Consumption uncovered

Is it possible that Chinese household consumption is rising by an amazing 17%, as signaled by the real rate of retail sales growth? In short, no – although consumption is still expanding at a healthy pace.

This growth is driven by several factors: a rapidly increasing level of consumer confidence, based on steady improvement in macroeconomic data; the absence of household debt; strong income growth in the past few years; mild deflation to mitigate the absence of wage growth this year; and strong demand for goods and services by the large number of Chinese who only recently earned enough money to move into the consuming class.

All told, household consumption is likely to account for about one-third of this year’s GDP growth. Retail sales – as the only broad indicator of consumption released by the National Bureau of Statistics (NBS) on a monthly basis – is widely relied upon as a bellwether for the speed and direction of this consumption.

However, the astonishing retail sales growth rates released recently have prompted doubts about the quality of the data. There is confusion as to how retail sales can grow so rapidly during an economic downturn – especially when same-store sales growth at many listed retailers has fallen to low single-digits from about 30% a year ago. The finger is often pointed at the role of government purchases and wholesale activity in all this.

There is inevitably some confusion as to what the NBS includes in its calculations. For instance, it says wholesale activity isn’t counted, but it does count retail sales by companies that hold a wholesaling license.

Similarly, the NBS says its data does include government purchases from retail shops, but it is unable to calculate the share of government sales in total retail sales. As a result, the national accounts government consumption data used to calculate the state consumption contribution to GDP comes from government agencies, not retailers. It shows that the growth rate of government consumption has usually been in line with the real growth rate of retail sales, which suggests that state procurement has no real distorting effect on retail sales figures.

Auto and gas sales and home purchases, on the other hand, are areas that certainly do have a distorting effect on retail sales. Yet the extent of the distortion is unclear.

The NBS only provides a detailed breakdown of sales for larger stores – those with annual sales above US$733,000 – that account for 31% of total retail sales. With autos responsible for one-quarter of nominal sales at these larger outlets, and contributing a surprising 42% of first-quarter sales growth, the impact on overall sales is clear. It does not make sense that autos accounts for a larger share of total household consumption than food, beverages and tobacco (16% of the total and 16% of growth), even at larger stores.

But there are two reasons why this distortion may not be as large as it seems. First, we have no data on sales of autos – or any other product category – at the smaller-size stores that account for 69% of all retail sales. It is safe to assume, though, that autos play a much smaller role.

Second, it is clear that consumer auto sales have risen sharply – from a base of close to zero – over the past decade. Autos accounted for only 5% of retail sales by larger firms in 1997, with the share rising to 14% in 2003 and 25% in the first four months of this year. And a far greater share of these autos – about 85% – are now being bought by individuals. In 2000, according to one industry source, 80% of all sales were to government agencies and companies, with individuals accounting for just 20% of purchases.

Home sales are not included in China’s retail sales data, but purchases of construction and decorating materials by individuals (not by building companies) are included. Given that home sales have risen sharply so far this year, it is worth asking whether a consequent uptick in these sales is distorting the overall trend. Apparently not. Constructions and decorating materials accounted for less than 1% of total retail sales at larger stores in the first quarter. The contribution to retail sales growth was only 3%, up from zero contribution in the first quarter of 2008.

It is possible to cross-check the accuracy of China’s retail sales figures against consumption data. This information is also provided by the NBS, but comes from a quarterly survey of household spending patterns. There are various disparities between the two. Retail sales calculations use numbers collected from point of sale and therefore include purchases by governments, companies and other organizations. The household survey doesn’t include this information. Neither does it include construction materials bought by individuals.

Equally importantly, the very fact that the household survey talks directly to consumers means it is able to capture spending on services like education, health care and entertainment, as well as consumption of goods produced and consumed in the same household, such as farmers who grow their own food.

Put side by side, retail sales and household consumption exhibit the same basic trends. However, the slowdown in the survey data is much sharper (growth was 52% slower in March 2009 compared to the June 2008 peak) than the slowdown in retail sales (growth was 38% slower in March 2009 compared to June 2008). The difference between the two has also widened recently, retail sales having grown on average 65% faster than household expenditure over the last four quarters. During the same four quarters one year ago, the difference was 41%.

The last piece in the jigsaw is consumer sentiment, which can be gauged from independent studies conducted by CLSA. While Chinese consumers became much more pessimistic in the second half of 2008, they now appear to be increasingly optimistic. CLSA found that the share of households expecting business conditions to improve over the next six months rose to 44% in May, up from 14% in December. Meanwhile, the share of households expecting their family income to decrease over the next six months fell to 16% in May from 38% in December.

This growing confidence is tied to a steady improvement in the macro data. The CLSA Purchasing Managers’ Index was above 50 – indicating expansion – for the second consecutive month in May. At 51.2 in May, the index is now approaching its historical average of 51.7. There has also been a sharp reduction in layoffs. Indeed, in April manufacturers expanded headcount for the first time in nine months, and this stabilized at 50 in May.

All this is consistent with improving retail sales data. Factor in the various distortions – some of which are severe and some relatively tame – and evidence provided by the household survey, and Chinese consumption is not hitting the lofty heights suggested by the real rate of retail sales.

But I do believe urban household consumption is growing in the still very healthy 7-9% range, while rural household consumption is expanding a percentage point or two faster (although again, not as quickly as the 18% suggested by real rural retail sales figures). This is in line with my GDP growth forecast for 2009 of 8%.