China’s 1994 Tax-Sharing Reform and the Rise of Local Government Financing Vehicles

On January 1, 1994, China implemented the Tax-Sharing System reform (Fenshuizhi, 分税制), a sweeping fiscal overhaul engineered by Vice Premier Zhu Rongji to rescue the central government in Beijing from a decade-long erosion of tax capacity. Together with the 1994 Budget Law, the reform succeeded in raising the central government’s share of national tax collections from 22 percent in 1993 to nearly 56 percent in 1994 while legally prohibiting provincial and municipal governments from running budget deficits or issuing municipal bonds. Yet because subnational authorities remained responsible for more than 70 percent of China’s total public expenditure, the 1994 settlement created a deep vertical fiscal imbalance. Forced to fund rapid urbanization and infrastructure without direct bond market access, Chinese cities turned first to monopolized land-lease sales (“land finance”) and, after Beijing’s RMB 4 trillion stimulus in November 2008, to thousands of off-balance-sheet corporate conduits known as Local Government Financing Vehicles (LGFVs, or chengtou)—accumulating an estimated RMB 50 to 60 trillion or more in hidden subnational debt. ^[1]

The Collapse of the Fiscal Contracting System and the “Two Ratios” Crisis#

To understand why Beijing recentralized fiscal power so abruptly in 1994, one must look at the unintended fiscal consequences of China’s post-1978 market transition. During the 1980s, China dismantled the Soviet-style unified revenue-and-expenditure system (tongshou tongzhi) and replaced it with a decentralized Fiscal Contracting System (caizheng baogan), colloquially known as “eating in separate kitchens” (fenzao chifan). Under negotiated multi-year contracts, provincial governments collected nearly all taxes within their jurisdictions and remitted only a fixed lump sum or negotiated percentage to Beijing, retaining the surplus locally. ^[2]

While fiscal contracting gave local officials strong incentives to promote township and village enterprises and regional growth, it starved the central treasury. Provincial authorities routinely sheltered booming local revenue streams in extrabudgetary accounts and granted generous tax exemptions to local state-owned enterprises to keep funds out of the shared revenue pool. Economist Hu Angang and researcher Wang Shaoguang famously warned China’s leadership of the collapsing “Two Ratios”: total government budgetary revenue as a share of gross domestic product plunged from roughly 28–31 percent in 1979 to just 10.8 percent in 1993, while the central government’s share of total fiscal revenue fell from 40.5 percent in 1984 to a historic low of 22.0 percent in 1993. By 1993, Beijing was struggling to fund national defense, interprovincial infrastructure, and macroeconomic stabilization without resorting to inflationary credit from the People’s Bank of China. ^[3]

Zhu Rongji’s 1994 Tax-Sharing Reform (Fenshuizhi) and the Budget Law#

Appointed Vice Premier in charge of economic and financial affairs and concurrent Governor of the People’s Bank of China in July 1993, Zhu Rongji spent the autumn of 1993 flying from province to province—most famously engaging in tough negotiations with coastal powerhouses such as Guangdong and Shanghai—to forge consensus on the 1994 Tax-Sharing System reform. Taking effect on January 1, 1994, the reform fundamentally restructured China’s fiscal architecture along three pillars: ^[4]

• Clear Assignment of Central, Local, and Shared Taxes: Taxes were divided into central taxes (such as customs duties, consumption taxes, and income taxes of central state-owned enterprises), local taxes (such as business taxes on services, urban land-use taxes, and local enterprise and personal income taxes), and shared taxes.
• The 75–25 Value-Added Tax (VAT) Split: A modernized, broad-based Value-Added Tax (zengzhishui) levied at 17 percent became the single largest pillar of the tax system, with 75 percent of all VAT receipts assigned directly to the central government and only 25 percent retained by local governments (followed in 2002 by a separate reform that also shifted 50 percent, and later 60 percent, of corporate and personal income taxes to Beijing).
• Dual Tax Administrations: To prevent local officials from diverting central revenues, Beijing established an independent State Administration of Taxation (SAT) network in every province and county to collect central and shared taxes directly, completely separate from local tax bureaus.

Crucially, the fiscal overhaul was accompanied by the Budget Law of the People’s Republic of China, adopted by the National People’s Congress in March 1994 (effective January 1, 1995). Article 28 of the 1994 Budget Law mandated that local budgets at all levels be strictly balanced according to the principle of “keeping expenditures within the limits of revenues” and explicitly prohibited local governments from issuing local government bonds or borrowing directly from banks, except as specifically authorized by the State Council. ^[5]

Vertical Fiscal Imbalance and the Turn to “Land Finance” (1994–2008)#

The 1994 reform achieved its primary objective almost overnight: in 1994 alone, the central government’s share of national budgetary revenue leaped from 22.0 percent to 55.7 percent, and over the next two decades total government fiscal revenue recovered from under 11 percent of GDP to more than 20 percent of GDP. Yet while the 1994 reform recentralized tax revenues upward to Beijing, it left expenditure responsibilities decentralized—and even pushed new spending mandates downward as state-owned enterprise restructuring shifted pension, healthcare, unemployment, education, and municipal infrastructure burdens onto city and county budgets. ^[6]

By the early 2000s, subnational governments collected less than half of national budgetary revenues (roughly 45 to 48 percent) yet accounted for more than 70 percent of total public expenditure—and over 80 percent by the 2010s. Although central fiscal transfers and tax rebates partially bridged the gap, transfers were often earmarked or unevenly distributed, while local cadres’ career promotions under the Communist Party cadre evaluation system depended heavily on delivering high local GDP growth and visible urban infrastructure. ^[7]

Barred by the 1994 Budget Law from issuing municipal bonds, local governments discovered a lucrative extrabudgetary revenue stream following the 1998 housing privatization: “land finance” (tudi caizheng). Because urban land was state-owned while rural land was collectively owned, municipal governments held a statutory monopoly on converting rural farmland into urban commercial and residential land. By expropriating rural land at low agricultural compensation rates and auctioning long-term land-use rights (typically 40 to 70 years) to real estate developers, cities generated massive land-transfer fees that soon rivaled their formal tax budgets. Furthermore, as early as the 1990s in Shanghai and Chongqing, pioneering municipalities established state-owned urban construction investment corporations (chengtou gongsi)—the first Local Government Financing Vehicles—injecting municipal land parcels as equity capital so these corporate entities could legally borrow from the China Development Bank and commercial banks without violating the letter of the 1994 Budget Law. ^[8]

The November 2008 RMB 4 Trillion Stimulus and the LGFV Explosion#

What had been a manageable workaround transformed into a systemic debt machine in late 2008. When the Global Financial Crisis caused Chinese export orders to plummet in the autumn of 2008, the State Council announced a massive RMB 4 trillion (roughly $586 billion) two-year infrastructure and economic stimulus package on November 9, 2008. However, the central government in Beijing committed to funding only RMB 1.18 trillion (less than 30 percent) of the stimulus out of the central budget, instructing provincial, municipal, and county governments to finance the remaining RMB 2.82 trillion.

Because the 1994 Budget Law still prohibited local governments from borrowing directly on their balance sheets, the People’s Bank of China and the China Banking Regulatory Commission (CBRC) issued a joint policy directive in March 2009 actively encouraging local governments at all levels to establish qualified Local Government Financing Vehicles (LGFVs) to raise bank loans and corporate bonds for stimulus projects. Commercial banks flooded the system with credit: as documented in National Bureau of Economic Research (NBER) studies by Chong-En Bai, Chang-Tai Hsieh, Zheng Michael Song, and Zhiguo He, the number of LGFVs nationwide exploded from roughly 3,000 before 2008 to more than 8,000 by the end of 2009 and over 10,000 in subsequent years.

The financial mechanics of an LGFV relied on a circular symbiosis between municipal land values and bank credit. A city government transferred land-use rights, toll roads, or utility shares into its LGFV; the LGFV pledged that land as collateral to obtain commercial bank loans and issued “chengtou bonds” (chengtouzhai) in China’s interbank and exchange bond markets; and the city supported the vehicle’s debt service through future land-transfer revenues and budget subsidies. Because banks and bond investors assumed that municipal governments would never allow a state-owned LGFV to default—an “implicit government guarantee” (yinxing danbao)—credit flowed freely into low-return highways, industrial parks, subways, and new urban districts even when the projects themselves generated almost no operating cash flow.

Shadow Banking, the 2015 Budget Law Revision, and the RMB 60+ Trillion Debt Overhang#

When Beijing sought to rein in bank lending to LGFVs after 2010 to curb inflation and credit risks, local financing vehicles did not stop borrowing; instead, as NBER researchers Zhuo Chen, Zhiguo He, and Chun Liu demonstrated, maturing stimulus bank loans were rolled over through China’s rapidly expanding shadow banking system—including trust loans, wealth management products (WMPs), entrusted loans, and urban construction bonds.

Recognizing that prohibiting formal municipal bonds had merely driven borrowing underground, the National People’s Congress amended the Budget Law in August 2014 (effective January 1, 2015) to adopt an “open the front door, block the back door” strategy. For the first time since 1994, provincial governments were granted legal authority to issue official local government bonds within quotas approved by the State Council and the NPC, and the Ministry of Finance launched a massive RMB 12+ trillion bond-swap program between 2015 and 2018 to refinance high-cost LGFV liabilities into lower-interest official municipal bonds.

Yet because the fundamental vertical fiscal mismatch established in 1994 remained intact—with local governments still carrying over 80 percent of public spending responsibilities—off-balance-sheet LGFV debt continued to climb. According to International Monetary Fund Article IV assessments and academic estimates, total LGFV interest-bearing debt rose from roughly RMB 5 trillion in 2008 to an estimated RMB 50 to 60+ trillion (approximately $7 to $9 trillion, or nearly 50 percent of China’s GDP) by 2022–2023, on top of roughly RMB 35 to 40 trillion in explicit official local government bonds. When China’s property market downturn after 2021 caused municipal land-transfer revenues to plunge by more than 30 to 40 percent, the thirty-year fiscal model born from the 1994 Tax-Sharing Reform and the 2008 stimulus reached a historic reckoning.

Theoretical & Balance-Sheet Transmission Mechanism#

In the Kindleberger–Minsky taxonomy of financial instability (Kindleberger, 1978; Minsky, 1986) and Irving Fisher’s (1933) debt-deflation framework, speculative asset booms become systemic crises when credit expansion shifts balance sheets from hedge finance (cash flows covering principal and interest) to speculative and Ponzi finance (relying on rising asset prices and short-term margin roll-overs).

Once monetary tightening or an exogenous shock halts price appreciation, forced margin liquidations trigger Fisherian debt-deflation: fire-sale asset discounts erode bank collateral values, contracting credit supply across the real economy.

Sources and Further Reading

1. Zhuo Chen, Zhiguo He, and Chun Liu, “ The Financing of Local Government in China: Stimulus Loan Wanes and Shadow Banking Waxes,” National Bureau of Economic Research (Working Paper No. 23598), 2017. Peer-Reviewed / Academic Research
2. Chong-En Bai, Chang-Tai Hsieh, and Zheng Michael Song, “ The Long Shadow of a Fiscal Expansion,” National Bureau of Economic Research (Working Paper No. 22801 / Brookings Papers on Economic Activity), 2016. Peer-Reviewed / Academic Research
3. Ehtisham Ahmad, Li Keping, Thomas Richardson, and Raju Singh, “ Recentralization in China?,” International Monetary Fund, 2002. International Financial Institution
4. International Monetary Fund, “ People’s Republic of China: Selected Issues—Local Government Financing Vehicles,” International Monetary Fund (IMF Country Report No. 22/22), 2022. International Financial Institution
5. David Dollar and Bert Hofman, “ Intergovernmental Fiscal Reforms, Expenditure Assignment, and Governance in China,” World Bank, 2008. International Financial Institution
6. Australian Centre on China in the World, “ China’s Local Government Debt: Fallout from a Perfect Storm,” The China Story (Australian National University), 2023. Historical & Financial Reference
7. Irving Fisher — "The Debt-Deflation Theory of Great Depressions," Econometrica 1, no. 4 (FRASER Federal Reserve Archive, 1933): 337–357 Central Bank / Monetary Authority
8. Charles P. Kindleberger & Robert Z. Aliber — Manias, Panics, and Crashes: A History of Financial Crises (Palgrave Macmillan) Peer-Reviewed / Academic Research

Have an idea for another financial history documentary or article? Request a story →