Predicting When a State Will Centralize

A 2015 paper by Koyama, Moriguchi and Sng (KMS) builds a model based on modern China and Japan's different experiences, suggesting a mechanism for how external threat and internal political and geographic characteristics drive centralization, as in Japan's case - or makes the state collapse, as in China's case. The small states either centralize to defend against the external threat or cease to exist, and due to their size, centralizing is not so difficult. The large states might not centralize, since there may be local disincentives that the central ruler cannot overcome. Consequently by the early twentieth century, the Japan (feudal a mere century before) emerged as a great power, and China (a centralized rich empire a century before) collapsed. KMS give statistics about railway-miles built per decade as a proxy for overall modernization, of course, referring to the dismantling of the Woosung Railway by the Qing.

This is intriguing, because it is not only consistent but complementary with the "labor-intensive agriculture drives state formation" hypothesis, and gives us a fuller picture of the appearance of the first states and their subsequent spread. The marginal agriculture hypothesis of state formation (MAHSF, pronounced "MASH-F") observes, again counterintuitively, that the first states (originating in a political analog of "spontaneous generation") tend to emerge in places which are actually quite marginal for agriculture. Like: the Nile Delta in the Sahara, which has to be irrigated in a time-sensitive and labor-intensive way; or, the Fertile Crescent, fertile only (again) if it is irrigated. Or the dry central Mexican highlands, or the Peruvian Andes. Not the plains of southeast Asia, or the Pacific Northwest, where even hunter-gatherers could develop a rich material culture.

But on consideration, MAHSF has an obvious flaw, which is China. (Any theory of state formation that does not account for the emergence of China is not a theory of state formation>) China was founded on fertile wet plains still today used for rice cultivation, and the exception shows that it's not a marginal physical environment, but rather the requirement for centralized labor, which played chicken-and-egg with centralized state capacity. Consequently a better account is the coordinated labor in agriculture hypothesis of state formation (CLAHSF, proncounced "CLASH-F.") It doesn't matter whether you live along a river in the desert and have to centrally organize labor because of the marginal environment, or you unfortunately have chosen a "tyrannical" crop like rice which places similar demands on its harvesters. In such a setting, advantage goes to the group that can organize labor, and the population and the state grows. CLAHSF explains why states emerge in certain places, and KMS provide a mechanism for how state formation "spreads" from those original nuclei. In the Old World, there are basically two nuclei and three vectors: from the Middle East to Europe and the Mediterranean; from the Middle East east along the Indian Ocean; and from China to the rest of East Asia. You'll notice the absence of Egypt, but modern Europeans, North Africans and Middle Easterners are much more the inheritors of statecraft and culture in general from the Fertile Crescent than from Egypt.

KMS apply their theory beyond China and Japan and look at both Anglo-Saxon England (ultimately unified by Alfred in response to Norse invasions) and Muscovy (in response to the Mongols; and Russians still psychologically think of Ukraine as a buffer state - the "Wild Fields" - since the Mongols easily crossed prairies but halted at the boundary of the boreal forest.) A third example that cries out for such a treatment would be the first unification of India after Alexander's arrival at its frontier; not to mention the seeming counterexample, why the Greeks, even after a long-standing threat from the Persians, never did manage to unify and remained a set of city-states - to be over-run by Alexander.

An entirely different question is why China, which to this day can be seen as the amalgamation of ethnicities it is (with the Han dominating in the east) unified so early, and essentially stayed unified under different rulers (as if the Roman Empire had never fallen), when Europe had one such period early on, which was never repeated (a Warring States period that never ended.) Thought experiment reversing this here.


Koyama M, Moriguchi C, Sng Tuan-Hwee. Geopolitics and Asia's Little Divergence: A Comparative Analysis of State Building in China and Japan after 1850. GMU Working Paper in Economics No. 15-54, 19 Nov 2015.