This is a repository copy of Constraining the Samurai: Rebellion and Taxation in Early Modern Japan. White Rose Research Online URL for this paper: http://eprints.whiterose.ac.uk/137080/ Version: Accepted Version Article: Steele, A, Paik, C and Tanaka, S orcid.org/0000-0001-6246-3332 (2017) Constraining the Samurai: Rebellion and Taxation in Early Modern Japan. International Studies Quarterly, 61 (2). pp. 352-370. ISSN 0020-8833 https://doi.org/10.1093/isq/sqx008 © 2017, The Author. Published by Oxford University Press on behalf of the International Studies Association. This is an author produced version of a paper published in International Studies Quarterly. Uploaded in accordance with the publisher's self-archiving policy. Reuse Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of the full text version. This is indicated by the licence information on the White Rose Research Online record for the item. Takedown If you consider content in White Rose Research Online to be in breach of UK law, please notify us by emailing [email protected] including the URL of the record and the reason for the withdrawal request. [email protected] https://eprints.whiterose.ac.uk/ Constraining the Samurai: Rebellion and Taxation in Early Modern Japan* Abbey Steele† University of Amsterdam Christopher Paik‡ NYU Abu Dhabi Seiki Tanaka§ University of Amsterdam Abstract On the eve of the Meiji Restoration in 1868, the nearly 300 semi-autonomous domains across Japan had widely varying tax rates. Some handed over 70 percent of their rice yield to the samurai ruler of the domain, while others provided 15 percent. This variation existed in spite of the similar fiscal demands that the domain rulers faced within the Tokugawa regime, the feudal system that governed Japan between 1603 and 1868. This period was remarkably stable, with no foreign or domestic wars, which allows us to focus on the impact of pressure from below on taxation. We study the extent to which peasant- led rebellions and collective desertion (“flight”) lowered the subsequent tax rate imposed by samurai rulers. From newly compiled data on different types of peasant-led political mobilization from petitions to insurrections, we find that large-scale rebellions and flight are associated with lower tax rates. We interpret the results as evidence of rebellious or mobile peasants' ability to constrain their rulers, while the more complacent fail to win concessions. Our findings suggest that peasant mobilization played a role in restricting state growth in early modern Japan through tax concessions. * Previous versions were presented at the Advancing the Scientific Study of Conflict and Cooperation: Alternative Perspectives from the UK and Japan, Essex, March 2012, the Annual Convention of the Midwest Political Science Association in Chicago, April 2012, and the Annual Conference of the Social Science History Association, Vancouver, November 2012. We thank the participants of the Empirical Studies of Conflict Workshop at Princeton University and of the Political Science Research Workshop at Syracuse University for their comments, as well as Matthew Cleary, Dominika Koter, Frances Rosenbluth, Jacob Shapiro, Ryan Sheely, Paul Staniland, Brian Taylor, two anonymous reviewers, and the editors of International Studies Quarterly for helpful feedback. We also thank Akari Yanada, Wakana Narisako, Kaori Mitsushima, and Cosmo Usami for research assistance, and Akira Nishizawa for sharing his shapefiles. The authors acknowledge support from the Air Force Office of Scientific Research (AFOSR) under Award No. FA9550-09-1-0314. † Email: [email protected] ‡ Corresponding author. Email: [email protected] § Email: [email protected] 1. Introduction On the eve of the Meiji Restoration in 1868, the 267 semi-autonomous domains across Japan had widely varying tax rates. Some villages had to hand over 70 percent of their rice yield to the ruler of the domain, while others only had to provide 15 percent. This variation existed in spite of the similar fiscal demands that the samurai rulers (daimyo) of the domains faced within the Tokugawa regime, the feudal system headed by a shogun that governed Japan between 1603 and 1868.1 Relative to Western Europe, the Tokugawa era was more stable, free from both internal and external wars, and even threats, until the mid-19th century. The daimyo were free to set their own tax rates, and to send their retainers (lower-ranked samurai) to collect the revenue from the peasants in their realms. If each ruler aims to maximize extraction (Levi, 1988), what explains the tax rate variation we observe? Though variation in levels and forms of taxation across and within autocratic regimes remains puzzling (Cheibub, 1998), the literature does offer some theoretical expectations.2 A classic political economy model theorizes that an autocrat determines an optimal tax rate so that he can maximize his payoff in the long run, by allowing his subjects to retain necessary resources for continued economic activity into the future. In this framework, a 1 Here we follow Ikegami (1995, 179), who argues that “...Tokugawa society can be regarded as a version of feudalism from almost every angle, but [...] it still differs from the ideal types generated by the European medieval experience – particularly in its political structure.” We return to these structural differences below. 2 Another stream of literature focuses on the comparison between autocrats and rulers in representative institutions. The latter experience a more efficient allocation of resources and economic growth (Lake, 1992), mobilize popular support for war better (Reiter and Stam, 2002), convert mass mobilization for war into progressive taxation (Scheve and Stasavage, 2010), and deliver more successful public policies (Bueno de Mesquita et al., 1999). They also have easier access to credit and are able to finance prolonged wars, as they are more likely to be credible in repaying debt (Schultz and Weingast, 2003). Here, we restrict the discussion to variation among autocracies. Slater (2010) ties rebellion and urban unrest to taxation (and regime type) within autocracies, depending on the extent to which the disorder incentivizes elites to tax themselves. As we explain below, the forms of rebellion we study were not aimed at overthrow of the regime, so did not trigger new progressive tax schemes among elites. We focus on the narrower question of the effect peasant resistance had on taxation of the peasants. 1 ruler avoids over-extraction because the autocrat has sufficient information to calculate the point at which the tax rate becomes harmful to the economy (McGuire and Olson, 1996).3 Citizens generally have little influence over a ruler’s decisions on revenue and spending, partly because the autocrat has overwhelming coercive power relative to citizens, and partly because citizens face a collective action problem (e.g., Olson, 1971). However, if citizens can overcome the collective action problem and rebel or threaten to rebel against high tax rates, then they could also influence the tax rate (See also Acemoglu and Robinson, 2006; Besley and Persson, 2009). As Levi (1988, 19) put it, the ruler’s imperative is to maximize revenue extraction while avoiding “fight or flight” of the taxed, or he will lose resources and possibly the ability to govern. Indeed, we have many examples of tax rebellions over time and across countries (see, for example, Kiser and Linton (2002) on France, Bush (1991) on Tudor England, and Rapoport (2004) on 14th-century Egypt under Mamluk rule). Te Brake (1998, 8) notes that in early modern Europe, resistance to tax increases was “widespread and predictable,” and could “bend and shape public policy in significant ways [...].” However, evidence of whether or not resistance systematically altered rulers’ extraction has been elusive. In this article, we aim to fill this empirical gap by exploiting sub-national variation in peasant rebellions and migrations and systematic data from Japan to analyze if and how peasants were able to constrain their powerful samurai rulers’ taxation. In the case of Tokugawa Japan, historical records indicate that peasants did band together and rebel in some form 1,787 times across the domains between 1603 and 1868, when the regime collapsed. According to Aoki (1971), 497 of those instances specifically involved resistance to taxation. Peasants collectively fled 35 times to avoid complying with a tax, of 161 total collective desertions.4 The structure of village life and collective taxation fostered collective action. But was it effective? Were rebellious and mobile villagers able to win tax concessions from their rulers? Or did rulers crackdown in these domains? In other words, do we observe lower tax rates where peasants were able to engage in fight or flight? Furthermore, were larger scale rebellions more successful, or was frequency of resistance more effective? This paper sets out to answer these questions. The period of the Tokugawa shogunate, also known as the Edo bakufu, is an 3 See also Besley and Persson (2008) for their evolutionary political economic model of taxation. 4 As we explain below, collective desertions were different from typical migration. In Tokugawa Japan, it was a sanctioned form of resistance that involved entire villages abandoning land to avoid working it temporarily, thereby denying tax payments to the ruler. For shorthand, we refer to this as “flight.” 2 especially interesting setting in which to explore the relative impact of internal pressure on taxation, because external war was not a critical factor during the period. While each daimyo was technically under the rule of the shogunate, he was free to set his own tax rate for his subjects, experienced no threat from powers outside of Japan, and was unable to engage in warfare with other daimyo within Japan.5 This is a key difference between early modern Japan and Europe.