Territoriality and Conflict: an Economist’s View

By Radek Szulga May 17, 2014 8 Comments

Dominic Johnson and Monica Duffy Toft analyze the importance of territoriality in their article Bringing “Geo” Back into Politics as well as in the related piece in International Security, “Grounds for War” (2013/14). Peter Turchin, also stresses the importance of territory for understanding conflict in the response article (see also Turchin, 2014). While Ginges, Atran and Iliev depart from the focus on territory to some extent, they do not abandon it entirely.

From a strictly economic point of view territory matters because it is an asset. It provides a stream of resources which can be used for consumption, acquisition of geopolitical power, or evolutionary success. In aggregate it is in fixed supply. This suggests that conflicts over territory are likely to be negative-sum contests. This leads to limitations of a narrowly focused economic approach. Rational actors should be able to avoid conflicts which destroy net value. Regardless who happens to be in possession of the territory at a given moment, if side payments are possible, a peaceful settlement which makes both parties better off should obtain. Some elements of a particular situation – asymmetric information, negotiation costs, indivisibilities, imperfect commitment – might make it more difficult to arrive at a resolution. However, as has been recognized for quite some time, and as the authors note, the very frequency of conflicts necessitates an explanation which goes beyond simple pecuniary cost/benefit calculations even with these standard modifications.

The authors provide that extension by invoking evolutionary pressures which make actors more aggressive – territorial – when they are already in a possession of a territory. Peter Turchin incorporates the notion of Sacred Land in the context of the Devoted Realism strategy.

However, while evolutionary theory certainly adds to our understanding of present day social phenomenon, and indeed it is complementary to the economic analysis of conflict (as the authors aptly put it “Evolution is less an alternative to economics than an example of it”, Johnson and Duffy Toft, 2013/14) its application to this subject matter is by no means straight forward. The difficulties are both theoretical and applied.

On the theoretical side, there actually isn’t really anything specific about territory in the Hawk/Dove/Bourgeois (henceforth H/D/B) game. Yes, the story woven around the game-theoretic model is a narrative about individuals defending territory they own more aggressively than territory they seek to acquire. And yes, the Bourgeois strategy has an evolutionary advantage. But the source of that advantage is not possession of territory per se; the model assumes that the payoffs to ownership are symmetric. Rather, it is the ability of species to correlate their strategies based on an external event that gives Bourgeois the edge. But this event could really be anything. One could construct a different narrative to argue, for example, that conflicts are more likely when it rains than when it’s sunny. An individual plays Hawk if it’s raining and Dove if the sun is out. This will also be an evolutionary stable strategy (ESS), and could be used as an argument for making weather a candidate for explaining observed human conflicts. Another often cited anomalous example is the anti-Bourgeois strategy, which involves the incumbent playing Dove and the intruder playing Hawk and which has the same exact payoff as the territorial Bourgeois strategy.

Hence, the territorial narrative of the H/D/B model is constructed ad-hoc to make it fit facts which we already suspect to be true, but this narrative itself is neither a prediction of the model nor is it in any way a necessary component.

Of course, weather is a pretty poor predictor of conflicts. Likewise, the existence of the so-called endowment effect – the fact that we tend to value something more, simply because we already possess it – has been well documented in the literature (but see below). The theory is in need of other features which will make territory central.

I should note that in “Grounds for War” the authors discuss associated issues in much more depth but here space limitations constrain my ability to comment upon them fully. In particular, they invoke Herbert Gintis’ (2006) work on the existence of what he refers to as private property equilibrium, which corresponds to the Bourgeois strategy. The authors cite a result from the paper which shows that if there is a (small) cost of transferring the territory then the Bourgeois strategy is the only ESS. However, Gintis also extends his model in the other direction, where the anti-Bourgeois strategy is the only viable one.

The second difficulty lies in the application of evolutionary theory to modern social phenomenon. We can take states as the relevant actors although it is not immediately obvious that this is the appropriate mapping. Regardless, this requires that we follow through with the logic of the underlying model; it is the current possession of territory by a particular state at the outset of a conflict that matters. This is where the application of the model to the Crimean dispute runs into problems.  If we take the H/D/B model literally, then we should have actually predicted that it would’ve been the Ukrainian state which aggressively defended what was their territory at the outset. Of course the opposite happened. To make the situation fit the predictions of the model the authors are forced to stretch the definition of “ownership” in order to place it on the Russian side. But the fact that historically Russia has been in possession of Crimea for longer should not matter; past ownership is not part of the model. Likewise, the ethnic composition of the peninsula’s population, or a moral right to the territory, should not play a role either.

There are some complicating factors. The presence of Russian naval bases in Crimea suggests that true ownership was actually split. But that too is a different model. The purpose here is to point out the pitfalls of trying to map evolutionary theory into modern social interactions.

Peter Turchin’s concept also has evolutionary logic behind it, with the initial possession of land replaced by the Sacred designation. Putting aside the theoretical drawbacks of the H/D/B approach discussed above, the idea of Sacred Land appears as a macro version of some findings in recent research on the endowment effect, which link its presence to identify formation.

The original postulate of the endowment effect goes back to Richard Thaler’s 1980 paper, itself based on earlier work by Kahneman and Tversky (1979) (see also Kahneman, Knetsch and Thaler, 1990 and 1991). It is important to note that the existence of the endowment effect, its source, or its generality, is not universally agreed upon. In a forthcoming American Economic Review article Capicella et al. show that the endowment effect is present among hunter-gatherers who have had contact with modern society, but not among those who have been relatively isolated from it. The endowment effect might be a modern construct, a product of sociocultural evolution, rather than something inherited from our distant evolutionary past.

Other authors have also argued that the presence of the endowment effect varies with culture. For example Maddux et al. (2010) present evidence that it is stronger for Western subjects than those from East Asia.

More directly related, the endowment effect has been, unsurprisingly, of interest to marketing researchers. The main finding is that it matters because possession is linked to identity formation (Loughran Dommer and Swaminathan, 2013). We begin attach extra value to something we already possess when it becomes part of our identity. This finding is probably not surprising to anyone who has had a favorite shirt, toy, blanket, or hang-out spot (territory!)

This is a similar phenomenon to that of Sacred Land although at the level of individuals and particular goods rather than territory and peoples. Its synthesis is probably the same. Land becomes sacred when it is inhabited, worked and possessed, when it becomes part of a peoples’ identity. While it is common for land to be named after a people, sometimes peoples are named after a land (Prussians or… Americans) reflecting the process of land based identity formation. While the development of Sacred Land may have primarily a historical and cultural basis, that does not invalidate its importance.

The application of the idea of Sacred Land to Crimea, and the role it played in the outcome, also appears to be evident. Still, while incorporating Sacred Land into models of conflict can potentially increase their explanatory power, the analysis inherits all the limitations of the basic economic model discussed at the beginning of this essay.

Economists do understand the importance of non-pecuniary incentives and we do not actually confuse value with price. If economic theory can be summarized by a single simple postulate it is that “the demand curve slopes downward” – the more costly something is, the less of it you will do. What that something is left open ended. The claim is only that if the cost of fighting over territory is higher than actors will be less likely to do it – holding the “Sacredness” of the land constant. This means however that while an important new variable is identified, if the degree of relative Sacredness is known to the parties involved, the situation with Sacred Land isn’t conceptually distinct from the set up where monetary value of a piece of land is higher to one party. And the big question – the frequency of conflict over territory – is still on the table.

Recalling that indivisibility is one of the things which can impede a peaceful resolution, I tend to sympathize with Ginges, Atran and Illiev’s argument that Sacredness may very well be an all or nothing variable in the utility. This may actually be an advantage of the concept although it may also make its quantitative application more difficult.

My own preferred explanation of conflict and territoriality is some form of modified standard economic approach, although the unspecified model I have in mind involves equilibria which are very sensitive to parameter specification, initial conditions, and maybe are non-unique. Many disputes do get resolved peacefully, although because violence is more tantalizing, as social scientists and historians we tend to focus on the “interesting” cases which make for dramatic stories. Many disputes appear to linger in a kind of a limbo (for example, Cyprus or Taiwan) where a kind of steady state is achieved. This persists until some exogenous shock alters the parameters (game theoretic models are notorious for their sensitivity to specification, and applications of evolutionary theory to modern social phenomenon tend to inherit this “over-abundance of riches”) and opens up a window of opportunity, as well as a period of adjustment. Either conflict results or the dispute is resolved. An example of the latter could be the finalization of the German-Polish border in the treaty of 1990, motivated by the shock “fall of communism” and German desire for reunification. The present situation in Ukraine, where an internal shock, the overthrow of Yanukovich, altered the cost-benefit calculus of international geopolitics, is an obvious example of the former.

I admit that from the perspective of a social scientist interested in developing frameworks which have predictive power, this view of the world is somewhat pessimistic since these kinds of theoretical non-robustness imply that “all kinds of things can happen”. But while cautious, I do believe that it is worthwhile to chip away at the problem and the approaches discussed here contribute towards that end.


Apicella, C.L., Azevado, E.M., Fowler, J.H. and Christakis, N.A. (forthcoming). Evolutionary Origins of the Endowment Effect: Evidence from Hunter-Gatherers. American Economic Review.

Gintis, H. (2007). The Evolution of Private Property. Journal of Economic Behavior and Organization, 64(1), 1-16.

Johnson, D.D.P. and Duffy Toft, M. (2013/14). International Security. 38(3), 7-38.

Kahneman, D. and Tversky, A. (1979). Prospect Theory: An Analysis of Decision Under Risk. Econometrica, 47(2), 263-291.

Kahneman, D., Knetsch, J.L. and Thaler, R.H. (1990). Experimental Tests of the Endowment Effect and the Coase Theorem. Journal of Political Economy, 98(6), 1325-1348.

Kahneman, D., Knetsch, J.L. and Thaler R.H. (1991). Anomalies: The Endowment Effect, Loss Aversion and Status Quo Bias. Journal of Economic Perspectives, 5(1), 193-206.

Loughran Dommer, S. and Swaminathan, V. (2013). Explaining the Endowment Effect through Ownership. The Role of Identity, Gender, and Self-Threat. Journal of Consumer Research, 39(5), 1034-1050.

Maddux, W.W., Yang, H., Falk, C., Hajo A., Adaur, W., Endo, Y., Carmon, Z. and Heine, S.J. (2010). For Whom Is Parting With Possessions More Painful? Cultural Differences in the Endowment Effect. Psychological Science, 21(12), 1910-1917.

Thaler, R.H. (1980). Towards a Positive Theory of Consumer Choice. Journal of Economic Behavior and Organization, 1(1), 1-16.

Turchin, P., (2014). Russia’s Sacred Land. Aeon Magazine. April 3.

Published On: May 17, 2014

Radek Szulga

Radek Szulga

Specialties: Economic growth and development, labor markets in developing countries, economics of immigration


  • John Weber says:

    territoriality is inherent in life itself. Plants put down poisons to stave off other plants. Watch a bird feeder sometime. We simply cloak our behavior in the human clothing of verbage.

    • radek says:

      Interestingly, Herbert Gintis brings up his own bird feeder as an example, but of the anti-private property (anti-Bourgeois, anti-territorial) equilibrium. It seems birds at a feeder play Dove to the newcomers’ Hawk.

  • Jeremy Ginges says:

    The suggestion that sacred values are extensions of the general endowment effect is interesting but almost certainly incorrect. Sacred values are more than just things (whether territory, ideas or rights) that belong to us and which are valued highly. Simply being endowed with something does not make it sacred. Most things we own, even things that are dear to our identity (say our family home), can be valued along a monetary scale. The endowment effect simply increases the cost or material value of something. The very definition of sacred values is that they can have no material value. To measure them along a monetary scale is taboo (Ginges et al., 2007, 2011). In this way they are not fungible with material values – even those that have become very valuable as a consequence of the endowment effect.

  • Peter Turchin says:

    Radek’s commentary is very interesting because it highlights one of the most problematic aspects of an economist’s approach to solving conflicts such as the one over the Temple Mount between Jews and Arabs. The economist would find out how much the Temple Mount is ‘worth’ and then pay this sum as compensation to whichever party that will give it up.

    Atran and Ginges’ point is that, of course, that’s not how it works in the real world. Offering to one of the sides a monetary compensation is indignantly rejected. Sacred Values cannot be monetized.

    So far so good, but in reality any Sacred Value still enters a (quasi)rational calculation. A party that claims Sacred Value on a piece of territory, or some other object or symbol, still must decide whether it is going to go all out to acquire it now, or wait for a more propitious moment. There are many examples where a party would rationally wait for a moment when their power to win conflict is maximized WRT their opponent.

    So what I am saying is that while Sacred Values may not be monetized, still a more sophisticated version of the rational choice model may be of interest in exploring the dynamics of conflict. I say this not because I am a big supporter of the Ration Choice Theory, but to be fair to the sophisticated proponents, such as, I’d imagine, Herb Gintis might be.

    • Jeremy Ginges says:

      I agree that it is vital to understand how people manage to do this – to negotiate sacred values that come into conflict, to make temporal trade-offs, and to balance their values with their responsibilities. We are only beginning to understand this from a behavioral perspective. We know certain types of choices are taboo (such as overt material trade-offs) yet flexibility is possible in other contexts. Modeling decisions over sacred values is important and valuable – we just need to get the assumptions right.

      • Peter Turchin says:

        Agreed. The big question is what kind of modeling framework we can use, which would allow us to handle how sacred values are balanced against other considerations.

  • Concom says:

    Perhaps it makes sense to recall the main difficulties of this evolutionary game approach. It is assumed that foreign policy is based solely on the bilateral relations between independent states, which in turn are guided by national interests. This is perhaps outdated (pre-globalization) views on international politics. For example, in the conflict between Russia and Ukraine, very little depends on the national interests of Russia (since Russia is only an intermediate player in a “big game” between US-China-EU) and nothing depends on Ukraine (whose current government legitimized only by the U.S. Embassy, and whose current president is, de facto, Joe Biden and Victoria Nuland ). In such mixed local-and-global situations, many seemingly extra-rational decisions might be just an extra-national ones. As for the annexation of the Crimea, there is reason to believe that it was a joint decision of China and Russia.
    This does not mean that there are no sacred values; it means that for their accounting needs approach comprising at least two (global and national) hierarchical levels. Perhaps a good framework for such models can be hierarchical cooperative games (like Hawk/Dove/Bourgeois within flock and flock of hawk vs. flock of dove, etc).

  • radek says:

    Jeremy, thank you for commenting.

    On the endowment effect I do think that what differentiates it from Sacred Values is that with the latter there is some discontinuity or threshold. And after all, people can and do consider some material objects sacred; a family heirloom, a book, an icon (the Iconoclast lost that battle). One is just a more extreme extension of the other. And since both the endowment effect and sacradness work through identity, it makes sense that they have a common socio-evolutionary origin.

    The way you and your co-authors describe and define Sacred Values sounds a lot like Lexicographic Preferences from economic theory (which are rational, in the economic sense, but cannot be represented by a continuous utility function). The main difficulty that arises with the ability to bargain if people have these, is when they are also symmetric, which in this case translates to the situation where both parties consider the same piece of land sacred. The Israeli-Palestinian conflict, which you and your co-authors consider in your papers, is probably a good example of just such a situation, although I’m not sure what portion of all interstate conflicts are of this nature.

    in the asymmetric case (there’s some details here I’m leaving out) there could be room for deal making. As a quick example, consider that the flip-side of “I will not give up this land for any price” is “I will pay any price to keep this land”. The land may be non-fungible for one party, but deal-making is not. The party which does not consider the land sacred, provided it can make a credible threat and there is no problem with commitment could say “I know you consider this land sacred and will fight to the death to keep it, making it costly for me to take it. So how about you just pay me off to go away?” I can imagine that that kind of proposal would induce a lot less moral outrage than a proposal to sell the Sacred Land, and there are numerous historical examples of such deals.

    I wholeheartedly agree that we know very little about how these kinds of preferences are formed and what exactly they imply (even though economists note the possibility of lexicographic preferences, they are almost never used in actual models or applications, outside of pure theory). I also very much agree with the importance of the phenomenon.

Leave a Reply

<textarea name="ak_hp_textarea" cols="45" rows="8" maxlength="100" style="display: none !important;">

This site uses Akismet to reduce spam. Learn how your comment data is processed.