China's Fiscal Regressivity: A Structural Analysis of Revenue and Expenditure

1. Introduction

Fiscal regressivity, in its technical sense, describes a fiscal system in which the net combined effect of taxation and public expenditure imposes a larger burden — or delivers a smaller benefit — on lower-income households relative to their incomes than it does on higher-income households. A tax is regressive if its effective rate falls as income rises; a spending program is regressive if its benefits accrue disproportionately to the non-poor. A fiscal system is regressive in aggregate when the progressive and redistributive elements are insufficient to offset the regressive ones, leaving post-fiscal income inequality higher than it would be under a neutral system.

Why this matters is not merely distributional in the academic sense. Fiscal systems that are regressive in aggregate undermine the social contract, reduce human capital formation among lower-income groups, generate political fragility, and suppress domestic consumption — a concern of particular salience in China, where rebalancing growth toward consumption has been an official policy priority for more than a decade. When the tax burden falls heaviest on those with the highest marginal propensity to consume, and when public services are rationed in ways that exclude the same groups, the fiscal system operates as a drag on the structural transformation it is ostensibly meant to support.

China is a worthwhile case for several reasons beyond its economic scale. It combines a nominally progressive statutory tax code with fiscal outcomes that are, in aggregate, less progressive than the code implies — a divergence that is instructive precisely because it is not produced by obvious policy failures but by structural features of the fiscal architecture. China has achieved extraordinary poverty reduction by absolute income measures while maintaining among the highest Gini coefficients of any major economy. Its fiscal system is not the only explanation for that divergence, but it is a central one. Understanding why the fiscal system redistributes as little as it does requires examining both the revenue and expenditure sides, and the structural forces that shape each.

The analysis proceeds in five steps. Section 2 examines the revenue side: the dominance of VAT, the limited coverage of personal income tax, the regressive structure of social insurance contributions, and the near-absence of wealth taxation. Section 3 examines land transfer revenues as a structurally regressive and fiscally distortionary funding mechanism. Section 4 turns to the expenditure side, focusing on the hukou system's role in rationing public services, the implicit subsidies embedded in state enterprise support, and the geographic skew of infrastructure investment. Section 5 identifies the structural drivers — the 1994 tax-sharing reform, local government fiscal pressure, and the hukou system — as a coherent institutional complex rather than a collection of separate policy failures. Section 6 examines distributional consequences through available data on the Gini coefficient and the hidden inequality in access to in-kind services. Section 7 concludes with a political economy assessment of reform paths and their constraints.

2. The Revenue Side: Regressivity by Design

VAT as the dominant indirect tax

Value-added tax is the single largest source of tax revenue in China, consistently accounting for roughly 30 to 35 percent of total tax receipts. It is levied at a standard rate of 13 percent on most goods and 9 percent on certain goods and services including food, with a reduced 6 percent rate for most services. The 2016 reform that replaced the business tax with VAT across all sectors expanded the base significantly, making VAT the fiscal backbone of the central government's revenue architecture.

VAT is regressive by construction. Because it is a tax on consumption rather than income, its effective incidence relative to income depends on the household savings rate: higher-income households save a larger fraction of their income and are thus exposed to VAT on a smaller share. Lower-income households in China, as elsewhere, consume almost all of their income — in rural areas and among urban migrant workers, saving rates are often low by necessity. The effective VAT burden as a share of income is therefore structurally higher for lower-income groups. Reduced rates on food and basic goods provide some mitigation, but the broad base and high headline rate mean that the overall incidence is nevertheless regressive. International comparisons reinforce the point: most OECD economies with VAT systems of comparable rates rely more heavily on income and wealth taxes to offset the regressive incidence; China's revenue mix does not.

Personal income tax: narrow coverage and structural limits

China's personal income tax is, in principle, progressive: the 2018 reform established a seven-bracket rate schedule with a top marginal rate of 45 percent on comprehensive income above approximately RMB 960,000 per year, alongside a raised basic exemption threshold of RMB 5,000 per month. In practice, the tax reaches a remarkably small fraction of the labor force. Official data and independent analyses suggest that only around 20 percent of the working population — concentrated in the urban formal sector — pays any personal income tax at all. This is not a function of deliberate exemption policy alone; it reflects the structure of the labor market.

A large share of the labor force consists of self-employed rural residents, agricultural workers, informal urban workers, and small business owners — categories whose income is difficult to assess, weakly reported, and administratively costly to tax. Capital income — interest, dividends, rental income, capital gains on equity — is subject to a flat 20 percent withholding rate, a substantial concession relative to earned income. Capital gains on owner-occupied residential property are largely exempt. The result is a tax that falls heavily on upper-middle and high earners in formal urban employment, leaves low-income workers untouched (appropriately), but also leaves capital income and wealth accumulation at the top substantially undertaxed relative to labor income at comparable levels.

The 2018 reform introduced additional deductions for housing loan interest, rent, children's education, continuing education, and care of elderly parents. The political framing was pro-household and pro-middle class. The distributional effect, however, was that these deductions disproportionately benefited middle and upper-middle income households with mortgages, children in urban schools, and elderly parents receiving formal care — a profile that does not describe the median Chinese household, particularly not migrant workers or rural residents.

Social insurance contributions: regressive at the base

China's social insurance system requires contributions to five programs: pension, medical insurance, unemployment insurance, occupational injury insurance, and maternity insurance — collectively known as the "five insurances." Contribution rates are set as a share of wages, split between employer and employee, and are nominally proportional. But several structural features render them regressive in practice for lower-income contributors.

Most schemes apply contribution ceilings — typically set at three times the local average wage — above which additional income is exempt from contributions. This means that high earners effectively face a declining marginal contribution rate above the ceiling, while those at or below average wages face the full statutory rate on all earnings. The incidence is thus regressive relative to income across the distribution of formal sector workers. More significantly, coverage is fractured between formal and informal employment. Formal sector workers in urban areas contribute to the more generous urban employee schemes; rural residents and informal workers are enrolled (if at all) in lower-contribution, lower-benefit resident or rural schemes. Since the higher-contribution urban employee schemes offer substantially better benefits, the system simultaneously taxes formal sector workers more heavily and delivers more value to them — a coupling that reinforces rather than offsets income stratification.

The near-absence of wealth taxation

China has no inheritance tax, no general net wealth tax, and no broad capital gains tax. Real estate — the dominant form of household wealth, particularly for the urban middle and upper classes — was subject to recurrent property taxation only in the limited pilot programs in Shanghai and Chongqing, which covered narrow categories of residential property and were widely understood to apply below the threshold of meaningful wealth in major tier-one cities. National property tax legislation has been in preparation for over a decade; progress has been slow, and the political dynamics are unfavorable.

The consequence is that wealth accumulation, particularly through real estate, proceeds with minimal fiscal offset. In a context where urban land values have compounded at rates far exceeding income growth over the past two decades, the absence of a recurrent property tax or capital gains tax represents a structural subsidy to asset holders — a group that skews heavily toward higher-income and higher-wealth households. Piketty, Yang, and Zucman's analysis of Chinese national accounts data finds that the top 10 percent income share in China rose from around 27 percent in 1978 to approximately 41 percent by the mid-2010s, while the bottom 50 percent share fell from 27 percent to 15 percent. The fiscal system's failure to tax wealth accumulation is a central part of this story.

3. Land Finance: A Regressive Revenue Pillar

The 1994 tax-sharing reform (fenshuizhi) restructured central-local fiscal relations in ways that remain constitutive of Chinese public finance thirty years later. The reform assigned the dominant tax revenues — VAT and corporate income tax — to a sharing arrangement heavily weighted toward the center, while local governments retained responsibility for the majority of public expenditure including education, healthcare, infrastructure, and social welfare. The structural gap between local revenue rights and local expenditure obligations created a persistent pressure on local governments to find alternative funding sources.

Land transfer revenues emerged as the primary solution. Under China's constitutional arrangement, all land is state-owned; rural collective land can be converted to urban construction land only through state appropriation, and urban land use rights can be sold by local governments for terms of typically 40 to 70 years. Local governments acquire rural land at administratively set compensation rates — keyed to the agricultural value of the land, not its development value — and sell the resulting urban land use rights at market prices to developers. The spread between acquisition cost and sale price constitutes government fund revenue (zhengfu xing jijin shouyi), which falls outside the formal tax system but has become the primary revenue source for many local governments.

At its peak, land transfer revenues accounted for upward of 40 percent of local government general and fund revenue combined in highly urbanized areas. Even at the national aggregate level, land revenues have consistently represented 30 to 40 percent of local government revenue in recent years. This is not a marginal revenue source; it is structurally load-bearing.

The distributional consequences are multiple. The forced conversion of rural collective land to urban use at below-market compensation prices represents an implicit tax on rural residents and collectives — a transfer from the less powerful to the state that is regressive by any reasonable standard. The proceeds fund urban public goods — infrastructure, schools, parks — that are not accessible to the rural populations whose land generated the revenue. In areas undergoing redevelopment, urban residents facing demolition of existing structures are compensated at administered rates that frequently fall below market values, with the gap captured by local governments and developers. Resistance is legally constrained and practically difficult.

Land finance also produces a structural incentive for local governments to keep urban land prices high. The revenue model depends on land values; anything that would suppress land prices — such as increasing land supply, allowing urban-rural land market integration, or expanding the leasehold term to reduce scarcity — threatens the revenue base. High land prices translate directly into high housing costs, which transfer purchasing power from urban renters and prospective homebuyers — disproportionately younger, lower-income, and migrant workers — to local government coffers. The fiscal architecture and the housing affordability crisis are two faces of the same structural arrangement.

4. The Expenditure Side: Rationed Access and Skewed Benefits

The hukou system and the geography of public services

The household registration system (hukou) was established in 1958 as a mechanism for managing population movement and controlling rural-urban migration during the planned economy period. It assigned every citizen a registration location — urban or rural, and within urban areas, tied to a specific city — and conditioned access to public services on registration status. Despite decades of reform, the hukou system continues to determine eligibility for the most consequential public services: compulsory education, healthcare reimbursement through local insurance schemes, public housing, social welfare, and pension access.

The distributional logic is straightforward and stark. Approximately 300 million workers have migrated from rural registration areas to urban employment — a labor force transfer of historic scale. These workers contribute to local economies through production, consumption, and social insurance contributions. But they do so as non-local residents, ineligible for the full suite of public services available to locally registered residents. In the largest cities — Shanghai, Beijing, Shenzhen, Guangzhou — obtaining local hukou requires satisfying conditions that systematically favor high-skilled, high-income migrants: educational credentials, stable formal employment, property ownership, and contribution histories. Low-wage factory workers, construction workers, and domestic service workers — the demographic backbone of China's urban economy — are largely excluded.

The expenditure consequence is that the fiscal resources generated partly by migrants' economic activity are deployed on services that migrants cannot fully access. Compulsory education funding flows to locally registered students; migrant children are legally entitled to attend local public schools in theory but face quota restrictions, documentation barriers, and quality differentials in practice. Healthcare reimbursement portability remains limited; social insurance contributions made in one city are often not portable to another, and withdrawal or transfer of accumulated pension entitlements involves significant administrative friction.

SOE subsidies and capital over labor

State-owned enterprises receive implicit and explicit subsidies that represent a significant expenditure-side transfer toward capital. These include preferential access to bank credit at below-market rates, subsidized land allocation, tax preferences, regulatory protection from competition, and implicit government guarantees that lower borrowing costs. The beneficiaries of these arrangements are SOEs and their capital owners, shareholders, and — to some extent — their often higher-compensated formal sector workforces. The costs are borne broadly through credit misallocation, suppressed returns on household savings deposited in state banks, and the crowding out of private sector investment.

The distributional consequence is a transfer from the broader population — including depositors, private sector workers, and households — to state capital and those employed in the most capital-intensive state sector industries. This is not a direct fiscal transfer in the sense of a budget line, but it operates through the fiscal-financial nexus in a way that is analytically equivalent to a subsidy. To the extent that SOE employment skews toward urban formal sector workers and away from informal, rural, or migrant workers, the implicit subsidy reinforces rather than offsets existing inequality.

Infrastructure investment: coastal and urban concentration

Public infrastructure investment has been one of the defining features of China's development model, and by headline measures it has been extraordinary: high-speed rail, urban metro systems, highways, ports, and digital infrastructure at a pace and scale with few international comparables. But the geographic distribution of this investment has been systematically skewed toward coastal and urban areas, particularly during the high-growth decades of the 1990s and 2000s.

The logic is partly endogenous to the development model: infrastructure investment is most financially productive where economic density is already high, and high-speed rail connections between major coastal cities generate higher returns than equivalent investment in the interior. But it is also partly a consequence of the revenue architecture: land finance revenues, which fund much local infrastructure, are highest in areas where land values are highest — which are coastal and urban areas. The fiscal capacity to invest correlates with the level of existing development, reinforcing spatial concentration rather than reducing it.

The distributional effect is that residents of interior and rural areas — who are generally lower-income — have received less public infrastructure investment per capita than coastal urban residents, despite arguably facing greater infrastructure deficits. This is not an absolute decline; infrastructure investment has reached most of the country over time. But relative to need, and relative to the distribution of the fiscal resources that fund it, the skew has been persistent.

5. Structural Drivers: Fiscal Federalism and Its Discontents

The features described in the preceding sections are not independent; they are connected by a common structural logic rooted in the 1994 fiscal reform. Before 1994, local governments retained a large share of tax revenues and remitted a negotiated sum to the center — an arrangement that proved fiscally unsustainable as local governments under-remitted and the central government share of revenue fell to levels incompatible with macroeconomic management. The 1994 reform centralized revenue collection while maintaining a broad assignment of expenditure to local governments, creating the structural gap that land finance subsequently filled.

The local revenue pressure this created has been the single most powerful driver of fiscal behavior at the sub-national level. Land finance, local government financing vehicles (LGFVs), and off-balance-sheet borrowing are all responses to the same underlying constraint: local governments are responsible for funding education, healthcare, social welfare, and infrastructure, but the revenue instruments available to them — primarily the local share of VAT, corporate income tax, and land revenues — are insufficient, volatile, or require revenue strategies (high land prices) that impose costs on residents.

The hukou system interacts with this revenue pressure in a specific way. Extending full public services to non-local migrants imposes fiscal costs on the receiving local government without a corresponding increase in fiscal capacity. Local governments that have invested in public services calibrated for their registered population face a direct fiscal disincentive to extend those services to migrants. The hukou system is, in part, a rationing mechanism for locally funded public goods — and as long as local fiscal capacity is constrained relative to expenditure obligations, the incentive to maintain hukou-based rationing persists regardless of central government rhetoric about migration liberalization.

6. Distributional Consequences: Gini Trends and Hidden Inequality

China's official Gini coefficient, as reported by the National Bureau of Statistics (NBS), peaked at 0.491 in 2008, declined to approximately 0.462 by 2015, and has oscillated in the range of 0.46 to 0.47 since. These figures, already among the highest for major economies, are widely understood to understate true inequality for several reasons. Survey-based Gini estimates typically undercount income at the top of the distribution; high-income households are systematically under-sampled and under-reporting of capital and business income is documented. Studies by Li Shi, Luo Chuliang, and others using corrected survey data and income tax records suggest that the true Gini may be 5 to 10 percentage points higher than the official figure.

More importantly, the Gini coefficient as conventionally measured captures monetary income inequality but not the inequality in access to in-kind public services, which constitutes a significant component of effective living standards. A complete measure of fiscal incidence would include the imputed value of education, healthcare, and social insurance access — and would find that inequality in these dimensions is substantial and correlated with monetary inequality. The migrant worker in Shenzhen and the locally registered resident with identical cash incomes do not have equivalent effective incomes once the public service differential is incorporated.

The World Bank's estimates of fiscal incidence for China — drawing on the methodology of the Commitment to Equity (CEQ) project — find that China's fiscal system reduces the market Gini by a smaller margin than comparable economies at similar income levels. Brazil, Mexico, and South Africa — countries with their own severe distributional problems — reduce inequality through fiscal operations by more than China does relative to their market distributions. China's fiscal system is not uniquely punishing, but it is notably less redistributive than the statutory structure and the scale of public spending would suggest.

7. Conclusion: Structural Outcome, Not Policy Failure

The distributional features documented in this essay are best understood as structural outcomes of a coherent development model rather than as a collection of policy failures amenable to incremental correction. The developmental state model that China pursued from the 1980s onward required rapid capital accumulation, high investment rates, and spatial concentration of economic activity in export-oriented coastal manufacturing zones. A fiscal system oriented around consumption taxation, minimal wealth taxation, land-financed infrastructure, and public service rationing through hukou was well suited to those imperatives. It mobilized resources without imposing the static allocative costs of income redistribution; it financed infrastructure without requiring the fiscal capacity that a more redistributive system would demand; it managed migration without requiring the public service expansion that free movement would entail.

The question for the current period is whether the model can adapt. The "common prosperity" agenda announced with greater prominence from 2021 onward signals awareness at the highest levels that distributional sustainability is now a political as well as economic priority. But the structural obstacles to reform are formidable. Property taxation faces resistance from the urban middle class, whose political quiescence is a priority for social stability management. Land finance reform requires either finding alternative revenue sources for local governments — which means either raising central grants (fiscal transfers) or allowing local governments to levy new taxes — or reducing local expenditure obligations, neither of which is straightforward. Hukou liberalization in major cities imposes real fiscal costs on receiving governments without a clear mechanism for funding the expanded service obligations.

The reform paths that would reduce fiscal regressivity — a recurrent property tax, inheritance tax, broadened capital gains taxation, portability of social insurance entitlements, decoupling of public services from registration status, fiscal equalization sufficient to fund services regardless of land revenue capacity — each face a political economy constraint rooted in the interests of groups that have accumulated advantage under the existing arrangement. In China's political system, the absence of electoral competition means that these constraints operate differently than they would in a democratic polity, but they are not absent. Local government fiscal interests, property-owning urban households, and state capital beneficiaries all have stakes in the status quo that shape the politics of reform even within an authoritarian system.

The analytic conclusion is that fiscal regressivity in China is not a residue of underdevelopment, nor a correctable inefficiency, nor a sign that the right policies have not been tried. It is the distributional signature of a specific development model — one that prioritized accumulation, investment, and growth at the expense of redistribution, and that built institutional structures capable of sustaining those priorities over decades. Whether those structures can be reformed without dismantling the political economy that sustains them is the central question of China's fiscal future.