Warlords at Work: Four Crucial Realms and Four Dynamics of State Building in Republican China, 1916–1937

During the years 1916–1937, a time of critical political disunity for China, statebuilding efforts by central authorities were not the only ones that were significant and worthy of attention. The contributions of local power-holders (warlords) were also crucial. Contrary to the usual depiction of warlords as concerned only with their personal wealth and power, I argue that many had a strong concern for state building. Warlords' achievements were particularly significant in four realms: transportation infrastructure, education, economic planning, and statistics. Besides reappraising the warlords' contributions in these four realms, it is of equal importance to explain the remarkable convergence of state-building politics at the national scale. The four different realms under consideration are interesting not only for their intrinsic importance in state building but also because they are representative of four more general "dynamics" that explain such a convergence.

One of the reasons that warlords' contributions to state building during the Republic has not been fully valued has to do with the fact that almost all related studies deal with one particular warlord. Many warlords were fascinating figures, and this has inclined scholars to opt for a biographical approach. Few attempts have been made to piece together the actions of the respective warlords and get a general picture of their real impact on state-building efforts in Republican China.¹ One obvious reason for that lacuna is that to examine the achievements of multiple warlords is an enormous task that requires tapping into a vast quantity of primary sources and secondary literature.

For that very reason, I decided to address the issue of warlords' contributions to modernization and state building during the 1916–1937 period at a China-wide scale but, at the same time, to restrict my focus to four different areas in which their accomplishments in terms of state building strike me as particularly outstanding. By drawing mostly on secondary literature, I focus on the four realms of (1) transportation infrastructure, (2) education, (3) economic planning, and (4) statistics. Although it was a key to warlord activities and ate up major proportions of their budgets, I omit the area of military modernization, which has already been the focus of many studies.²

The 1926–1928 Northern Expedition was a watershed in the period under our consideration, as it opened the way to the nominal reunification of China under the aegis of the Guomindang (GMD) and ushered in a period of stronger central government (the Nanjing Decade, 1928–1937). But it also dramatically altered the composition of the set of warlords opposing the central government. The Northern Expedition, in fact, created opportunities for former high-ranking GMD military officials—such as Chen Jitang (陳濟棠), Han Fuju (韩复榘), and He Jian (何健)—to rise to the position of local power-holders. These new warlords joined others already in place before 1928, like Feng Yuxiang (馮玉祥), Yan Xishan (閻錫山), and the New Guangxi clique, who had all been allies of the GMD but not part of its army during the Northern expedition.³ Though in some instances they severed their relations with the GMD, all these men nonetheless remained ideologically aligned with it.⁴ After 1928, warlords not connected to or very loosely associated with the GMD and its ideology were in short supply. They could be found only in the western parts of China: the several named Ma (馬) in the northwestern provinces and Long Yun (龍雲) in Yunnan.

As a rule, warlords' state-building achievements appeared to be less impressive (but by no means nonexistent) during the era of the Beiyang governments (1916–1928) than during the Nanjing Decade: the new commitment of warlords to state building during the Nanjing Decade derived from a reformist objective they shared, in its broad outlines, with the GMD.

While it is important to explore local power-holders' achievements in the four selected realms of state building and to (re)evaluate them against those of the central government, it is of equal importance to characterize the factors that, despite significant differences in the warlords' overall styles and objectives, explain the remarkable convergence of their state-building policies with that of Nanjing.

The four different realms under focus are interesting not only for their intrinsic importance in state building but also because they are representative of four dynamics that explain such a convergence, namely, (1) the "cunning of reason" (List der Vernunft), (2) a common set of values inherited from the past about what constitutes "right government," (3) zeitgeist, and (4) emulation.

Transportation Infrastructure and Cunning of Reason

Studies examining state building often tend to oppose, on the one hand, state-building efforts by the central government, which are deemed positive because they pursue the strengthening of China as a whole, and, on the other hand, state-building efforts by local power-holders, which are viewed as negative efforts to preserve their independence. By resorting to the dynamic of the cunning of reason,⁵ it is possible to reconsider this misguided opposition.

As was accurately pointed out by Frank Dikötter, China as a whole made spectacular progress during the Republic in improving transportation infrastructure.⁶ The development of railways was considerable during the Republic, increasing from 9,600 km of track in 1912 to 25,000 km in 1945.⁷ Shanxi warlord Yan Xishan was famous for his commitment to railway development, and mining and agriculture benefited greatly from the new railways. Zhang Zuolin built the Shenyang–Hailong (瀋陽–海龍) line between 1925 and 1928 and was also committed to road building.⁸ Similar efforts in road building were witnessed in Guangxi under the New Guangxi clique, as well as in the province of Sichuan.⁹ During the Nanjing Decade, the total length of the highway network expanded from 1,000 km to 109,000 km to form a reasonably comprehensive network.¹⁰ Waterways were no exception to the general trend. As Anne Reinhardt has demonstrated, in the early 1930s the relatively minor Sichuan warlord Liu Xiang (劉湘) was strongly committed to the creation of a purely Chinese navigation company (Minsheng 民生) and determined to oust foreign-owned navigation companies from the Yangzi River basin.¹¹

For Yan Xishan, Zhang Zuolin, and other warlords, the development of transportation infrastructure was closely connected to strategic interests: an effective transportation network is a precious asset, as it enables the rapid and efficient movement of troops and military gear. In their efforts to improve transportation in their regions, warlords no doubt aimed to reinforce their own power and independence. These considerations notwithstanding, their achievements connected to produce a nationwide effort, with spectacular results.

In connecting transportation infrastructure to a dynamic of the cunning of reason, I point out that in that realm, although warlords who sought to extend transportation systems were motivated mainly by concerns about their own military strength, their actions converged to create a significant expansion of China's network of transportation and communications.

Education and Right Government

In recent decades, historians of education during the Late Qing and Republican period were mostly concerned with the issue of Western and Japanese influences on pedagogy and overall organization of the Chinese system of education, with a strong emphasis on university and higher education.¹² This historiography makes clear that an impressive growth of secondary schools and higher-level education took place during the Republican era.¹³

Despite a general scarcity of studies on primary education, many sources hint at its rapid development during the Republican period. In Guangdong, for example, the number of primary schools totaled 18,969 in 1931 (for 1,098,965 students) and 22,754 in 1934 (for 1,469,406 students).¹⁴ According to national statistics released in 1931, from 1912 to 1930, the number of modern-type primary schools in China as a whole rose from 86,318 to 250,840. The persistence of traditional private schools (sishu 私塾) during the whole period has often been seen as a manifestation of failed modernization in primary education. As pointed out by Thomas Curran, however, sishu-type and modern-type primary school complemented one another, in particular because sishu were "simply better suited to rural conditions and customs than modern schools were. For those not destined for higher education, they offered an elementary education that was functional and in tune with rural habits."¹⁵

Therefore, it seems safe to state that the educational sector at all levels, from primary education to universities, experienced remarkable growth during the period under our consideration.

The important role warlords played in this important development has been generally overlooked. Zhang Zuolin (張作霖), for example, is better known for his threats and attacks on Peking University, made in 1926,¹⁶ than he is for his founding in 1923 of Dongbei daxue (東北大學), Northeast China's first university.¹⁷

According to various Japanese sources that can hardly be suspected of exaggeration, the efforts of the warlord Chen Jitang in Guangdong in the early 1930s yielded impressive results, as the figures for this province's primary schools presented above confirm.¹⁸ Results were even more spectacular for secondary education, with the number of secondary schools jumping from 178 to 318 over a four-year period (1929–1932).¹⁹ Yan Xishan was very much committed to developing primary education in his province; so were the three powerful Muslim warlords of Northwest China, Ma Fuxiang (馬福祥), Ma Qi (馬麒) and Ma Bufang (馬步芳),²⁰ as well as Zhang Xueliang (張學良) during the brief period of his leadership in the three northeastern provinces (1927–1931).²¹ Warlords of lower caliber were not to be outdone. Chen Quzhen (陳渠珍), a local strongman in western Hunan, made similar efforts during the 15 years preceding the Second Sino-Japanese War.²² On the whole, it seems that warlords' efforts often focused on primary education, which might be why their contribution has been downplayed.

So universal a commitment to developing education among local power-holders across China is all the more remarkable given that educational circles were decidedly unsympathetic to warlords. Much of the criticism of warlord rule came from within educational circles. This was true of national luminaries such as Cai Yuanpei (蔡元培) or Lu Xun (魯迅) and also of common students and schoolteachers. The students of Beijing University made warlords a favorite target of their demonstrations in the 1920s.

Why, then, were many warlords such committed sponsors of education? I would suggest that their commitment in such matters stemmed from a traditional concern among the local Confucian elite for development of education as a characteristic of good government. It would be a great mistake to believe that illiterate warlords like the notorious bogeyman Zhang Zongchang (張宗昌) were the norm. While it is true that few warlords were educated abroad or had even received a "progressive" (i.e., Western-style) education, there was no shortage of warlords who had been highly educated in the traditional manner. Wu Peifu (吳佩孚) was only a xiucai (秀才), but he liked to pose as a literatus and was said to write poems on the battlefield, in close emulation of Zhuge Liang (諸葛亮). Ma Fuxiang was a juren (舉人) and Tan Yankai (譚延闓) had received the jinshi (進士), the highest degree in the imperial examinations.²³ Their exposure to traditional education inclined these men to be proactive in the realm of education.

Thus, unlike transportation, regarding which their actions were guided by self-interest, the educational sector exemplifies what can be considered warlords' unselfish (and not, in the short term, rewarding) commitment to state-building efforts via a dynamic of right government.

Economic Planning and Zeitgeist

The apparent success of the Soviet Union's First Five-Year Plan (1928–1932) inspired a worldwide interest in economic planning. China was no exception. The idea that the state should be a driving force in the industrial development of China was treasured by Sun Yat-sen. The prominent leaders of the GMD who competed to succeed him after his death in 1925—Hu Hanmin (胡漢民), Wang Jingwei, and Chiang Kai-shek—shared the same faith. Moreover, all three were convinced that planning was an indispensable tool. The Nanjing regime had three agencies in charge of the implementation of economic planning.²⁴

Among local power-holders, Yan Xishan launched a provincial 10-year plan of economic reconstruction out of admiration for the success of the Soviet First Five-Year Plan.²⁵ Another adept of government-led planning was Chen Jitang. He launched a threeyear plan (1932–1935) that achieved remarkable results.²⁶ Even if they did not resort to economic planning, other warlords were no less committed to the idea that the state should be a leading force in the realm of economic development. Miao Jiaming (繆嘉銘), who was educated in the United States, tirelessly worked for the industrialization of Yunnan, one of the most backward provinces of China, with the full support of the local warlord, Long Yun. In 1934, Miao created an ad hoc administrative agency to monitor economic development: the Yunnan Economic Commission (雲南全省經濟委員會 Yunnan quansheng jingji weiyuanhui).²⁷ The most spectacular achievements of planning in Yunnan were the expansion of the textile industry and the expansion of tin mining. Numerous other examples (like Zhang Zuolin's activities in the textile industry) can be pointed to in order to challenge the mistaken idea that warlords' investments were only intended to develop military-related industrial capacity to meet their need for armaments.²⁸

Interestingly, during the Nanjing Decade, the scope of state intervention was not necessarily more considerable in the regions under the direct administration of Nanjing than in the regions beyond its reach. Emily Hill has stated that in the mid-1930s Guangdong's government invested more substantially in industrialization than the central government did.²⁹

What is remarkable about economic planning, therefore, is not the absence of a nationwide plan. True, despite the claims to draw plans at a nationwide level made by the Nanjing regime's three agencies in charge of the implementation of economic planning, in the context of a lack of political unity, such plans were implemented only in the areas Nanjing actually controlled. But the case of economic planning shows that, drawing their inspiration from the same zeitgeist, the different power-holders shared the goal of industrial development and to a large extent followed similar methods in implementing their state-building policies. A precious insight by Elizabeth Remick is that, in contrast to the more familiar story that European states arose because of "war and preparation for war" (Charles Tilly), "the impetus behind local Chinese state building was essentially political and ideological."³⁰ Remick thereby reminds us that the local actors' subscription to modernizing ideologies was a major driving force for state building in Republican China.

Most warlords of the period after 1928 (when the best results in terms of state building took place), after all, belonged to the same generation. In many instances, they had graduated from the same academies.³¹ All deeply resented China's "weakness," and they shared the same commitment to strengthening China through a modernizing political agenda.

Statistics and Emulation

Connected to zeitgeist is emulation. By "emulation," however, I do not mean the act of emulating foreign or modern models; rather, I mean a dynamic of competition among China's various power-holders to showcase their achievements and enhance their legitimacy in the forum of national public opinion.

Emulation played a crucial role in explaining a craze for statistics that encompassed the political disunity of China. The Nanjing Decade, in particular, was characterized by the publication of an enormous volume of statistics.³² Among these publications, those of the Bureau of Statistics of the central government are well known, such as the valuable Summary of Statistics of the Republic of China (中華民國統計提要 Zhonghua minguo tongji tiyao) published in 1936.³³ As a rule, all central government administrations published accurate and detailed statistics, like the Ministry of the Interior Yearbook (內政年鑑 Neizheng nianjian) and the Yearbook of Chinese Economy (中國經濟年鑑 Zhongguo jingji nianjian). Provinces and municipalities under the central government also issued statistical publications. The GMD did not lag behind, publishing a yearbook in 1930, entitled Guomindang Yearbook for the Year 1929 (民國十八年中國國民黨年鑑 Minguo shibanian Zhongguo guomindang nianjian).

As Tong Lam has demonstrated,³⁴ the passion for scientifically grounded facts and modern statistics was general among the political and intellectual elites during the late Qing and Republican periods. It therefore comes as no surprise that in regard to statistics the warlord regimes did not fall short of the GMD authorities either in quality or quantity.

In that regard, Guangzhou (Canton) is a case in point. This city had been the stronghold of the GMD during the 1920s. Yet the city's best yearbook was not published during the time when the city was under GMD administration but in 1935, four years after the city passed into the control of a local warlord, Chen Jitang.³⁵ A quick glance at the previous yearbooks published under the GMD administration, such as Statistical Yearbook of the Guangzhou Municipal Government (廣州市市政府統計年鑒 Guangzhoushi shizhengfu tongji nianjian), which appeared in 1929, suffices to show that they were no match for the 1935 yearbook in either quality or comprehensiveness.

A yearbook of similar quality, more than 1,000 pages in length, was published during the same period in Hunan, a province under the domination of another warlord, He Jian.³⁶

Relatively remote and backward provinces like Shanxi followed suit and could claim similar records in terms of statistical publications. The province of Guangxi published two yearbooks of a high quality in close succession (in 1933 and 1935) while under the control of the New Guangxi clique.³⁷

The case of the New Guangxi clique, which was extremely vocal in opposition to the central government, points to the fact that the motivation for the publication of accurate and up-to-date statistical publications (a costly endeavor) was that it allowed warlords to showcase their most spectacular achievements in the area of state building and garner positive public opinion.

Conclusion

During a time of division in China, the contribution of warlords appears to have been particularly outstanding in the four realms under consideration in this article. To build a more comprehensive picture of state building would of course require investigation of other crucial issues like taxation, city governance, or welfare. Indeed, such topics are the focus of other articles in this issue.³⁸

Figure 1.

Connections among the four realms (left) and four dynamics (right) of state building in Republican China. Diagram by Xavier Paulès.

It is also necessary to underline that each of the four crucial realms under our scrutiny was connected not just to one dynamic but to several, as shown in Figure 1. For example, warlords' commitment to developing transportation infrastructure stemmed not only from the dynamic of cunning of reason but also from zeitgeist and emulation.

As to the relationship between local power-holders and the central government, scholars have so far focused their attention on demonstrating how local power-holders successfully carved out realms of independence. Their unanimous conclusion has been that warlords were "semi-independent," "relatively independent," "almost independent," or in French "semi-indépendance" and in Chinese banduli (半獨立), and so forth. Against such a backdrop, works analyzing the ways in which power-holders were semi-dependent are sorely needed.³⁹

State building offers an excellent example of that semi-dependence. An overlooked dimension of the process of state building during the 1916–1937 period is that, notwithstanding the four dynamics under our focus, the remarkable convergence of local power-holders and central government policies was also the outcome of active coordination and cooperation among them. For example, important health improvement programs were implemented by the GMD in Northwest China in cooperation with local power-holders.⁴⁰ The completion of the Hankou–Canton section of railroad in 1936, a decisive step in the making of a comprehensive national railway network, is another case in point. The section crossed the territories of two warlords: He Jian and Chen Jitang. Its construction therefore could not have taken place without a great deal of cooperation among these men and the central government. In its first pages, the official publication designed to celebrate the event displays calligraphy contributed by key central government figures like Lin Sen (林森) and Chiang Kai-shek but also by He Jian and by Hu Hanmin, who at that time was a staunch opponent of the central government and had sought refuge in Chen Jitang's domain. A few pages later, the publication states explicitly that "central and local government and society have all united their forces in full sincerity" (中央與地方，政府與社會，莫不同心協力).⁴¹

Future research should therefore envision state building under the Republic as a process involving also a fair deal of cooperation between competing power-holders.