premia-parnters logo
Has the World Underestimated China’s Household Consumption by more than 10% of GDP?
Has the World Underestimated China’s Household Consumption by more than 10% of GDP?
2024/5/8
China’s household consumption appears to have been massively underestimated in international comparisons, because of differences in data definitions and valuation methodologies. The two big areas of differences in international comparisons are: 1) Social transfers in kind, which could be worth some 6% of GDP; and 2) The value of housing services provided by owner-occupied homes, which could be worth another 5% of GDP. In this article, our Senior Advisor Say Boon Lim discusses why the criticisms of China’s growth model and "underconsumption" look flawed.

The independent American think tank, the Peterson Institute for International Economics (PIIE) recently published an article which argued that international comparisons of China’s household consumption against those of the OECD economies have overlooked social transfers in kind worth some 6% of GDP (or more than US$1 trillion a year).

Meanwhile, a 2023 book by UBS economist Dr. Wang Tao suggests that China’s usage of historical construction cost, rather than imputed rent, to value the services provided by owner-occupied homes may have resulted in another area of underestimation, by up to some 5% of GDP.

In any event, despite all the pessimism about China’s consumption, the household consumption share in GDP has been increasing significantly from the low in 2010. Further, the focus on the household consumption share of GDP misses something very important – the massive, absolute Dollar amounts of increases in China’s consumption.

Even based on official OECD and IMF data, which contains the above-mentioned underestimations, China’s household consumption increased by US$6.6 trillion between 2010 and 2021. That increment is equivalent to the total household spending of Germany, Japan and South Korea put together last year. 

Analyses of China’s “underconsumption” are flawed: They are based on massive underestimations of household consumption leading to incompatible international comparisons. In summary, Chinese statistics for household consumption in China do not include “social transfers in kind” (STIK). Messing up the international comparisons, OECD data for its member countries do include STIK in calculations of household consumption as a percentage of GDP. 

American think tank says this underestimation is worth more than US$1 trillion a year. China’s household consumption has been underestimated in international comparisons by some 6% of GDP, according to the independent American think tank, the Peterson Institute for International Economics (PIIE). In an article on its website published in April, PIIE pointed out that international estimates of China’s household consumption had not included “social transfers in kind”, which would have amounted to over US$1 trillion a year.

The PIIE article said conclusions that China’s low household consumption would inhibit economic growth were based on “flawed” analysis.

“It overlooks data on household consumption that includes social transfers in kind, which have amounted to about 6% of GDP in recent years. Including transfers in kind makes Chinese consumption data comparable with that of other countries in the Organization for Economic Cooperation and Development (OECD),” PIIE said.

China’s household consumption should be around 45% of GDP, not 38% as commonly reported, on just the inclusion of STIK. Let’s start with that 6% of GDP that PIIE reckons is unreported in official Chinese household consumption data. In 2022, the OECD reported household spending as a percentage of GDP was 56% for Japan, 51% for Germany, 50% for Australia, and 48% for South Korea. Adjusting for the 6% of GDP not included in Chinese data, China’s household consumption would have been around 45% of GDP in 2021, which we should note was a low point in household consumption due to the pandemic. In 2019, before the pandemic, the figure would have been around 46% of GDP. Of course, this is not high by OECD standards, but it is hardly dysfunctional as some commentators have been portraying.

After adjusting for unreported STIK, criticisms of China’s growth model look flawed. As PIIE wrote: “Some analysts argue that China must raise the share of consumption in GDP, while reducing the share of investment, to sustain its economic growth. One recent paper argues that to achieve this goal, China needs to reverse the pattern in which consumption growth lags GDP growth — and that doing so will be possible only if Beijing shifts 1 to 2 percent of GDP annually to households. Failing that, according to this argument, China’s GDP growth will likely fall to the low single digits.” If so, the addition of 6% of GDP in unreported STIK alone could change the picture quite significantly.

The Economist magazine made a similar point about STIK late last year from an income perspective. The Economist quoted a figure of unreported household disposable income in China equivalent to 7% of national income. Disposable income is of course before savings. Nevertheless, both The Economist and PIIE are pointing to large, systemic underestimation of household income and consumption as a result of the exclusion of STIK in calculations, although obviously not by the same amount in each instance.

The Economist pointed to China’s statisticians’ failure to account for “goods and services that governments provide to individuals at little or no cost.”

It added: “These transfers include education and health care, not least reimbursements for medicines. They also encompass cultural amenities and subsidised food…..According to international standards, these goodies should appear in the official statistics as “social transfers in kind” …..They can then be added to household income and consumption to provide a fuller “adjusted” picture…..In the past, it has not reported them cleanly or separately, shovelling them into other parts of the national accounts, including government consumption.”

China’s household disposable income to GDP ratio is almost the same as in the Euro Area if STIK is included. If social transfers in kind are also stripped out of the household disposable income of other countries, their numbers will look more like China’s, according to The Economist. The figure for the Euro Area would be less than 64% in 2020. For the same year, the figure for China was 62%. The Economist rated China’s household disposable income share of GDP at just below Germany and ahead of Sweden, South Korea and Denmark. “A dozen” European countries had a smaller household income share than China, it said.

Then you might want to add another 5% of GDP for the different valuation methodologies for housing services provided by owner-occupied homes. China’s NBS’ (National Bureau of Statistics) use of the original cost of the construction of homes rather than imputed rents to estimate the value of owner-occupied housing services has likely led to China’s household consumption being “seriously underestimated”, according to UBS Chief China Economist Wang Tao, in her 2023 book “Making Sense of China’s Economy”. OECD countries use imputed rents to calculate GDP and personal consumption expenditures. That underestimation of housing services is likely worth another 5% of GDP for China’s household consumption.

Dr Wang pointed to the anomaly of official statistics putting housing services provided by owner-occupied housing at just 3%–4% of China’s GDP, despite more than 80% of Chinese owning their own homes. This is less than half the level in developed economies. “Overall housing services”, including rents and utilities, is around 7% of GDP in China, compared to approximately 13% in the US and the Euro Area.

“Using alternative estimates of housing services, China’s household consumption as a share of GDP would be about 5 percentage points higher than the official data suggest,” Dr Wang wrote.

Whether we land at 6% of GDP or 11% or somewhere in between, there is evidently massive underestimation of China’s household consumption. These are huge variances, and we recognise the complexities of arguments which may arise from these assertions. However, whether it’s 6% of GDP as asserted by PIIE, or 11% if we include the 5% of GDP as suggested by Dr Wang, or somewhere in between, our takeaway is there is massive underestimation of China’s household consumption relative to the OECD economies. 

And that is before adding paid services provided by family members. Something else worth considering is the phenomenon of young people in China preferring to be paid to do household chores for their own families, rather than joining the formal workforce. We had written about these “full time children”, who are paid by their families, in an insight published last year. Although this phenomenon will likely result in an understatement of both GDP and household consumption, the net impact will still be some measure of understatement in the ratio.

While talking about ratios, don’t forget the importance of the absolute levels of consumption. Beyond the underestimation, there also seems to be a lack of appreciation of the phenomenon of developing economies enjoying higher percentage gains in household consumption when the share of household consumption in GDP is declining.  The reverse is also true: That is, economies where the household consumption share has been stable have seen much lower rates of growth in consumer spending. In this regard, it is worth remembering that it is not the “percentage share of GDP” that will necessarily drive revenue and earnings: It is the absolute amount of spending. 

Stable shares for household consumption in GDP or higher growth in absolute levels of consumption? Take your pick. Research published in Huang Yukon’s book “Cracking the China Conundrum” suggests that growth in real consumption is much faster in economies which experienced significant declines in the consumption share in GDP. Mr. Huang is an economist at the Washington-based Carnegie Endowment for International Peace. The above argument is summarised in this table from his book.

Country

Average annual growth in real consumption (t to 30 years) %

Change in C/GDP share, %

Korea

7.0

-31.7

Taiwan

8.8

-13.4

China

7.8

-15.7

Philippines

3.7

+4.9

Mexico

3.9

-4.9

Brazil

4.3

-4.3

"t” = 1960 for Korea and Taiwan; “t” = 1980 for China and Philippines
 Source: Huang Yukon (2017), Cracking the China Conundrum, New York, Oxford University Press

The logic of GDP growing faster than household consumption lies in urbanisation. The countries with the highest household consumption to GDP ratios are some of the poorest in the world. In the process of urbanisation and development off low levels of national income, the logical outcome is that the value of GDP inevitably grows faster than that for household consumption. The faster growth in the denominator in the household consumption/GDP ratio reflects the greater value add for workers’ labour as they migrate from subsistence levels of income in rural areas to urban economies which generate greater productivity. Part of that increased productivity goes to companies in profits, part of that to pay other resource owners who provide goods and services to the company, part of that in taxes to the State, and part in higher wages to workers. In a successful economy, even as workers consume much more in absolute terms, the percentage they consume of their income declines, as they can then afford savings, which was previously unattainable. And their consumption as a percentage of GDP declines even more.

Rather than bemoaning the rapid growth of the denominator (GDP), it would make more sense to celebrate the growth of the numerator (consumption).

The household consumption share of GDP reverses at a more mature phase of urbanisation. Indeed, the turning point for China’s household consumption share of GDP may have been in 2010, when the share (adjusted for STIK) hit 39%. It then reversed up steadily and very significantly to 46% in 2019. The disruption of the pandemic nudged that back down to 45% in 2021. It should be remembered that China reopened from COVID only in 2023. With growing confidence after the COVID shock – and hopefully stabilisation in the property sector – the rise in the household consumption share of GDP from 2010 should resume. The trend  of consumers moving away from international travel to domestic tourism will also help.

US$6.6 trillion in additional consumption in a decade – likely much more if missing consumption data is included. Using unadjusted data for household consumption, China added another US$6.6 trillion in household consumption between 2010 and 2021. That’s the total household spending of Japan, Germany and South Korea put together for 2022. Including the STIK as identified by PIIE, the incremental household consumption would have been US$7.8 trillion. This is already a big enough delta, without including the 5% of GDP from the undervaluation of housing services.

  • 林哲文

    資深指導顧問

以下翻譯結果由阿里雲機器翻譯引擎生成,僅供參考。 睿亞資產對內容的準確性或適當性不承擔任何責任,如與英文原文含義不同,以原文為準。

China’s household consumption appears to have been massively underestimated in international comparisons, because of differences in data definitions and valuation methodologies. The two big areas of differences in international comparisons are: 1) Social transfers in kind, which could be worth some 6% of GDP; and 2) The value of housing services provided by owner-occupied homes, which could be worth another 5% of GDP. In this article, our Senior Advisor Say Boon Lim discusses why the criticisms of China’s growth model and "underconsumption" look flawed.

The independent American think tank, the Peterson Institute for International Economics (PIIE) recently published an article which argued that international comparisons of China’s household consumption against those of the OECD economies have overlooked social transfers in kind worth some 6% of GDP (or more than US$1 trillion a year).

Meanwhile, a 2023 book by UBS economist Dr. Wang Tao suggests that China’s usage of historical construction cost, rather than imputed rent, to value the services provided by owner-occupied homes may have resulted in another area of underestimation, by up to some 5% of GDP.

In any event, despite all the pessimism about China’s consumption, the household consumption share in GDP has been increasing significantly from the low in 2010. Further, the focus on the household consumption share of GDP misses something very important – the massive, absolute Dollar amounts of increases in China’s consumption.

Even based on official OECD and IMF data, which contains the above-mentioned underestimations, China’s household consumption increased by US$6.6 trillion between 2010 and 2021. That increment is equivalent to the total household spending of Germany, Japan and South Korea put together last year. 

Analyses of China’s “underconsumption” are flawed: They are based on massive underestimations of household consumption leading to incompatible international comparisons. In summary, Chinese statistics for household consumption in China do not include “social transfers in kind” (STIK). Messing up the international comparisons, OECD data for its member countries do include STIK in calculations of household consumption as a percentage of GDP. 

American think tank says this underestimation is worth more than US$1 trillion a year. China’s household consumption has been underestimated in international comparisons by some 6% of GDP, according to the independent American think tank, the Peterson Institute for International Economics (PIIE). In an article on its website published in April, PIIE pointed out that international estimates of China’s household consumption had not included “social transfers in kind”, which would have amounted to over US$1 trillion a year.

The PIIE article said conclusions that China’s low household consumption would inhibit economic growth were based on “flawed” analysis.

“It overlooks data on household consumption that includes social transfers in kind, which have amounted to about 6% of GDP in recent years. Including transfers in kind makes Chinese consumption data comparable with that of other countries in the Organization for Economic Cooperation and Development (OECD),” PIIE said.

China’s household consumption should be around 45% of GDP, not 38% as commonly reported, on just the inclusion of STIK. Let’s start with that 6% of GDP that PIIE reckons is unreported in official Chinese household consumption data. In 2022, the OECD reported household spending as a percentage of GDP was 56% for Japan, 51% for Germany, 50% for Australia, and 48% for South Korea. Adjusting for the 6% of GDP not included in Chinese data, China’s household consumption would have been around 45% of GDP in 2021, which we should note was a low point in household consumption due to the pandemic. In 2019, before the pandemic, the figure would have been around 46% of GDP. Of course, this is not high by OECD standards, but it is hardly dysfunctional as some commentators have been portraying.

After adjusting for unreported STIK, criticisms of China’s growth model look flawed. As PIIE wrote: “Some analysts argue that China must raise the share of consumption in GDP, while reducing the share of investment, to sustain its economic growth. One recent paper argues that to achieve this goal, China needs to reverse the pattern in which consumption growth lags GDP growth — and that doing so will be possible only if Beijing shifts 1 to 2 percent of GDP annually to households. Failing that, according to this argument, China’s GDP growth will likely fall to the low single digits.” If so, the addition of 6% of GDP in unreported STIK alone could change the picture quite significantly.

The Economist magazine made a similar point about STIK late last year from an income perspective. The Economist quoted a figure of unreported household disposable income in China equivalent to 7% of national income. Disposable income is of course before savings. Nevertheless, both The Economist and PIIE are pointing to large, systemic underestimation of household income and consumption as a result of the exclusion of STIK in calculations, although obviously not by the same amount in each instance.

The Economist pointed to China’s statisticians’ failure to account for “goods and services that governments provide to individuals at little or no cost.”

It added: “These transfers include education and health care, not least reimbursements for medicines. They also encompass cultural amenities and subsidised food…..According to international standards, these goodies should appear in the official statistics as “social transfers in kind” …..They can then be added to household income and consumption to provide a fuller “adjusted” picture…..In the past, it has not reported them cleanly or separately, shovelling them into other parts of the national accounts, including government consumption.”

China’s household disposable income to GDP ratio is almost the same as in the Euro Area if STIK is included. If social transfers in kind are also stripped out of the household disposable income of other countries, their numbers will look more like China’s, according to The Economist. The figure for the Euro Area would be less than 64% in 2020. For the same year, the figure for China was 62%. The Economist rated China’s household disposable income share of GDP at just below Germany and ahead of Sweden, South Korea and Denmark. “A dozen” European countries had a smaller household income share than China, it said.

Then you might want to add another 5% of GDP for the different valuation methodologies for housing services provided by owner-occupied homes. China’s NBS’ (National Bureau of Statistics) use of the original cost of the construction of homes rather than imputed rents to estimate the value of owner-occupied housing services has likely led to China’s household consumption being “seriously underestimated”, according to UBS Chief China Economist Wang Tao, in her 2023 book “Making Sense of China’s Economy”. OECD countries use imputed rents to calculate GDP and personal consumption expenditures. That underestimation of housing services is likely worth another 5% of GDP for China’s household consumption.

Dr Wang pointed to the anomaly of official statistics putting housing services provided by owner-occupied housing at just 3%–4% of China’s GDP, despite more than 80% of Chinese owning their own homes. This is less than half the level in developed economies. “Overall housing services”, including rents and utilities, is around 7% of GDP in China, compared to approximately 13% in the US and the Euro Area.

“Using alternative estimates of housing services, China’s household consumption as a share of GDP would be about 5 percentage points higher than the official data suggest,” Dr Wang wrote.

Whether we land at 6% of GDP or 11% or somewhere in between, there is evidently massive underestimation of China’s household consumption. These are huge variances, and we recognise the complexities of arguments which may arise from these assertions. However, whether it’s 6% of GDP as asserted by PIIE, or 11% if we include the 5% of GDP as suggested by Dr Wang, or somewhere in between, our takeaway is there is massive underestimation of China’s household consumption relative to the OECD economies. 

And that is before adding paid services provided by family members. Something else worth considering is the phenomenon of young people in China preferring to be paid to do household chores for their own families, rather than joining the formal workforce. We had written about these “full time children”, who are paid by their families, in an insight published last year. Although this phenomenon will likely result in an understatement of both GDP and household consumption, the net impact will still be some measure of understatement in the ratio.

While talking about ratios, don’t forget the importance of the absolute levels of consumption. Beyond the underestimation, there also seems to be a lack of appreciation of the phenomenon of developing economies enjoying higher percentage gains in household consumption when the share of household consumption in GDP is declining.  The reverse is also true: That is, economies where the household consumption share has been stable have seen much lower rates of growth in consumer spending. In this regard, it is worth remembering that it is not the “percentage share of GDP” that will necessarily drive revenue and earnings: It is the absolute amount of spending. 

Stable shares for household consumption in GDP or higher growth in absolute levels of consumption? Take your pick. Research published in Huang Yukon’s book “Cracking the China Conundrum” suggests that growth in real consumption is much faster in economies which experienced significant declines in the consumption share in GDP. Mr. Huang is an economist at the Washington-based Carnegie Endowment for International Peace. The above argument is summarised in this table from his book.

Country

Average annual growth in real consumption (t to 30 years) %

Change in C/GDP share, %

Korea

7.0

-31.7

Taiwan

8.8

-13.4

China

7.8

-15.7

Philippines

3.7

+4.9

Mexico

3.9

-4.9

Brazil

4.3

-4.3

"t” = 1960 for Korea and Taiwan; “t” = 1980 for China and Philippines
 Source: Huang Yukon (2017), Cracking the China Conundrum, New York, Oxford University Press

The logic of GDP growing faster than household consumption lies in urbanisation. The countries with the highest household consumption to GDP ratios are some of the poorest in the world. In the process of urbanisation and development off low levels of national income, the logical outcome is that the value of GDP inevitably grows faster than that for household consumption. The faster growth in the denominator in the household consumption/GDP ratio reflects the greater value add for workers’ labour as they migrate from subsistence levels of income in rural areas to urban economies which generate greater productivity. Part of that increased productivity goes to companies in profits, part of that to pay other resource owners who provide goods and services to the company, part of that in taxes to the State, and part in higher wages to workers. In a successful economy, even as workers consume much more in absolute terms, the percentage they consume of their income declines, as they can then afford savings, which was previously unattainable. And their consumption as a percentage of GDP declines even more.

Rather than bemoaning the rapid growth of the denominator (GDP), it would make more sense to celebrate the growth of the numerator (consumption).

The household consumption share of GDP reverses at a more mature phase of urbanisation. Indeed, the turning point for China’s household consumption share of GDP may have been in 2010, when the share (adjusted for STIK) hit 39%. It then reversed up steadily and very significantly to 46% in 2019. The disruption of the pandemic nudged that back down to 45% in 2021. It should be remembered that China reopened from COVID only in 2023. With growing confidence after the COVID shock – and hopefully stabilisation in the property sector – the rise in the household consumption share of GDP from 2010 should resume. The trend  of consumers moving away from international travel to domestic tourism will also help.

US$6.6 trillion in additional consumption in a decade – likely much more if missing consumption data is included. Using unadjusted data for household consumption, China added another US$6.6 trillion in household consumption between 2010 and 2021. That’s the total household spending of Japan, Germany and South Korea put together for 2022. Including the STIK as identified by PIIE, the incremental household consumption would have been US$7.8 trillion. This is already a big enough delta, without including the 5% of GDP from the undervaluation of housing services.