## CHINA’S HIGH SAVINGS: DRIVERS, PROSPECTS, AND POLICIES ## Source details **Canonical URL:** [CHINA’S HIGH SAVINGS: DRIVERS, PROSPECTS, AND POLICIES](https://www.imf.org/-/media/files/publications/cr/2017/cr17248.pdf) ## Other formats - [Markdown version](/-/media/files/publications/cr/2017/cr17248.pdf.md) - [Structured JSON version](/-/media/files/publications/cr/2017/cr17248.pdf.json) --- ### A. Introduction and Overview - China has one of the highest levels of national savings in the world, historically around 35-40 percent of GDP. - After WTO entry in 2001, savings surged to peak at 52 percent of GDP in 2008. - Since the Global Financial Crisis (GFC), national savings have gradually come down to 46 percent in 2016. - Global average national savings is around 25 percent of GDP. - High savings contributed to large external imbalances pre-GFC: the current account surplus surged to 9 percent of GDP in 2007; after the GFC the current account surplus declined to 2 percent of GDP. - Investment ratio is elevated at 44 percent of GDP; private consumption accounts for only 39 percent of GDP (compared to the global average of around 60 percent). - China’s household savings today are 23 percent of GDP, which is 15 percentage points higher than the global average. ### B. Composition of National Savings - Sectoral drivers: - Households: 23 percent of GDP — main driver of high national savings. - Corporates: surged in the 2000s, moderated after the GFC; modestly higher than global average. - Government: fiscal savings volatile and on average a small part of national savings; recent years higher than global average due to low current spending and very high capital spending. ### C. Household Savings — Phases, Drivers, and Empirical Effects - Three phases of household savings surge (as percent of disposable income): - 1980s: rose from 5 to 20 percent following the one-child policy and decollectivization. - 1990s: rose to 25 percent amid SOE reform and breakdown of social safety net. - After 2001: rose to 30 percent amid an export-driven boom; since 2010 plateaued and gradually started to decline. - Demographics: - One-child policy led fertility decline from 6 to below 2 and an extremely low youth dependence ratio. - Demographic changes account for about half of the rise in household savings (overlapping generations model). - Social safety net and precautionary savings: - Health care coverage of urban workers declined by 17 percentage points between 1990 and 2000. - Average replacement rate for urban workers dropped from close to 80 percent to below 50 percent. - Recent schemes: new rural health care and pension schemes and the basic pension scheme for non-working urban residents; government spending remains below international standards; migrant worker access impeded by slow “Hukou reform”. - Micro-level empirical elasticities (median effects preserved): - Urban households: a 1 percent increase in social security spending is associated with a 0.06-0.08 percent increase in household spending. At medians, an 80 yuan increase in social security spending (about 10 percent of median monthly social security per capita) increases consumption by 90 yuan. - Rural households: a 1 percent increase in health care spending raises rural consumption by 0.11 to 0.19 percent. At medians, a 50 yuan increase in health spending per capita (about 10 percent of the median monthly healthcare spending per capita) would increase rural consumption by between 70 and 120 yuan. - Implication: social security transfers are more effective in urban areas; health spending more effective in rural areas. - Income inequality and savings inequality: - Gini index rose from 0.3 in the 1980s to about 0.5 in 2010 (moderated recently). - Difference between saving rates of the richest and poorest decile often as large as 20 percentage points. - In 2013, top earners saved close to 50 percent of their income; households in the bottom 10 percent saved about 20 percent. - Chinese households save more at any income decile compared to other countries; the gap largest for the poor. - Housing reforms and effects: - 1988 privatization and 1998 end of enterprise-supplied housing transformed market. - Homeownership surged from 20 percent in 1988 to 90 percent in 2007 and has been stable since. - Micro evidence on ownership and saving dynamics: - Prior to 1995 reform: no systematic tenant/homeowner savings difference. - After reform: homeowners saved more (mortgage effect); by 2013 homeowners saved marginally less than tenants (wealth effect); tenants’ savings remained high to finance down payments. ### D. Financial Repression, Interest Rates, and Interpretation - Target-saving hypothesis: - Nabar (2013): a 1 percent interest rate decline increases household savings rate by about 0.5 percentage point. - IMF analysis: effect disappears when controlling for cross-region income differences. - With rapid interest rate liberalization and proliferation of higher-yield non-deposit products, financial repression is no longer a major driver today. - Overall interpretation: rising household savings largely due to demographics, rising incomes and widening inequality, breakdown of social safety net, and housing reforms; some factors have started reversing (rebuilding safety net, moderation in inequality). ### E. Corporate Savings and Ownership Patterns - Aggregate and cross-country comparability: - Corporate savings rate surged in early 2000s, moderated after GFC. - Median gross saving ratio in China about 4 percent of assets (vs 5-6 percent in rest of world). - Net savings ratio about -0.01 percent (global average 0 – 0.1 percent). - SOE vs private firms: - SOEs save less in gross terms (low operating efficiency/profits) but more in net terms (lower investment ratios) after controlling for size, Tobin’s q, sector, macro variables and time dummies. - SOEs held 57 percent of total corporate debt or 72 percent of GDP in 2016 and contributed to almost 60 percent of the rise in total corporate debt during 2008–16. - SOE productivity on average about 25 percent lower than private firms (controlling for industries). ### F. Government Savings, Fiscal Aggregates, and Perimeter - Trends and magnitudes: - Fiscal savings trended up since 2000, peaked near 6 percent in 2008, stabilized around 5 percent of GDP in recent years. - In 2014, Chinese government savings were 5.6 percent of GDP. - Composition and reasons: - Strong focus on public investment: government investment to GDP ratios of 4 percent on-budget and 12 percent including off-budget investment. - Augmented fiscal deficit of 10 percent in 2014; fiscal savings still higher than other countries due to skew toward investment. - Social security moved from net savings engine in 1990s/early 2000s to deficit in 2013; pension spending expected to rise from 5 percent of GDP in 2015 to 10 percent in 2050. - Fiscal perimeter and debt measures: - Narrowly-defined government debt: 37 percent of GDP by end-2016. - “Augmented+” government debt: around 62 percent of GDP in 2016; staff projects it rising to 92 percent of GDP in 2022 under augmented measures. - Under augmented definition, primary balance needs improvement by 4 percentage points of GDP by 2022 to stabilize augmented debt at around 100 percent of GDP. ### G. Role of the State and Macro Consequences - Policy-induced structural drivers of high savings: - One-child policy, breakdown of social safety net during 1990s transition, WTO entry and export boom with undervalued exchange rate, skewed government spending toward investment, housing reform and rising housing prices, and rising income inequality. - Macro consequence: - High household savings rate and falling household share of national income depressed household consumption; consumption-to-GDP ratio fell sharply in early 2000s. - Recommendation: less skewed distribution of growth benefits between state and households, more private sector income-earning opportunities to boost consumption. ### H. Policy Recommendations to Reduce Savings and Boost Consumption - Overarching goals: strengthen social safety net, reduce income inequality, raise household incomes. - Specific measures: - Make income tax more progressive and family-friendly: - Current tax structure regressive for the very poor; personal income tax has high exemption threshold and flat nominal social contribution results in effective tax rate higher than 40 percent for the poor. - Consider tax allowance based on family size to boost fertility. - Increase social transfers to poor households: - Close large household savings gaps for the poor; many countries’ bottom-income earners have negative savings due to transfers. - Increase expenditure on health care, pensions and education: - Higher health care and pension spending increase consumption and reduce precautionary savings. - Continue Hukou reform to ensure migrant worker access. - Education spending reduces future income inequality. - Shift spending toward public services (“soft infrastructure”) requiring higher spending to attract talent and address staffing shortages. - Finance higher social spending via SOEs: - Bringing SOE dividend payment ratio to announced target of 30 percent could increase budget revenues by about 1 percent of GDP per year. - Transferring SOE shares to social security funds could help address actuarial imbalance. - Improve access to formal financing for private firms: - Remove implicit SOE guarantees and redirect credit to more efficient private sector users to reduce firms’ reliance on internal savings. - Service sector liberalization: - Open restricted service sectors to private and foreign investment to boost productivity and household income opportunities. ### I. Prospects and Scenarios for Savings and Consumption - Demographics and social safety net strengthening: - Old-age dependency ratio projected to rise from 15 in 2015 to 50 in 2050. - Model simulations: demographic changes would reduce household savings rate (in percent of disposable income) by 6 percentage points by 2030. - With continued social safety net strengthening (assumption: government budget health expenditure-to-GDP ratio increases from 1.8 percent in 2016 to 2.5 percent by 2022), household savings, in percent of GDP, are forecast to fall by 3 percentage points by 2022. - Changing consumer behavior of younger cohorts and slower wage growth may lead to faster decline in household savings rate. - National savings outlook: - Lower household and fiscal savings expected to reduce overall national savings by about 4 percentage points by 2022. - Corporate savings likely to fall moderately as capital returns decline and labor intensity rises with services shift. - Proactive scenario: - Policies: increase on-budget healthcare spending to 3.1 percent of GDP by 2022 (vs 2.5 percent in baseline) and moderately reduce income share of top 10 percent via redistribution. - Outcome: reduce household savings by 5 percent of GDP by 2022 (vs 3 percent in baseline); household consumption would be boosted by an extra 2 percent of GDP. ### J. Credit Boom, Financial Risks, and Macroprudential Policy - Credit trajectory and magnitudes: - Nominal credit to nonfinancial sector more than doubled in last five years; total domestic nonfinancial credit-to-GDP ratio increased by 60 ppt to about 230 percent in 2016. - Credit gap about 25 percent of GDP, above the 10 percent BIS threshold for maximum counter-cyclical buffer. - Banks’ claim on private nonfinancial sector (narrowest): about 155 percent of GDP as of 2016. - Total social financing (TSF) stock: about 209 percent of GDP as of end-2016; households accounted for about 44 percent of TSF. - Total domestic nonfinancial sector credit: estimated about 234 percent of GDP. - Augmented public debt treating non-recognized LGFV debt as contingent liabilities: “Augmented” total public debt 62 percent of GDP as of 2016; remaining nonfinancial sector debt (private debt) 172 percent of GDP. - Cross-country evidence on credit booms: - Identified 43 cases where credit-to-GDP rose by >30 percentage points over 5 years; only 5 cases ended without major slowdown or crisis. - All credit booms that began when ratios were above 100 percent—as in China’s case—ended badly. - Efficiency deterioration: - Credit efficiency fell: in 2007-08 new credit of about RMB 6½ trillion raised nominal GDP by about RMB 5 trillion/year; in 2015-16 it took RMB 20 trillion in new credit. - If nonfinancial private sector credit-to-GDP had increased only by 10 ppt (no private sector credit gap in 2016), average real GDP growth for 2012–16 would have been 5.3 percent instead of actual 7.3 percent (alternative estimate: 5.9 percent). - China-specific buffers and limits: - Strong external position and low external debt reduce typical external funding crisis risk but do not eliminate crisis risk. - Narrow loan-to-deposit ratio: 72 percent in 2016; total nonfinancial domestic credit-to-GDP: 118 percent (comparator ratios much higher in other crises). - Official general government debt: less than 40 percent of GDP as of 2016; augmented debt projected to rise to more than 90 percent over medium term. - PBOC can provide liquidity and capital controls remain partially effective, but quick central bank action may be insufficient given system size/complexity. - Policy implications: - De-emphasize high and hard GDP targets to reduce credit expansion driven by targets. - Comprehensive strategy to increase credit efficiency by reducing demand for least efficient uses. - Financial reforms: close regulatory loopholes, rein in leverage, increase transparency of nonbank institutions and wealth management products. - Macroprudential policies should be primary instrument to prevent instability; monetary policy should not substitute except when costs/benefits warrant. ### K. Real Estate Risks and Recommendations - Residential real estate dynamics: - Real estate investment rose from about 4 percent of GDP in 1997 to peak of 15 percent in 2014; residential investment > two-thirds of total housing investment. - Nationwide housing inventory ratio declined to about 18-20 months from 30 months at peak in 2014. - Residential property sales volume: 12 percent y/y in May 2017. - Average loan-to-value (new mortgage loans to property sales) rose from around 15 percent in 2012 to 48 percent in [incomplete original sentence]. - Risks: - Downside: further price rises, spread to smaller cities, bubble expansion, and higher cost of sharp correction. - Nationwide house prices deviate only 5 percent from long-term trend; tier-1 cities about 10-15 percent. - A correction of house prices by 10-15 percent would reduce GDP growth by around 0.9 percentage points through impacts on property investment and household consumption. - Household debt rose from less than 20 percent of GDP in 2008 to more than 40 percent in 2016; mortgages account for more than half of outstanding household debt. - Policy recommendations: - Continue and expand macro-prudential and city-specific policies; differentiated LTV limits appropriate. - Expand toolkit: consider debt servicing-to-income (DSTI) caps (current cap 50-55 percent; recommend gradual tightening to below 45 percent) and extend to non-bank household loans. - Sectoral capital requirements: increase risk weights or LGD floors on banks’ real estate exposure. - Introduce recurrent property taxes (national property registration system by 2017 welcomed) to dampen volatility and provide local revenue. - Increase real estate supply in higher-tier cities, complement with Hukou reform and social security reforms. - Reduce reliance on administrative measures over medium term; de-emphasize quantitative GDP growth targets that foster short-term stimulus and credit-driven cycles. ### L. SOE Debt, Zombies, Overcapacity, and Restructuring - SOE scale and vulnerabilities: - SOE industrial output share declined from over 40 percent to about 15-20 percent over last 15 years. - SOEs held 57 percent of total corporate debt or 72 percent of GDP in 2016; contributed to almost 60 percent of rise in corporate debt 2008–16. - Productivity gap: SOE productivity on average 25 percent lower than private firms. - Policy measures and assessment: - Deleveraging and restructuring: guidelines for debt-equity swaps and creditor committees issued; positive first steps but lack details on loss recognition and operational restructuring. - Cleaning up zombies: government identified over 2,000 central SOE zombies (assets ~4 percent of GDP) and over 7,000 local SOEs; reportedly 20 percent of central SOE zombies were resolved in 2016 (details limited). - Reducing overcapacity: coal and steel target cuts of 10–15 percent of 2015 output and workforce reduction of 1.8 million over 3–5 years; targets exceeded in 2016 and on-track in 2017 but debt in overcapacity sectors has not fallen. - SOE reforms: consolidation, phasing out social functions, transferring about ½ percent of state-owned equity to social security funds, pilot mixed-ownership and employee stock ownership programs; transfer of SOE profits to budget well below the 30 percent target. - Recommended actions (holistic, time-bound): - Restructuring corporate debt: conduct targeted asset quality review; reinforce accounting/audit rules; raise appraiser standards; develop credit registers. - Recognize and allocate losses: strengthen loan classification, bank capital, collateral valuation, and prudential reporting. - Operational restructuring: develop plans for weak SOEs; ensure viability for creditors and equity holders. - Refining mechanisms: align creditor committees with international best practices; precede debt-equity swaps with asset quality review and operational restructuring; transfer bank loans at economic/market value. - Removing zombies: suspend implicit support; transfer essential social functions to fiscal budget; publicly identify nonviable zombies and use liquidation; clear timetable to resolve identified zombies within 1–2 years. - Reducing overcapacity: raise net reduction targets; shift from administrative to strict enforcement; phase out energy subsidies and impose resource/environmental taxes; provide targeted social policies to mitigate layoffs (estimated 2.8 million workers). - Reforming SOEs: emphasize operational restructuring, harden budget constraints, reduce implicit support and privileged access to credit, enforce profit transfers to reach 30 percent by 2020, facilitate market entry and open protected markets. ### M. Monetary Policy: Market-Based Approach and Operational Independence - Key summary: - Implementation becoming more market-based; new interest rate corridor main instrument. - Significant pass-through from PBC’s interest rates to bank lending rates, government yields and other market rates. - VAR analysis: 100bps interbank repo rate increase yields cumulative output loss about 1.7 percent after two years; modest reduction in CPI headline inflation about 0.2 percentage points after two years (not statistically significant). - Taylor rules suggest monetary policy accommodative through early 2017; next major reform: introduce an inflation target or range with operational independence for PBC. - Pass-through and empirical results: - Regressions: high coefficient about 70 percent for the PBC benchmark rate and about 30 percent for the interbank repo rate on average lending rate. - Bivariate VAR: almost complete pass-through from PBC benchmark rate changes to 2-year government bond yields after about 6 months; interbank rate pass-through to 2-year yield about 25 percent. - Changes in M2 or aggregate loan volume generally do not produce significant response in GDP or inflation in most VAR specifications. - Operational recommendations: - PBC should gain operational (instrument) independence; State Council to define long-term goals (e.g., medium-term inflation target) and PBC accountable for achieving them. - Declare the 7-day interbank repo rate as operating target and phase out benchmark rates. - Separate macroprudential responsibilities and monetary policy while recognizing interactions. - Rely on price-based instruments via interest rate corridor; price/access to PBC lending facilities based on clear collateral rules. ### N. Capital Account Opening, CFMs, and Policy Sequencing - Opening and sequencing: - Capital inflows liberalized before outflows; FDI liberalized early 1990s (services and strategic sectors still restricted); ODI opened gradually from 1999; portfolio access expanded since 2002 and 2015; cross-border lending partially relaxed 2008-10. - De jure vs de facto openness and outflows: - Chinn-Ito index: very closed; Quinn index: gradually opened. De facto openness greater than de jure indicated by capital outflows. - Net capital outflows: around $648 billion in 2015 (5.8 percent of GDP); almost US$640 billion in 2016 (5.7 percent of GDP). - Assessment and recommendations on CFMs: - Sequencing broadly aligned with IMF integrated approach but supporting reforms lagged. - CFMs tightened since mid-2016 to curb outflows; PBC intervention and improved growth outlook helped stabilize reserves slightly above USD 3 trillion. - When CFMs used: implement consistently, communicate clearly, implement transparently through written rules, avoid restricting current international payments and transfers. - Authorities should expeditiously implement reforms: strengthen financial supervision, macroprudential framework, SOE reforms, and monetary policy framework. Liberalize cautiously in tandem with reforms; prioritize FDI in services, financial services, and health. ### O. Fiscal Framework, Intergovernmental Reform, and Revenue Assignment - On-budget expansion and augmented aggregates: - On-budget net borrowing expanded by 2¾ percent of GDP in 2014-2016; off-budget activities broadly unchanged after perimeter expansion. - Fiscal space narrow vs augmented: - Narrow government debt 37 percent of GDP end-2016; median EME government debt 48 percent of GDP. - Augmented+ around 62 percent of GDP in 2016. - Ageing could increase yearly structural deficit by average 2.5 percent of GDP until 2050; present value liability ~90 percent of 2015 GDP. - Under stylized contingent liability shock of 10 percent of bank claims, debt reaches 69 percent of GDP in 2022. - Intergovernmental reform recommendations: - Formulate medium-term fiscal framework including all LG infrastructure spending; create incentives (carrots and sticks) for LG units. - Increase capacity for monitoring and enlarge perimeter of monitoring. - Publish detailed provincial budget accounts and contingent liabilities yearly; publish medium-term fiscal projections. - Centralize pensions and unemployment insurance financing for better risk pooling and portability. - Move toward rules-based general transfers, increase equalization grants, rationalize targeted transfers. - Allow higher local borrowing quotas to bring off-budget spending onto budget. - Revenue assignment: - Consider allowing provinces to impose PIT surcharge within centrally-approved limits (say 5 to 10 percent). - Maintain national scope of VAT; review revenue-sharing arrangements to reduce compliance costs. - Adopt recurrent market-value based property tax as ideal local revenue source; define tax base nationally and allow local rate-setting within central bands. - Use revenue-side measures (PIT progressivity, property tax) plus expenditure-side (education, health, social assistance) to reduce inequality. *Source: IMF staff report content (PEOPLE’S REPUBLIC OF CHINA — CR17248) as provided in the source document.* ### References _______________________________________________________________________________ 16 ### CHINA’S HIGH SAVINGS: DRIVERS, PROSPECTS, AND POLICIES ### A. Introduction and Overview - China has one of the highest levels of national savings in the world, historically around 35-40 percent of GDP. - After WTO entry in 2001, savings surged to peak at 52 percent of GDP in 2008. - Since the Global Financial Crisis (GFC), national savings have gradually come down to 46 percent in 2016. - Global average national savings is around 25 percent of GDP. - High savings contributed to large external imbalances pre-GFC: the current account surplus surged to 9 percent of GDP in 2007. - After the GFC, the current account surplus declined to 2 percent of GDP, and excess savings morphed into a growing internal imbalance. - Investment ratio is elevated at 44 percent of GDP; private consumption accounts for only 39 percent of GDP (compared to the global average of around 60 percent). - China’s household savings today are 23 percent of GDP, which is 15 percentage points higher than the global average. ### B. Composition of National Savings - Sectoral shifts over time: - Households: have trended up since the early 1990s; at 23 percent of GDP, households are the main driver of high national savings. - Corporates: surged in the 2000s but after the GFC narrowed to be modestly higher than the global average. - Government: fiscal savings have been volatile and on average only a small part of national savings; in recent years fiscal savings have been higher than the global average due to low current spending and very high capital spending. ### C. Household Savings — Phases and Drivers - Three phases of household savings surge: - 1980s: savings rate rose from 5 to 20 percent of disposable income following the one-child policy and decollectivization (with a temporary dip in the late 1980s). - 1990s: savings rose to 25 percent of disposable income amid SOE reform and the breakdown of the social safety net. - After 2001: savings rose to 30 percent of disposable income amid an export-driven boom; since 2010 household savings have plateaued and gradually started to decline. - Demographics - The one-child policy led to a rapid decline of fertility from 6 to below 2 and an extremely low youth dependence ratio. - Demographic changes account for about half of the rise in household savings (using an overlapping generations model). - Demographics increased savings via: less child-related expenditure and higher precautionary savings for retirement. - Social Safety Net - The transition of the 1980s/90s led to the breakdown of the social safety net and rising savings. - Health care coverage of urban workers declined by 17 percentage points between 1990 and 2000. - Average replacement rate for urban workers (pension benefits in percent of wages) dropped from close to 80 percent to below 50 percent. - Recent policy efforts: new rural health care and pension schemes and the basic pension scheme for non-working urban residents; government spending remains below international standards; access to health care remains an issue for migrant workers due to slow progress on “Hukou reform”. - Empirical findings on social spending and consumption: - Urban households: a 1 percent increase in social security spending is associated with a 0.06-0.08 percent increase in household spending. At medians, an 80 yuan increase in social security spending (about 10 percent of median monthly social security per capita) increases consumption by 90 yuan. - Rural households: a 1 percent increase in health care spending raises rural consumption by 0.11 to 0.19 percent. At medians, a 50 yuan increase in health spending per capita (about 10 percent of the median monthly healthcare spending per capita) would increase rural consumption by between 70 and 120 yuan. - The differential impacts imply targeted social spending: social security more effective in urban areas; health spending more effective in rural areas. - Income Inequality - Gini index rose from 0.3 in the 1980s to about 0.5 in 2010 (moderated in recent years). - Income inequality is associated with significant savings inequality: - Difference between saving rates of the richest and the poorest decile often as large as 20 percentage points. - In 2013, top earners saved close to 50 percent of their income, while households in the bottom 10 percent saved about 20 percent. - Chinese households save more at any income decile compared to other countries; the gap is largest for the poor (many countries have negative saving rates for the bottom 10-20 percent, while China’s bottom 10 percent saved about 20 percent). - Housing - Two major housing reforms transformed the market: 1988 privatization (sale of rental housing to SOE workers at low prices) and 1998 end of enterprise-supplied housing, moving to market-based provision. - Homeownership surged from 20 percent in 1988 to 90 percent in 2007, and has been stable since then. ### D. Key Findings and Implications - High national savings have contributed to both external imbalances (pre-GFC) and internal imbalances (post-GFC) through excessive investment and falling efficiency. - Household sector is now the principal source of high national savings. - Demographic forces will put downward pressure on national savings going forward. - Strengthening the social safety net and reducing income inequality are needed to reduce savings further and faster and to boost consumption. - Targeted social spending (social security in urban areas; health in rural areas) can effectively increase consumption. *Source: IMF staff report content (CHINA’S HIGH SAVINGS: DRIVERS, PROSPECTS, AND POLICIES) from the provided PDF content.* ### 12. Micro-level regressions show that the overall impact of household ownership on ### cr17248 - 12. Micro-level regressions show that the overall impact of household ownership on ### Household ownership and household savings - Prior to the housing reform in 1995, there was no systematic difference in the savings rates between tenants and home owners. - After the reform, home owners started to save more than tenants, as shown in the 2002 and 2007 surveys. - By 2013, the trend reversed, with home owners saving marginally less than tenants. - Household-level regressions controlling for micro characteristics (age, sex, education, occupation, income, household size, etc.) confirm these patterns. - Interpretation: - Early years: the mortgage effect dominated for home owners, raising their savings. - Later years (booming housing market): the wealth effect reduced home owners’ savings rate. - Tenants’ savings rate remained higher, reflecting the need to save for down payments. - Source for survey-based chart: Various waves of CHIP survey. ### Financial repression and interest rate liberalization - The “target saving” hypothesis: depressed deposit rates could lead to higher household savings. - Nabar (2013) finds that a 1 percent interest rate decline increases the household savings rate by about 0.5 percentage point. - IMF analysis: that small quantitative effect disappears once cross-region income differences are controlled for in the regression. - With rapid interest rate liberalization and proliferation of non-deposit financial products with much higher yields, financial repression is no longer a major driver of household savings today. - Overall interpretation: rising household savings were largely a result of demographic changes, rising incomes and widening inequality, and breakdown of the social safety net; housing reforms also contributed. Some of these factors have started reversing in recent years (including rebuilding the social safety net and slight moderation in income inequality). ### Corporate savings - Aggregate pattern: - Corporate savings rate surged in the early 2000s, but moderated after the GFC. - The net savings rate (gross savings minus investment as a ratio to assets) was negative throughout last decade, reflecting high investment in China. - As a result, the corporate sector in China is not a driver of the current account surplus. - Cross-country comparability: - Median gross saving ratio in China is about 4 percent of assets, slightly lower than the 5-6 percent in the rest of the world. - Net savings ratio is about -0.01 percent, lower than global average of 0 – 0.1 percent. - Ownership differences: - SOEs have lower gross savings but higher net savings compared to private firms. - After controlling for size, Tobin’s q, sector, macro variables and time dummies: Chinese SOEs save less in gross terms (low operating efficiency and profits) but more in net terms (lower investment ratios). - Exchange rate effects: - Regression analysis shows a significant impact of the exchange rate on corporate savings, larger in the tradable sector. - During 2005-2008, the RMB was assessed to be undervalued, contributing to the export boom and large corporate savings; in recent years, as the RMB moved in line with equilibrium, excessive savings of the export sector have largely been unwound. ### Government savings - Trends: - Since 2000, fiscal savings were on an upward trend and peaked near 6 percent in 2008. - Since 2008, government savings have stabilized at around 5 percent of GDP in recent years. - In 2014, Chinese government savings were 5.6 percent of GDP. - Reasons for relatively high government savings: - More focus on public investment than public services: - Government investment to GDP ratios of 4 percent on-budget and 12 percent including off-budget investment. - Augmented fiscal deficit of 10 percent in 2014, yet fiscal savings remain higher than other countries due to skewed composition towards investment. - Lower social spending compared to emerging markets, especially on public education, public health and social assistance. - Early stage of social security system: - Social security shifted from being an engine for savings in the 1990s/early 2000s to turning into deficit in 2013 as payments outpaced contributions. - Expected rise in pension spending from 5 percent of GDP in 2015 to 10 percent in 2050. ### Role of the state: drivers of high savings - Policy-induced structural changes driving high savings: - The one-child policy implemented in the late 70s reduced fertility and raised the savings rate. - 1990s transition from planned to market economy dismantled the social safety net, increasing precautionary household savings. - 2000s WTO entry led to export-oriented growth surge and higher corporate savings, partly owing to an undervalued exchange rate. - High government savings due to weak social spending and excessive investment. - Housing reform and rapidly rising housing prices increased household saving needs for down payments or mortgage payments. - Rising income inequality, partly due to limited income redistribution, led to higher savings as more income accrues to richer households with higher propensity to save. - Macro consequence: - High household savings rate and falling household share of national income depressed household consumption; the consumption-to-GDP ratio fell sharply in early 2000. - Recommendation: a less skewed distribution of growth benefits between the state and households, including more private sector income-earning opportunities, is essential to boost household consumption. ### Policy recommendations to lower savings and boost consumption - Strengthen social safety net, reduce income inequality, and raise household incomes. - Specific measures: - Make income tax more progressive and more family-friendly: - Current tax structure is regressive, especially for the very poor. - Personal income tax has a relatively high exemption threshold, but the flat nominal amount of social contribution at the bottom results in an effective tax rate higher than 40 percent for the poor. - Consider tax allowance based on family size to boost fertility. - Increase social transfers to poor households: - Gap in household savings is notably larger for poor households; many countries see bottom-income earners with negative savings due to transfers. - Call for further increase in social assistance spending in China to align with international norms. - Increase expenditure on health care, pensions and education: - Higher health care and pension spending significantly increase consumption and reduce precautionary household savings. - Continued Hukou reform needed to ensure migrant workers have same access to social safety net. - Education spending helps reduce future income inequality by improving access for the poor. - Increase general spending on public services: - Shift some focus from hard infrastructure to “soft infrastructure” (policy frameworks, delivery of public services), requiring higher spending to attract talent and address staffing shortages. - Finance higher social spending via SOEs: - Bringing SOE dividend payment ratio to the announced target of 30 percent could increase budget revenues by about 1 percent of GDP per year to fund higher social spending. - Transferring SOE shares to social security funds could help address projected actuarial imbalance. - Improve access to formal financing for private firms: - Remove the “implicit guarantee” for SOEs and redirect credit to more efficient private sector users; improved financing access will reduce private firms’ reliance on internal savings for investment. - Service sector liberalization: - Provide private and foreign sectors access to restricted service sectors to boost productivity, enhance income-earning opportunities, and increase household consumption. ### Prospects and scenarios for savings - Aging and social safety net strengthening - Old-age dependency ratio projected to rise from 15 in 2015 to 50 in 2050. - Model simulations: demographic changes would reduce household savings rate (in percent of disposable income) by 6 percentage points by 2030. - With continued strengthening of the social safety net (assumption: government budget health expenditure-to-GDP ratio increases from 1.8 percent in 2016 to 2.5 percent by 2022), household savings, in percent of GDP, are forecast to fall by 3 percentage points by 2022. - Changing consumer behavior of younger cohorts and slower wage growth may result in a faster decline in household savings rate. - National savings outlook - Lower household and fiscal savings are expected to reduce overall national savings by about 4 percentage points by 2022. - Corporate savings likely to fall moderately, reflecting falling capital returns as growth slows and rising labor intensity as the economy shifts more towards services and the labor share of income rises. - Proactive scenario - Policies: increase on-budget healthcare spending to 3.1 percent of GDP by 2022 (compared to 2.5 percent in the baseline) and moderately reduce the income share of the top 10 percent via income redistribution to those at the bottom. - Outcome: reduce household savings by 5 percent of GDP by 2022 (compared to 3 percent in the baseline); household consumption would be boosted by an extra 2 percent of GDP. ### Credit boom context (related content on credit-driven growth) - Recent trajectory and risks: - Nominal credit to the nonfinancial sector more than doubled in the last five years; total domestic nonfinancial credit-to-GDP ratio increased by 60 ppt to about 230 percent in 2016. - The credit gap is about 25 percent of GDP, above the 10 percent BIS threshold for the maximum counter-cyclical buffer. - Credit efficiency deterioration: - Credit to certain sectors (industrial), firms (SOEs), and regions (Northeast) is significantly higher than their value added, indicating inefficient credit use. - Illustration: in 2007-08, new credit of about RMB 6½ trillion was needed to raise nominal GDP by about RMB 5 trillion per year; in 2015-16, it took RMB 20 trillion in new credit. - Growth without excessive credit expansion: - First approach: if nonfinancial private sector credit-to-GDP ratio had only increased by 10 ppt (no private sector credit gap in 2016), average real GDP growth for 2012–16 would have been 5.3 percent rather than the actual average of 7.3 percent. - Second approach (Chen and Ratnovksi, 2017): using provincial panel data 2003–15 with lags, finds 2012–16 average real GDP growth would have been 5.9 percent. - Growth-subtracting effect of credit restraint could be partly offset by pro-rebalancing, on-budget fiscal stimulus and productivity gains from structural reforms. *Source: PEOPLE’S REPUBLIC OF CHINA — INTERNATIONAL MONETARY FUND (excerpts from cr17248).* ### 4. International experience suggests that China’s current credit trajectory is dangerous ### 4. International experience suggests that China’s current credit trajectory is dangerous ### Credit-boom evidence and implications - Identified 43 cases of credit booms in which the credit-to-GDP ratio increased by more than 30 percentage points over a 5-year period; among these, only 5 cases ended without a major growth slowdown or a financial crisis immediately afterwards. - All credit booms that began when the ratios were above 100 percent—as in China’s case—ended badly. - Complex and primarily short-term funding structures underpinning rapid credit growth are a key vulnerability: banks’ rapid asset expansion has relied increasingly on interbank markets, wealth management products, and complex interlinked networks of entities. Asset managers often finance illiquid long-term investments by rolling over short-term funding or pooling investment funds, implying a potentially bumpy deleveraging process. ### China-specific buffers and their limits - Strong external position - Persistent current account surplus and small external debt could reduce the possibility of a typical external funding crisis. - Caveat: Several countries experienced credit booms that ended badly despite current account surpluses/small external debt; funding crises can occur without foreign funding exposure (examples cited: U.S. savings and loan crisis in the 1980s, Japan’s banking crisis in 1997, U.S. and U.K. financial crises in 2008). - Domestic deposit base and loan-to-deposit ratio - Narrowly-defined loan-to-deposit ratio: 72 percent in 2016. - Total nonfinancial domestic credit-to-GDP ratio: 118 percent. - Comparators: 400+ percent in Korea (1998), 350+ percent in the US (2008), 250 percent in Japan (mid-1990s). - Caveat: Loan-to-deposit ratios do not capture total assets and liabilities; the ratio of non-loan assets to total assets is about 50 percent in 2016, higher than the median in the cross-country sample. - Strong asset side of corporate balance sheets - Corporate balance sheets have benefitted from rising asset values, so leverage as measured by the debt-to-asset ratio has been falling. - Caveat: Asset valuations could fall sharply if the boom ends; corporates’ liabilities are mostly financial liabilities, while a significant portion of assets are nonfinancial fixed assets (e.g., land) that may not be easily liquidated. More vulnerable firms hold fewer liquid assets and have lower buffers. - Fiscal space in a state-controlled economy - Official general government debt: less than 40 percent of GDP as of 2016. - Caveat: “Augmented” debt projected to rise to more than 90 percent of GDP over the medium term, with debt on an unsustainable path. The economy-wide cost of bailouts likely exceeds direct financial system costs due to indirect costs (slower GDP growth, lower tax revenue, higher government spending, higher interest payments, contingent liabilities). - Cross-country estimates cited: average direct fiscal cost estimated as 5-10 percent of GDP (Laeven and Valencia, 2010); indirect fiscal cost arising from contingent liabilities realization estimated to be about 6 percent of GDP during 1990-2014 (Bova, et al., 2016); average gross fiscal cost of systemic banking crises about 15 percent of GDP (Dell’Ariccia et al., 2012). - Liquidity provision and capital controls - The PBOC can provide liquidity against funding stress and capital controls are still effective. - Caveat: Quick central bank action may be insufficient given system size, complexity, and interconnectedness (example: the U.S. in the global financial crisis). Liquidity provision could lead to capital outflows and FX pressure and exacerbate moral hazard. Capital controls tightening could spur a growth slowdown and may lose effectiveness over time. - Growth and financial deepening - Rapid credit growth can reflect financial deepening. - Caveat: Efficiency of investment/credit is falling sharply; corporate financial performance—particularly of SOEs—is deteriorating, affecting bank asset quality. IMF SDN/15/08 suggests China’s financial deepening has exceeded the turning point that maximizes positive effects on growth; further deepening could drag growth lower. ### Quantitative measures of China’s credit (Box 1) - Banks’ claim on the private nonfinancial sector (narrowest measure): about 155 percent of GDP as of 2016. - Total social financing (TSF) stock: about 209 percent of GDP as of end-2016; households accounted for about 44 percent of TSF. - Official general government debt: about 37 percent of GDP. - Total domestic nonfinancial sector credit: estimated to be about 234 percent of GDP. - Former credit to LGFVs explicitly recognized as local government debt: about 17 percent of GDP (explains overlap between TSF and general government debt). - Under an alternative accounting that treats non-recognized LGFV debt and policy-guided funds as contingent government liabilities: - “Augmented” total public debt: 62 percent of GDP as of 2016. - Remaining nonfinancial sector debt (private debt): 172 percent of GDP. ### Illustrative proactive scenario and projections - A proactive scenario with faster structural reform (especially SOE reform) and improved resource allocation efficiency could allow credit growth to slow gradually while supporting medium-term growth. - Under the scenario, private credit growth of about 8½ percent would help stabilize the ratio of total domestic nonfinancial sector credit-to-GDP at about 270 percent in 2022. - Near-term growth could dip reflecting faster adjustment, but medium-term growth would rise driven by higher TFP growth. ### Policy implications and recommendations - Deemphasize high and hard GDP targets to reduce the excessive credit expansion driven by targets. - Adopt a comprehensive strategy to increase credit efficiency by reducing demand for least efficient/productive uses. - Financial reforms needed to bolster regulatory and supervisory frameworks, including: - Closing loopholes for regulatory arbitrage. - Reining in leverage. - Increasing transparency of nonbank financial institutions and wealth management products. - Macro-monetary stance and macroprudential roles (Box 2 summary) - Monetary policy should deviate from price/output stabilization only if costs are smaller than benefits; short-term costs of higher rates are lower output and inflation, while benefits accrue mainly in the medium term through reduced financial risks. - The case for “leaning against the wind” is generally limited due to detection, calibration, and implementation challenges, but benefits can outweigh costs in certain circumstances (e.g., clear transmission of rates to credit, large interconnected financial system, absence of targeted macroprudential measures). - Macroprudential policies should be the key instrument to prevent financial instability, as they can target imbalances closer to the source and allow monetary policy to focus on price stability. _International Monetary Fund — PEOPLE’S REPUBLIC OF CHINA, CR17248 (section 4)._ ### 6. While SOEs account for declining shares of industrial output (from over 40 percent to ### 6. While SOEs account for declining shares of industrial output (from over 40 percent to ### SOE shares, debt, and productivity - SOEs account for declining shares of industrial output (from over 40 percent to about 15-20 percent over the last 15 years). - SOEs held 57 percent of total corporate debt or 72 percent of GDP in 2016. - SOEs contributed to almost 60 percent of the rise in total corporate debt during 2008–16. - SOEs underperform private firms on average, with lower returns and productivity. - Productivity was also 25 percent lower than that in private firms on average, controlling for industries. ### Productivity gains from resolving weak firms - Converging to countries with a more efficient productivity distribution (measured by the top 75–90 percentile across countries) would generate productivity gains and raise long-term growth potential by 0.7–1.2 percentage points per year (IMF 2017; Lam and others 2017; Hsieh and Klenow 2009). ### Recent developments and policy measures — Deleveraging and restructuring debt - Government strategy and recent developments: - Strategy envisages a market and legal framework for debt restructuring—rather than state direction or bailouts—and aims to guard against systemic or regional risks. - An inter-ministerial group, led by the NDRC, was tasked to facilitate deleveraging. - Regulators renewed focus to resolve excessive leverage. - Guidelines for debt-equity swaps and creditor committees for claims of multiple creditors were issued. - Assessment: - Guidelines are positive initial steps but lack important details on loss recognition and on operational restructuring of weak firms. - Some SOE restructuring cases contain elements of operational restructuring, but details and time frames are not specified (at least not publicly). - A few debt-equity swaps appear to be equity in name but debt in essence. - Without concerted efforts to slow credit growth, there is an increasing risk of superficial financial restructuring to meet a deleveraging “target” without tackling underlying structural problems ("kicking the can down the road"). ### Recent developments and policy measures — Cleaning up zombies - Government strategy and recent developments: - Strategy is to allow zombies—mainly focused on SOEs—to exit through a menu of options, including asset transfers, consolidation, and liquidation. - Government identified over 2,000 central SOE zombies (with total assets of about 4 percent of GDP) and over 7,000 local SOEs. - Reportedly, 20 percent of the identified central SOE zombies were resolved in 2016, although without details. - Amendments were made to the regulation of nonperforming loans to expedite liquidation of zombies. - Assessment: - Renewed focus to allow exit of zombies is appropriate, but lack of resolution details makes it difficult to assess progress. - Zombie debt is estimated to be moderating due to improving profitability, but resolving them may remain difficult as zombies are about 30 percent more likely to remain so if they are state-owned. ### Recent developments and policy measures — Reducing overcapacity - Government strategy and recent developments: - Government intends to phase out low-technology and high-polluting capacity through market forces, legal, and administrative measures. - Coal and steel sectors set a medium-term target to cut capacity by 10–15 percent of 2015 output and to reduce employment by 1.8 million workers over 3–5 years. - Capacity reduction exceeded the target in 2016 and is on-track in 2017; workforce was down by 30 percent over the last 2-3 years without creating major unemployment (Appendix Table 2). - Government broadened cuts to other sectors such as coal-fueled power and building materials and strengthened the social safety net by a RMB 100 billion restructuring fund. - Assessment: - Reduction targets are frontloaded but could be more ambitious. - Under current cut targets, crude steel capacity would still be close to 2013 levels and account for nearly half of global capacity by 2018–20 due to previously-planned investment. - By some accounts, some closures have resumed production. - Debt in overcapacity sectors has not fallen, suggesting problem loans have not been fully recognized (banks do not classify those problem loans as nonperforming loans). - Government has relied on administrative measures (mergers, work day reductions, window guidance on prices); banks reportedly refinanced firms with payment difficulties (e.g., seven major coal firms in Shanxi). - Impact on production has been muted, with coal and crude steel output up by 3-8 percent since 2015. ### Recent developments and policy measures — Reforming SOEs - Government strategy and recent developments: - SOE reforms highlighted as key priorities with strategy to “integrate naturally” modern corporate governance and leadership of the Communist Party so that SOEs can raise efficiency while meeting national development goals. - Recent measures include: - Consolidating some central SOEs. - Phasing out SOEs’ social functions to workers. - Transferring about ½ percent of state-owned equity to social security funds. - Cutting central SOE losses. - Individually incorporating the subsidiaries of central SOEs by 2017. - Implementing pilot employee stock ownership programs to align incentives. - Bringing in other investors under mixed-ownership pilot reforms and committing to open up sectors (such as travel, medical care, electricity, and power and utilities) to private and foreign investment. - Assessment: - SOE reform implementation has lagged other reforms and has not yet raised growth potential. - Transfer of SOE profits to the budget has been well below the target 30 percent level. - Preliminary classification suggested that only less than 60 percent of SOEs were considered commercially competitive—the category in which SOEs will face direct market competition—raising concerns about achieving significant productivity gains. ### Policy implications — summary and recommended actions - Corporate debt vulnerabilities cut across zombie companies, overcapacity firms, and SOEs; progress is complicated by reactivation of closed capacity, piecemeal exit of zombies, and limited SOE reform progress. - Resolving excessive debt in weak firms requires a holistic, coordinated approach with time-bound actions; government-led process should allow market forces to operate, with complementary operational restructuring to raise efficiency. - Restructuring corporate debt (paragraph 9): - Assessing firm viability: - A targeted asset quality review would be useful. - Reinforce accounting and audit rules for timely and accurate financial information. - Raise standards of appraisers for asset valuation. - Develop efficient credit registers. - Recognizing and allocating losses: - Regulators should strengthen reviews of loan classification, bank capital, collateral valuation, and prudential reporting to foster banks’ proactive NPL resolution. - Operational restructuring: - Operational restructuring plans should be quickly developed for weak SOEs. - Creditors and new equity holders need reasonable assurance restructured firms will be viable on a going-concern basis. - Empirical results support that corporate governance reforms (possibly divestment and a change of management), deleveraging, and tighter budget constraints will help distressed firms return to viability. - Refining the restructuring mechanism: - Creditor committees should align with international best practices (for example, the insolvency principles for multi-creditor workouts, INSOLs), allowing a sufficient standstill period, information sharing, and restraints on debtors from weakening firm value. - Debt-equity swaps should be preceded by an asset quality review and supported by operational restructuring. - Bank loans should be transferred at economic/market value (rather than face value) with independent assessment. - Ensure sufficient resources for bankruptcy courts and professionals on valuation and overcome remaining hurdles in the insolvency framework. - Clarify the role of the public sector: - Joint-ministerial council should be given an explicit mandate to guide expectations, particularly for identification of zombies and SOE reform. - Government should lead assessment of SOE viability, which should exclude state support. - Public creditors such as tax authorities will need to acknowledge a loss in debt restructuring (under strict conditions). - Removing zombies (paragraph 10): - Suspend implicit support and allow banks to manage their claims on zombie SOEs to reallocate assets to productive uses. - Essential social functions provided by zombies should be transferred to fiscal budget. - Nonviable zombies should be publicly identified and subject to greater use of liquidation. - Complement with a clear timetable to resolve all identified zombies within 1–2 years. - Reducing overcapacity (paragraph 11): - Raise the net reduction target and shift reliance from administrative measures to strict enforcement of a larger, more ambitious net reduction of capacity. - Set indicative targets for other overcapacity sectors and more ambitious net reduction terms. - Phase out energy subsidies and impose resources tax: - Greater use of (coal) resource and environmental taxes, reducing energy subsidies, and stricter enforcement of regulatory standards (safety and environmental) will help phase out substandard capacity. - Provide social safety net: - Capacity closures and debt restructuring will involve social welfare costs, such as layoffs (estimated at 2.8 million workers). - Targeted social policies can complement local social security to mitigate social costs. - Reforming SOEs (paragraph 12): - Operational restructuring: - Emphasize operational restructuring of weak SOEs and recognizing (and stopping) losses. - State should neither “window-dress” by merging them with sound SOEs nor encourage creditors to refinance. - Hardening budget constraints: - Reduce credit access and implicit support to SOEs to address existing debt overhang and improve efficiency of new credit. - Enforce transfer of individual SOE profits to the fiscal budget to reach 30 percent by 2020. - Facilitating market entry: - Reduce entry barriers and phase out restrictions that give SOEs a privileged role to level the playing field and make markets more contestable. - Implement commitments to open up protected markets in state-dominated services (such as logistics, finance, and telecommunications) and break up administrative monopolies. ### A market-based monetary policy approach for China (opening points) - Key summary bullets at start of the monetary section: - The implementation of China’s monetary policy is becoming more market-based and the new interest rate corridor has become the main instrument of monetary policy. - There is significant pass through from the PBC’s interest rates to bank lending rates, government yields and other market funding or loan rates. - A VAR analysis with output and prices suggests a significant impact of interest rates on economic activity, and with some lag, on prices. - Taylor rules suggest that the monetary policy stance has been accommodative through early 2017. - The next major reform should be the introduction of an inflation target or a range together with operational (instrument) independence for the PBC. - Context and recent developments: - Money growth (M2) remains the official intermediate target (12 percent y/y for 2017). - PBC uses the 7-day repo rate (DR007) as the appropriate operating target; since about mid-2015 this rate has stayed in the interest rate corridor defined by the PBC’s repo bid rate (lower bound) and the SLF rate (upper bound). - In early 2017, the PBC raised its reverse repo operations bid rate and the rate on its standing lending facilities twice by 10 bps in February and March. - The PBC also lifted rates of longer-term repo operations and medium-term lending facilities, though these are infrequent and mainly offset the liquidity drain from FX sales. *PEOPLE’S REPUBLIC OF CHINA INTERNATIONAL MONETARY FUND* ### 5. There is a significant pass through from the PBC’s policy rates to the average bank ### 5. There is a significant pass through from the PBC’s policy rates to the average bank lending rate ### Pass-through to average bank lending rates - Regressions reveal a high coefficient of about 70 percent for the PBC benchmark rate and about 30 percent for the interbank repo rate on the average lending rate. - The historically-close link reflects that the PBC’s rates used to define the floor (abolished in 2013) for bank lending rates. - The interbank rate spike in 2013 pushed up the average bank lending rate as banks passed on higher funding costs to protect margins, despite the unchanged benchmark lending rate, indicating the increasing importance of the interbank repo rate. ### Impact on short-term government yields - All policy rates are positively correlated with the 2-year government bond yield. - Bivariate VAR analysis shows an almost complete pass-through from PBC benchmark rate changes to government bond yields after about 6 months. - The interbank rate exhibits a lower pass-through to the 2-year government bond yield of 25 percent. ### VAR analysis: output and prices - A quarterly VAR finds a significant GDP response to changes in the interbank repo rate. - Cumulative output loss is about 1.7 percent after two years following a 100bps rate increase. - There is a modest reduction in CPI headline inflation of about 0.2 percentage points after two years, but this response is not statistically significant. - Changes in the PBC benchmark rate have a modest but significant impact on inflation; output declines as well but the impulse response falls within the two-standard error band. - Repo rate changes have a significant impact on both output and prices when CPI inflation is replaced with the GDP deflator in the quarterly VAR. - Note: To account for significant autocorrelation in the lending rate, its lagged values (2) were included in the regressions and results were robust across specifications. ### Quantitative measures of monetary stance (M2, aggregate loans) - Changes in M2 or aggregate loan volume do not produce a significant response in either GDP or inflation/deflator in most VAR specifications. - When M2 or bank loans are added to the VAR with output, prices and policy rates, or replace the latter, their changes generally do not have a significant impact on output and prices. - Positive policy rate changes (rate hikes) trigger a decline in M2 and bank loans. - These findings align with other studies finding that interest rate changes have substantial impacts on activity and inflation, while shocks to M2 or lending levels alone do not. ### PBC’s interest rate corridor and operational instruments - The PBC’s new interest rate corridor encompasses key effective policy rates: the PBC’s 7-day repo rate (frequent liquidity provision), the rate on excess reserves, and the rate on the 7-day standing lending facility (SLF). - Since the “de facto” introduction of the corridor (mid 2015), the PBC has steered the 7-day interbank repo rate (DR007) close to its own repo rate but has allowed somewhat more market rate volatility within the 100bps corridor defined by the PBC’s 7-day repo and SLF rate. - The PBC could declare the 7-day interbank repo rate as its new operating target and phase out publication of benchmark rates, which are difficult to interpret after full liberalization of bank lending and deposit rates. ### Assessing appropriate policy levels and neutral real rate - Estimating appropriate policy rates in China is difficult because price and output stability coexist with exchange rate, capital flow and financial stability considerations. - Over the sample period 2005-16, both the short-term repo rate and headline inflation averaged slightly below 3 percent implying a real (CPI-deflated) rate of zero on average. - These historically low real rates were influenced by very high national savings, financial repression, significant exchange rate appreciation, and gradual liberalization of the capital account and financial markets, limiting their usefulness for current/future neutral rate estimates. - He and others (2015) estimate the long-term neutral rate in the range of 4.0 to 4.5 percent. - By comparison, U.S. longer-term (beyond 2020) real rate is currently seen by the U.S. Federal Reserve Board as close to 1 percent. - It seems likely China’s longer-term neutral real interest rate should be considerably above zero and somewhat above levels observed in other major advanced countries as China continues catching up with faster output growth. ### Policy stance as of early 2017 and Taylor rules - Regressing the PBC’s benchmark lending rate on the output gap and inflation provides a reasonably good fit once persistence in effective policy rates is modeled with an appropriate lag structure. - Simple Taylor rules point to an expansionary monetary stance at the end of 2016, with policy being somewhat more accommodative than implied by the rule (by 100 bps). - The fit for the repo rate (used as a “shadow policy rate” over the past decade) was worse; the output gap did not have a significant impact on the repo rate, possibly reflecting occasional repo rate hikes for financial stability concerns. - Taylor rules implicitly assume a constant neutral rate, which may not have been the case over the past decade; China’s neutral real interest rate may rise as financial liberalization continues and remaining credit constraints for households and private firms (and preferential credit conditions for SOEs) are removed. A lower national savings rate would likely work in the same direction. ### Recommendations for transition to a market-based monetary policy framework - The PBC should gain operational (instrument) independence for conduct of monetary policy: - The PBC’s benchmark lending/deposit rates are determined by China’s State Council; the PBC should be fully and independently in charge of setting the rates for its new interest corridor. - The PBC could declare the 7-day interbank repo rate as its new operating target and phase out publication of benchmark rates. - Operational independence should be accompanied by a more narrowly defined mandate: the State Council would define long-term goals (such as a medium-term inflation target) and the PBC would be accountable for reaching these goals, implying greater transparency about plans, objectives and policy decisions. - While the PBC could play a major role in macroprudential policy, there should be a clear separation with monetary policy, recognizing occasional interactions. - Monetary policy should rely on price-based instruments, primarily the new interest rate corridor. - Pricing and access to the PBC’s lending facilities should be based on clearly defined collateral rules and not supervisory criteria. *IMF staff analysis as presented in the source PDF.* ### 1. The capital account has been opened gradually amid evolving policy priorities. High ### 1. The capital account has been opened gradually amid evolving policy priorities. High ### Opening and sequencing of the capital account - High domestic savings searching for yield and diversification, interest rate differentials and exchange rate expectations have driven capital in and out of China. - Major policy drivers of capital flows over time include “Going Global”, “One Belt One Road”, “RMB internationalization”, and “Made in China 2025”. - General sequencing and milestones: - FDI liberalization: FDI into several sectors was liberalized in the early 1990s, but services and some strategic sectors remain closed to FDI. - ODI (Overseas Direct Investment): opened gradually starting with the “Going Global” initiative in 1999, partly to secure commodities, climb the value chain, and alleviate RMB appreciation pressures. - Portfolio investment liberalization: - Access to select Chinese stock exchanges for qualified foreign institutional investors starting in 2002. - Extended to the Chinese interbank market for foreign sovereign and private entities in 2015 and 2016 respectively. - Domestic qualified institutional investors (QDII) permitted to make global portfolio investments starting 2007. - Cross-border lending and borrowing: partially relaxed during 2008-10. - Chinese commercial banks permitted to lend abroad in 2008. - Qualified domestic enterprises allowed to lend to overseas subsidiaries in 2009 and to borrow short-term in 2010. - In response to large capital outflows in 2015/16, authorities allowed some currency depreciation, used FX intervention, and applied a wide range of measures to stem outflows and depreciation pressures. - Overall: Capital inflows were liberalized before outflows, and FDI flows before portfolio and other flows. ### De jure versus de facto openness and recent outflows - De jure indicators diverge: - Chinn-Ito index shows that China’s capital account is very closed. - Quinn index shows that it has gradually opened. - Even the most generous indicator shows China’s capital account openness still lags that of advanced economies and large emerging markets. - De facto openness is greater than de jure measures indicate, evidenced by the surge of capital outflows in 2015 and 2016. - Flow-scale comparison: capital flows into China (as a percent of GDP) are of similar magnitude as those into comparable emerging markets. - Stock basis: because China commenced opening later, on a stock basis capital account openness still lags peers. - Net capital outflows: - Reached a record of around $648 billion in 2015 (5.8 percent of GDP), mostly driven by other investments and errors and omissions. - In 2016, outflows amounted to almost US$64 0 billion (5.7 percent of GDP), driven by all categories of the financial account and errors and omissions. - These outflows resulted in significant depreciation of the RMB, partially cushioned by FX intervention. ### Assessment of liberalization, reforms, and tightening of CFMs - Sequencing broadly in line with the IMF’s integrated approach, but supporting reforms to build resilience to capital flow volatility did not keep pace. - Slow progress on needed reforms (Appendix II): - Financial stability risks have accumulated and complicate monetary policy; modest interest rate increases could stress smaller and over-leveraged financial institutions. - Gaps in supervision/regulation and in the macro prudential framework. - High corporate debt and insufficient SOE reform raise systemic risk concerns. - Quasi-fiscal pressures from local governments have contributed to increasing debt and vulnerabilities. - Faster progress on exchange rate flexibility and development of the monetary policy framework would have strengthened management of capital flows. - Given limited reform progress and growing vulnerabilities, authorities relied heavily on tighter enforcement of existing restrictions during recent bouts of capital outflows (Appendix I). - Since mid-2016, authorities stepped up enforcement of existing CFMs and tightened a few others, contributing to curbing outflow pressure since late 2016. - Subsequently, PBC intervention subsided and headline reserves increased modestly to stabilize at slightly above USD 3 trillion. - Other factors (better growth outlook in early 2017, anchored exchange rate expectations) also played an important role. - Concerns about tightened CFMs: - Macro policies should play a more central role; CFMs cannot substitute for macroeconomic adjustment. - Effectiveness and costs: - Long-term effectiveness of CFMs is questionable; outflows largely driven by residents adept at moving capital (errors and omissions were 2 percent of GDP in 2016 vs annual average of 0.7 percent of GDP in 2010-14). - Monitoring and enforcing extensive CFMs impose heavy administrative burdens. - Administrative measures could impede economic efficiency (e.g., curtailing ODI in non-core businesses may lead to suboptimal government decisions). - Inconsistent enforcement across time and locations creates uncertainty and can harm the business climate; risk of breaching IMF obligations not to restrict current international payments and transfers. - Use of moral suasion and variable enforcement could erode confidence in the regulatory system and damage the business climate. - Risk that reliance on CFMs delays reforms and increases risks of asset price bubbles and financial instability. ### Policy recommendations on CFMs and reforms - When CFMs are warranted: - Implement consistently over time and across locations. - Communicate clearly to market participants. - Implement transparently, preferably through written rules; avoid reliance on window guidance. - Design and implement CFMs so they do not restrict current international payments and transfers. - Authorities should expeditiously implement key reforms to strengthen resilience to capital flows: - Financial sector reforms, including supervision and the macro prudential policy framework. - Further strengthen the monetary policy framework; regulatory/supervisory tightening in the financial sector starting early 2017 is critically important and should continue. - SOE reforms to increase efficiency and contain financial sector vulnerabilities; impose hard budget constraints to dispel perceptions of state guarantees. - Liberalization should proceed cautiously, assessing benefits and costs while accelerating supporting reforms: - Pace and phase liberalization in tandem with necessary reforms. - Avoid liberalizing inflows solely to provide immediate relief against outflow pressure. - Areas where benefits may outweigh risks include FDI in service sectors, financial services, and health. - Opening to foreign entities in the financial sector, including credit rating agencies, could improve capital allocation and financial sector efficiency. - Consider eliminating or gradually reducing the reserve requirement for onshore FX hedging to support a more flexible exchange rate; if derivatives give rise to stability concerns, consider macro prudential measures. - The “impossible trinity” challenge: - As the economy grows, it is desirable to move toward an independent monetary policy. - Authorities must choose between the extent of capital account openness and exchange rate flexibility. - Greater exchange rate flexibility helps absorb shocks and reduces the need for a wide range of CFMs. - Deep reforms enabling greater macroeconomic policy adjustment should be expedited to benefit from greater financial openness while mitigating risks. ### Appendix I — Selected measures tightened since mid-2016 - ODI: closer scrutiny of certain activities (e.g., investment in non-core businesses); SAFE circular (January 2017) requiring companies to explain to banks sources and purposes of investment funds and present board resolutions; PBC urged commercial banks to tighten verification of authenticity of certain ODI. - Offshore RMB lending by non-financial institutions: commercial banks urged to enforce existing rules strictly (including prudential lending limit of 30 percent of lender’s equity); banks must examine suitability of borrower business scale to loan size and the authenticity and reasonableness of use of outbound loan. --- ### Snapshot of residential real estate risks and policy guidance (selected findings) - Real estate rebound and vulnerabilities: - Real estate investment grew from about 4 percent of GDP in 1997 to the peak of 15 percent of GDP in 2014; residential investment accounted for over two-thirds of total housing investment. - Housing cycles are more pronounced in top-tier cities but smaller cities constituted over half of residential real estate investment. - Market distortions: local governments’ control on land supply and reliance on land sales (land sales accounted for about 30 percent of local government revenue in 2016); demand-side attraction of housing as investment due to robust capital gains, high savings, negative real deposit interest rates, lack of alternative financial assets, and capital account restrictions. - Recent developments: - After a slowdown in 2014-2015, the market rebounded sharply; tightening measures since late 2016 seem to have dampened activity but house prices and sales remain strong, particularly in smaller cities. - Residential property sales remain at 12 percent y/y in May 2017 in terms of volume of floor space sold. - Nationwide housing inventory ratio declined to about 18-20 months from 30 months at the peak in 2014. - Down payment ratios: 40 percent of buyers of a first home have a down payment ratio of 25 percent or higher. - Average loan-to-value (calculated as new mortgage loans to property sales ratio) increased from around 15 percent in 2012 to 48 percent in - Risks and recommended policies: - Downside risks significant: further price rises, spread to smaller cities, and bubble expansion would raise likelihood and costs of a sharp correction, weaken growth, undermine financial stability, reduce local government spending room, and spur capital outflows. - Continue and expand macro-prudential and city-specific policies; expand toolkit to include more active use of debt servicing-to-income (DSTI) caps and capital requirements on banks’ exposure to real estate. - Longer-term measures: introduce recurrent property taxes, resolve land supply constraints in large cities, mitigate local governments’ reliance on land sales, and accelerate reforms of social security and “hukou” systems. *Prepared by Calixte Ahokpossi (SPR); Ding Ding and W. Raphael Lam (residential real estate section).* ### 2016. Household balance sheets have remained ### 2016. Household balance sheets have remained ### Household balance sheets and indebtedness - Household debt increased from less than 20 percent of GDP in 2008 to more than 40 percent, with mortgages accounting for more than half of outstanding household debt. - Household debt ratio is still far below the average OECD level of 102 percent of GDP, but is already higher than in some other large emerging economies. - Underlying demand in major cities remains robust and household balance sheets are described as still solid, though buffers are eroding. ### Indicators of overheating and government measures - As signs of overheating emerged, the government implemented a range of measures differentiated across cities, including: - tighter down-payment requirements; - home purchase restrictions; - higher mortgage rates; and - financing restrictions for property developers. - Evidence these measures have been effective: - moderation in price rises (in sequential month-on-month basis) in large cities; - moderation in new mortgage loans and real estate sales; - empirical estimations confirm changes to down payment requirements have been effective in dampening price cycles (Annex I). - Given modest recovery in real estate investment, robust demand in major cities, and still solid household balance sheets, the impact of overheating should be contained, especially if government action continues and broadens. ### Risks in the real estate market - Main near-term risks: - house prices rising further beyond “fundamental” levels; - bubble expanding to smaller cities, increasing likelihood and costs of a sharp correction. - Price deviations: - Nationwide house prices deviate only 5 percent from their long-term trend (based on HP filtering). - Price deviation for tier-1 cities is about 10-15 percent. - Historical precedent: - In previous cycles, this level of overvaluation was followed by a slowdown in real estate activity. - Impact of a sharp correction: - A correction of house prices by 10-15 percent (roughly the magnitude in previous cycles) would reduce GDP growth by around 0.9 percentage points through the impact on property investment and household consumption. - Real estate investment may start to slow in the coming months reflecting intensified regulatory and supervisory tightening. - A sharp correction would weaken local public finances; even if local governments maintain prices by restricting land supply, tighter government spending would have knock-on effects on growth. - Financial stability risks from a sudden correction: - Likely increase in impaired loans and deterioration in profitability and capitalization of financial institutions. - Impact through the mortgage channel is likely limited given still low household leverage. - Highly leveraged small real estate developers and small city-level banks exposed to weak developers are likely the weakest link in the short run. - A sharp decline in valuations and household and corporate borrowing would have knock-on effects on the real economy, exacerbating asset quality problems and amplifying the initial shock. - Banks and other intermediaries would be simultaneously hit by credit and market losses. - China has never experienced a significant decline in house prices, so the scenario is uncertain. - The government retains much control over real estate markets and the financial system and will likely take further administrative measures to stabilize in response (as it did in 2015). ### Policy implications and recommendations - Continue macro-prudential policies tailored to local conditions as first line of defense: - Differentiated LTV limits by borrowers (first-time home buyers versus investors) and by regions are appropriate to target riskier segments. - Caution: changes in LTV limits can be pro-cyclical and allow borrowers to borrow more during boom cycles, potentially increasing future credit losses. - Expand the macro-prudential toolkit: - Consider more active use of debt servicing-to-income (DSTI) caps. - Current cap is 50-55 percent; recommendation is to gradually tighten to international norms of below 45 percent and extend to other types of household loans including loans from non-bank financial institutions. - Stress testing of household debt servicing capacity to interest rate and income shocks to gauge potential risks in adverse scenarios. - Consider sectoral capital requirements for banks: - Increase risk weights or loan given default (LGD) floors on banks’ exposure to the real estate sector. - Tighten these requirements during market booms to increase cost of funding for property developers and build additional buffers. - Step up efforts to collect information beyond aggregate credit and house prices, such as indicators on borrowers and speculative activities. - Reduce reliance on administrative measures over the medium term: - Home purchase restrictions and funding restrictions are effective short-term but can have excessively abrupt impacts, more distortions, and circumvention compared to market-based or macro-prudential measures. - Home purchase restrictions tend to disproportionately affect new migrants, undermining urbanization efforts. - Introduce recurrent property taxes: - Overcome hurdles on registration of properties and legislation procedures; a nationwide system on property registration by 2017 is welcomed. - Benefits: - dampening impact on house price volatility (cross-country evidence); and - provide revenue sources for local governments to finance local public services and avoid excessive dependence on land sales, thereby dampening boom/bust cycles. - Implementation can be gradual (but decisive) given potentially large economic impact. - Increase real estate supply in line with demand, especially in higher-tier cities: - Increase land supply and build higher-density housing in major cities to attract migrants and facilitate smooth market transition. - Complement with reforms on social security and household registration (hukou) to give migrants greater access to public services. - Address macroeconomic drivers: - Reduce excessive domestic savings and continue gradual and cautious opening of the capital account to help reduce pressure on housing prices and propensity for recurrent asset price boom/bust cycles. - De-emphasize quantitative GDP growth targets: - Quantitative targets foster short-term, low-quality stimulus (infrastructure spending, real estate activity, credit) during downturns to meet targets, creating perception of housing as an aggregate demand instrument and reinforcing boom/bust cycles. - De-emphasizing targets would align real estate dynamics more with fundamental demand and supply, allowing prudential policies greater role in guarding against macro-financial risks. ### Empirical analysis (Annex I) — panel regression findings - Model and sample: - Panel regression using methodology in Igan and Loungani (2012). - Annual data for 68 cities for which the National Bureau of Statistics reports monthly house prices. - Provincial-level proxies used for bank credit to the private sector and local government fiscal deficit where city-level data unavailable. - Key regression findings: - Changes to down payment requirements have a significant impact on house prices, especially in tier 1 cities. - A tightening of down payment requirement by 10 percentage point in tier 1 cities corresponds to a decline in real residential house prices by 3.5-4.5 percent, other things being equal. - Increasing land supply per capita is found to have a negative impact on house price increases. - Selected coefficient estimates and statistics (dependent variable: real residential house prices, change): - Affordability, lagged: -0.1419**, -0.1568***, -0.1726***, -0.2117*** - Income per capita, change: 0.0225, 0.0126, 0.0225, 0.1424** - Bank credit, change: 0.0938, 0.0905*, 0.0900*, 0.0476 - Downpayment ratios, change, lagged: -0.0709*, -0.0431, -0.0221, -0.1313 - Interacted with Tier-1 dummy: -0.3528***, -0.4680***, -0.4369*** - Interacted with Tier-2 dummy: -0.0252, -0.0283, -0.0503** - Land supply per capita, change, lagged: -0.0023, -0.0023, -0.0055*, -0.0020 - Interacted with Tier-1 dummy: 0.0850***, 0.0621** - Local mortgage rate, change: 0.0107, 0.0121**, 0.0116**, -0.0029 - Local government fiscal balance, change: -0.4279, -0.3997, -0.4450, 0.0065 - Stock prices, change: 0.0744***, 0.0724***, 0.0706*** - RMB per USD, change: -2.0645***, -2.1142***, -2.1473*** - Interbank interest rate, change: -0.0170, -0.0194*, -0.0230*** - Dummy for housing cycle: -0.0661***, -0.0652***, -0.0635*** - Constant: -0.7315, -0.8016***, -0.8748***, -1.0282*** - Year dummies (selected): 2008 -0.0701***; 2009 -0.0918**; 2010 -0.0897***; 2011 -0.0916***; 2012 -0.1548***; 2013 -0.1071***; 2014 -0.2207***; 2015 -0.2289*** - R-sq (overall): 0.2974, 0.3086, 0.3026, 0.3928 - Number of observations: 418 - Number of groups: 68 - Significance notation: ***, **, * denote significance at the 1, 5, and 10 percent level, respectively. *Source: PEOPLE’S REPUBLIC OF CHINA — International Monetary Fund (2017).* ### 6. Even after expanding the perimeter of augmented aggregates, all of the expansion of ### cr17248 - 6. Even after expanding the perimeter of augmented aggregates, all of the expansion of ### On-budget expansion and augmented aggregates - Even after expanding the perimeter of augmented aggregates, all of the expansion of China’s fiscal stance in 2014-16 was on-budget. - On-budget net borrowing expanded by 2¾ percent of GDP in 2014-2016, meaning off-budget activities are estimated to have remained broadly unchanged after expanding the perimeter of augmented aggregates. ### Treatment of quasi-fiscal units (GFSM 2014) - GFSM 2014 classification: - Entities classified on public vs private sector units; classification inside public sector based on government control using GFSM 2014 Box 2.2 eight indicators. - Public units are classified as market or non-market based on nature of activities; public non-market producers should be included in government; public market producers are classified outside government as public corporations. - GFSM 2014 2.69-75 test: operating revenues must exceed 50 percent of production costs to classify as market producers. - Entities outside government can have part of activities reflected in government accounts if acting on behalf of or under instruction of government (GFSM 2014 3.28 on rerouting). - Treatment of SCFs and GGFs: - These funds should be regarded as public sector units since the government (National Development and Reform Commission, Ministry of Finance or local governments) retains control (even if the government is a minority shareholder). - These funds are akin to venture capital funds, but do not appear to make market-based decisions in exclusive pursuit of commercial returns; staff concludes they should be included in government. - Treatment of social capital contributions as debt is based on GFSM 2014 7.143, 7.150 and 7.166. - Perimeter of government in IMF surveillance: - IMF surveillance frequently extends coverage of fiscal data to include parts of the wider public sector (examples provided for other countries). - Many countries, particularly in the E.U., use a definition of general government consistent with GFSM and thus the perimeter goes well beyond budgetary units. ### Fiscal space: narrow versus augmented definitions - Key headline debt measures: - Narrowly-defined government debt was 37 percent of GDP by end-2016 (official on-budget general government debt). - Median EME government debt: 48 percent of GDP (for comparison). - Broadly defined debt (Augmented+) around 62 percent of GDP in 2016 (adds remaining LGFV debt and SCF and GGF government liabilities to general government debt). - Financing environment and needs: - Participation of foreigners in domestic sovereign bond market limited; external financing requirements around 7 percent of GDP. - Gross financing needs projected to stay elevated at above 12 percent of GDP both in 2017 and 2018. - Fiscal space assessments: - Narrow definition: - If fiscal space is defined as the ability to use fiscal policy temporarily without losing market access, China certainly has space, which could support reform if limited in time to avoid faster debt accumulation. - A permanent 0.9 p.p. of GDP expansion over the 2016 primary balance stabilizes debt at around 100 percent of GDP (75th percentile of AEs’ debt-to-GDP). - China could likely tolerate significantly higher levels of government debt than almost all EMs due to high savings, strong external position, capital controls, strong state control and confidence. - Augmented definition: - Under the “augmented” definition of debt/deficit, China would have only very limited space for a temporary loosening, and negative space for a permanent expansion. - Staff estimates “Augmented+” government debt around 62 percent of GDP in 2016, rising to 92 percent of GDP in 2022. - Under the augmented definition, the primary balance needs to be improved by 4 percentage points of GDP by 2022 to stabilize augmented debt at around 100 percent of GDP. - Ageing and structural balance: - Ageing could increase the yearly structural deficit by an average of 2.5 percent of GDP under current policies until 2050. - Applying this increase to China’s projected primary balance after 2017 leads to government debt under the narrow definition of 57 percent of GDP in 2022 compared to 42 percent without this adjustment. - The 2022 primary balance is nearly 3 percentage points higher than its debt stabilizing level. - The ageing calculation corresponds to an additional liability of around 90 percent of 2015 GDP in present value terms (constant equivalent flow of additional health and pension expenditures under UN medium-variant population projections against baseline of constant population for 2015-2050). - Contingent liabilities and shocks: - Under the DSA’s stylized contingent liability shock of 10 percent of bank claims against non-public entities, China’s debt reaches 69 percent of GDP in 2022, compared to 42 percent of GDP (narrow definition) without this shock, and does not stabilize. ### If fiscal stimulus is needed: composition matters (Box 2) - Two stimulus packages considered for restoring growth under shocks: - Investment-based package: - Entirely based on scaling up investment. - Assumes output multiplier decreases as investment exceeds baseline levels (diminishing multiplier). - Rebalancing-based package: - Based on revenue and spending rebalancing measures, capped at 2 and 1 percent of GDP respectively; any additional stimulus needed comes from investment. - Rebalancing measures cannot be scaled up indefinitely. - Findings: - Under a temporary mild shock the two packages would roughly have the same size. - If the shock is large and protracted, a very large amount of investment would be needed to lift GDP back to baseline due to diminishing multiplier; instead, a smaller package of rebalancing measures would suffice and have a smaller impact on debt. - Assumptions on compensation and permanence (IMF staff estimates): - The rebalancing package assumes only 30 percent of cost of revenue and spending measures is compensated in 2017, 70 percent in 2018, and 90 percent in 2019. - The permanent cost is fully compensated in 2020. ### Policy implications and recommendations (Section C) - Authorities’ progress: - Authorities have taken important steps to stem off-budget borrowing in the last two years; intensified efforts, new laws and regulations, and stepped up enforcement. - Bringing all LG infrastructure spending on-budget requires further fiscal framework improvements in three key areas: - Planning: - Formulate a medium-term fiscal framework, including a medium-term integrated capital financing, that includes all LG infrastructure spending. - Create an incentive system for LG units using a combination of carrots (higher bond quotas, expedited approval of infrastructure projects) and sticks (enforcement, legal actions). - Monitoring: - Increase capacity and resources at the provincial government level for effective monitoring of lower level units. - Crucially, enlarge the perimeter of monitoring. - Transparency: - Publish detailed and easily accessible budget accounts and overall debt levels by province at least yearly, including contingent liabilities by type. - Publish medium-term fiscal projections and an assessment of the riskiness of broadly defined public debt by province or district regularly. ### Intergovernmental fiscal reform in China: main findings and recommendations - Main findings: - China has the largest share of local government spending in the world; revenue autonomy at the local level is very limited. - Public service delivery is improving and more people covered by social safety net, but social spending on public education, health care and social assistance is low and regional disparities are growing. - Large and growing unfunded mandates caused local governments to accumulate around 40 percent of GDP in debt by 2015, of which 22 percent was transferred to the government’s balance sheet. - An intergovernmental fiscal reform plan was announced in 2016; a new budget law introduced in 2014 and efforts stepped up to ensure all fiscal activities conducted on-budget. - Policy recommendations (intergovernmental focus): - Determine appropriate level of decentralization for social spending to ensure development and rebalancing. - Pensions and unemployment insurance: - Centralize policies and financing, with some local autonomy for administrative functions. - Centralization reduces cost of risk pooling, improves portability, ensures equal benefits across regions. - Taxes and transfers: - A recurrent market value-based property tax would be ideal for local governments. - Consider allowing provinces to impose a surcharge in addition to the national personal income tax. - Rules-based general transfers could eventually replace revenue-sharing and tax rebate transfer programs. - Increase size of equalization grants to reduce fiscal disparities. - Borrowing quotas: - Increase borrowing quotas for local governments to ensure all off-budget fiscal spending can be brought on budget. _Italic: IMF staff report content as provided in the source document._ ### 10. The broad principles for optimal assignment of revenues are the same as for spending. ### 10. The broad principles for optimal assignment of revenues are the same as for spending. ### Principles and trade-offs - Decentralization of revenues can better match the tax system to local preferences and promote more accountability for policy makers. - Centralization of taxation is warranted by economies of scale, risk-sharing, mobility of tax bases and externalities (IMF 2009). ### Personal Income Tax (PIT) - Proposal: China could consider allowing provinces to impose a surcharge in addition to the national personal income tax (PIT). - Design features: - Provincial surcharges could be allowed within a centrally-approved upper limit (of say 5 to 10 percent). - This would increase provincial autonomy while efficiently increasing the revenue contribution of the PIT, which is low by international standards. ### Value-Added Tax (VAT) - Principle: The national scope of the VAT should be maintained. - Practical issues: - Significant difficulties exist in monitoring border flows between local jurisdictions. - Country experiences with subnational VATs have been mixed (Perry 2009). - Recommendation: Review revenue-sharing arrangements to reduce compliance costs for taxpayers with multiple business locations. - Consider simple allocation rules such as population (Germany) or aggregate consumption (Japan). ### Property tax - Recommendation: Adopt a recurrent market-value based property tax as an ideal tax for local governments. - Rationale: - Closely tied to public service delivery through property values. - Base is immobile. - Highly visible and payable by households on a recurrent basis, improving accountability of local officials. - Broadly viewed as progressive because high-income households tend to have higher property wealth (Norregaard 2013). - Design: Tax base should be defined following national guidelines; local governments could set tax rates within bands set by the central government. ### Intergovernmental transfers (section 11) - Need: Review current system of intergovernmental transfers. - Problems with current system: - Revenue-sharing and tax rebate programs reduce clarity and predictability and contribute to pro-cyclicality of local government funding. - Fiscal disparities across areas remain and should be reduced further. - Targeted transfers are fragmented and complex: the current system has over 200 different transfers, increasing administrative costs at central and local levels (World Bank and DRC 2014). - Equalization grants currently account for less than half of total transfers (World Bank and DRC 2014). - Recommendations: - Move toward a rules-based general transfer instead of the current revenue-sharing and rebate programs to improve clarity, predictability, and reduce pro-cyclicality. - Increase the size of the funding pool for equalization grants. - Rationalize and simplify targeted transfers. - Establish stronger conditionality on outputs or quality of services rather than inputs to promote efficiency in public service delivery. ### Local government borrowing and fiscal transparency (section 12) - Principle: Borrowing quotas for local governments should be set high enough to ensure all fiscal spending can be brought onto the budget. - Rationale: - Despite gains in management of local government debt, there remains a need to improve transparency, resource allocation and medium-term fiscal planning. - All fiscal activities by LGFVs and government-guided funds should be fully reflected in the budget to make explicit fiscal and aggregate demand implications of current policies. ### Inequality context and implications for revenue assignment and fiscal policy (sections A–E) - Key facts: - China’s rapid economic growth since 1990 lifted an estimated 731 million people out of poverty by 2013. - The Standardized World Income Inequality Database (SWIID) estimates the Net Gini coefficient for China at 50 points as of 2013. - Income for the bottom 10 percent rose by as much as 63 percent between 1980 and 2015. - The population in poverty fell from 88 percent in 1981 to 2 percent in 2013 (headcount ratio, $1.90 a day at 2011 PPP). - Drivers of inequality: - Structural factors—demographics, the urban/rural divide, and education/skills—are major drivers and are likely to push inequality higher. - The rural-urban gap explains a large share of inequality; its contribution has been declining due to rapid urbanization and specific government programs. - Projections and scenario analysis: - A cross-country panel regression attributes most of the rise in inequality until 2010 to structural factors (urbanization and demographic changes). - Using projections of structural variables and keeping other variables constant, inequality is predicted to rise further due to demographic changes. - Under an illustrative proactive policies scenario—gradual adoption of fiscal policies bringing China to levels of the most proactive G7 countries by 2050—inequality flattens out after 2010 rather than increasing as under unchanged policies. - Spending on social protection and health together decrease the Gini by 3 points compared to that observed in 2010 (after controlling for redistribution). - Fiscal policy recommendations to reduce inequality: - Revenue-side: - Increase reliance on personal income tax and enhance its progressivity. - Lower the current high basic personal allowance, transform it into a tax credit, and redesign tax brackets to increase contributions from middle and high income households. - Remove imputed minimum earnings for social security contributions, which have been found to be regressive. - Consider adopting a recurrent market-value based property tax. - Expenditure-side: - Boost social spending on education, health and social assistance—China still lags other emerging economies and OECD countries. - Address unequal provision of public services—liberalize the hukou residency system to allow more migrants to contribute to and benefit from the social safety net. *IMF staff summary based on the provided chapter content.* ### 15. Provincial and regional inequalities in public service provision and access have also ### 15. Provincial and regional inequalities in public service provision and access have also ### Main findings on provincial and regional inequalities - Provincial and regional inequalities in public service provision and access have been growing in recent years, with richer provinces outpacing poorer areas. - The recently announced reform plans by the State Council aim to reduce regional disparities by increasing transfers to poorer regions. - Recentralization of social insurance is expected to improve equality, risk sharing and labor mobility. ### Policy recommendations and reform priorities - Increase the pool of funds used to finance equalization grants. - Move from the current ad hoc process used in annual budget preparation to a more rules-based system for intergovernmental transfers. - Reform the overly complex system of conditional transfers, with a stronger focus on outcomes as opposed to inputs, to support improvement in public service delivery. - Recentralize social insurance to strengthen equality, risk sharing and labor mobility. ### Appendix I — Cross-Country Regression (model specification and sample) - Regression form: - y_it = X'_it β + P'_it γ + Z_it δ + μ_i + ε_it - where X is a vector containing the structural variables, P includes policy variables, Z are the controls and μ the country-fixed effects. - Sample includes: Argentina, Australia, Brazil, Bulgaria, Canada, China, Denmark, Hungary, India, Italy, Japan, Mexico, Netherlands, New Zealand, Norway, Panama, Philippines, Poland, Portugal, Korea, Singapore, Spain, Sweden, Switzerland, Thailand, the United Kingdom, the United States of America, and Venezuela. - Number of Observations: 646 - Adjusted R-squared: 0.962 - Country Fixed Effects: Yes - t statistics in parentheses - + p<0.1, *p<0.05 ### Appendix I — Selected regression results (Net Gini Coefficient) - Share of Employment in Services: -0.356* (-2.42) - Share of Employment in Services Squared: 0.00228+ (1.76) - Share of Employment in Industry: -1.842* (-11.11) - Share of Employment in Industry Squared: 0.0283* (9.37) - Age Distribution D1: 71.59* (7.80) - Age Distribution D3: -9.693* (-8.12) - Age Distribution D3 (squared?): 0.389* (8.43) - Share of Population living in Urban Areas: 1.756* (15.46) - Share of Population living in Urban Areas squared: -0.0134* (-15.98) - Share of Population without Education: 0.0469 (1.04) - Share of Population with some Primary Education: -0.0238 (-0.75) - Share of Population with some Secondary Education: -0.00337 (-0.12) - Public Social Protection Expenditure as Share of GDP: -0.0305 (-0.86) - Public Health Expenditure as Share of GDP: -0.670* (-3.22) - Public Health Expenditure as Share of GDP Squared: 0.0808* (3.72) - Absolute Redistribution: -0.175* (-4.65) - Property Tax Revenue as a Share of GDP: -0.617* (-3.18) - Individual Income Tax Revenue as Share of GDP: -0.0426 (-0.58) - Top Personal Income Tax Rate: -0.0553* (-4.76) - Relative GDP per Capita: -9.715* (-3.30) - Relative GDP per Capita squared: 5.278* (3.52) - Trade Openness: -0.0171* (-3.50) *Italic line: PEOPLE’S REPUBLIC OF CHINA — INTERNATIONAL MONETARY FUND* --- _Source: https://www.imf.org/-/media/files/publications/cr/2017/cr17248.pdf_