IMPROVING MEASUREMENT OF ONLINE RETAIL TRADE
MEASUREMENT OF ONLINE RETAIL TRADE
Although online retail trade activity is conceptually included in macroeconomic accounts, there are limited data sources available for estimating the value of online transactions, particularly for low-value goods imports and services delivered online (e.g. downloads of computer software, audio-visual material, e-books and the provision of telecommunications and information services). Therefore, the ABS is investigating ways to improve measurement and coverage of online retail trade activity; both domestic online retail sales and imports from overseas.
ENHANCEMENT OF THE RETAIL BUSINESS SURVEY
With respect to domestic online retail sales, the ABS has enhanced the monthly Retail Business Survey by adding a new online retailing question to the survey. This new question has enabled the ABS to estimate the proportion of Australian retail turnover which is derived from online retail sales.
Preliminary results from the enhanced survey show that 1.8% of the May 2013 estimate of Australian retail turnover was derived from online retail sales. Pure-play online retailers contributed approximately 40% of this estimate. Multi-channel retailers contributed the remaining 60%, of which half was food-related (i.e. derived from the food retailing and food services industries).
If the same proportion of online retail sales (1.8%) is applied to the annual Australian retail turnover estimate for 2011-12, then the annual turnover of domestic online retail sales would be $4,554m in 2011-12.
Results from the enhanced survey will only reflect the value of domestic online retail trade activity which is within the existing scope of the Retail Business Survey. As described earlier, non-employing businesses and non-resident businesses are outside the scope of the survey. The survey also excludes online retail sales from businesses with a non-retail ANZSIC classification, such as businesses which sell predominantly to other businesses (which are predominantly wholesalers) and businesses which produce goods for direct selling to consumers from the same premises (which are predominantly manufacturers) . These scope limitations may explain why the preliminary results from the enhanced survey are lower than other estimates which have been published externally in the media. Complete results from the enhanced Retail Business Survey are expected to be available from November 2013.
Further enhancements to improve the coverage and variability of estimates of domestic online retail sales are the subject of ongoing investigation by the ABS. These include (but are not limited to) potential enhancements to the methodology of the Retail Business Survey which would optimise the survey design for estimating both online and total retail turnover, rather than the current design which is optimised for estimating total retail turnover only.
ENHANCEMENTS TO THE WIDER MACROECONOMIC ACCOUNTS
In addition to enhancing the Retail Business survey, the ABS is investigating data gaps in the wider macroeconomic accounts created by online retail trade activity. Imports of goods above the Low Value Threshold (LVT), currently set at $1000, are currently included in International Merchandise Trade statistics. As a result, these are reflected in both the international and national accounts. However, data gaps exist for imports of goods below the LVT, including low-value goods imports from overseas online retailers.
There are three main channels through which goods are imported under the LVT into Australia:
1. As part of a consignment of goods with a total value above the LVT. These are required to be included on a Full Import Declaration (FID);
2. Through Self-Assessed Clearances (SACs). SACs are consignments that are generally carried and delivered by freight forwarding companies; and
3. As parcels conveyed through the international postal network.
Imported goods under the LVT which are reported to the Australian Customs and Border Protection Service (Customs) on an FID are not included in International Merchandise Trade statistics, although estimates of these goods are currently included in Balance of Payments statistics and the national accounts using information supplied by Customs. However, there are currently no estimates for imports of goods under the LVT which are not included on an FID. Therefore, a large number of parcels imported under the LVT, including low-value purchases from overseas online retailers, are not currently included in the macroeconomic accounts.
To improve measurement of online transactions in the wider macroeconomic accounts, reliable data on the value of SACs valued under the LVT arriving via air cargo and the number of consignments received has been sourced from Customs data, covering the financial years 2008-09, 2009-10 and 2010-11. The volume of SACs valued under the LVT arriving via sea cargo was too insignificant to separately estimate. Estimates of the number of parcels arriving through the international postal network were also sourced from the Productivity Commission report on the Economic Structure and Performance of the Australian Retail Industry, released on 9 December 2011. This was in respect of the financial year 2010-11 only. The value of post parcels for 2010-11 was then derived by multiplying the number of parcels received by an estimate of the average value of a parcel in that same year. This average value was then derived by the ABS using value information relating to comparable SACs imports. The SACs and post parcel estimates were then backcast to 1997-98 under a range of assumptions. It is important to note that the method of purchase, online or other, was not relevant in the context of compiling these estimates.
Based on this investigation, the ABS estimates the value of goods imported under the LVT which are not currently included in the macroeconomic accounts was $5,375m for 2010-11 and $6,226m for 2011-12. These estimates are made using a range of assumptions. More reliable sources of data may become available in future to replace these assumptions and improve the quality of the estimation process.
The ABS anticipates that these estimates will be published for the first time as part of the Balance of Payments coverage adjustment in the August 2013 issue of International Merchandise Imports, Australia (cat. no. 5439.0) and all subsequent Balance of Payments estimates of Goods Debits including the August 2013 issue of International Trade in Goods and Services, Australia (cat. no. 5368.0) and the September quarter 2013 issue of Balance of Payments and International Investment Position, Australia (cat. no 5302.0).
Estimates of the value of goods imported under the LVT are also being incorporated into the supply-use tables which underpin the national accounts. As a result, they will be included in the accounts for the first time with the release of the 2012-13 issue of Australian System of National Accounts (cat. no. 5204.0) and the September quarter 2013 issue of Australian National Accounts: Income, Expenditure and Product (cat. no. 5206.0). Detail on the sources, methods and assumptions used to estimate the value of goods imported under the LVT will be published in a Technical Note before these publications are released.
There is a comparable low value threshold for the reporting of exports of goods to Customs, which is currently $1,999. There are fewer reliable sources of information to enable estimation of this trade, and it is expected to be much less significant. As a result, no estimates of goods exports under the low value threshold have been made at this time.
The ABS is also investigating methods to estimate the value of intangible products consumed by households. These are services that are delivered online, such as e-books, films, music, apps, online newspaper and magazine subscriptions, and gambling services.