BSUG/
FSUG
PRICE
INDICES

Report by Ulric Spencer of meeting of 13 March 2002

The Bank of England was the venue for a joint meeting of BSUG and FSUG, chaired by Duncan McKenzie (IFSL), at which four topics related to Price Indices were addressed. 

First, Nicholas Palmer and Keith Jones (ONS) outlined the work on the Corporate Services Price Index (CSPI) Development Project. A staff of 17are engaged on four work areas – the quarterly data collection and resultant outputs; development; quality assurance; and an expansion programme. The rationale of the CSPI is to provide deflators for UK national accounts, and to fill a gap in the UK inflation picture, by monitoring inflation in the service sector. The key customers for the index are ONS, the Bank of England, the Treasury and industry. 

The Index measures quarterly movements in prices of services provided by businesses to other business or government customers, which accounts for about 25 per cent of GDP. Only services provided by UKbased business are at present covered. Data are collected from nearly 1000 contributors and yield over 4000 price quotations covering about 130 service activities or products. The experimental quarterly results cover 28 industries. 

Indices currently under development include: computer services, business use of hotels, rail freight, accountancy & management consultancy, architectural services, advertising placement & creation, legal services, office machinery rental, banking (commercial loans and deposits), and non-life insurance services. 

A variety of price mechanisms is used: actual transaction prices, list prices, model prices, fee-based-model fees, fees per period, percentage-based fees, or price data purchased from commercial suppliers. There are difficulties; for example, with model prices for accountancy, contracts may not be considered ‘typical’, prices charged depend on several different and changing factors; prices may be ‘loss-leaders’; or projects could be more broadly-based than just accountancy. 

Future plans embrace: continuing activities – development of new industries; quality assurance of existing industries; assessing more CSPIs for use as deflators; and liaising with international counterparts; and additional activities: review of samples; doubling the number of contributors; rebasing to year 2000; development of a price computer system, further links with consumer price index; strategy for quality adjustment; and maximising coverage of corporate services. 

Then, Louise Morris (ONS) discussed developments in the Producer Price Index (PPI) and Damon Wingfield (ONS), deputising for David Blunt, described the compilation of the Retail Price Index (RPI). The PPI had a new sample design with 7000 companies (compared to 3500 previously), rotated annually, selected from Prodcom with better weights, represented small businesses better and had optimal allocation. 

Finally, Alan Mankikar (Bank of England) gave an account of the use of retail prices at the Bank. The MPC’s remit is to keep RPIX inflation at 2.5% at all times; it has one target and one instrument to achieve it – the base rate. Consequently RPI data are crucial. Key tasks are to understand the movements in inflation and to forecast inflation. It is necessary to distinguish movements in relative prices from movements in prices that reflect the effects of changes in aggregate demand, as policy response will crucially depend on this. To do this, the range of price information examined includes: retail price components, derivative measures – core inflation, the distribution of price changes, domestically-generated inflation including whole economy (eg PGDP), sectoral prices (eg PPI & CSPI), other measures (eg consumption deflator & HICP), and foreign prices (eg world export prices). Survey data (eg BRC & CBI) are also used as are theory-based models (macro-model and supply chain models). In forecasting inflation, there are two horizons – short-term one quarter ahead and medium term. The short-term uses ‘bottom-up’ 25 disaggregated components projecting monthly changes based on four sources: seasonal patterns, ‘errors’ in the previous forecast, market intelligence and reports from the bank’s 12 regional Agents. The Bank’s ‘top-down’ medium-term macromodel has 20 behavioural equations and 130 variables. RPIY is the final behavioural equation. RPI is calculated by identity from RPIY, duty and council tax components to give RPIX and then MIPs to get RPI.