Poor Chinese Consumption Data Creates Savings Myth

Jun Zhang, Zhu Tian - April 2013

Speed Read
  • The generally held belief that China’s consumption is too low is a myth based on inadequate theory, a misreading of official statistics and the use of market exchange rates for making international comparisons.
  • Chinese official statistics underestimate consumption expenditure on housing, they omit consumption paid for as benefits by the corporate sector and there are a number of problems with the household expenditure surveys employed.
  • An adjustment for statistical issues suggests that the rate of consumption is 60-65% of GDP and not the 48% based on the widely quoted official statistics figures.
  • Using purchasing power parity adjustments China had a rate of consumption of 60.9% of GDP in 2010 quite similar to the level experienced by other East Asian economies.

Slowing Chinese economic growth and a sluggish world economy is leading to increasing calls for China to boost its domestic consumption. The Chinese consume too little and save too much, and because of this, China has had to rely on investment and exports for its phenomenal growth in the past. Indeed, according to official statistics, consumption makes up only 48% of China’s GDP, which means a gross savings rate of 52%. These savings finance not only domestic investment at the rate of 48% of GDP but also, in the form of capital outflow, net exports at 4% of GDP. The worldwide rate of consumption is 80% of GDP, with the United States at 88% and European Union at just above 80%. China’s consumption rate is much lower than the rest of the world. It is also lower than it was domestically 20 years ago when it was over 60% of GDP. China’s reluctant consumers, according to the conventional wisdom, have contributed to global imbalances, and China must adopt policies to stimulate consumption so as to not only contribute to global rebalancing but also make China’s own growth sustainable.



Demand does not determine long-term growth

This nearly universal belief about China’s consumption being too low is a myth based on a wrong theory and a superficial reading of the official statistics. First, it assumes wrongly that demand drives growth. Demand may determine a country’s current or short-term growth rate when the economy is operating under less than full capacity. That is the traditional Keynesian theory. We have no argument with that. But a country’s long-term economic growth and development depends on its capacity to produce, which in turn depends on the accumulation of the country’s physical and human capital as well as the speed of its technological progress. It is investment in both physical and human capital, and not consumption, that is an engine of sustained economic growth.


The reverse side of a low consumption rate is a high saving rate, which makes a high level of investment possible without heavy borrowing from foreigners. Indeed, one of the most important proximate factors behind China’s rapid economic growth in the past few decades has been its high rate of saving and investment. From 1990 to 2010, China’s real GDP grew at an average rate of almost 10.5% a year while real consumption grew at 8.6%. India, another fast-growing economy, saw its consumption grow at an average annual rate of 5.8% in the same period, also below its average annual GDP growth rate of 6.5%. It may sound paradoxical, but China’s comparatively low consumption rate (and in turn, high saving rate) is precisely a reason that the growth of China’s consumption has been so high.


Those who are familiar with growth theory may readily agree with our criticism of the popular view on consumption-driven growth, but some may still point to the fact that China’s consumption rate at less than 50% is unusually low. Few economies, it seems, have had such a low consumption rate in history for a sustained period of time. In particular, even the high income East Asian economies did not seem to have such a low rate of consumption during the periods of their own rapid growth. If true, then China’s investment efficiency may be too low, presumably a result of too much investment. This argument leads to our second point: China’s consumption has been significantly underestimated by official statistics. China’s true rate of consumption can be 10 to 15 percentage points higher than the official figure, reaching 60-65% of GDP, a normal as well as desirable ratio for a fast-growing economy.



Official Statistics underestimate consumption

There are several sources of underestimation of consumption in China’s official statistics. First, Chinese statistics has significantly underestimated housing consumption. Economists, including ourselves, have written on this issue before, but have not caught much attention. Even the National Bureau of Statistics (NBS) recognizes this, but there have been no adjustments made. Housing consumption is the sum of expenditures on utilities, decorations and rentals. Rentals include actual rentals paid by tenants and more importantly, imputed rentals for owner-occupied homes. In theory, the imputed rental of a self-owned home should be equal to the rent the owner-occupiers would have needed to pay if they had rented the home on the market. In practice, calculating imputed rentals is not an easy task.


The NBS uses construction costs multiplied by a fixed depreciation rate (2% in urban areas and 3% in rural areas) for a rough estimate. While this method is easy, it greatly underestimates actual housing consumption. First, construction costs, which do not even include land costs, greatly underestimate the market values of housing, and the 2% depreciation rate also underestimates the rental rate of return. Using this method, China’s housing consumption made up only 6% of GDP in 2009. In contrast, housing consumption in high income OECD countries such as USA, Japan, UK, Germany, France, and Canada took up around 14% of GDP in the same year, while in Mexico and Turkey, it accounted for 11% and 16.5% of GDP respectively. In India, housing consumption made up 8.6% of GDP in 2009.


According to our calculations, if China’s urban housing had an average price of RMB4000 (US$640) per square meter and the average gross rental rate of return was 3% in 2009, the resulting housing consumption in China would add up to 10% of GDP that year. That’s 4 percentage points larger. If rural housing consumption was underestimated by 1% of GDP, that makes up a total underestimation of 5% of GDP. (Incidentally, because imputed rental is also a source of income, this means that China’s GDP is also underestimated by at least 5%). Given the unusually high housing price in Chinese cities, it is very possible for China’s housing consumption to make up 11% of GDP or more.


Another source of underestimation of consumption is that official statistics cannot account for private consumption that is paid for by company accounts and thus treated as either business costs or, in the case of durable goods, as investment expenditures. There is a lot of such consumption in China. For example, many business owners or executives purchase private cars on company accounts. Instead of being counted as private consumption, the costs of these cars are counted as investment expenditures. We suspect that most of imported luxury cars fall into this category. This practice of private consumption on company accounts is widespread in China and extends to vacations, dining and wining, mobile phones, gasoline, and almost any household goods. It saves significantly on corporate as well as personal income taxes. It is also a benefit or in-kind consumption for executives at state-owned firms. Of course it’s impossible to know exactly how much of such consumption is unaccounted for. If it makes up, say, 10% of total household consumption, then China’s rate of consumption would increase by 3.5%.


A third source of underestimation of consumption has to do with the household survey method. This method relies on randomly sampled households to record all monthly income and expenditures. These households are paid only a small nominal compensation for participating. The aggregate household consumption expenditures are calculated by multiplying population with the average consumption expenditure per person. Several problems arise. First, high income households are known to be under-represented because they have no incentives to report their income or bother with keeping expenditure records. This would underestimate average consumption per person. Second, households may easily miss recording some purchases, especially small ones, as there are no explicit incentives for them to keep an accurate record. Third, in-kind consumption may be underestimated either because people fail to record it or use low prices to calculate its value. We hasten to add that we are not saying that China’s statistical authority is doing a lousy job, but that we as users of statistics would do better if we know what they measure and how they are collected.



Distortions from inter-country comparisons

Finally, academic economists normally do not simply use official national account statistics for cross-country comparisons, but use what is called the Penn World Table (or Summers-Heston Dataset) instead. The Penn World Table (PWT) is developed by the University of Pennsylvania. It differs from the official national income statistics in that it makes purchasing power parity (PPP) adjustment to consumption and investment price levels as well as official (or market) exchange rates. According to the most recent version of PWT, and without including the missing consumption expenditures mentioned above, China’s consumption ratio stood at 60.9% of GDP in 2010, the latest year the table covers, not the official 47.4%. The gist of the PWT adjustment is that the real value of China’s consumption measured in “international prices” is much higher than that measured in local Chinese prices because it is relatively cheaper to buy consumption goods and services (including government services) than investment goods in China. Specifically, in PWT, the 2010 price level for household consumption in China was at 42% of the US level while the price for investment was at 68% of the US level.


Furthermore, the PWT figure for China’s rate of consumption in 1990 was 58.9% while the official figure was 62.5%. Thus, instead of a significant decline in the consumption ratio over the past 20 years as demonstrated in unadjusted official statistics, the PWT figures show a very stable ratio around 60% over the period. (Technically, this is due to the fact that the price level of consumption in China rose relatively more slowly than that of investment during the period). So, if China's consumption rate at 60% is too low, then, at least according to PWT, it has always been this low for the past 20 years. What reason is there to worry now that there is a problem with China’s low consumption?


The 60% consumption rate is generally comparable to the level of consumption in the East Asian tiger economies based on figures from PWT. Korea’s consumption has been around 60% of GDP since mid 1980s, Hong Kong’s consumption rate has also stayed around 60% since 1960, and Singapore’s consumption has never exceeded 60% of GDP since mid 1970s, with an average of 45% over the past 20 years, much lower than China’s. All three places have gone on to become high income developed economies. Taiwan has had a slightly higher consumption rate, but it has been below 70% for most years since mid 1970s, reaching 60% in mid 1980s. Similarly, another high saving economy Japan has seen its consumption rate below 70% for most years since mid 1960s, going as low as 61% in some years.


We do not have to put all our faith in the PWT figures. But combined with the fact that a substantial amount of consumption expenditures is not accounted for in China’s official statistics, it seems safe to say that the ratio of China’s real consumption measured in comparable international prices should be no less than 60%, quite similar to the level experienced by other East Asian economies. Indeed, probably because of this similarity in the relatively low propensity to consume and high propensity to save, China may be on its way to becoming the next high income East Asian economy.