Annex 1: Estimating the output gap for mainland Norway

Norges Bank. Monetary Policy Report, Sep 2003

The output gap is an expression of total resource utilisation in the economy, and can be used as an indicator of domestic inflationary pressures. The output gap is defined as the difference between the actual output level in the economy and the level that is consistent with stable inflation over time. This level is often called trend output. If the output gap is positive, the level of activity in the economy is so high that it generates pressures on economic resources, resulting in rising price and cost inflation. An output gap of zero indicates an activity level that is consistent with stable price and cost inflation.

There is uncertainty associated with quantifying output gaps. Trend output is not observable, and has to be estimated. Different methods yield different results. The charts show the output gap as estimated by Norges Bank, Statistics Norway, the IMF and the OECD. Despite some differences in calculation methods, the charts present a fairly similar picture of the output gap. Following the downturn in around 1990, the output gap gradually widened up to 1998. Thereafter, the gap narrowed and is approaching zero in 2002-2003. In the following we take a closer look at the calculation methods used for the different output gaps.

Norges Bank

Norges Bank uses a Hodrick-Prescott filter (HP filter) to estimate trend output. This is a widely used method based on the assumption that a time series can be decomposed into a trend component and a cyclical component. The method allows gradual changes in trend output over time, while more short-term fluctuations are assumed to reflect cyclical variations in demand. One advantage of this method is that only figures for actual output are required to calculate trend output. One weakness of the method is that it is necessary to choose a priori a parameter ([lambda]) that expresses the degree of variance in trend output - i.e. how rigid the trend should be17. Another weakness is that economic fluctuations towards the end of the estimation period are heavily weighted and can make it difficult to interpret the output gap in real time. This problem can be avoided to some extent by extending the data series.

Norges Bank's calculations of trend output are based on national accounts figures for mainland GDP for the period 1970-2001, extended by our own estimates for the period 2002-2005. Following discretionary assessments, Norges Bank has chosen to use [lambda] = 100 when calculating the output gap using annual data. To prevent the end point being too heavily weighted, the series is extended with GDP growth equal to the long-term trend. Norges Bank uses an HP filter to calculate trend growth up to the end of 1999. A discretionary estimate for trend growth is made for subsequent years, in part to take account of known supply side factors that are not captured in an estimate using a basic model. The increase in the number of vacation days in 2001 and 2002 reduced growth potential of the economy by an estimated 1Xz percentage point in each of these years. As a result, the output gap will remain at a somewhat higher level than would have been the case if the HP filter were used mechanically. In Inflation Report 2/03, Norges Bank estimated that the output gap will narrow from 1.4 per cent of trend GDP in 2002 to approximately zero in 200318.

Statistics Norway (SN)

SN estimates cyclical deviation - the difference between actual GDP and trend GDP - by means of an HP filter. They use quarterly national accounts figures for mainland GDP and extend the series with their projections for the forecast period. Annual growth is thereafter set equal to trend growth to avoid the problem of placing excessive weight on the end point. SN calculates the cyclical deviation using quarterly figures with [lambda]=40 000 (the size of [lambda] is influenced by the frequency of the data).

SN's cyclical deviation shows roughly the same path as Norges Bank's output gap, but is somewhat lower towards the end of the period. This is partly because SN uses the HP filter more directly, and does not correspondingly adjust for the reduction in potential growth in 2001 and 2002 which was a result of additional vacation days. According to SN's estimates in the Economic Survey in September 2003, the output gap will be negative towards the end of 2002 and trough in mid-2003. The average output gap for the Norwegian economy for the year is estimated at 0.5 per cent of trend GDP in 2002 and -1.2 per cent in 2003.

Ministry of Finance

The Ministry of Finance does not publish regular estimates of the output gap in Norway. The most recently published figures for trend GDP were in the National Budget for 2001. Although figures for trend GDP are not published regularly, the concept is often used in the Ministry of Finance, as trend GDP is used in calculating the budget indicator.

OECD

The OECD calculates the output gap by means of the production function method. With this method, trend levels for labour, capital and available technology are inserted into a specified production function. Potential production is then determined by trend growth in factor inputs. The production function method is a less mechanical method than the HP filter method, and may be more relevant for macroeconomic assessments. The drawbacks of this method are that it assumes that the economy can be described by means of a simple aggregate production function (for example, a Cobb-Douglas function), and that it is data-intensive and requires estimates of normal total factor productivity and the level of equilibrium unemployment.