Social Evolution Forum
FIND sef:
Institutions and Efflorescences in the Pre-Modern World
Reconstruction of Roman-era Carthage

Having received a PhD in Classics and now working primarily in the social sciences, I have been eagerly following Peter Turchin’s blog series on institutional development (Institutions and the Efflorescence of classical Greece; Institutions: What Are They Good for?, Institutions Are Software; Organizations Are ‘Meatware’, and That Cultural Background). In particular, the most recent post about Josh Ober’s work on classical Greece was quite fascinating to me, and I wanted to take the opportunity to share my thoughts on the issue of institutions and economic growth in antiquity and on the use of social science methods in ancient history.

In my work with Peter at SESHAT: Global History Databank, we are taking very much the same approach as Josh Ober and some of his peers from Stanford – Joe Manning (our colleague who is now a professor of Classics and History at Yale), Ian Morris, and Walter Scheidel – in attempting to treat ancient historical evidence as data that can be used to test various hypotheses about the functioning of ancient states. While I greatly admire Ober’s work, the way that he approaches the topic and some of the conclusions he draws seem to me more specious than accurate. Mainly, I worry that it is not doing enough to convince ancient historians and scholars in the Humanities more generally of the utility of employing hypothesis-testing and causal analyses with historical evidence (and believe me, most people working in the field today do indeed need convincing). Hopefully, this is something that Peter, myself, and the other scholars involved with the SESHAT project can rectify in the coming years.


Sign up for our newsletters

I wish to receive updates from:

Ober’s argument is, essentially, that Athens during the classical period (roughly 500-300 BCE) developed democratic institutions that promoted human capital accumulation, lower transaction costs, urban development, and equitable wealth distribution which, in turn, fostered intensive economic growth (that is, growth of GDP per capita). Many people have found fault with Ober’s data, especially the evidence he cites for intensive economic growth and the wealth equality. While some of the details of Ober’s analysis are suspect – especially his assumption-laden argument that 42-58% of the Athenian citizens lived at ‘middle class’ level of wealth (see his 2010 article “Wealthy Hellas”, a copy of which can be found here) – on the whole, I find his conclusions about economic growth, urbanization, and institutional stability both convincing and a rather refreshing take on the ancient Greek economy.

The problem for me comes with Ober’s claims about the cause of Athens’s impressive and relatively unusual (for the preindustrial world) economic success. Namely, that classical Athenian democracy produced all of the ‘good things’ that Ober highlights in his work. The issue is, in essence, a methodological one; Ober sees a fairly clear correlation (classical Athens experienced intensive economic growth // classical Athens had democratic institutions), but he reads it as causation (classical Athens experienced intensive economic growth because it had democratic institutions). I understand that it is a very compelling argument, especially given the current global reality of extreme income disparity and persistent authoritarian regimes. But the myopic scope of this sort of analysis, focusing exclusively on the ancient Greek world, does not allow Ober’s hypotheses to be properly tested or falsified. A more fruitful way to test this sort of hypothesis is to look at a wider spectrum of pre-modern societies to determine if there were other instances of intensive economic growth. If so, were they all attended (and preceded) by the adoption of democratic institutions comparable to those of Athens?

This kind of hypothesis-testing is precisely what we are doing with SESHAT. Although still in its preliminary phases (stay tuned for details from Peter on this blog!), our work so far strongly suggests that Athens was not as unique in terms of significant economic growth as Ober and many others think. Indeed, nearly all of the main indicators employed by Ober to demonstrate Athens’ achievements—increased urbanization, monetization, marketization and trade, sizable median housing, expanding productivity and occupational specialization—were known also in numerous other pre-industrial societies: Rome during the late Republic–early Imperial periods (roughly the 1st c. BCE to 1st c. CE); early imperial China during approximately the same period as well as another Chinese ‘golden age’ during the Song dynasty (10th-12th c CE); Egypt during the Ptolemaic period (3rd – 1st c. BCE) and likely during parts of the Middle Ages (under the Mamluks, for instance); northern Europe in the later Medieval period; Tokugawa Japan (17th–19th c CE); and several other such pre-modern economic ‘efflorescences.’

The crucial point here is that the causal link between democracy and economic growth on which Ober’s arguments are pinned simply does not hold up to in-depth, cross-cultural scrutiny. All of these polities experienced what looks like intensive economic growth similar to, if perhaps not quite as dramatic as, classical Athens, but they each featured some form of autocratic rule, including the ostensibly democratic republic of Rome. Further, they all employed very different legal institutions and property rights regimes (the central pillars of the New Institutional Economics approach to which Ober and many others ascribe).

The obvious follow-up questions are, of course, if democratic governance was not the primary cause of these periods of increased economic performance, then what was? And what, then, of the still-appealing correlation between Athens’s economy and democratic governance? One intriguing possible connection between these pre-industrial efflorescences is whether they all shared certain cultural traits that have been associated with high institutional performance, notably trust of other members of the society, the ability for collective action, and a general willingness to engage in prosocial behavior (the evolution of these traits through group-level selection is part of what SESHAT is exploring). For example, Italian economic historian Guido Tabellini makes the case that trust-promoting elements of culture determine how well institutions function, rather than the reverse scenario that Ober suggests (see a recent article making this argument here).

These are all important issues that we will continue to explore with the SESHAT project. For, only by gathering massive amounts of data about a wide variety of societies will it be possible truly to test different theories about economic growth, equality, quality of life, and a host of other vital issues that affect both our understanding of these historical societies, and remain pressing concerns in the modern world.

Our approach is important because such cross-cultural and interdisciplinary work is often met with skepticism, if not outright prejudice, by historians and other humanities scholars. These critics raise familiar tropes about the inapplicability of historical evidence as analyzable data as well as a weariness about extending research outside of one’s area of expertise. And when you’re working with thousands of data-points from dozens of different societies and time-periods from across the globe, as we are doing with SESHAT, going outside of your area of expertise is simply unavoidable. Yet, it is becoming increasingly clear that such interdisciplinary and cross-cultural engagement, although certainly susceptible to many methodological pitfalls, is the only way to advance our knowledge and avoid the correlation-to-causation trap that even learned and careful thinkers can so easily fall into. Our ability to reap the rewards of this ambitious approach will only be decided once we have gathered sufficient data and run our analyses. But, as a wise man once said, the proof will be in the pudding.

social science


Join the discussion


  1. Mike Alexander says:

    This post is off-topic for the present article but on-target for the series. I just submitted a paper for Cliodynamics Journal that involves the parameter S used in a variety of models for secular cycles. S is termed “state resources” by Turchin and co-workers, but it seems to be more than that.
    Population began rising in England in the late 11th century. Why then and not before? Cliodynamic models explain this in terms of rising S. It is hard to see how William the Conqueror or his sons possessed some “state resources” unavailable to Cnut the Great or his sons. According to folks like Thomas Bisson or Harold Berman what we would call “government” did not emerge in either Normandy or England until the late 12th century as the earliest. In the absence of government does the concept “state resources” even make sense?

    Could “S” in this particular instance be a cluster of beliefs that produced a “normative pacification” that protected new settlements from the effects of warlords’ private wars.

    Such an effect might result from a reluctance for Christian warlords to directly attack an *authentic* religious establishment (i.e. one who might be occupied by “proto-saints”). That is, a person who if killed will immediately appear in heaven where he can employ spiritual power against you (if you are a believer). And if you aren’t you still have to worry about warlords who are believers who may decide to kill you as penance for their sins, particularly if their bishop suggests it. So maybe you leave the independent monasteries (i.e. those not associated with a political faction; those seen as holy) strictly alone.

    Such independent monasticism appeared as a new “meme” in AD 910 with the founding of the monastery of Cluny. The explicit intervention of the Saints was promoted by a series of church councils and open air meetings beginning in 975 (link to

    The Normans brought these French cultural innovations with them to England and population growth followed.

    This development seems to me to be an example institutional evolution as being discussed in this series and thought I would post on it before the series ends.

    • Daniel Hoyer says:

      This is an interesting comment, and I think definitely fits with the general themes of this series on institutions. I agree that the sort of cultural norms that you point out, insofar as they actually affect people’s behaviour and choices, do indeed count as part of the resources or factors driving the outcome of state activity. Many people (including myself) would include cultural norms and traits as part of the overall institutional system of a society, even if culture is a more informal institutional than something like property law and is certainly harder to measure.

      I guess the question is how would such a ‘meme’ like a taboo against attacking independent monasteries evolve in a competitive military/political environment, and how did it interact with other traits (an overall incentive to attack and conquer nearby territory, etc)? Perhaps one of the more fruitful ways of approaching these sorts of questions is through comparative work; looking at how similar religious political communities acted in similar scenarios, or even how different sorts of communities acted, could be an interesting way to pull apart what was unique about these Medieval Christian warlords from what is more generally true about small groups facing intense military competition.

      In any case, I look forward to seeing the article!

  2. […] now Peter Turchin and Daniel Hoyer report that the Stanford classicist Josiah Ober is coming out with a new book called The Rise and […]