^* I would like to extend special thanks to William Kirby and R. Bin Wong for their helpful comments on earlier versions of this article.

¹ K. S. Lieuw, Struggle for Democracy: Sung Chiao–jen and the 1911 Revolution (Berkeley: University of California Press, 1971), esp. pp. 127–201, and Ernest Young, The Presidency of Yuan Shih–kai (Ann Arbor: University of Michigan Press, 1977) are the best examples.

² For a relatively sympathetic view of the KMTs state–building efforts (as well as the role of Western financial and technical knowledge in aiding those efforts), see Arthur Young, Chinas Nation Building Effort: The Financial and Economic Record, 1927–37 (Stanford: Hoover Institution Press, 1971), Arthur Young, China and the Helping Hand, 1937–45, (Cambridge, MA: Harvard University Press, 1963). Another somewhat later sympathetic view is expressed in Maria Hsia Chang, Facism and Modern China, The China Quarterly, No. 79 (September 1979).

³ For critical views, see Chien Tuan–sheng, The Government and Politics of China (Cambridge, MA: Harvard University Press, 1950), Lloyd Eastman, The Abortive Revolution: China Under Nationalist Rule, 1927–37 (Cambridge, MA: Harvard University Press, 1974), and Tien Hung–mao, Government and Politics in Kuomintang China, 1927–37 (Stanford: Stanford University Press, 1972).

⁴ An analysis of the widespread phenomenon of penetration of the state by society in developing countries can be found in Joel Migdal, (Princeton: Princeton University Press, 1988). An impressive example of a state with many parallels to Republican China is Atatiirk s Turkey in the 1920s and 1930s: like Chiang Kai–shek, Atatiirk proclaimed himself the head of an anti–imperialist movement of national liberation, and like the KMT in the 1930s, the Republican Peoples Party attempted to build a strong central state from the top down in urban areas, with little capacity and still less interest in radically restructuring the life of the Anatolian peasantry.

⁵ This shift in emphasis began in the early to mid–1980s, with the publication of William Kirbys Germany and Republican China (Stanford: Stanford University Press, 1984). Preliminary research on my forthcoming book, Strong Institutions in Weak Polities: Personnel Policies and State Building in China, 1927–40 (Oxford: Oxford University Press, 1998) was conducted in 1987–89. At present there is a significant cadre of graduate students from programmes in history and political science at Harvard, Berkeley, and Washington University, St. Louis who are in some stage of writing or research on Ph.D. dissertations that address such different aspects of Republican government as constitution–drafting, the KMTs efforts to combat low–level corruption, and water conservancy. Even within the Peoples Republic, scholars are also beginning to reconsider the Republican period in a more sympathetic light: Nanjing University Press began publication of Minguo yanjiu (Studies on Republican China) with an editorial board of members from China and abroad, in summer 1995.

⁶ The xinzheng era, important as it is for the subsequent evolution of the 20th–century Chinese state, remains crucially understudied. Some of the more important works on this period include: Mary C. Wright, China in Revolution: The First Phase, 1900–11 (New Haven: Yale University Press, 1968); Ernest Young, The Presidency of Yuan Shih–kai (Ann Arbor: University of Michigan Press, 1977), chs. 1–2; Esther Morrisson, The modernization of the Confucian bureaucracy, unpublished dissertation, Radcliffe College, 1959; and Paul Hickey, Fee taking, salary reform and the structure of state power in late Qing China, 1909–11, Modern China, Vol. 17, No. 3, pp. 389–417

⁷ There was much about the xinzheng reform package that was either ambivalent or unclear, and this lack of clarity particularly stood out in the delicate issue of relative degrees of centralization and decentralization, an issue that continued to reverberate through the remainder of the 20th century. The late Qing xinzheng reforms allowed for the creation of provincial assemblies and even drew up a schedule for the gradual implementation of constitutional monarchy. Yuan Shikai, while hostile to any body that he could not directly control, toward the end of his life seemed quite desperate to make links with any source of popular support and launched a doomed monarchical revival, and the Chiang Kai–shek regime, while no less suspicious, allowed the existence of a Legislative Yuan and National Assembly, and paid at least lip service to the idea of village serf–government. On the xinzheng constitutional programme, see Morrisson, The modernization of the Confucian bureaucracy. On Yuan Shikai, see Ernest Young, Yuan Shikai as a conservative modernizer, in Charlotte Furth, The Limits of Change: Essays on Conservatism in Modern China (Cambridge, MA: Harvard University Press, 1976), pp. 171–190.

⁸ Young, The Presidency of Yuan Shih–kai, pp. 148–155. On warlord actions in the capital, see Chien Tuan–sheng, The Government and Politics of China, pp. 65–76, and on the structural systems behind warlord behaviour, see Chi Hsi–sheng, Warlord Politics in China, 1916–28 (Stanford: Stanford University Press, 1976), pp. 185–195.

⁹ Young, The Presidency of Yuan Shih–kai, pp. 164–167. On the creation of the Sino–Foreign Salt Inspectorate and the revenues it remitted to the Beiyang governments, see S. A. M. Adshead, The Modernization of the Chinese Salt Administration, 1900–1920 (Cambridge, MA: Harvard University Press, 1970), pp. 61–117. Prasenjit Duara in Culture, Power and the State: Rural North China, 1900–42 (Stanford: Stanford University Press, 1988), pp. 59–85, also suggests that the very late imperial and Republican proliferation of local tax bureaus and tankuan led to significant increases in the rate of real taxation, virtually none of which was passed up to the central government.

¹⁰ The relative weights of the perceived importance of good institutions and good men in late imperial Chinese statecraft varied substantially, and is a vast topic in its own right. For views in English, see Benjamin Schwartz, The World of Thought in Ancient China (Cambridge, MA: Harvard University Press, 1985), pp. 102–105 for a discussion of the importance of good men. For the importance of good institutions, see Madeleine Zelins excellent study of the rationalizing fiscal reforms of the Yongzheng Emperor, The Magistrates Tael: Rationalizing Fiscal Reform in 18th Century China (Berkeley: University of California Press, 1984).

¹¹ Wolfgang Franke, The Reform and Abolition of the Traditional Chinese Examination System (Cambridge, MA: Harvard Monographs, 1960) and Benjamin Elman, Delegitimation and decanonization: the trap of civil service examination reform, 1860–1910, paper prepared for the First International Conference of Ching Intellectual History, National Chung–shan University, Kaohsiung, Taiwan, 19–21 November 1993.

¹² For the idea of an institutional breakthrough, I slightly adapt Kenneth Jowitts concept of a revolutionary breakthrough for the early developmental stage of Leninist systems. In contrast to revolutionary breakthroughs, which are highly specific to Leninist regimes attempting to consolidate and carry out a heroic programme of modernization through collectivization and transformation of the basic units of society, institutional breakthroughs are more generic, and can be directed towards any number of different ends (including socially quite conservative ones). However, both revolutionary and institutional breakthroughs are sought after by executive elites wherever institutions are weak and fragmented. See Kenneth Jowitt, Revolutionary Breakthroughs and National Development: The Case of Romania (Berkeley: University of California Press, 1971), pp. 94–95.

¹³ This list of attributes is drawn from Max Weber, Economy and Society (Berkeley: University of California Press, 1978), pp. 223–26.

¹⁴ This discussion draws on my article, Symbol and reflection of the reconstituting state: the Examination Yuan in the 1930s, Modern China, Vol. 20, No. 2 (April 1994). Many of the original details can be found in Qian Shipu (ed.), Beiyang zhengfu shiqide zhengzhi zhidu (The Political System of the Beiyang Government Period) (Beijing: Zhonghua shuju, 198

¹⁵ This segment is based on details inch. 5 of Strauss, Strong Institutions in WeakPolities, with the original documentation from the No. 2 Historical Archives, files 1027/176 (2), 1027/179, 1027/181, and 1027/188 (1).

¹⁶ The fact that the Beiyang government in Beijing was the officially recognized central government of China meant that control of it provided automatic access to a range of financial resources not available elsewhere (primarily customs and salt surpluses as well as the possibility of foreign loans). This ironically increased the attractiveness of the city and rump central government administration as a target for warlord depradations during the 1920s.

¹⁷ The analysis in this and the following section of the Salt Inspectorate presents an overview of ch. 3, Effective institution building: the case of the Sino–Foreign Salt Inspectorate, in Strauss, Strong Institutions in Weak Polities. See also S. A. M Adshead, The Modernization of the Salt Administration, 1900–20 (Cambridge, MA: Harvard University Press, 1970), especially ch. 4.

¹⁸ These figures are taken from Kwei Chungshu (ed.), The Chinese Year Book, 1935–36 (Shanghai: Commercial Press, 1936), p. 1298. The currency in which these figures were calculated was the standard silver dollar.

¹⁹ Yanwu renshi guize (Regulations on Salt Affairs Personnel), Ministry of Finance, n.d. (c. 1948). Interviews with: Zhong Liangzhe, Taipei, Taiwan, 20 January 1989; Lin Jiyong, Tainan, Taiwan, 24 January 1989; Chen Guisheng, Taipei, Taiwan, 16 January 1989.

²⁰ Interviews, Chen Guisheng, Taipei, Taiwan, 16 January 1989; and Zhou Weiliang, Tainan, Taiwan, 24 January 1989.

²¹ In fact, when the National Government was established in 1927–28, political leaders j varied quite dramatically between wishing to abolish the Inspectorate as a tool of imperialism jjj insulting to Chinas sovereignty and desperately needing access to the funds that the | Inspectorate could reliably supply.

²² These figures are taken from the table Distribution of revenues in Young, Chinas % Nation Building Effon, p. 73. In 1929, the Inspectorate only contributed 9% of the central governments revenues, but by 1930 the figure was up to 25.2%.

²³ Song Ziwen (T. V. Soong), 1931 annual financial report, reproduced in China Year Book, 1933 (Shanghai: Kelley and Walsh, 1933), p. 338

²⁴ Liu Foding, interview, Tianjin, 24 June 1988.

²⁵ Young, Chinas Nation Building Effort, Distribution of revenues from 1929–37, p. 73.

²⁶ 26. For the National Resource Council, see William C. Kirby Continuity and change in Modern China: economic planning on the mainland and on Taiwan, 1943–58, Australian Journal of Chinese Affairs, No. 58 (July 1990), pp. 121–141. On water conservancy, David Pietz has an excellent dissertation in progress; on the Ministry of Personnel, particularly its attempt to objectify and depersonalize the criteria for advancement and promotion in the civil bureaucracy, see Strauss, Symbol and reflection of the reconstituting state, and on the activities of technocratically oriented administrative reformers, see Julia Strauss, The cult of administrative efficiency: myth and statecraft in the Republic of China, unpublished paper presented at the Association for Asian Studies Annual Meeting, 25–28 March 1993, Los Angeles.

²⁷ For the original formulation of this fact/value dichotomy, see Herbert Simon, Administrative Behavior, 2nd ed. (New York: Macmillan, 1961), pp. 4–8.

²⁸ Aaron Wildavsky, Speaking Truth to Power: The Art and Craft of Policy Analysis (Boston: Little, Brown, 1979).

²⁹ Sun Yat–sen, Three Principles of the People and Fundamentals of National Reconstruction (Taipei: China Cultural Service, 1953), Chinese and English text.

³⁰ I take the idea of a combat/heroic Party from Ken Jowitt, who uses the term to describe Leninist parties of all types. See Kenneth Jowitt, The Leninist phenomenon, reproduced in New World Disorder: The Leninist Extinction (Berkeley: University of California Press, 1992). There were of course significant differences in the methods of the KMT and the Chinese Communist Party: the former shied away from violent social revolution and class struggle while the latter embraced it. But both were definitively stamped by Leninist principles of discipline and mobilizing the heroic commitments of those from below, as well as a faith in the party to effect an institutional breakthrough by transforming beliefs.

³¹ C. Martin Wilbur and Julie Lien–ying How, Missionaries of Revolution: Soviet Advisors and Nationalist China (Cambridge, MA: Harvard University Press, 1989), Document 29 The training of the National Revolutionary Army for war, Document 37 Political work in the National Revolutionary Army, Document 39, Program of political lessons for enlisted personnel and Document 40 How to carry on the political education of officers are full of examples.

³² See Chiang Kai–shek, Gemingjun teshu de jingshen he zhanshu (The National Revolutionary Armys special esprit and military techniques), speech given at Whampoa on 9 November 1924, and Junshi jiaoyu zhi yaozhi yu junji de genyuan (The basic principles of military education and the origin of military discipline), speech given at Whampoa on 14 April 1925, in Zhang Qijun (ed.), Xian zongtong jianggong chuanji, Vol. 1 (Taipei: Wenhua daxue and Zhonghua xueshuyuan, 1984), pp. 475–481 and 493–96.

³³ The term in Chinese is Zhongyang zhixing weiyuanhui, which can also be translated as Politburo or, more loosely, Central Executive Committee.

³⁴ The nature of the one–party government, and the dual roles that much of the senior leaders played have led to an understandable tendency in much of the English–language literature to conflate the Party with the National Government as the KMT Government or KMT Party–state. The two were closely related, particularly at the very top, but were analytically and organizationally separate, particularly for middle and lower levels of the government bureaucracy.

³⁵ This section is drawn from my article, Wartime stress and Guomindang response: xunlian as a means to statebuilding, delivered at the Annual Meeting of the Association for Asian Studies, 11–14 April 1996, Honolulu, Hawaii.

³⁶ Liu Xianyun, who was then a relatively low level ganshi, was on site for the 1937 xunlian session, and attested that while the majority of the individuals undergoing xunlian were from the KMT Party and military, there was some number of high level non–party civilians involved with either education or cultural activities who were in attendance. Interview, Taipei, 17 May 1995. See also Wang Dongyuan, in speech given on 9 December 1939 in which he reviewed the previous occasions of xunlian, reproduced in Ganbu xunlian wenti (Questions on the Training of Cadres) (Chongqing: Zhongyang xunlianban di wuqi jiangyan lu, 1939). See esp. pp. 8–13

³⁷ Jiaotong bu nianjian (Ministry of Communications Yearbook) (Nanjing: Jiaotong bu, 1935), pp. 106–113 and Caizheng bu nianjian (Nanjing: Caizheng bu, 1935), pp. 67–69 and 72–73.

³⁸ As one suggestive example, the Ministry of Personnel, which before 1937 was an exceptionally small organization with under 100 staff, quintupled in size. Documents published by the Examination Yuan in the early 1950s state explicitly that the Ministry of Personnel grew five–fold over the course of the Sino–Japanese War. See See Niu Yongzhen (ed.), Xingxianhou kaoshiyuan chengli sanzhounianjiantao huiyi tekan {Special Publication: The Examination Yuans Self–Criticism Meeting on the Third–Year Anniversary of the Implementation of the Constitution), published by the Kaoshiyuan Secretariat, 1951, p. 11. It is a reasonable hypothesis that other organizations may have grown even more quickly, as it became very common to make daili (acting) appointments that escaped registry and scrutiny with the Ministry of Personnel.

³⁹ Lloyd Eastman, Nationalist China during the Sino–Japanese War, 1937^15, in Lloyd E. Eastman etal.. The Nationalist Era in China, 1927–49 (Cambridge: Cambridge University Press, 1991, 1993), pp. 155–56, 165–67. For the range of monopoly control the wartime government attempted to institute on products as diverse as matches, salt, sugar and tobacco see Kangzhan shiqi zhuanmai shiliao {Historical Materials on Monopolies During the Sino–Japanese War) (Taipei: Guoshiguan, 1992).

⁴⁰ Wang Dongyuan, Zenyang shixing xunshi? (How Should Training Instruction be Implemented!), pamphlet (Chongqing: Zhongyang xunliantuan, 1940), p. 8.

⁴¹ For the mechanics of xunlian, see ibid. For critical views of those who underwent the training session, see Liu Yaozhang, Guomindang zhongxuntuan de lailong qumai (The origins and development of the KMTs Central Training Corps), Wenshi ziliao xuanji (Selected Materials on Civil History) No. 74; Cai Duan (ed.), Zhongguo renmin zhengzhi xieshang huiyi quanguo weiyuanhui wenshi ziliao yanjiu weiyuanhui (Beijing: Wenshi ziliao chubanshe, 1981) pp. 1–21; and Zhao Puju, Zhongyang xunliantuan genggai (A synopsis of the Central Training Corps), pp. 23–41, in same.

⁴² Liu Yaozhang, ibid., p. 2.

⁴³ Tongyi gedi xunlian jiguan banfa quanguo ge xunlian liguan xunlian gangling (An Outline for the Unification of Each Regions Organization and Methods for Training) (Chongqing: Guomindang zhongyang zhixing weiyuanhui xunlian weiyuanhui, January 1940), pp. 1–3.

⁴⁴ For a partial listing of the names of the non Central Training Corps xunlian organizations that sprang up during the first years of the Sino–Japanese War, see Wang Dongyuan, Ganbu xunlian wenti, pp. 4–5.

⁴⁵ Although this probably varied from case to case, it appears that at least some individual organizations subtly fought to maintain their technocratic yewu orientations to the greatest degree possible given the extraordinary politicization and militarization of the war years. For example, the Examination Yuan, while in form following the schedule of activities laid out by the Central Training Institute model, in practice modified the contents of many of the lectures and the orientation of the small group discussions to focus on the practical problems of personnel administration under wartime conditions. This information comes from a long document entitled Disi renshi xingzheng renyuan xunlian ban de baogao (The Fourth Personnel Administration Personnel Training Report), which was written in February 1942. This report, which runs to nearly 100 pages, reviews the previous xunlian sessions held by the Examination Yuan, and can be found in Guoshiguan microfilm reel No. 20000000A 0313/8050.01–01, pp. 1155 and passim. The degree to which other organizations did the same is an empirical question that awaits further research. However, whatever the quiet resistance, the trend to conflate military cum political xunlian with functionally oriented technocratic xunlian during the Sino–Japanese War is clear.

⁴⁶ See Eugene Levich, The Kwangsi Way in Kuomintang China, 1931–39 (Armonk: M.E. Sharpe, 1993), and John Fitzgerald, Awakening China: Politics, Culture and Class in the Nationalist Revolution (Stanford: Stanford University Press, 1997).

⁴⁷ Christian Heoriot, Shanghai, 1927–37: Municipal Power, Locality, and Modernization (Berkeley: University of California Press, 1993), pp. 232–38.

⁴⁸ For the original, and still extremely important Western analysis of the Red and Expert dichotomy in the Peoples Republic, see Franz Schumann, Ideology and Organization in Communist China, 2nd ed. (Berkeley: University of California Press, 1966, 1968), pp. 162–172 and 75–76.