been more efficient than small tracts used for family subsistence (I'm not sure if successful co-ops were formed; I doubt it).

The color balancing is different between the images. The lakes are very uniform between the images, leaving the possibility of a color correction explanation or even sun glare "whiting out" the water. The after lakes still show as bluer than the whiter land plots.

There are almost no exceptional plots of land. If the plots on the right were showing degradation of care, it seems unlikely that the degradation would be so uniform unless they were all completely fallow.

There may very well be a degradation going on, as suggested by the clear demarcation in the before image, but these two images don't appear to both be apples.

You just hate the poor. ((snark))

Are you kidding? If I understand what's going on the picture, there's no basis to infer anything about the effects of redistribution here. For that, we would have to have some areas in the "after" shot that were not redistributed. Then, inferences about the effects of redistribution could come from looking at how changes in the non-redistributed areas compared to changes in the redistributed area (this is basic "difference-in-differences"). But we don't have those non-redistributed areas in the "after" photo. So we cannot disentangle the effects of redistribution from any other adverse event that occurred between the "before" and "after" photos (e.g. drought, different times of year, different photo technology, etc.).

While I'm sympathetic to the sentiments, I'm not sure the before/after picture actually provides any evidence about tragedy of the commons. Other commenters have already raised the issue of whether these were in fact communal farms, but there appears to be another issue. Just looking at the photos, the land on the left seems to be mountainous and not flatish farmland (as on the right).

The rivers appear to flow in branch-like patterns as you would expect when flowing down mountainsides. This is especially true when you think about how water flows, and try to discern what could make a river suddenly branch up to 90 degrees.

Rivers flowing through farmland, in contrast, tend to either snake along, often curving back on themselves, or in completely straight lines because of being diverted through culverts and ditches used for irrigation. You would also expect to see more ponds and lakes in the foothills of the mountains into which these rivers would eventually flow, much as you see on the right.

If I am right, then the stark contrast between the left and right images would likely be attributable to the lesser amount of vegetation on rocky mountainside, as compared to farmland, and the before/after images are probably the difference between dry and rainy seasons.

Although I am not suggesting that this is definitely the case, the differences in terrain immediately jumped out at me.

So, land redistribution by force is bad if it results in lower utility, but land redistribution by force is good if utility increases?

Land redistribution is bad because large numbers of small farms operating at subsistence levels (or nearly so) are a recipe for persistent, endemic poverty. No developed country is characterized by large number of people farming small plots of land.

I think the artifacts pointed out by Dmitri are likely the result of Google Earth's photo stitching process. That seems like a more plausible explanation than deliberate photo-shopping, especially if you look closely at similar google satellite images. Same with the water in the ponds. A lot of water in close-up google satellite images is blanked out, I think because of its high reflectivity. The implication is similar, though--the variations between the before and after may be due to differences in the photos themselves more than actual differences on the ground.

The more I look at this, the more I think its just photographic differences. The changes on both sides of the dividing line appear just about equivalent. The color balance just changes from one photo to the next from more blue and green to more brownish--the change is even all over. As for the lakes, I think the sun was likely at a different angle, causing the blue to be washed out by reflection. The outlines of all the ponds are still there.

Congratulations, Alex. You've made exactly the same argument my great-grandfather made about the difference between his family's farm and the farms of his family's freed slaves after the Civil War.

I guess it's just one of those tragedy of the commons things when you plop someone with zero economic resources on a patch of land next to a patch of land still in the hands of someone with farm equipment, lines of credit, longstanding relationships with buyers, market savvy, lifetimes of experience managing a farm instead of just laboring on one, etc., etc., etc.

Like you my great-grandfather said it was all about white vs. black moral and mental character plus the capacity for pride of ownership. Of course he was a vicious racist who died before the Great Depression but, hey, he'd agree with you 100%.


Also, about that "tragedy of the commons" thing. The paradox being that before the tragic loss of centuries-old commons could become a metaphor for individual selfishness there had to be commons for centuries. I'd have more patience with claims of big-L Libertarianism if one of you guys could come up with an explanation that didn't rely on distinctly non-economic explanations like social dislocation and the breakdown of law and order.


Like Doug I agree it looks like the land on both sides of the dividing line became more deteriorated over time. This could be because the Mugabe government is a bunch of murderous thugs who ran the whole country into the dirt -- a phenomenon that would apply to all farmers. It could be because the whole area has been drying out (either in terms of rainfall or overuse of the water table for irrigation) such that marginal land is dropped from production. In small-l libertarian terms it could be that even though they're in a very definite and perilously ethnic minority, Zimbabwe's whites are both better organized and better armed than recently-relocated "communal" farmers and consequently better able to resist the very real and very persistent graft and extortion by agents and allies of national and local government. It would also, for instance, be interesting to see satellite photos of the entire region to see if this was a localized effect. In other words, Alex, I'm not saying you're wrong. I'm just saying you usually want to be right so badly you're inclined to jump on outlier data and say "ahah, vindication!"


Finally, re "there were two theories to explain this difference." While I know you're all excited about getting this "proof" of the tragedy of the commons into economics classrooms, if I was instead putting together powerpoints for logic and rhetoric classes I might want to use your slides to demonstrate the logical fallacy of False Dilemmas. Yes, if your only possible choices are A and B then if the answer isn't B (better land) then it must be A (tragedy of the commons.) Yet number of other theories, including political, geological, and economic theories, had been proposed here in comments even before I stumbled along. Seriously, dude, you should talk more to Robin Hanson. I believe his office is just down the hall from Tyler's. (If you're in a hurry just read the title of Hanson's blog.)


Let me remind readers that they can find more information about this episode in the link that I gave in the post:

The tragedy of the commons is only one of the many tragedies to beset Zimbabwe. Craig Richardson who collected the photos from Google Earth has written a book, The Collapse of Zimbabwe, which I recommend, which discusses the flight of foreign capital, the attenuation of property rights, the uncertainty, the fear and many other factors which led to this problem.

"Like you my great-grandfather said it was all about white vs. black moral and mental character plus the capacity for pride of ownership. Of course he was a vicious racist who died before the Great Depression but, hey, he'd agree with you 100%."

Woah. He never said it was all about white vs. black moral and mental character. That strikes me as a pretty outrageous and slanderous remark--and not a very fair or honest debating tactic. He said is was about communal vs. private ownership (hint, black people can and do also own private property, and white people can and do share communal property). You and mulp are making some pretty uncalled for remarks regarding race in these comments.

"Its SANCTIONS silly"

In as much Ido not support the white ownership of land or black ownership in Zimbawe. I would prefer land reform for the benefit of all Zimbabweans both white or black without the colour factor. The problem in Zimbawe right now is we have two extremist groups the so called white commercial farmers and Zanu Pf technocrats. There was transfer of power from white supremacists to black minority.Smith racist regime to Mugabe racist regime.Land reform was necessary in Zim but not for Agric purposes for living standards only. Some of us we are used to rural life where I just need a piece of land for my 10 herd of cattle, my 2 wives and kids not a commercial farm.Whilst I am working as a teacher when I retire I have to go back to the ruralarea sothat i can relax and enjoy my life we do like this city life.The mugabe land reform as just another 1930 land apportionment act.Give us small pieces of land rather where we will come and stay when we retire.

It would have been interesting to see what would have happened if rather than requiring every business have a black co-owner, they had required an appropriate number of black apprentices to each business.

So the end of slave capital based plantation system replaced by inefficient share cropper system in which individual poor white and blacks ran small plots was a step backward in a developed nation like the late 19th and early 20th century US? To reduce poverty, slave property and large plantations should have been sustained?

It's not the 19th century any more. Perhaps you're aware that from the beginning to the end of the 20th century, the U.S. went from the farming sector accounting for something like 40% of the labor force to around 2-3%. Would you want to reverse that process with 'land reform' -- breaking up and redistributing large commercial farms? If not, why on earth would you recommend such an approach for other countries?

Or you may have noticed what's going on China, yes? Tens of millions of formerly rural people leaving small villages and farm plots for opportunities in the modern economy -- the effect of which has been fundamental transformation of the lives of the people in China. If you were a parent in China, would you want your children and your children's children to stay in the village and work a small plot like the generations before them?

When I was down in SA a couple of years ago, I spoke to some people that had relocated from Zimbabwe. They had traveled with farm owners that had left the country and moved to SA. They said that once it was clear that the Zimbabwe government was going to take their land by force or pay sub market prices for the land, the private farm owners ruined the land out of anger. They knew they were going to be kicked off the land, so they dumped lime and various chemicals on the land to render it unproductive for a few years.

You realize that the thing you linked to has no overlap at all with the years the pictures cover, right Marty?

I would be wary of using processed aerial imagery for this type of assessment. Where is the "line" in the below link from Google maps?,29.475632&spn=0.396089,0.628967&t=h&z=11

Zoom in and out and you will see clouds split in half, or images from different sources, dates, and/or wavelength sensors. The stark "green" line seen could be a result of post-processing of different flight swaths, at different times, with differing adjustments to make a CIR image (color-infrared) look closer to a natural color one.

Thanks for the idea. I think I'll do the same thing with the refrigerator in the break room at work. Before: Just cleaned. After: One week later.

Hurray for the commenters here, who have note the rather automatic grasp for a dogmatic explanation. Two pictures and a reference to private virtue and public tragedy - all done, now. Thank you all for making sure this nonsense didn't stand.

Below is what I posted at DeLong's place, when he linked to this piece --

We "knew" for a long time that the Tragedy of the Commons was a fact, and that regulation was called for. Then, we "discovered" that the Tragedy of the Commons was a theory that didn't stand up to scrutiny. Then, we learned that private ownership is obviously implied as the right response to deterioration of common property, which we suddenly "remembered" to be supported by the data after all. So, after decades of simplistic, dogma-illuminated understanding of what-ever-the-heck it is that happens on publicly-held resources, Elinor Ostrom comes along and shows us that culture and tradition play an enormous role in determining how commonly held resources are treated, and that simple bumper-sticker explanations fall short of reflecting the complexity of common resource use.

If land was taken away from local folk generations back and given to colonials, then turn back over to (usually utterly different?) local folk, do we have reason to think that stable cultural norms and traditions are at work?

Don't suppose anybody ever coined a term like "The Tragedy of Overturned Norms" or "The Tragedy of Changing the Rules" or "The Tragedy of Blinkered Nonsense"?

C'mon, Ostrom's Prize wasn't granted so long ago that we have forgotten her, was it? She has given us a new set of tools, and what do we do? Same old crap. Please...

You're begging the question Charles. I think the point that is trying to be made is that communal rights regimes suffer from structural flaws that will repeatedly and predictably lead to waste, whether that comes in the form of slash and burn agriculture, overfishing, loss of open space, etc. You can disagree with that argument, but simply assuming it is incorrect is no more valid than assuming it is correct.

Reading through the comments, I notice several people referring
to this as "land redistribution." That misses the point that
Alex is making. The issue is communal ownership and the impact
of having many families share the use of the same piece of land.
To the best of my knowledge in every society that has ever done
this, and there have been many such examples, the land in question
has degraded significantly. This is what is meant by the Tragedy
of the Commons. The most dramatic example currently is the oceans,
which are communally owned, and despite their enormous size are
degrading quite rapidly.

Like some others here, I assume that a significant part of the
land taken from the white farmers was not given to communal
ownership but is instead in the hands of powerful regime supporters.
Supporters that are also likely incompetent farmers. But given
the marxist ideology of the regime, I also imagine that most of the
land was given to groups of people.

I know of two countries where land redistribution was very
successful. I suspect there are more such examples. In neither
was the land given to groups. In Taiwan, the KMT gave the land
to the landowner's tenants. That is they gave ownership to the
people that were actually farming it. Taiwan's agricultural
productivity shot up. This was a side-effect, productivity was
almost certainly not the KMT's motivation for the land

In the U.S. land redistribution of the late 1700s and early
1800s; part of the process of acquiring legal title to someone
elses was land was actually making living on it. Short of
demonstrating that one could farm it, the local community was
unlikely to award an invader with title to it. As with Taiwan,
the point of land redistribution was not to increase agricultural
productivity, but it was in fact a side effect.

Stefan Geens sees a problem with Craig Richardson's images:

"His before-and-after images are derived from the same original raw Landsat image. They are from the same point in time... All the fallow fields in one image are fallow in the other. Not a single field has been tilled, planted, harvested, abandoned, burned or overgrown, in whole or in part, in the supposed interval.

As anybody who's stared long enough at satellite images knows, the state of the fields is like a fingerprint. If they match, the images match. The main difference between these two identical images is that one is yellower than the other, and that in one the blue hues have been turned way up, while in the other they have not. Instead of scorched earth and dried up reservoirs, we have two different post-processing settings."

More at this link:

10 year is a short period of time to evaluate the results. However, even 10 year is enough to conclude that the reform is a failure. I think they should educate farmers before giving precious land.

I must say I find highly amusing commenters who seem to think that Ostrom's work "refutes" the notion of the tragedy of the commons.

It's always good when some country is quick to capitalize upon a neighbor's stupidity. One would hope that this disincentives stupidity, but we seem to have rather short memories.

Comments for this post are closed