Ernst & Young ITEM Club Endorse UK NGDP Target
We have a fantastic high-profile endorsement of UK nominal GDP targeting this week, titled “The UK needs a new approach to monetary policy”, from the Ernst & Young ITEM Club. Their piece nails many of the arguments and they also dive straight in to the measurement debate:
This textbook argument for the superiority of money GDP over inflation targets seems compelling in current circumstances. Moreover, the money value of output is not just easier for the MPC to control, it is also much easier than inflation for the ONS to measure. We naturally look at our income and expenditure in pounds and pence, and companies do the same for their management, as well as the taxman and shareholders. Splitting these values into real and price components is relatively straightforward for a manufacturing firm producing standardised products like cars and computers that can be measured as physical quantities, but even there it is difficult to take account of technological improvement and quality change. However this is notoriously difficult (and arguably futile) for service sector firms that have sales invoices and perhaps figures for payroll and hours worked but few tangibles to go on. Digital products present even more formidable measurement problems.
The problems of measuring inflation in the consumer sector have been highlighted by the ONS proposals to change the methodology used to construct the RPI. Technically, it is very difficult to devise indices that do not suffer from an in-built upward bias. The RPI is particularly prone to this problem and has the awkward property that it shows an increase in prices if they fluctuate during a period but return to where they were at the beginning! Indeed, the advent of the new CPIH and the new RPI measures proposed by the ONS in their attempt to duck the issue of index bias has left us with no less than nine official inflation measures. It is also very difficult for the ONS statisticians to allow for discounting. The problem is that the general move to more frequent and pervasive discounts will bias measures of consumer prices up and bias measures of real spending down.
They miss a discussion of the importance of targeting the level rather than rate, to make up for past errors, but it is still an superb endorsement. Bravo!