## APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS

## Source details

**Canonical URL:** [APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS](https://www.imf.org/-/media/files/publications/wp/2018/wp18277.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2018/wp18277.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2018/wp18277.pdf.json)

---

### Major findings — national and structural context
- China’s national savings rate historically: around 35–40 percent of GDP since the 1980s; surged to a peak of 52 percent of GDP in 2008; declined to 46 percent in 2017.
- Global comparisons: global average savings rate of 20 percent; 15 percent for emerging economies.
- Internal demand structure (percent of GDP):
  - Investment ratio at 43 percent of GDP.
  - Private consumption at 38 percent of GDP.
  - Global average private consumption is 60 percent of GDP.
- Comparison statement: If Chinese households consumed comparably to Brazilian households, Chinese consumption levels would be more than double (statement in source).
- Current account: surplus down to 2 percent of GDP since the GFC; high domestic savings have been absorbed by elevated investment.
- Sectoral composition shifts:
  - Households: household savings peaked at 25 percent (of disposable income) in 2010; household savings at 23 percent of GDP (current, as described) are 15 percentage points higher than the global average.
  - Corporates: corporate savings were low in the 1990s, surged in the 2000s, and after the GFC narrowed back to be in line with the global average.
  - Government: fiscal (budgetary) savings are volatile and on average a small portion of national savings; in recent years China’s fiscal savings have been higher than the global average, reflecting high capital spending.
- Household disposable income share of GDP: fell from 67 percentage points in the late 1990s to 58 percentage points in 2008; increased to 61 percentage points in 2014.
- Repressed deposit rates historically about 4 percent during high-growth period.

### Household savings — historical phases and empirical magnitudes
- Three phases of household savings (in percent of disposable income):
  - 1980s: rose from 5 to 20 percent (following the one-child policy and agricultural de-collectivization, with a temporary dip in the late 1980s).
  - 1990s: rose to 25 percent (post-SOE reform and social safety net transformation).
  - 2000s (post-WTO entry): rose further to 30 percent during an export-driven boom; since 2012 household savings have plateaued and begun to decline.
- Aggregate household savings measured in percent of GDP increased less than savings rate due to falling household income share of GDP.

### Drivers analyzed and quantitative contributions
- Demographics:
  - Fertility decline: fertility fell from about 6 to below 3 in 1970s policies, then below 2 after the one-child policy in 1980.
  - Youth-dependence ratio: China has one of the lowest youth-dependence ratios in the world (chart noted in source).
  - Mechanisms: fewer children reduce child-related expenditures and increase need to save for old-age support.
  - Micro evidence: households with twins save about 10 percentage points less than households with one child (Choukhmane et al., 2014).
  - Quantitative result from overlapping generations model (Curtis et al., 2015): demographic shifts alone account for about half of the rise in household savings (holding income growth and interest rates constant). Adding income and interest rates allows the full model to explain the broad savings trend.
- Social safety net and precautionary savings:
  - Large-scale layoffs during SOE reform: over 27 million during 1997–2002.
  - Health coverage and pension changes:
    - Urban worker health care coverage 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.
  - Out-of-pocket shares:
    - Health care out-of-pocket: rose from 20 percent in 1978 to a peak of 60 percent in 2000.
    - Education out-of-pocket: rose from 2 percent in 1990 to 13 percent in 2001.
  - Precautionary motive: He et al. (2017) find precautionary motives (from reduced pension generosity) contributed a 5.2 percentage point increase in the household savings rate from 1995 to 2009.
  - Policy response: since the early 2000s significant policy efforts have been made to rebuild the social safety net (e.g., New Rural Cooperative Medical Scheme in 2003; New Rural Social Pension Scheme introduced).

### Housing, financial repression, and inequality
- Housing:
  - Ownership rose from 20 percent in 1988 to 90 percent in 2007 and has been stable since.
  - Housing reform contributed about a quarter of the increase in household savings since the 1980s.
  - Empirical findings: homeownership increased the savings rate significantly by 3 percentage points during the first 10 years of the 2000s; by 2013 homeowners saved 2 percentage points less.
- Financial repression and interest rates:
  - Repressed deposit rates can raise savings via the “target savings” hypothesis.
  - Average real deposit rate in the past two decades has been below 3 percent when GDP growth was 10 percent.
  - Cross-provincial evidence: Nabar (2011) finds a 1 percent interest rate decline increases the household savings rate by about 0.5 percentage point, but once cross-region income differences are controlled this effect disappears in the authors’ regressions.
  - Deposit rate cap removed by 2015; wealth management products (WMP) and private lending proliferated (about 5 percent returns for WMP, typically over 10 percent for private lending).
  - WMP stock rose to 120 percent of GDP in 2016.
  - Conclusion in source: financial repression is arguably no longer a driver of household savings today.
- Income inequality and savings inequality:
  - Growth and poverty reduction: China’s economic growth averaged over 10 percent over the past three decades, lifting 600 million out of poverty and reducing the poverty headcount ratio from 66.6 percent in 1990 to 1.9 percent in 2013.
  - Income Gini rose from 0.3 in the 1980s to about 0.5 in 2010.
  - Top 10 percent’s income share rose from 27 to 41 percent between 1978 and 2015; bottom 50 percent’s share dropped from 27 to 15 percent.
  - Savings inequality:
    - CHIP microdata: differences between savings rates of richest and poorest deciles often as large as 20 percentage points.
    - 2013: top earners saved close to 50 percent of their incomes, bottom 10 percent saved about 20 percent.
    - Composition effect: from 1978 to 2015 the simple composition effect contributed about 3 percentage points to the increase in the aggregate savings rate.
  - International comparison: China’s household savings rate is higher at every income decile compared to other countries; the gap is particularly large for the poor, where many countries show negative savings rates but China’s bottom decile saves about 20 percent.

### Corporate savings — empirical analysis summary
- Aggregate and firm-level patterns:
  - Corporate gross savings surged in the early 2000s, moderated after the GFC.
  - Median gross savings ratio about 4 percent of assets in China, slightly lower than 5–6 percent in the rest of the world; net savings ratio about -0.01 percent (global average 0–0.1 percent).
  - SOEs: lower gross savings (linked to weaker profits) but higher net savings than private firms; average dividend payout ratio for SOEs is 22 percent, about 10 percentage points below that of private firms.
  - Industry patterns: sharpest decline in savings rates since the GFC in agriculture and mining; construction, financial and real estate also saw sizable declines; manufacturing and retail/wholesale relatively stable.
  - Exchange rate role: undervalued RMB during 2005–2008 contributed to export boom and high corporate savings (savings-to-asset ratio surged from 0.01 in 2005 to 0.15 in 2008).
- Firm-level regression specification (from APPENDIX IV):
  - Savings_{i,j,t} = c + β1 size_{i,j,t} + β2 Qratio_{i,j,t} + β3 SOE dummy_{i,j,t} + β4 Sector_{j} + β5 Year_{t} + β6 ER_{t} * Tradable Sector Dummy_{i,j} + ε_{i,j,t}
- Key regression coefficient highlights (Table 4):
  - Size:
    - Gross Saving Ratio: 0.4595***
    - Investment Ratio: 0.4726***
  - Q ratio:
    - Gross Saving Ratio: 0.7892***
    - Net Saving Ratio: 0.5868***
    - Profit Ratio: 1.0405***
  - SOE dummy:
    - Gross Saving Ratio: -0.4221***
    - Net Saving Ratio: 0.7656***
    - Investment Ratio: -1.0884***
  - Exchange rate depreciation (interacted with tradable sector dummy):
    - Net Saving Ratio: 0.1851***
    - Investment Ratio: -0.1120***
  - R-sq (overall) examples:
    - Gross Saving Ratio: 0.1397
    - Net Saving Ratio: 0.0734
    - Investment Ratio: 0.0923
  - Number of observations examples:
    - Gross Saving Ratio: 12,246
    - Dividend Ratio: 23,509
- Interpretation:
  - Ownership structure matters: SOEs exhibit lower gross saving but higher net saving due to weaker investment activity.
  - Exchange-rate movements, particularly RMB undervaluation in 2005–2008, amplified export-led profits and corporate savings for private, export-oriented firms; normalization of the RMB has reduced that excess saving over time.
  - Decomposition shows changes in investment-to-asset ratios can offset profit-driven increases in gross savings.

### Government savings and fiscal composition
- Evolution and levels:
  - Government savings stabilized since 2008 at around 5 percent of GDP; pre-2000 averaged around 3 percent of GDP; peaked near 6 percent in 2008.
  - In 2014, at 5.6 percent of GDP, Chinese government savings were among the highest in major economies; this largely reflects lower government consumption rather than higher revenues.
- Drivers:
  - Greater focus on public investment than public service: government-investment-to-GDP ratios noted as 4 percent on-budget and 12 percent off-budget; “augmented” fiscal deficit cited as 10 percent.
  - Lower social spending compared to other emerging markets in public education, public health, and social assistance.
  - Social security: became a deficit in 2013 as payments outpaced contributions; pension spending expected to rise from 4 percent of GDP in 2015 to 10 percent in 2050.
- Policy implication: need for decisive SOE reforms and policies encouraging larger transfers of SOE dividends to the budget to reduce SOE savings and fund social spending.

### Modeling approach, calibration, and scenarios
- Model design and key mechanisms:
  - Model source: Curtis et al. (2015, hereafter CLM15) and Curtis et al. (2017, hereafter CLM17).
  - Type: Overlapping generations (OLG) model with Barro-Becker (1989) households.
  - Three key features:
    - Parents value children’s consumption; children are dependent on parents up to a certain age.
    - Demographic composition matters: savings depend on whether an agent is a dependent child, a working adult, or a retired adult.
    - Family transfers beyond pay-as-you-go pensions are important; intrafamily transfers were an important cultural feature in China historically.
  - Population and lifecycle:
    - 95 overlapping cohorts, with a new cohort born every year.
    - Only those above 20 years old make savings and consumption decisions.
    - Retirees receive family transfers and a pension in a pay-as-you-go system.
  - Equilibrium scope: partial equilibrium; wages and interest rates are exogenous and taken from the data.
- Data inputs and calibration:
  - Household savings rate updated using NBS household survey per capita disposable incomes and consumption expenditures (2013–2016).
  - Baseline demographics use the medium-variant scenario from the July 2015 UN Population Prospects; alternatives use low- and high-variant scenarios.
  - Interest rates: input series Hodrick-Prescott filtered; CLM15 interest series preferred to capture returns to national capital stock; using CLM15 interest series results in a larger equilibrium savings rate than using deposit rates.
  - Wages: wage growth taken from CLM15; CLM15 wage series estimated using survey data—excluding capital income and family transfers.
  - Sensitivity:
    - Higher future interest rates lead to slower declines in savings rates.
    - Higher future wages depress both the future path of savings and the CLM15 model’s implied savings rates considerably.
- Parameter choices and deviations from CLM17:
  - Parametrization follows CLM17’s Table 1 except:
    - Retirement age set at 59 rather than 64.
    - Use of the age-earning profile estimated in CLM15.
- Model limitations noted:
  - Income growth and interest rates introduced exogenously; model does not capture general equilibrium effects of savings on interest rates or wages.
  - Interest rate changes contribute less to savings variations than income dynamics.

### Prospects for savings and projection scenarios
- Main projection results:
  - National savings-to-GDP ratio expected to fall by 4.5 percentage points by 2022 and by close to 10 percentage points by 2030.
  - Structural forces and policy support should lead to continued decline in household and government savings; corporate savings likely to fall with lower capital returns and rising labor income shares.
- Demographics:
  - Old-age dependency ratio projected to rise from 15 percent in 2015 to 50 percent in 2050.
  - Model: demographic changes would reduce household savings rates (percent of disposable income) by 6 percentage points by 2030.
- Social spending projections:
  - Baseline assumes government budget health expenditure to GDP increases from 1.7 percent in 2016 to 2.4 percent by 2022; this would lower household savings by about 1 percentage point.
  - Combined with aging, household savings in percent of GDP predicted to fall by 3 percentage points by 2022.
- Cohort effects:
  - Younger cohorts (born after 1990) show much lower savings rates; shifting cohort behavior may exert additional downward pressure on aggregate savings as they age.
- Comparative evidence:
  - Japan: household savings peaked at about 25 percent in 1974 and fell to almost zero.
  - Korea: peaked at 27 percent in early 1990s, about 15 percent today.
  - Taiwan POC: peaked about 30 percent in 1993, stabilized later at about 20 percent.
- Proactive scenario:
  - If government increases healthcare spending to 3 percent of GDP (compared to 2.5 percent in the baseline) and moderately reduces the top 10 percent’s income share via redistribution, household savings would be reduced by 5 percent of GDP by 2022 and household consumption would increase by an additional 2 percent of GDP relative to the baseline.
- Risks:
  - Corporate savings volatility could alter trajectory due to PPI swings and cyclical factors.
  - Household savings may not fall as projected if social safety net remains weak or aging impacts differ.

### Policy recommendations and state roles
- Identified policy-induced drivers of high savings:
  - One-child policy reduced fertility and raised savings.
  - 1990s transition and social safety net transformation increased precautionary savings.
  - 2000s WTO entry led to export-oriented growth and higher corporate savings, partly from an undervalued exchange rate.
  - Inadequate social spending and high public investment increased government savings.
  - Housing reform and rising housing prices increased household savings for down payments/mortgages.
  - Rising income inequality increased aggregate savings because richer households have higher propensities to save.
- Priority policy directions to lower savings and boost consumption:
  - Make income tax more progressive and family friendly: current tax structure described as regressive for the very poor; effective tax rate for the bottom can exceed 40 percent; tax allowances should be based on family size to boost fertility.
  - Increase social transfers to poor households to reduce poor households’ saving rates and align with international norms.
  - Increase expenditure on health care, pensions, and education; continue Hukou reform to ensure migrant workers’ access to social safety nets.
  - Increase general spending on public services (“soft infrastructure”) to improve delivery and staffing.
  - Finance higher social spending via larger SOE dividend payments or asset transfers:
    - Government target to increase SOE dividend payment ratio to 30 percent by 2020; de facto payment ratio still below 10 percent.
    - Raising payment ratio to target could increase budget revenues by 30 percent of GDP per year (text statement).
  - Improve access to formal financing for private firms by removing implicit SOE guarantees and redirecting credit to private sector to reduce private firms’ reliance on internal savings.
  - Service sector liberalization to allow private and foreign access to restricted service sectors to boost productivity and household income opportunities.

*Source: APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS — content excerpt from "China’s High Savings: Drivers, Prospects, and Policies" (wp18277).*

### APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS _________________________ 25

### APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS

### Major findings — national and structural context
- China’s national savings rate historically: around 35–40 percent of GDP since the 1980s; surged to a peak of 52 percent of GDP in 2008; declined to 46 percent in 2017.
- Global comparisons: global average savings rate of 20 percent; 15 percent for emerging economies.
- Internal demand structure (percent of GDP): investment ratio at 43 percent of GDP; private consumption at 38 percent of GDP; global average private consumption is 60 percent of GDP.
- If Chinese households consumed comparably to Brazilian households, Chinese consumption levels would be more than double (statement in source).
- Current account: surplus down to 2 percent of GDP since the GFC; high domestic savings have been absorbed by elevated investment.
- Sectoral composition shifts:
  - Households: household savings peaked at 25 percent (of disposable income) in 2010; household savings at 23 percent of GDP (current, as described) are 15 percentage points higher than the global average.
  - Corporates: corporate savings were low in the 1990s, surged in the 2000s, and after the GFC narrowed back to be in line with the global average.
  - Government: fiscal (budgetary) savings are volatile and on average a small portion of national savings; in recent years China’s fiscal savings have been higher than the global average, reflecting high capital spending.
- Household disposable income share of GDP: fell from 67 percentage points in the late 1990s to 58 percentage points in 2008; increased to 61 percentage points in 2014.
- Repressed deposit rates historically about 4 percent during high-growth period.

### Household savings — historical phases and empirical magnitudes
- Three phases of household savings (in percent of disposable income):
  - 1980s: rose from 5 to 20 percent (following the one-child policy and agricultural de-collectivization, with a temporary dip in the late 1980s).
  - 1990s: rose to 25 percent (post-SOE reform and social safety net transformation).
  - 2000s (post-WTO entry): rose further to 30 percent during an export-driven boom; since 2012 household savings have plateaued and begun to decline.
- Aggregate household savings measured in percent of GDP increased less than savings rate due to falling household income share of GDP.

### Drivers analyzed and quantitative contributions
- Demographics:
  - Fertility decline: fertility fell from about 6 to below 3 in 1970s policies, then below 2 after the one-child policy in 1980.
  - Youth-dependence ratio: China has one of the lowest youth-dependence ratios in the world (chart noted in source).
  - Mechanisms: fewer children reduce child-related expenditures and increase need to save for old-age support.
  - Micro evidence: households with twins save about 10 percentage points less than households with one child (Choukhmane et al., 2014).
  - Quantitative result from overlapping generations model (Curtis et al., 2015): demographic shifts alone account for about half of the rise in household savings (holding income growth and interest rates constant). Adding income and interest rates allows the full model to explain the broad savings trend.
- Social safety net and precautionary savings:
  - Large-scale layoffs during SOE reform: over 27 million during 1997–2002.
  - Health coverage and pension changes: urban worker health care coverage 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.
  - Out-of-pocket shares:
    - Health care out-of-pocket: rose from 20 percent in 1978 to a peak of 60 percent in 2000.
    - Education out-of-pocket: rose from 2 percent in 1990 to 13 percent in 2001.
  - Precautionary motive: the dismantling of the social safety net contributed materially to higher household savings, particularly during the 1980s and 1990s; He et al. (2017) find precautionary motives (from reduced pension generosity) contributed a 5.2 percentage point increase in the household savings rate from 1995 to 2009.
  - Policy response: since the early 2000s significant policy efforts have been made to rebuild the social safety net (e.g., New Rural Cooperative Medical Scheme in 2003; New Rural Social Pension Scheme introduced).

### Modeling approach and scope (from source)
- Overlapping generations model used (Curtis et al., 2015) captures:
  - Expenditure channel: children’s consumption included in parents’ utility function.
  - Transfer channel: old-age support modeled as a constant share of children’s wages.
- Model limitations noted in source:
  - Income growth and interest rates are introduced exogenously; the model does not capture general equilibrium effects of savings on interest rates or wages.
  - The model explains broad trends when income dynamics are included; interest rate changes contribute less to savings variations.

### Key implications highlighted
- The future trajectory of savings will shape China’s rebalancing:
  - If investment and credit slow and savings decline, external balance could be maintained and consumption could become the new engine of growth.
  - If high savings persist while investment slows, large current account surpluses could resume.
  - Sustained high savings with household and firm diversification needs could increase capital outflows as capital account liberalizes, complicating exchange rate management.
- Sectoral differences matter for policy design: households, corporates, and government each have different drivers and thus require targeted policy responses.

*Source: APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS — content excerpt from "China’s High Savings: Drivers, Prospects, and Policies" (source PDF content provided).*

### 2009. Both schemes are heavily subsidized by direct

### wp18277 - 2009. Both schemes are heavily subsidized by direct

### Social spending, household savings, and urban–rural differences
- China established a mandatory insurance scheme for formal sector urban workers in 2009, funded mostly via social security contributions from employers and employees; an insurance scheme for non-working residents was also established in 2009 requiring a minimum household contribution with a significant government subsidy. A basic pension scheme was introduced in 2010. As of the paper’s date, China has broadly achieved universal coverage for old-age pensions, although the benefits level remains low for rural households.
- Access to health care remains an issue for migrant workers, reflecting only gradual progress in “Hukou reform.”
- Empirical cross-sectional analysis for 287 prefectural municipalities finds a significant positive relationship between regional government social spending and household consumption:
  - Urban households: a 1 percent increase in social security spending is associated with a 0.04–0.07 percent increase in household spending. At the median, a 100 yuan increase in social security spending (about 13 percent of median monthly social security per capita in 2013) is associated with an increase in urban consumption by about 40 yuan in 2013.
  - Rural households: a 1 percent increase in health spending is associated with a 0.11–0.19 percent rise in rural consumption. At the median, a 100 yuan increase in health spending per capita (about 19 percent of the median of 532 yuan per capita spending in 2013) is associated with an increase in rural consumption by about 35 yuan in 2013.
- Interpretation:
  - Urban: social security (pensions) matters more for consumption.
  - Rural: health spending matters more for consumption.
  - Policy implication: better targeting of social spending between urban and rural areas is warranted.

### Income inequality and savings inequality
- Macro and distributional facts:
  - China’s economic growth averaged over 10 percent over the past three decades, lifting 600 million out of poverty and reducing the poverty headcount ratio from 66.6 percent in 1990 to 1.9 percent in 2013.
  - Income Gini rose from 0.3 in the 1980s to about 0.5 in 2010.
  - Top 10 percent’s income share rose from 27 to 41 percent between 1978 and 2015; bottom 50 percent’s share dropped from 27 to 15 percent.
- Savings inequality:
  - CHIP microdata: differences between savings rates of richest and poorest deciles often as large as 20 percentage points.
  - 2013: top earners saved close to 50 percent of their incomes, bottom 10 percent saved about 20 percent.
  - Composition effect: from 1978 to 2015 the simple composition effect contributed about 3 percentage points to the increase in the aggregate savings rate.
- International comparison:
  - China’s household savings rate is higher at every income decile compared to other countries; the gap is particularly large for the poor, where many countries show negative savings rates but China’s bottom decile saves about 20 percent.

### Housing and its effects on savings
- Institutional/historical facts:
  - 1988 reforms privatized much rental housing; 1995 reforms ended enterprise-supplied housing and moved to market-based provision.
  - Housing ownership rose from 20 percent in 1988 to 90 percent in 2007 and has been stable since.
- Channels through which housing affects savings: down payment effect, mortgage effect, wealth effect.
- Empirical findings:
  - Ownership’s impact on savings was positive in earlier years and turned negative by 2013.
  - Regression results: homeownership increased the savings rate significantly by 3 percentage points during the first 10 years of the 2000s; by 2013 homeowners saved 2 percentage points less.
  - Interpretation: mortgage effect dominated earlier; by 2013 wealth effect and rising down-payment needs for tenants led to changes. Housing reform contributed about a quarter of the increase in household savings since the 1980s.

### Financial repression and interest rates
- Theory: Repressed deposit rates can raise savings via the “target savings” hypothesis.
- Empirical evidence:
  - Average real deposit rate in the past two decades has been below 3 percent when GDP growth was 10 percent.
  - Cross-provincial evidence: Nabar (2011) finds a 1 percent interest rate decline increases the household savings rate by about 0.5 percentage point, but once cross-region income differences are controlled for this effect disappears in the authors’ regressions.
- Financial liberalization and alternatives:
  - Deposit rate cap removed by 2015; wealth management products (WMP) and private lending proliferated (about 5 percent returns for WMP, typically over 10 percent for private lending).
  - WMP stock rose to 120 percent of GDP in 2016.
  - Conclusion: financial repression is arguably no longer a driver of household savings today.

### Corporate savings
- Aggregate and firm-level findings:
  - Corporate gross savings surged in the early 2000s, moderated after the GFC.
  - Median gross savings ratio about 4 percent of assets in China, slightly lower than 5–6 percent in the rest of the world; net savings ratio about -0.01 percent (global average 0–0.1 percent).
  - SOEs: lower gross savings (linked to weaker profits) but higher net savings than private firms; average dividend payout ratio for SOEs is 22 percent, about 10 percentage points below that of private firms.
  - Industry patterns: sharpest decline in savings rates since the GFC in agriculture and mining; construction, financial and real estate also saw sizable declines; manufacturing and retail/wholesale relatively stable.
  - Exchange rate role: undervalued RMB during 2005–2008 contributed to export boom and high corporate savings (savings-to-asset ratio surged from 0.01 in 2005 to 0.15 in 2008).
- Policy implication: need for decisive SOE reforms and policies encouraging larger transfers of SOE dividends to the budget to reduce SOE savings and fund social spending.

### Government savings and fiscal composition
- Evolution and levels:
  - Government savings stabilized since 2008 at around 5 percent of GDP; pre-2000 averaged around 3 percent of GDP; peaked near 6 percent in 2008.
  - Fiscal savings volatility mainly driven by tax revenues.
  - In 2014, at 5.6 percent of GDP, Chinese government savings were among the highest in major economies; this largely reflects lower government consumption rather than higher revenues.
- Drivers of low government consumption and high fiscal savings:
  - Greater focus on public investment than public service: government-investment-to-GDP ratios noted as 4 percent on-budget and 12 percent off-budget; “augmented” fiscal deficit cited as 10 percent.
  - Lower social spending compared to other emerging markets in public education, public health, and social assistance.
  - Social security: became a deficit in 2013 as payments outpaced contributions; pension spending expected to rise from 4 percent of GDP in 2015 to 10 percent in 2050.

### Prospects for savings and projection scenarios
- Main projection results:
  - National savings-to-GDP ratio expected to fall by 4.5 percentage points by 2022 and by close to 10 percentage points by 2030.
  - Structural forces and policy support should lead to continued decline in household and government savings; corporate savings likely to fall with lower capital returns and rising labor income shares.
- Demographics:
  - Old-age dependency ratio projected to rise from 15 percent in 2015 to 50 percent in 2050.
  - Model: demographic changes would reduce household savings rates (percent of disposable income) by 6 percentage points by 2030.
- Social spending projections:
  - Baseline assumes government budget health expenditure to GDP increases from 1.7 percent in 2016 to 2.4 percent by 2022; this would lower household savings by about 1 percentage point.
  - Combined with aging, household savings in percent of GDP predicted to fall by 3 percentage points by 2022.
- Cohort effects:
  - Younger cohorts (born after 1990) show much lower savings rates; shifting cohort behavior may exert additional downward pressure on aggregate savings as they age.
- Comparative evidence:
  - Japan: household savings peaked at about 25 percent in 1974 and fell to almost zero.
  - Korea: peaked at 27 percent in early 1990s, about 15 percent today.
  - Taiwan POC: peaked about 30 percent in 1993, stabilized later at about 20 percent.
- Proactive scenario:
  - If government increases healthcare spending to 3 percent of GDP (compared to 2.5 percent in the baseline) and moderately reduces the top 10 percent’s income share via redistribution, household savings would be reduced by 5 percent of GDP by 2022 and household consumption would increase by an additional 2 percent of GDP relative to the baseline.
- Risks:
  - Corporate savings volatility could alter trajectory due to PPI swings and cyclical factors.
  - Household savings may not fall as projected if social safety net remains weak or aging impacts differ.

### Policy recommendations and state roles
- Identified policy-induced drivers of high savings:
  - One-child policy reduced fertility and raised savings.
  - 1990s transition and social safety net transformation increased precautionary savings.
  - 2000s WTO entry led to export-oriented growth and higher corporate savings, partly from an undervalued exchange rate.
  - Inadequate social spending and high public investment increased government savings.
  - Housing reform and rising housing prices increased household savings for down payments/mortgages.
  - Rising income inequality increased aggregate savings because richer households have higher propensities to save.
- Priority policy directions to lower savings and boost consumption:
  - Make income tax more progressive and family friendly: current tax structure described as regressive for the very poor; effective tax rate for the bottom can exceed 40 percent; tax allowances should be based on family size to boost fertility.
  - Increase social transfers to poor households to reduce poor households’ saving rates and align with international norms.
  - Increase expenditure on health care, pensions, and education; continue Hukou reform to ensure migrant workers’ access to social safety nets.
  - Increase general spending on public services (“soft infrastructure”) to improve delivery and staffing.
  - Finance higher social spending via larger SOE dividend payments or asset transfers:
    - Government target to increase SOE dividend payment ratio to 30 percent by 2020; de facto payment ratio still below 10 percent.
    - Raising payment ratio to target could increase budget revenues by 30 percent of GDP per year (text statement).
  - Improve access to formal financing for private firms by removing implicit SOE guarantees and redirecting credit to private sector to reduce private firms’ reliance on internal savings.
  - Service sector liberalization to allow private and foreign access to restricted service sectors to boost productivity and household income opportunities.

*Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18277.pdf*

### APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS

### APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS

### Model design and key mechanisms
- Model source: Curtis et al. (2015, hereafter CLM15) and Curtis et al. (2017, hereafter CLM17).
- Type: Overlapping generations (OLG) model with Barro-Becker (1989) households.
- Three key features:
  - Parents value children’s consumption; children are dependent on parents up to a certain age. As the ratio of parents to children rises, consumption falls and savings increase.
  - Demographic composition matters: the savings rate depends on whether an agent is a dependent child, a working adult, or a retired adult. Savings are larger if the share of working adults in the population is higher.
  - Family transfers (intrafamily transfers) beyond pay-as-you-go pensions are important: savings increase if working adults who are single children need to support their parents. Intrafamily transfers were an important cultural feature in China historically.
- Model population and lifecycle:
  - 95 overlapping cohorts, with a new cohort born every year.
  - Only those above 20 years old make savings and consumption decisions.
  - Agents of the same age are identical and take the current and future age distribution as given.
  - Agents face a cohort- and age-specific probability of death which rises to 1 when agents turn 95.
  - Retirees receive family transfers and a pension in a pay-as-you-go system.
  - Social security contribution falls on wages and is adjusted period-by-period to fund a constant replacement rate pension, specified as a percentage of the wage in the last working year.
- Equilibrium scope:
  - Partial equilibrium: wages and interest rates are exogenous and taken from the data.

### Data inputs and calibration choices
- Household savings rate update: actual household savings rate was updated between 2013–2016 using per capita disposable incomes and consumption expenditures of all residents from the NBS household survey.
- Demographics:
  - Baseline scenario uses the medium-variant scenario from the July 2015 update of the UN Population Prospects database (same as CLM).
  - Alternative scenarios use the low- and high-variant scenarios from the same UN database (see figure referenced in Section VI).
- Interest rates:
  - Input interest series are Hodrick-Prescott filtered to remove high-frequency movements.
  - The authors prefer the interest rate series constructed in CLM15 (to capture returns to national capital stock), noting marginal returns to capital have been considerably higher than deposit rates since the mid-1990s.
  - Using the CLM15 interest rate series results in a larger equilibrium savings rate than using deposit rates.
  - Recent proliferation of wealth management products and private lending generally pay larger interest rates than bank deposits and are likely better proxies for returns to cash.
- Wages:
  - Wage growth is taken from CLM15 rather than CLM17.
  - The CLM15 wage series is estimated using survey data—excluding capital income and family transfers—to match the model’s wage concept and differs from real wage or real disposable income growth.
- Sensitivity to future wages and interest rates:
  - Scenarios with higher future interest rates lead to slower declines in savings rates, without significantly affecting the CLM15 model’s implied savings rates.
  - Higher future wages depress both the future path of savings and the CLM15 model’s implied savings rates considerably.

### Parameter choices and deviations from CLM17
- Parametrization follows CLM17’s Table 1 except for:
  - Retirement age set at 59 rather than 64 (64 deemed too high historically).
  - Use of the age-earning profile estimated in CLM15.

*Source: APPENDIX I: OVERLAPPING GENERATIONS MODEL SIMULATIONS (wp18277)*

### APPENDIX IV: CORPORATE SAVINGS: EMPIRICAL ANALYSIS

### APPENDIX IV: CORPORATE SAVINGS: EMPIRICAL ANALYSIS

### Methodology
- Estimated firm-level regressions of corporate savings following the specification similar to Bayoumi et al. (2010), with an additional exchange-rate term:
  - Savings_{i,j,t} = c + β1 size_{i,j,t} + β2 Qratio_{i,j,t} + β3 SOE dummy_{i,j,t} + β4 Sector_{j} + β5 Year_{t} + β6 ER_{t} * Tradable Sector Dummy_{i,j} + ε_{i,j,t}
- Size denotes firm assets; Qratio denotes Tobin’s q.
- Controls: year dummies (global trend) and sector dummies (three-digit U.S. Standard Industrial Classification).
- Included SOE dummy to test ownership differences.
- Interaction of bilateral RMB/dollar exchange rate and tradable-sector dummy to capture exchange-rate impact on tradable firms’ savings.
- Additional regressions replaced savings ratio with firm-level profits, dividends, and capex (each measured against assets) to decompose drivers of corporate savings.

### Data
- Sample: Datastream’s annual balance sheet data for more than 1,000 publicly listed Chinese firms combined with Worldscope’s firm ownership data.

### Key Findings
- Ownership effects:
  - Chinese SOEs tend to save less in gross terms but more in net terms; this reflects SOEs’ lower investment ratios relative to private firms.
- Exchange-rate effects:
  - The exchange rate influences corporate investment, with a measurable indirect impact on net savings in tradable-sector firms.
  - During 2005–2008, the RMB was found to be undervalued, contributing to an export boom and large corporate savings, especially for the export-oriented private sector.
  - Although corporate investment increased in 2005–2008, the investment-to-asset ratio actually declined.
  - In more recent years, as the RMB moved closer to equilibrium, the export sector’s excessive savings have largely been unwound.

### Regression Results (Table 4: coefficients and statistics)
- Dependent variables: Gross Saving Ratio | Net Saving Ratio | Profit Ratio | Dividend Ratio | Investment Ratio
- Size:
  - Gross Saving Ratio: 0.4595***
  - Net Saving Ratio: 0.1171
  - Profit Ratio: 0.5495***
  - Dividend Ratio: 0.0672
  - Investment Ratio: 0.4726***
- Q ratio:
  - Gross Saving Ratio: 0.7892***
  - Net Saving Ratio: 0.5868***
  - Profit Ratio: 1.0405***
  - Dividend Ratio: 0.2532***
  - Investment Ratio: 0.3062***
- SOE dummy:
  - Gross Saving Ratio: -0.4221***
  - Net Saving Ratio: 0.7656***
  - Profit Ratio: -0.6302***
  - Dividend Ratio: -0.2128***
  - Investment Ratio: -1.0884***
- Exchange rate depreciation (interacted with tradable sector dummy):
  - Gross Saving Ratio: 0.0156
  - Net Saving Ratio: 0.1851***
  - Profit Ratio: 0.0349
  - Dividend Ratio: -0.0047
  - Investment Ratio: -0.1120***
- Sector: YYYYY (included)
- Year: YYYYY (included)
- Constant:
  - Gross Saving Ratio: -4.8461***
  - Net Saving Ratio: -5.3110***
  - Profit Ratio: -5.1614***
  - Dividend Ratio: -0.1780
  - Investment Ratio: -1.6544*
- R-sq (overall):
  - Gross Saving Ratio: 0.1397
  - Net Saving Ratio: 0.0734
  - Profit Ratio: 0.1627
  - Dividend Ratio: 0.0679
  - Investment Ratio: 0.0923
- Number of observations:
  - Gross Saving Ratio: 12,246
  - Net Saving Ratio: 12,236
  - Profit Ratio: 12,252
  - Dividend Ratio: 23,509
  - Investment Ratio: 23,492
- Number of groups:
  - Gross Saving Ratio: 2,508
  - Net Saving Ratio: 2,508
  - Profit Ratio: 2,508
  - Dividend Ratio: 2,536
  - Investment Ratio: 2,536
- Significance notation: ***, ** and * denote significance at the 1, 5 and 10 percent level, respectively.

### Interpretation and implications
- Ownership structure matters for corporate saving behavior: SOEs exhibit lower gross saving but higher net saving due to weaker investment activity.
- Exchange-rate movements, particularly RMB undervaluation in 2005–2008, amplified export-led profits and corporate savings for private, export-oriented firms; normalization of the RMB has reduced that excess saving over time.
- The decomposition (profits, dividends, investment) highlights that changes in investment-to-asset ratios can offset profit-driven increases in gross savings.

*Source: APPENDIX IV: CORPORATE SAVINGS: EMPIRICAL ANALYSIS (wp18277).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2018/wp18277.pdf_