RPI/CPI User Group

last person joined: 5 months ago 

To foster co-operation and the exchange of information between the ONS, its advisory bodies and other users. Please see 'Aims of the RPI/CPI User Group' in the library for further details.

1.  Web scraped clothing data

Posted 17 days ago

Yesterday we published a couple of research articles on using web scraped data to calculate alternative price indices for clothing. The data are from WGSN, a global trend authority specialising in fashion. The first article, Analysis of product turnover in web scraped clothing data, and its impact on methods for compiling price indices, takes a detailed look at the level of product turnover in the data and how this impacts on the calculation of different price indices. There are some interesting results around how different types of shop behave (for example, online retailers retain their stock for much longer than high-street retailers). The second article is more in line with previous research articles that we have produced on web scraping, and calculates a number of different indices from the data (including applying the new CLIP methodology to this data source): Research indices using web scraped price data: clothing data.

Any thoughts or comments on these articles are welcome.

Best wishes
Tanya Flower (ONS)

2.  RE: Web scraped clothing data

Posted 16 days ago

This needs handling with care.  One member of the Boskin Commission in the 1990s, which claimed that the US CPI had a substantial upward bias, had previously produced price indices for various items based on mail-order catalogues.  These showed lower price rises than the official CPI, which the commission regarded as proof that the CPI was biased.

3.  RE: Web scraped clothing data

Posted 15 days ago
What action would ONS intend to take in the light of this and all other information?


4.  RE: Web scraped clothing data

Posted 15 days ago

Hi Gareth,

Thanks for your query.

This work is part of a wider work stream looking at the use of alternative data sources in the development and production of consumer price statistics. The work is still currently at the research stage and a decision on next steps will not be made until this stage of the project is finished.

Best wishes,

5.  RE: Web scraped clothing data

Posted 13 days ago
  |   view attached
Hi Tanya

Attached for interest are actual CPI and RPI figures for clothing and footwear.

The only year which we can really compare is 2014 to 2015.  The CPI shows virtually no rise over this period while the RPI shows an 8% growth.  The CLIP methodology is closer to the RPI than the CPI.



6.  RE: Web scraped clothing data

Posted 4 days ago
Hi Gareth,

Thanks for your reply.

The indices in the article are based on only 9 of around 80 clothing and footwear items in the consumer prices basket. In Figures 7, 8 and 9 of the research, we've provided special aggregates of these CPIH item indices to allow for a more direct comparison. The CPIH all clothing special aggregate index remains relatively stable over the period, with a slight upward trend (between August 2014 and October 2015, prices for all clothing increased by around 7%). This is also the case for the men's and women's clothing aggregates. The CLIP matches the CPIH trend (between August 2014 and October 2015, prices for all clothing measured by the CLIP index increased by around 9%) but also appears to be more volatile.

Following your response, we also did a quick estimate of what an RPI special aggregate measure would look like using the item indices and weights published in our micro-data release for the same 9 items. While higher than the published CPIH data (as expected) it was not noticeably closer to the trend in the CLIP index, in fact it ended up slightly higher.

Nonetheless, there are many reasons why we would not draw a direct comparison between the price indices presented in this research and other published consumer price statistics. These include differences in data sources and methodology used. Further information on these differences is given in Research indices using web scraped data (Breton. R. et al, 2016). However, I agree that it is still a useful exercise to examine the trends shown in the different indices.

Best wishes