Economic Forecasting
Nicholas Carnot, Vincent Koen and Bruno Tissot, Palgrave Macmillan, 2005, xix + 325 pages, £65.00 (paper cover £22.50).
This book successfully fills the void between specialist textbooks with a heavy statistical content and more general economics textbooks. As such, it is a useful addition to the armoury of economic forecasters and non-specialists alike who wish to better understand the field of economic forecasting.
The book can be split into two parts. For those wishing to refresh their approach to forecasting Chapters 4 through to 10 describe and discuss the most widely-used forecasting tools, covering time-series methods, macroeconomic modelling and financial and sectoral forecasting. The inclusion of a stand-alone chapter devoted to medium- to long-run forecasting clearly highlights the shift towards models that can capture supply-side issues such as population and technology within a growth-accounting framework and which can address concerns such as the ageing population.
The remaining chapters are somewhat self-contained and can be read in isolation, and are consequently more amenable to the non-specialist. Chapter 2 presents a refreshingly concise description of the complex national accounts framework that underpins macroeconomic forecasting before moving on to set out the virtues and limitations of macroeconomic forecasts: data revisions, annual chain linking and the complete disregard of environmental and welfare considerations when measuring growth.
Chapter 3 on incoming news and near-term forecasting is a must-read for any economic analyst or economic commentator as understanding and interpreting data is fundamental. It should certainly be read before you start forecasting.
The age-old issue of forecast accuracy and the idiosyncratic methods of measuring accuracy from a statistical standpoint – can it be done, how do you measure it and so what if you can? – is successfully tackled in Chapter 12.
The section on communication challenges interested me most in my capacity as a forecaster. Many users often treat economic forecasts as given and misuse or simply misinterpret what an economic forecast is saying. It is most helpful and refreshing to have a response to the scepticism that all forecasts are inaccurate, unhelpful or simply wrong. Conveying the uncertainty and risk surrounding a forecast will benefit users of forecasts no end.
On full reading, the book shows that economic forecasting is not simply a number-crunching exercise: judgement and experience combined with sound statistical methods underpinned by theory are needed to ensure consistency of story and a good forecast.
In summary, the book achieves what it sets out to do: it provides a useful compendium that will not go amiss in the offices of economic forecasters and policy makers alike.
Nick Stewart
Experian Business Strategies Division