## _wp1093

## Source details

**Canonical URL:** [_wp1093](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1093.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1093.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1093.pdf.json)

---

### I. Introduction — key facts and empirical preview
- China’s private consumption as a share of GDP declined from around 55 percent in the early 1980s to around 37 percent in 2008.
- General explanations: higher savings as households move away from subsistence and greater capital accumulation needs during early development.
- Specific channels cited in the literature:
  - Precautionary savings due to limited healthcare, pensions, and education benefits (Blanchard and Giavazzi (2005), Kujis (2005), Modigliani and Cao (2004)); Chamon and Prasad (2008) find the elderly save the most in China.
  - Gender imbalances increasing household savings, particularly where the unique child is male (Wei and Zhang (2009)) — attributed to about half the increase in household savings rate during 1990–2007.
  - Decline in households’ income as a share of GDP driving the fall in private consumption rather than a higher savings rate (Aziz and Cui (2007)).
  - Statistical reclassification effects explaining part of the decline in the labor share of income (Bai and Qian (2009)); a reclassification after 2004 explains 58 percent of the decline in the labor share during 1995-2004.
- Key empirical finding preview:
  - Around one-third of the fall in private consumption from 2000 to 2007 can be directly attributed to a fall in household income; the remaining two-thirds are due to other factors affecting the household savings rate or income.
  - The model does not require a China-specific dummy to fit China’s consumption profile.

### II. Empirical framework and estimation
- Dependent variable: private consumption as a share of GDP.
- Main explanatory variables (shares of GDP where applicable): household income, public consumption; plus per capita GDP, real GDP growth, real interest rates, CPI inflation, change in the terms of trade, old-age dependency ratio, change in the real effective exchange rate, share of employment in the services sector, measure of past foreign financing, and a measure of financial development.
- Structural interpretations:
  - Yh/Y (household income share) links income-share changes to private consumption.
  - Per capita GDP and real GDP growth capture stage-of-development effects.
  - Real interest rate (r) captures substitution/income effects; inflation (π) has ambiguous effects; terms of trade changes (λ) matter depending on permanence; old-age dependency ratio (D) captures demographic effects.
  - Public consumption (G) may substitute for or crowd out private consumption.
  - Real effective exchange rate changes (e), share of employment in services (Ls), foreign savings (s*), and financial development influence consumption through purchasing power, labor income composition, financing availability, and alternative savings instruments.
- Specification and estimation:
  - Estimated reduced-form includes nonlinear components and an exponential term.
  - Estimation method: Generalized Method of Moments (GMM) on an unbalanced panel of 39 economies, total 515 observations; lagged regressors used as instruments; real interest rates, changes in terms of trade, old-age dependency ratio and share of employment in services treated as exogenous.
  - Sample: 39 economies including China; longest sample spans 1980–2008.
  - Estimation note: All main regressors except inflation have a significant impact on private consumption.

### III. Model fit and decomposition of China’s consumption decline
- Model performance:
  - The model tracks the trend decline in Chinese consumption reasonably well, though fit is weaker in the early sample years.
  - The small unexplained component suggests conventional economic forces largely account for China’s declining consumption share rather than China-specific cultural/historical factors.
- Identified conventional forces:
  - China’s rapid economic growth.
  - Declining labor share of income.
  - Relatively low level of financial development.
  - Relatively capital-intensive production.
  - Low level of service employment.
- Decomposition results:
  - For China, around one-third of the change in consumption behavior from 2000-2007 is directly attributable to a decline in household income.
  - For 1992-2007, household income contribution rises to around 45 percent (consistent with Aziz and Cui (2007)).
- International contrasts:
  - Japan and Korea: falling household income shares were largely offset by falls in the savings rate.
  - United States: increase in private consumption largely attributed to a lower saving rate while household income changed little as a share of GDP.

### IV. The role of the household share of income — components and magnitudes
- Aggregate decline:
  - China’s household share of income declined from about 59 percent in 1992 to 53 percent in 2007.
- Labor income:
  - Labor share fell to 48¾ percent in 2008 from around 54 percent in 1997 (flow of funds data).
  - Using provincial data, the fall is from 51 percent to 40 percent.
  - Causes: sluggish employment growth; wages generally outpaced productivity over the past 10 years; export-oriented, low-cost-of-capital growth favors capital-intensive production.
  - Sectoral employment shares: primary 40¾ percent, secondary 26¾ percent, tertiary 32 1/3 percent.
  - China’s share of employment in services is far lower than expected given fundamentals; shifting labor from agriculture to services could raise the aggregate labor share.
  - Comparative examples: Japan’s services employment rose from around 38 percent in 1955 to around 60 percent by 1987; Korea’s services employment rose to around 65 percent in 1995 from around 30 percent in 1961.
- Investment income:
  - Households’ investment income fell to 2 1/3 percent of GDP, down from about 5 percent in the early 1990s.
  - Investment income is mainly interest income, which accounts for around 80 percent of investment income; dividends and other property income are minor.
  - Decline in interest income largely driven by falling deposit rates; administered rates (deposit ceiling, lending floor) favor investment (particularly capital-intensive sectors) at expense of consumption, effectively subsidizing borrowers.
  - Low share of property income by international standards suggests scope for expansion via alternative savings instruments and higher deposit rates.
  - Policy implication: interest rate liberalization could raise interest rates, discourage marginal investment, improve intermediation and monetary transmission, and enhance access to financial services.
- Transfers:
  - Net transfers to households have fallen to 1/3 percent of GDP, down from 2½ percent of GDP.
- Summary: falling shares of labor income, investment income, and transfers have held back households’ disposable income and private consumption.

### V. Savings rate determinants (after controlling for per capita income)
- Employment in services:
  - Consumption share could increase by 3½ percentage points for every 10 percentage point increase in the share of employment in the services sector.
  - Mechanisms: higher labor intensity, higher wages when accompanied by productivity-enhancing reforms, and supply-side effects (availability/quality of services raising consumption).
  - Policy implications: structural reforms to level playing field between tradable and non-tradable sectors, increase contestability, and improve access to financing could raise productivity and service quality.
- Real exchange rate:
  - Independent effect: around 2 percentage points increase in the consumption share for every 10 percent appreciation in the real effective exchange rate.
- Financial development and real interest rates:
  - Greater financial development tends to increase consumption by offering alternative investment instruments and improving access to financing.
  - China’s financial system: large but not very developed; dominated by banks; state-owned bank dominance implies savings are not efficiently used, with banking sector assets lent primarily to SOEs.
  - Stock markets: market capitalization $3½ trillion on average in 2008; only one company for every million persons was listed, compared with around 40 on average for the sampled economies.
  - Bond markets remain relatively underdeveloped.
  - Developing capital markets (corporate bond markets, mutual funds, broader equity ownership) could broaden savings instruments and insurance products, facilitating higher consumption.
  - The real interest rate is estimated to have a positive impact on the share of private consumption; liberalizing interest rates would promote private consumption via higher deposit rates.
- Public consumption:
  - On average across selected economies, private consumption is substitutable with public consumption: private consumption share declines with public consumption share.
  - Alternative interpretation: crowding out from deficit-financed public spending.
  - Provincial evidence for China: government spending on health (but not education) reduces urban household saving — a one yuan increase in government health spending translates into a two yuan increase in households’ consumption (Barnett and Brooks, 2009). Effects for rural households are more mixed.
- Demographics:
  - Across economies, consumption share rises with the dependency ratio as older people draw down lifetime savings.
  - China has a lower dependency ratio than other countries, which depresses its private consumption share relative to others.
  - Despite an expected rise in China’s dependency ratio, it is unclear this will raise consumption because average urban household saving rates in China are highest among the youngest and the oldest households—contrary to theory and other countries’ behavior.
  - Policy implication: pension and healthcare reform to reduce precautionary savings among the elderly is important.
- The neglected role of pensions:
  - Pension coverage appears to explain about 20 percent of the residual during transition to fuller coverage.
  - Relationship between the residual and the replacement rate has an R2 of around 13 percent.
  - Expanding pension coverage to include rural and urban workers could help lift private consumption.

### VI. Policy implications and recommended reforms
- Policies likely to raise consumption:
  - Raise household income through greater employment creation and a higher labor share.
  - Increase the share of employment in the services sector to raise labor intensity of growth.
  - Develop capital markets and financial sector: liberalize interest rates and create alternative savings instruments to raise households’ investment income and provide higher-return saving options.
  - Improve healthcare and pension systems to reduce precautionary savings, especially by elderly households.
- Rationale:
  - The absence of a China-specific unexplained factor suggests conventional economic and social reforms (service-sector employment expansion, financial development, interest rate liberalization, social safety net improvements) can materially affect private consumption behavior.

### VII. Conclusion — synthesis of findings
- China’s share of private consumption in GDP is low compared with benchmarks from international comparators, but much of this can be explained by:
  - Relatively low household income share.
  - Factors affecting household saving rates: level of development, employment share in services, financial development, and real exchange rate changes.
- The current level of consumption is not pre-determined by historical or cultural factors; targeted policies can raise consumption toward levels consistent with countries at similar development stages.
- Key actionable reforms:
  - Increase households’ income.
  - Develop capital markets, liberalize interest rates, and create alternative savings instruments.
  - Raise the share of employment in the services sector via structural reforms that increase productivity and service quality.
  - Expand pension coverage and undertake pension and healthcare reforms to reduce precautionary savings.

*Italic: Source — _wp1093 - References (PDF) — content as provided.*

### References .............................................................................................................

### _wp1093 - References

### I. Introduction
- China’s private consumption as a share of GDP declined from around 55 percent in the early 1980s to around 37 percent in 2008.
- General explanations for declining consumption shares at early development stages: increased savings as households move away from subsistence and greater capital accumulation needs.
- Specific explanations cited in the literature:
  - Precautionary savings due to limited healthcare, pensions, and education benefits (Blanchard and Giavazzi (2005), Kujis (2005), Modigliani and Cao (2004)); Chamon and Prasad (2008) find the elderly save the most in China.
  - Gender imbalances increasing household savings, particularly where the unique child is male (Wei and Zhang (2009)) — attributed to about half the increase in household savings rate during 1990–2007.
  - Decline in households’ income as a share of GDP driving the fall in private consumption rather than a higher savings rate (Aziz and Cui (2007)).
  - Statistical reclassification effects explaining part of the decline in the labor share of income (Bai and Qian (2009)); authors note a reclassification of individual business owners’ income to capital income after 2004 explaining 58 percent of the decline in the labor share during 1995-2004.
- Key empirical finding preview:
  - Around one-third of the fall in private consumption from 2000 to 2007 can be directly attributed to a fall in household income; the remaining two-thirds are due to other factors affecting the household savings rate or income.
  - The model does not require a China-specific dummy to fit China’s consumption profile.

### II. Empirical framework
- Dependent variable: private consumption as a share of GDP.
- Main explanatory variables (all as a share of GDP where applicable): household income, public consumption; plus per capita GDP, real GDP growth, real interest rates, CPI inflation, change in the terms of trade, old-age dependency ratio, change in the real effective exchange rate, share of employment in the services sector, measure of past foreign financing, and a measure of financial development.
- Structural interpretation of regressors:
  - Yh/Y (household income share) central to linking income-share changes to private consumption.
  - Per capita GDP and real GDP growth capture stage-of-development effects (Modigliani, 1966).
  - Real interest rate (r) captures substitution/income effects.
  - Inflation (π) may increase consumption via the Pigou effect or lower it through interest rate and measured income effects.
  - Terms of trade changes (λ) can affect consumption depending on permanence.
  - Old-age dependency ratio (D) captures demographic effects; private consumption generally increases as dependency ratio rises.
  - Public consumption (G) enters via potential substitution/complementarity or crowding-out effects.
  - Real effective exchange rate changes (e) affect household purchasing power.
  - Share of employment in services (Ls) raises labor income and service availability affecting consumption.
  - Foreign savings (s*) and financial development may influence availability of financing and alternative savings instruments.
- Estimated reduced-form specification includes nonlinear components and an exponential term (as presented in equation (1) of the source).
- Estimation method: Generalized Method of Moments (GMM) on an unbalanced panel of 39 economies for a total of 515 observations; lagged regressors used as instruments; real interest rates, changes in terms of trade, old-age dependency ratio and share of employment in services treated as exogenous; instrument set includes country dummies though regressions exclude country dummies.
- Sample: 39 economies including Argentina, Australia, Austria, Belgium, Brazil, Canada, Hong Kong SAR, Chile, Colombia, Denmark, Egypt, Finland, France, Germany, Greece, Iceland, India, Indonesia, Ireland, Israel, Italy, Japan, South Korea, Malaysia, Netherlands, New Zealand, Norway, Peru, Philippines, Portugal, Singapore, South Africa, Spain, Sweden, Switzerland, Taiwan Province of China, United Kingdom, United States, and China; longest sample spans 1980–2008.
- Estimation result note: All main regressors except inflation have a significant impact on private consumption.

### III. Fit of the model and decomposition of changes
- The model tracks the trend decline in Chinese consumption reasonably well, though fit is weaker in the early sample years.
- The small unexplained component suggests conventional economic forces largely account for China’s declining consumption share rather than China-specific cultural or historical factors.
- Identified conventional forces include:
  - China’s rapid economic growth.
  - Declining labor share of income.
  - Relatively low level of financial development.
  - Relatively capital-intensive production.
  - Low level of service employment.
- Decomposition result:
  - For China, around one-third of the change in consumption behavior from 2000-2007 is directly attributable to a decline in household income.
  - For the period 1992-2007, a similar calculation would raise the household income contribution to around 45 percent (consistent with Aziz and Cui (2007)).
- International contrasts:
  - Japan and Korea: falling household income shares were largely offset by falls in the savings rate.
  - United States: increase in private consumption largely attributed to a lower saving rate while household income changed little as a share of GDP.

### IV. The role of the household share of income
- Aggregate decline: China’s household share of income declined from about 59 percent in 1992 to 53 percent in 2007.
- Components of the decline:
  - Labor income:
    - Labor share fell to 48¾ percent in 2008 from around 54 percent in 1997 (based on flow of funds data).
    - Using provincial data, the fall is from 51 percent to 40 percent.
    - Causes: sluggish employment growth; wages generally outpaced productivity over the past 10 years; export-oriented growth and low cost of capital favor capital-intensive production over employment creation.
    - Sectoral employment shares: primary 40¾ percent, secondary 26¾ percent, tertiary 32 1/3 percent.
    - China has a far lower share of employment in the service sector than expected given its fundamentals (Guo and N’Diaye, 2009); shifting labor from agriculture to services could raise the aggregate labor share.
    - Comparative examples: Japan’s share of employment in services rose from around 38 percent in 1955 to around 60 percent by 1987; Korea’s services employment rose to around 65 percent in 1995 from around 30 percent in 1961.
  - Investment income:
    - Households’ investment income fell to 2 1/3 percent of GDP, down from about 5 percent in the early 1990s.
    - Investment income is mainly interest income, which accounts for around 80 percent of investment income; dividends and other property income play a minor role.
    - Decline in interest income largely driven by falling deposit rates; deposit and lending rates are administered (deposit rates subject to a ceiling; lending rates subject to a floor).
    - Low administered rates favor investment (particularly capital-intensive sectors) at the expense of consumption, effectively subsidizing borrowers.
    - Low share of property income in GDP by international standards suggests scope for expansion via alternative savings instruments and higher deposit rates.
    - Policy implication: interest rate liberalization could raise interest rates, discourage marginal investment, improve intermediation and monetary transmission, and enhance access to financial services (as argued by Porter and others (2009)).
  - Transfers:
    - Net transfers to households have fallen to 1/3 percent of GDP, down from 2½ percent of GDP (text ends at this point in the provided content).

### V. Policy implications and suggested reforms
- Policies likely to raise consumption include:
  - Raising household income (through greater employment creation and higher labor share).
  - Increasing the share of employment in the services sector to raise labor intensity of growth.
  - Developing capital markets and financial sector development: liberalizing interest rates and creating alternative savings instruments to raise households’ investment income and provide higher-return saving options.
  - Improving healthcare and pension systems to reduce precautionary savings, especially by elderly households.
- The absence of a China-specific unexplained factor suggests that a focused set of economic and social reforms (service-sector employment expansion, financial development, interest rate liberalization, and social safety net improvements) can materially affect private consumption behavior.

*Italic: Source — _wp1093 - References (PDF) — content as provided.*

### 1992. Looking at gross transfers,

### 1992. Looking at gross transfers,

### Summary of income and consumption trends
- China stands well below international comparators in gross transfers; this results largely from the SOE reform of the late 1990s, which created a void in social safety nets that the government is filling gradually.
- Falling shares of labor income, investment income, and transfers have held back households’ disposable income and private consumption.
- China’s share of private consumption falls outside the range of estimates based on equation (1) (highlighted area in figure 9) and the gap has been widening over time, indicating a rise in households saving rate.
- On average across countries, the consumption share declines steadily as countries get richer, and eventually stabilizes when per capita GDP is around US$2,500–US$3,000 on a PPP basis.
- China’s level of private consumption in 1980 was on the low side but broadly consistent with comparators at similar development; consumption then fell faster than suggested by the cross-country model, leaving China well below other countries by 2007.
- Similar rapid reductions in consumption as a share of GDP occurred in Korea, Japan, and the United States as their income per capita rose, though their initial consumption levels in 1980 were much higher than China’s.

### The role of the savings rate — main factors (after controlling for per capita income)
- Employment in services
  - Empirical results suggest the consumption share could increase by 3½ percentage points for every 10 percentage point increase in the share of employment in the services sector.
  - Employment in services increases the labor share of income since the services sector is generally more labor intensive, raising aggregate consumption.
  - The effect is larger when higher employment in services is accompanied by higher wages brought about by structural reforms that raise productivity.
  - Employment-in-services also captures supply-side effects: increased availability and quality of services raise household consumption.
  - Policy implications: structural reforms that level the playing field between tradable and non-tradable sectors, increase contestability in markets, and improve access to financing could raise productivity and service quality.
- Real exchange rate
  - An appreciating exchange rate increases the share of private consumption even after controlling for effects on household income.
  - The independent effect is around 2 percentage points increase in the consumption share for every 10 percent appreciation in the real effective exchange rate.
- Financial development and real interest rates
  - Greater financial development tends to increase consumption, possibly by offering alternative investment instruments and improving access to financing for credit-constrained households.
  - China’s financial system is large but not very developed and is mostly dominated by banks; state-owned bank dominance suggests savings are not efficiently used, with banking sector assets lent primarily to SOEs.
  - Stock markets had a large market capitalization ($3½ trillion on average in 2008), but only one company for every million persons was listed, compared with around 40 on average for the sampled economies.
  - Bond markets remain relatively underdeveloped.
  - Developing capital markets (corporate bond markets, mutual funds, broader equity ownership) could broaden savings instruments and insurance products, facilitating higher consumption.
  - The real interest rate is estimated to have a positive impact on the share of private consumption; liberalizing interest rates would further promote private consumption by leading to higher deposit rates.
- Public consumption
  - On average across selected economies, private consumption is substitutable with public consumption: the share of private consumption declines with that of public consumption.
  - An alternative interpretation is crowding out, where households cut spending in response to higher deficit-financed public spending.
  - For China, provincial evidence suggests government spending on health (but not education) reduces urban household saving: a one yuan increase in government health spending translates into a two yuan increase in households’ consumption (Barnett and Brooks, 2009). Effects for rural households are more mixed.
- Demographics
  - Table 1 evidence suggests the consumption share rises with the dependency ratio on average across economies as older people draw down lifetime savings.
  - China has a lower dependency ratio than other countries, which depresses its private consumption share relative to others.
  - Despite an expected rise in China’s dependency ratio, it is not clear this will raise consumption because average urban household saving rates in China are highest among the youngest and the oldest households—contrary to theory and other countries’ behavior.
  - Policy implication: measures to reduce precautionary savings among the elderly (pension and healthcare reform) are important so that rising dependency could increase consumption over time.
- The neglected role of pensions
  - Most sampled economies introduced pension systems after World War II and have undertaken major reforms, particularly in replacement rates.
  - Data limitations prevented including a pension variable in equation (1), but figures show a positive relationship between the residual in equation (1) and pension features such as coverage ratios and replacement rates.
  - Pension coverage appears to explain about 20 percent of the residual during transition to fuller coverage.
  - The relationship between the residual and the replacement rate has an R2 of around 13 percent.
  - Expanding pension coverage to include rural and urban workers could help lift private consumption.

### Conclusion and policy recommendations
- China’s share of private consumption in GDP is low compared with benchmarks from international comparators, but much of this can be explained by its relatively low share of household income and factors affecting household saving rates: level of development, employment share in services, financial development, and real exchange rate changes.
- The current level of consumption is not pre-determined by historical or cultural factors; policy efforts can raise consumption toward levels consistent with countries at similar development stages.
- Key policy actions that can generate tangible results:
  - Increase households’ income.
  - Develop capital markets, including liberalizing interest rates and creating alternative savings instruments.
  - Raise the share of employment in the services sector through structural reforms that increase productivity and service quality.
  - Expand pension coverage and undertake pension and healthcare reforms to reduce precautionary savings.

*Source: _wp1093 - 1992. Looking at gross transfers,*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2010/_wp1093.pdf_