Select Committee on Treasury Appendices to the Minutes of Evidence


Memorandum from the British Retail Consortium


  BRC represents 90 per cent of the total retail trade in the UK, operating in excess of 320,000 shops and stores, occupying over 30 per cent of commercial property portfolio by floor space.

  Membership covers all forms of retailing, from the large multiples and department stores, through to the corner shop, from food and drink to furniture and DIY, from centre of town to rural, to mail order and electronic commerce.

  Retailing is a major engine of employment growth in the economy, creating 58,500 new jobs (41,400 of them being full-time positions) during 1999.

  The British Retail Consortium welcomes this opportunity to provide an input into the Treasury sub-Committee inquiry into the launch of National Statistics. Retailing is a significant part of the economy, accounting for 10 per cent (2.4 million employees) of all employment and 26 per cent of GDP by expenditure. Increased scrutiny by government of the retail sector makes reliable statistical data an essential requirement.


  BRC is broadly supportive of the Framework for National Statistics.

  If the National Statistics logo is going to be seen as a guarantee of quality; non-government producers of statistics should be able to apply for its use on approved series.

  Ministers should not have the power to prevent inclusion of statistics produced by their departments within the scope of National Statistics.

  BRC expects to see vast improvements in the accuracy and timeliness of data relating to the retail sector.

  National Statistics should be made available on a UK basis. The current situation where some series relate to GB and others to UK is unhelpful to users.

  National Statistics should be meaningful to the sectors to which they relate.

  BRC welcomes the commitment to focus on relevance of statistics produced.

  Users should be consulted before series are discontinued, and incompatible sectors should not be aggregated for statistical series.

  In order to minimise burdens on data suppliers, BRC suggests that data be accepted in the form in which it is convenient for the supplier to provide it, either electronically or on company printouts.

  For companies that operate in all regions of the UK, requests for data from the ONS and each of the devolved administrations would be unwelcome. We suggest the data is collected centrally and disseminated as required.


  BRC is supportive of the need for official statistics that are more user-focused, and particularly welcomes the commitment to review key outputs at least every five years. Not only do outputs need to be reviewed for quality, but also relevance.


  According to the Framework, National Statistics are to provide a comprehensive description of the UK economy and society. BRC welcomes the guarantee of the quality that the logo National Statistics will provide, but would question why only statistics produced by government departments and the devolved administrations should be eligible to display the logo. The National Statistician and the Statistics Commission could provide a valuable service in assessing the quality of statistics produced by non-government bodies. In cases where non-government statistics are in common use, for example the various consumer confidence or house price indices, the publishers of these statistics should be able to apply to the Statistics Commission for audit. If the quality of the statistics meets the standards set, then the right to display the National Statistics logo should be given. This would help users of statistics by providing a guarantee of quality, and would remove the need for the Office for National Statistics to begin publishing statistical series that the private sector already covers. If the Statistics Commission is to be a source of high quality and independent advice on statistical issues, whose advice is available to the wider public, it should not limit itself to advice on government statistics.

  BRC would also argue that it should not be for ministers to agree whether or not statistics produced by their department should be included within the scope of National Statistics, the decision should be made by the Statistics Commission or National Statistician on the basis of quality and relevance.


  This is a laudable aim for National Statistics, but at present retail statistics would fall far short on accuracy, timeliness, being comprehensive or meaningful. BRC assumes that the following issues will be addressed before statistics relating to retail are categorised as national Statistics.


  Data for the monthly retail sales index (RSI) is requested from retailers. The weeks comprising the month are currently a matter of choice for retailers, so long as they state which weeks are used on the form. Retail sales can vary hugely between weeks, especially around Bank Holidays or in the run-up to Christmas. Accurate information on the state of the market can only be obtained if the Office for National Statistics insists that the correct weeks are included on the monthly return.


  The "Annual" Business Inquiry relating to retail has just been published (September 2000). It contains 1998 data. The latest available data on retail outlet numbers is from 1996; several of the country's largest shopping centres have opened since then.


  The monthly retail sales index covers only Great Britain. Virtually all other important economic statistics cover the United Kingdom. It is not possible to calculate retail sales as a percentage of household spending when sales data are only available for GB, and spending data are only available for UK.


  BRC members do not see volume of sales as a meaningful measure. Retailers provide the ONS with the value of sales in each period, ie money taken through the till. The ONS then apply a deflator and seasonal adjustment. In a time of price deflation, an increase in sales volumes can simply reflect extra sales that are being bought at the expense of retailers' margins. In the short term, volume of sales could be increased almost infinitely by retailers giving away goods. Seasonal adjustments are also contentious. The seasonal adjustment applied to, for example December and January sales, is massive. Average weekly sales in January are around two thirds of their December level, and in some sectors, sales activity drops by more than half. Additionally, seasonal adjustment necessarily relies on a consistent pattern; retailing is constantly changing, as are patterns of consumer spending.


  BRC would welcome a greater focus on relevance and the needs of the user. This will entail complete rethinking on some ONS series related to retail. For example, manufacturing is measured in terms of amounts of each good produced; it would be logical for retail sales to be measured in the same way, ie amount/value of each good sold. Instead, the measure is sales by each category of retailer. There are two reasons that this does not provide meaningful data:

    (i)  Categories of retailer are largely irrelevant in the modern retail world. Food retailers are increasingly selling fashion and homewares, and clothing retailers are developing mail order offerings;

    (ii)  There are also some bizarre outcomes from the categorisation of retailers, rather than what they sell. For example, none of Boots sales get categorised in "pharmaceutical, medical, cosmetic and toilet goods" in the monthly publication of sales data (SDM 28). Equally the clothing category only includes sales by retailers categorised as clothing retailers; sales of clothing by department stores are not included.

  There is also an argument that retail should be recorded as productive activity in its own right, rather than simply in terms of the volume and price of goods sold.

  In terms of user focus, some consultation would be appreciated before statistics are arbitrarily dropped. A case in point is number of retail outlets, which has previously been included in the Annual Business Inquiry. The number of retail outlets has been dropped from the retail ABI in order to bring it into line with ABIs for other sectors. Whilst it is understandable that the number of sites is not of interest for some sectors; it is very much so for retail. Number of outlets is a key factor in level of competition in the sector.

  Additionally, BRC would ask that investment data be published for retail separately from hotels and restaurants. Users of statistics are not well served by grouping sectors that have little in common.


  In terms of minimising the burden on data providers, BRC would draw attention to the point that data providers and data users within retail companies are often in separate departments. The data users would want as much detailed information on sales, employment, store numbers etc as possible. To reduce the burden for data providers, ONS needs to be flexible in how output is received (print out, electronic transfer etc) and to give adequate notice of new data requirements to allow for necessary computer reprogramming. If ONS and other Government departments, such as DTI and Customs and Excise, shared data it would prevent retailers being asked for the same information by several different Government departments.

  One example of current duplication is that the 70 largest retailers provide ONS with a seven category split of their retail sales which is used in calculating quarterly household expenditure. A separate six category split from the 43 largest retailers is used to provide the commodity breakdown on the Retail Sales First Release. It appears that, the two sections of the ONS which collect the monthly and quarterly data from 43 of the same retailers do not communicate with each other.

  Small retailers, who do not have specialist data analysis departments, find form filling for ONS purposes a particular burden. The main reason for this is that the analysis required does not necessarily correspond with the records small retailers keep for their own purposes. This applies especially to partnerships, which are not obliged to submit returns to Companies House. As small firms are selected at random to submit data, they do not have in place the systems to collate the required data quickly. BRC suggests that ONS should publish and distribute details of the data they require and the way in which they collect it to all retailers; small firms would then be forewarned and better prepared when the demand for them to supply data arrived.


  BRC welcomes the aim of the Scottish Executive and Welsh Assembly "to ensure that the statistical needs of devolved areas are met and allow comparisons to be made". We look forward to seeing the publication of more timely statistics for the devolved regions.

  In terms of Northern Ireland, BRC would suggest that retail sales data from the Province must be included in the monthly releases. Detailed employment data should also be made available for the whole UK, not just GB.

  In terms of collection of data, BRC would urge against data requests being sent out from each of the ONS, the Scottish Executive, the National Assembly for Wales and the Northern Ireland Executive. Large retail companies will operate in all of the devolved regions, but data will often be compiled only at head office. BRC would therefore suggest that one request is sent, from ONS asking for all information required, and that the data for the devolved administrations is disseminated by ONS.


  BRC eagerly awaits the improvements the new framework has been designed to bring.

October 2000

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2001
Prepared 18 January 2001