¹ See, for example, John Gerring, Strom C. Thacker and Carola Moreno, ‘Centripetal Democratic Governance: A Theory and Global Inquiry’, American Political Science Review, 99 (2005), 567–81; Susanne Lohmann, ‘Federalism and Central Bank Independence: The Politics of German Monetary Policy, 1957–92’, World Politics, 50 (1998), 401–46; Gabriella Montinola, Yingyi Qian and Barry R. Weingast, ‘Federalism, Chinese Style: The Political Basis for Economic Success in China’, World Politics, 48 (1995), 50–81; Yingyi Qian and Gérard Roland, ‘Federalism and the Soft Budget Constraint’, American Economic Review, 88 (1998), 1143–62; Jonathan Rodden and Susan Rose-Ackerman, ‘Does Federalism Preserve Markets?’ Virginia Law Review, 83 (1997), 1521–72.

² Francis G. Castles, ‘Decentralization and the Post-War Political Economy’, European Journal of Political Research, 36 (1999), 27–53; Daniel Treisman, ‘Decentralization and Inflation: Commitment, Collective Action, or Continuity?’ American Political Science Review, 94 (2000), 837–57; Erik Wibbels, ‘Federalism and the Politics of Macroeconomic Policy and Performance’, American Journal of Political Science, 44 (2000), 687–702.

³ Jonathan Rodden, ‘Comparative Federalism and Decentralization: On Meaning and Measurement’, Comparative Politics, 36 (2004), 481–500.

⁴ Countries that are as different as Pakistan and Switzerland, for instance, are often grouped in the same category under the conventional legalistic classification scheme. Rodden, ‘Comparative Federalism and Decentralization’, p. 491.

⁵ William H. Riker, Federalism: Origin, Operation, Significance (Boston and Toronto: Little, Brown and Company, 1964), p. 131.

⁶ Jonathan Rodden and Erik Wibbels, ‘Beyond the Fiction of Federalism: Macroeconomic Management in Multitiered Systems’, World Politics, 54 (2002), 494–531; Erik Wibbels, ‘Federal Politics and Market Reform in the Developing World’, Studies in Comparative International Development, 36 (2001), 27–53.

⁷ Yasheng Huang, Inflation and Investment Controls in China: The Political Economy of Central–Local Relations during the Reform Era (New York: Cambridge University Press, 1996).

⁸ Geoffrey Brennan and James M. Buchanan, The Power to Tax: Analytical Foundations of a Fiscal Constitution (Cambridge: Cambridge University Press, 1980); Wallace E. Oates, Fiscal Federalism (New York: Harcourt Brace Jovanovich, 1972); Charles M. Tiebout, ‘A Pure Theory of Local Expenditures’, Journal of Political Economy, 64 (1956), 416–24.

⁹ Francesca Fornasari, Steven B. Webb and Heng-fu Zou, ‘The Macroeconomic Impact of Decentralized Spending and Deficits: International Evidence’, Annals of Economics and Finance, 1 (2000), 403–33; Jorge Martinez-Vazquez and Robert M. McNab, ‘Fiscal Decentralization and Economic Growth’, World Development, 31 (2003), 1597–616; Ernesto Stein, ‘Fiscal Decentralization and Government Size in Latin America’, Journal of Applied Economics, 2 (1999), 357–91; Jessica Seddon Wallack and T. N. Srinivasan, eds, Federalism and Economic Reform: International Perspectives (Cambridge: Cambridge University Press, 2006).

¹⁰ Barry R. Weingast, ‘The Economic Role of Political Institutions: Market-Preserving Federalism and Economic Development’, Journal of Law, Economics, and Organization, 11 (1995), 1–31.

¹¹ Douglass Cecil North and Barry R. Weingast, ‘Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth-Century England’, Journal of Economic History, 49 (1989), 803–32.

¹² Robert J. Barro and David B. Gordon, ‘Rules, Discretion and Reputation in a Model of Monetary Policy’, Journal of Monetary Economics, 12 (1983), 101–21.

¹³ This seems to be working in Germany where local representatives outnumber federal appointees in the Bundesbank Council making German monetary policy. See Lohmann, ‘Federalism and Central Bank Independence’.

¹⁴ Castles, ‘Decentralization and the Post-War Political Economy’.

¹⁵ Treisman, ‘Decentralization and Inflation’.

¹⁶ Mancur Olson, The Logic of Collective Action: Public Goods and the Theory of Groups (Cambridge, Mass.: Harvard University Press, 1971).

¹⁷ Wibbels, ‘Federalism and the Politics of Macroeconomic Policy and Performance’.

¹⁸ Castles, ‘Decentralization and the Post-War Political Economy’; Treisman, ‘Decentralization and Inflation’.

¹⁹ Treisman, ‘Decentralization and Inflation’.

²⁰ George Tsebelis, Veto Players: How Political Institutions Work (Princeton, N.J.: Princeton University Press, 2002).

²¹ Both Castles and Treisman adopt a binary scheme – countries with multiple levels of government sharing constitutionally-delineated power are federal, in Castles, ‘Decentralization and the Post-War Political Economy’, pp. 33–4, and Treisman, ‘Decentralization and Inflation’, p. 841. Wibbels uses a trichotomous definition of federalism with two additional categories: (1) whether sub-national units elect their own regional legislatures, and (2) whether sub-national units are represented in the national legislature (Wibbels, ‘Federalism and the Politics of Macroeconomic Policy and Performance’, p. 691).

²² Rodden, ‘Comparative Federalism and Decentralization’, p. 491.

²³ Riker, Federalism, pp. 129–31.

²⁴ ‘Governing political party’ is operationalized as the party of the national chief executive heading the executive branch of the central government (see Rodden and Wibbels, ‘Beyond the Fiction of Federalism’, p. 509; and Wibbels, ‘Federal Politics and Market Reform in the Developing World’, p. 37).

²⁵ Eliza Willis, Christopher da C. B. Garman and Stephan Haggard, ‘The Politics of Decentralization in Latin America’, Latin American Research Review, 34 (1999), 7–56.

²⁶ Rodden and Wibbels, ‘Beyond the Fiction of Federalism’; Wibbels, ‘Federal Politics and Market Reform in the Developing World’.

²⁷ Riker himself did not explicitly spell out the specific channels whereby the central government could exert such political control. While there are other avenues for exercising control, such as the authority relating to party finances, means of communications etc., in this study we focus on the aspect of personnel authority.

²⁸ Willis, Garman and Haggard, ‘The Politics of Decentralization in Latin America’.

²⁹ This observation is based on the assumption that the national chief executive (with the central government she leads) and the national leadership of her political party share similar policy preferences and can often act together. This should be especially likely in countries under single-party rule, the focus of this article, where there often exists a substantial personnel overlap between the two.

³⁰ See Barbara Geddes, ‘Authoritarian Breakdown: Empirical Test of a Game Theoretic Argument’ (paper presented at the annual meeting of the American Political Science Association, Atlanta, Ga., 1999), for an exhaustive list of countries ruled by such parties since the end of the Second World War.

³¹ While it is not clear whether the political party of its national chief executive enjoyed much sub-national personnel power, all the sub-national governments of Austria, a democracy, were ruled by one political party between 1987 and 1993, according to Rodden and Wibbels, ‘Beyond the Fiction of Federalism’, p. 510.

³² The Partido Acción Nacional (PAN) first won a governorship in Baja California in 1989. The opposition parties had since gradually increased their share of the Mexican governorships; by August 2000, one month after losing the presidential election for the first time, PRI only held the governorship in 19 of Mexico’s 32 provincial units (Rogelio Hernández-Rodríguez, ‘The Renovation of Old Institutions: State Governors and the Political Transition in Mexico’, Latin American Politics and Society, 45 (2003), 97–128, pp. 107–8).

³³ It has been argued that PRI’s nomination powers regarding the governors had been less unconstrained even during the heyday of its ‘hegemony’ than conventionally believed. For instance, between 1936 and 1994 about 21 per cent of governors could be considered local bosses (cacique governors) with a substantial local power base, according to Alberto Diaz-Cayeros, Federalism, Fiscal Authority, and Centralization in Latin America (New York: Cambridge University Press, 2006), chap. 3. It suggests that the ascendancy to governorship for these cacique governors might also reflect regional preferences as well as the nomination powers of PRI’s national office. Still, this does not deny the fact that PRI still held formal nomination powers over all the Mexican state governors, the overwhelming majority of which (nearly 80 per cent ) were more tightly controlled by the PRI national office during that period.

³⁴ One example of the exponential information required for coding political party decentralization bearing on candidate selection is to be found in Kenneth Janda, Political Parties: A Cross-National Survey (New York/ London: Free Press/Collier Macmillan, 1980). Janda collected data on a sample of political parties in 53 countries. Despite this monumental effort, his book with over 1,000 pages only has information for 1950–62 and many of the important countries – such as China, Mexico or Japan – are missing from his dataset. A second example is Richard S. Katz and Peter Mair, Party Organizations: A Data Handbook on Party Organizations in Western Democracies, 1960–90 (London: Sage, 1992), a data compendium that is 900 pages in length but covers only twelve Western democracies for 1960–90.

³⁵ Wibbels, ‘Federal Politics and Market Reform in the Developing World’.

³⁶ Rodden and Wibbels, ‘Beyond the Fiction of Federalism’.

³⁷ For an excellent discussion of the sub-national research strategy in comparative politics, see Richard Snyder, ‘Scaling Down: The Subnational Comparative Method’, Studies in Comparative International Development, 36 (2001), 93–110.

³⁸ Huang, Inflation and Investment Controls in China.

³⁹ For the case of China, see also Yumin Sheng, ‘Global Market Integration and Central Political Control: Foreign Trade and Intergovernmental Relations in China’, Comparative Political Studies, 40 (2007), 405–34. Such variation in political control by the central government over regional governments led by different types of regional officials in China is also present in other centralized systems such as PRI-controlled Mexico where some state governors were local bosses less sympathetic to central government policy preferences. See Diaz-Cayeros, Federalism, Fiscal Authority, and Centralization in Latin America, chap. 3. Presumably, even if it wields the requisite personnel authority, the central government in such systems does not have the luxury of being able to employ officials uniformly predisposed to high policy compliance in all regions due to such constraints as lack of enough qualified personnel (Politburo members or officials with the requisite central experiences in the case of China, as we will see below). The need to assign different types of officials to lead different regional governments inevitably results in varying effective political control by the central government.

⁴⁰ Barro and Gordon, ‘Rules, Discretion and Reputation in a Model of Monetary Policy’; North and Weingast, ‘Constitutions and Commitment’.

⁴¹ Olson, The Logic of Collective Action; Wibbels, ‘Federalism and the Politics of Macroeconomic Policy and Performance’.

⁴² Huang, Inflation and Investment Controls in China; Sheng, ‘Global Market Integration and Central Political Control’. More details will be provided in the next section.

⁴³ Huang, Inflation and Investment Controls in China.

⁴⁴ The link between the two is developed in János Kornai, Economics of Shortage (Amsterdam and New York: North-Holland, 1980).

⁴⁵ China’s national-level inflation data come from Guojia Tongji Ju, Zhongguo Tongji Nianjian 1999 (Beijing: Zhongguo tongji chubanshe, 1999); Guojia Tongji Ju, Zhongguo Tongji Nianjian 2000 (Beijing: Zhongguo tongji chubanshe, 2000); Guojia Tongji Ju, Zhongguo Tongji Nianjian 2004 (Beijing: Zhongguo tongji chubanshe, 2004).

⁴⁶ Loren Brandt and Xiaodong Zhu, ‘Redistribution in a Decentralized Economy: Growth and Inflation in China under Reform’, Journal of Political Economy, 108 (2000), 422–39.

⁴⁷ Lowell Dittmer and Yu-shan Wu, ‘The Modernization of Factionalism in Chinese Politics’, World Politics, 47 (1995), 467–94.

⁴⁸ Unless otherwise noted, all provincial-level inflation data in this study (and other economic data used for later regression analysis) for 1978–98 are from Guojia Tongji Ju, Xin Zhongguo Wushinian Tongji Ziliao Huibian (Beijing: Zhongguo tongji chubanshe, 1999). They are cross-checked and updated to 2003 with individual annual issues of the provincial statistical yearbooks (Guojia Tongji Ju, Provincial Statistical Yearbooks 2000–2004 (Individual Issues for Each Province/Year for 1999–2003 Data) (Beijing: Zhongguo tongji chubanshe, 2000–2004)), whose data are always used as a rule in cases of inconsistencies between the historical series in the single volume and the individual/annual volumes. It is important to note that averages of annual provincial-level data are broadly consistent with the annual national-level data in Table 1. In fact, the former are not always lower than the latter. Together with the substantial cross-provincial variation in annual inflation rates, this suggests little evidence that inflation data from the provincial sources have been systematically and artificially deflated for political purposes.

⁴⁹ According to the consumer price indices, China’s national inflation was 24.1 per cent in 1994. Tibet had the highest inflation at 28.3 per cent, while Yunnan, with the lowest inflation, registered only 19.2 per cent. We mainly rely on the data on retail price indices although it might be preferable to use the consumer price indices. Unfortunately, data on the latter are sparse. Still, available data on the two price series for the period under study are highly correlated (at about 0.97, p = 0.000, Pearson 2-tail test).

⁵⁰ The region-wide deflationary pressures since the late 1990s (after the Asian financial crisis) may have dented the salience of such preference divergence for some time. Nevertheless, central–provincial differences in policy preference regarding inflation remain valid even today.

⁵¹ Yunchen Huang, Jian Lu and Yu Fan, ‘1991–1992 Nian Hongguan Jingji Taishide Panduan Yu Zhengce Jianyi [Judgment on the 1991–1992 Macroeconomic Situation and Policy Suggestions]’, Tizhi Gaige Neibu Cankao (1992), 3–8.

⁵² Benedict Stavis, ‘The Political Economy of Inflation in China’, Studies in Comparative Communism, 22 (1989), 235–50; Andrew G. Walder, ‘The Political-Sociology of the Beijing Upheaval of 1989’, Problems of Communism, 38 (1989), 30–40.

⁵³ Olson, The Logic of Collective Action.

⁵⁴ For more details, see Huang, Inflation and Investment Controls in China, chaps 2–3.

⁵⁵ Nicholas R. Lardy, China’s Unfinished Economic Revolution (Washington, D.C.: The Brookings Institution, 1998).

⁵⁶ China has five levels of government – the centre, province, prefectures/municipalities, counties and townships. In all Communist countries, there is a parallel structure consisting of party committees and government units at all levels of the political system.

⁵⁷ Susan L. Shirk, The Political Logic of Economic Reform in China (Los Angeles and Berkeley: University of California Press, 1993), chap. 3.

⁵⁸ The following two paragraphs draw on Huang, Inflation and Investment Controls in China, pp. 210–11.

⁵⁹ Gailong Liao, Zhufeng Luo and Yuan Fan, Zhongguo Renmin Dacidian: Xianren Dangzhengjun Lingdao Renwujuan (Beijing: Waiwen chubanshe, 1994); Xueming Shen and Jianying Zheng, eds, Zhonggong Diyijie Zhi Shiwujie Zhongyang Weiyuan (Beijing: Zhongyang wenxian chubanshe, 2001); Zutang Wang, Zhongguo Renming Dacidian: Dangdai Renwujuan (Shanghai: Shanghai cishu chubanshe, 1992); and http://www.peopledaily.com.cn.

⁶⁰ Nathaniel Beck, ‘Time-Series-Cross-Section Data: What Have We Learned in the Past Few Years?’ Annual Reviews of Political Science, 4 (2001), 271–93.

⁶¹ Donald P. Green, Soo Yeon Kim and David H. Yoon, ‘Dirty Pool’, International Organization, 55 (2001), 441–68. Tests of possible redundancy for province/year fixed effects also suggest that a two-way fixed effects model is appropriate. Meanwhile, panel unit root tests indicate that provincial inflation data are largely stationary. Specifically, the Levin–Lin–Chu test, assuming a common unit root process, and the Im–Pesaran–Shin test, assuming an individual unit root process, both with automatic lag selection, either including only individual intercept or both individual trend and intercept, are conducted (G. S. Maddala and In-Moo Kim, Unit Roots, Cointegration, and Structural Change (New York: Cambridge University Press, 1998).

⁶² Provincial retail price indices are used as the GDP deflator to calculate the real value of the economic variables. RMB stands for renminbi, the Chinese currency.

⁶³ Wibbels, ‘Federalism and the Politics of Macroeconomic Policy and Performance’, p. 693.

⁶⁴ Rodden and Wibbels, ‘Beyond the Fiction of Federalism: Macroeconomic Management in Multitiered Systems’, p. 516.

⁶⁵ David Romer, ‘Openness and Inflation: Theory and Evidence’, Quarterly Journal of Economics, 108 (1993), 869–903.

⁶⁶ Sheng, ‘Global Market Integration and Central Political Control’. The rationale for including such variables is from Gary King, ‘ “Truth” Is Stranger than Prediction, More Questionable than Causal Inference’, American Journal of Political Science, 35 (1991), 1047–53, p. 1050.

⁶⁷ Provincial export data for 1978–85 are from Guojia Tongji Ju, Xin Zhongguo Wushinian Tongji Ziliao Huibian, data for 1986–2002 are from China Ministry of Foreign Trade and Economic Cooperation, Zhongguo Duiwai Jingji Maoyi Nianjian (Xianggang: Huarun maoyi zixun youxian gongsi, 1987–2003), and data for 2003 are from China Ministry of Commerce, Zhongguo Shangwu Nianjian (Beijing: Zhongguo shangwu chubanshe, 2004). Using the total provincial foreign trade will generate similar results, but the export data will allow us to have slightly more data observations. Because export data are in US dollars and provincial GDP data are in RMB, we use the average annual exchange rate (the middle rate) to calculate provincial export/GDP share. Exchange rate data for 1977–80 are from Nicholas R. Lardy, Foreign Trade and Economic Reform in China, 1978–1990 (New York: Cambridge University Press, 1992), pp. 148–9. Data for 1981–84 are from Guojia Tongji Ju, Zhongguo Tongji Nianjian 1999, those for 1985–2001 are from Guojia Tongji Ju, Zhongguo Tongji Nianjian 2002 (Beijing: Zhongguo tongji chubanshe, 2002), and those for 2002–3 are from Guojia Tongji Ju, Zhongguo Tongji Nianjian 2004.

⁶⁸ Rodden and Wibbels, ‘Beyond the Fiction of Federalism’; Treisman, ‘Decentralization and Inflation’; Wibbels, ‘Federalism and the Politics of Macroeconomic Policy and Performance’.

⁶⁹ Tao Zhang and Heng-fu Zou, ‘Fiscal Decentralization, Public Spending, and Economic Growth in China’, Journal of Public Economics, 67 (1998), 221–40. Central government budgetary spending, national population data and national retail price indices used as the GDP deflator for constructing data on real per capita budgetary expenditures by the central government for 1978–98 are from Guojia Tongji Ju, Zhongguo Tongji Nianjian 1999. Data for 1999–2003 are from Guojia Tongji Ju, Zhongguo Tongji Nianjian 2004. Central budgetary expenditures include interest payment for data since 2000, but not before. We have subtracted the interest payment for 2000–03.

⁷⁰ It is this rationale that motivated the World Bank to argue for fiscal recentralization. See World Bank, China: Macroeconomic Stability in a Decentralized Economy (Washington, D.C.: World Bank, 1995).

⁷¹ Hehui Jin, Yingyi Qian and Barry Weingast, ‘Regional Decentralization and Fiscal Incentives: Federalism, Chinese Style’, Journal of Public Economics, 89 (2005), 1719–42.

⁷² Brandt and Zhu, ‘Redistribution in a Decentralized Economy’.

⁷³ Robert H. Bates, Markets and States in Tropical Africa: The Political Basis of Agricultural Policies (Los Angeles and Berkeley: University of California Press, 1981).

⁷⁴ Lagging all economic control variables yields similar results for Political Decentralization.

⁷⁵ Beck, ‘Time-Series-Cross-Section Data’; Nathaniel Beck and Jonathan N. Katz, ‘What to Do (and Not to Do) with Time-Series Cross-Section Data’, American Political Science Review, 89 (1995), 634–47.

⁷⁶ As in time-series analysis, including the lagged dependent variables could make the estimation inconsistent in theory, but this seems ‘seldom a problem in practice’ (Beck, ‘Time-Series-Cross-Section Data’, p. 279).

⁷⁷ James Harding, ‘Shanghai Gang looking to call the shots’, Financial Times, 22 February 1997, p. 3; Cheng Li, ‘The “Shanghai Gang”: Force for Stability or Cause for Conflict?’, China Leadership Monitor, 2 (2002), available at http://www.hoover.org/publications/clm/issues/2906851.html, accessed on 26 August 2007. This emphasis on elite factionalism enjoys a long pedigree in the study of Chinese politics. See Dittmer and Wu, ‘The Modernization of Factionalism in Chinese Politics’. We thank one reviewer for raising the possible concern with Shanghai.

⁷⁸ Shanghai was led by a centralist in 1990, a localist for 1991–92 and a concurrent centralist for 1993–97. As seen from Table 2, Shanghai was one of the least decentralized provinces throughout this era.

⁷⁹ Dropping all observations from Shanghai yields almost identical results to those reported in Model 7.

⁸⁰ Here we follow the approach described in Robert J. Barro, ‘Democracy and Growth’, Journal of Economic Growth, 1 (1996), 1–27, p. 9. We thank Fengshi Wu for urging us to tackle this issue.

⁸¹ While the coefficient for the third dummy variable fails to be statistically significant at the conventional levels, the p value for the joint significance test of the three dummy variables is about 0.011.

⁸² Romer, ‘Openness and Inflation’.

⁸³ Sheng, ‘Global Market Integration and Central Political Control’.

⁸⁴ Rodden and Wibbels, ‘Beyond the Fiction of Federalism’; Wibbels, ‘Federalism and the Politics of Macroeconomic Policy and Performance’.

⁸⁵ Bates, Markets and States in Tropical Africa.

⁸⁶ Brandt and Zhu, ‘Redistribution in a Decentralized Economy’.

⁸⁷ C. W. J. Granger, ‘Investigating Causal Relations by Econometric Models and Cross-Spectral Methods’, Econometrica, 37 (1969), 424–38.

⁸⁸ Treisman, ‘Decentralization and Inflation’, p. 837.

⁸⁹ Rodden, ‘Comparative Federalism and Decentralization’, p. 491.

⁹⁰ Riker, Federalism.

⁹¹ Snyder, ‘Scaling Down’.

⁹² Huang, Inflation and Investment Controls in China. In addition to more cross-national data collection, one possible avenue for future research, while beyond the scope of this article, would be to look systematically at how political decentralization affects unwarranted local state bank lending in reform-era China. Other future research efforts, when data on provincial government policy making in China become available, could be directed at tracing with a few detailed case studies exactly how different types of provincial governments made economic policies in these areas that could affect provincial inflation rates.

⁹³ Rodden and Wibbels, ‘Beyond the Fiction of Federalism’; Wibbels, ‘Federalism and the Politics of Macroeconomic Policy and Performance’.