From BofAML:
Proposed changes in China’s GDP
According to reports by 21st Century Business Herald, the NBS has recently completed a preliminary draft of a new GDP accounting methodology. This draft will be submitted to the State Council for approval later this year and is expected to be implemented in 2015. Back in November 2013, a Deputy Head of the NBS, Mr. Xu Xianchun told the official Xinhua News Agency that adjustments are being made to the 2002 China National Account System. Mr Xu said that the revised historical GDP data will be released after the third national economic census.
Moving to the 2008 SNA
China’s proposed revision to GDP accounting approach mainly reflects the updated international guidelines for national economic account in the 2008 UN System of National Accounts (SNA). A number of countries such as the US, Australia, Canada and Korea have already implemented the changes. In particular, the NBS is planning to make the following changes:
- Recognition of R&D as gross fixed capital formation in GDP: under the current system, R&D expenditure is considered as “intermediate consumption” and not included in the calculation of final GDP. A number of surveys have been conducted on R&D activities carried out by the government, research institutions and companies. These surveys provide a good basis for including R&D expenditure in GDP calculations.
- Adopt a rent-based approach to estimate the services provided by owner-occupied housing:Currently, NBS uses a cost-based method to calculate the imputed rent of owner-occupied housing by adding depreciation of housing capital to maintenance and management costs. This method was established at the 2004 census when the rental market was not fully developed and data on rents were scarce. However, with recent development in the property market and rapid increase in both property price and rental price, the NBS plans to estimate the value of services of owneroccupied housing using data in the rental market.
- Adopt actual final consumption of households: The NBS currently measures household consumption using household final consumption expenditure. The main difference between household final consumption expenditure and actual final consumption of households is that the latter takes into account goods and services provided by the general government such as education and health services.
- Recognition of income resulting from the transfer of land contract and management rights: This change reflects the guidance given in the 2008 SNA to distinguish between legal ownership and economic ownership. According to China’s constitution, land in rural areas is collectively owned. However, based on the principle of economic ownership, farmers who obtained the contract and management rights are the economic owners of the land. Therefore, income received by the farmers from transferring the contract and management right to other individuals and/or institutions constitute households income. This will account for a significant proportion of some farmer’s income.
- Include stock options as employee compensation
Recognizing research and development
The NBS is planning to reclassify R&D as gross fixed capital formation rather than its current treatment as “intermediate consumption”. According to the NBS, expenditure on R&D was RMB1.2tn in 2013, and its share of GDP has grown steadily from 1.1% in 2003 to 2.0% in 2013 (Chart 1). Therefore, recognizing R&D as GFCF will boost the level and growth rate of nominal GDP.
However, it is not clear whether all of the R&D expenditure as reported by the NBS will be counted as GFCF. This is because the R&D expenditure could include spending on used equipment which should not be included in GDP as it is not produced in the current period. Moreover, some of the R&D conducted by government funded research institutions may have already been included in the GDP in items such as government spending.
For 2013, we estimate a 2% increase in the nominal GDP level and 0.1pp rise in nominal GDP growth assuming all of the R&D expenditure will be included in the GDP. By linearly extrapolating the growth rate in the R&D expenditure, we estimate that including all R&D expenditure will add 2.1% to nominal GDP level and less than 0.1pp to nominal GDP growth in 2014. The actual impact on nominal GDP level and growth could be even smaller due to reasons mentioned above. Thus, the impact of including R&D as GFCF on real GDP growth could be minimal once we take into account of inflation.
Looking at the experience of other countries that have included R&D expenditure, our US economics team estimated that the addition to real GDP growth from R&D is between 0-0.1pp during 2002-2007 when US R&D expenditure accounted for around 2.6% of GDP. In Korea, real GDP growth in 2013 was revised up by 0.2pp after the country adopted the 2008 SNA. R&D expenditure accounted for around 4.3% of GDP in 2013. So our estimates of the impact of including R&D expenditure on China’s real GDP growth is roughly in line with the data from other countries given China’ R&D expenditure’s share of GDP is relatively lower.
Adopting rent-based approach for owner-occupied housing
Services provided by owner-occupied housing are currently calculated using a cost-based approach by adding depreciation of housing capital to maintenance and management costs. In particular, the depreciation of housing capital is calculated based on construction cost of the property and the depreciation rate.
However, with the rapid increase in the property and rental prices in recent years, adopting a rent-based approach is likely to increase the value of services provided by owner-occupied housing, which will lead to a boost households’ consumption expenditure. Moreover, this change is probably more relevant for urban households than rural households as there is no active rental market in rural areas.
According to the latest available data published by the NBS, consumption expenditure by urban households on residence related services was at RMB2.2tn in 2011, account for 4.6% of GDP (Chart 2). Consumption expenditure on residence related services here includes spending on utilities, property management fees, rental expense and imputed rental value of owner-occupied housing. According to the 2010 Census, home ownership in cities and towns are around 79%. Therefore, imputed rental value of owner-occupied housing is likely to account for a significant proportion of the consumption expenditure on residence related services. However, it is difficult to estimate the increase in the nominal GDP level from the adoption of rent-based approach because we don’t know what rental data will be used. Moreover, residence related consumption expenditure has been growing at a double-digit pace in recent years, so the impact on nominal GDP growth from changing to rent-based approach could be rather limited.
