GDP data revision

The assessment of gross domestic product (GDP) in 2004 Q1 and Q2 in this and the previous Inflation Report is based on newly adjusted time series of the GDP component indicators. The revised quarterly national accounts figures have been calculated using a new methodology meeting the requirements of the European System of Accounts (ESA 95). The overall revision of the national accounts data at both constant and current prices for the period 1990-2003 should be completed in 2005.

The revision involves a number of methodological changes modifying the previously published GDP data. These changes include a revision of the estimate of GDP formation, the inclusion of public sector fixed capital consumption under services (i.e. government consumption), a change in the imputed rentals methodology, a revision of the estimate of wages and salaries in kind, a revision of the calculation of insurance services and a revision of the estimate of household consumption. Also of great importance was a change to a new method of converting data from current prices to constant prices. In contrast to the former "base" method, i.e. the conversion of time series to the prices of a base year a new method is now applied whereby the indicators are first converted to average prices of the previous year and only then into base year prices using chained year-on-year implicit deflators ("chain indices"). With this method, however, the aggregated variables no longer equal the sum of their individual components. In the case of GDP this means that total GDP at constant prices does not equal the sum of its expenditure components. The new method should better express the changes ongoing in the structure of the Czech economy. Chart 1 (Box), showing this indicator's development under both the former and new methodology, gives an idea of how the ESA methodology affects the past GDP profile at constant prices.

Graph 1 (Box)
The view regarding the course of the cycle was essentially unchanged by the revision to the GDP data (annual percentage changes, constant prices)

The comparison of year-on-year GDP growth under the former and new methodologies shows that the introduction of the ESA methodology has generated no fundamental changes in the past GDP profile. Table 1 (Box) provides a better picture of the effect of the methodological changes on GDP on the demand side specifically in 2003. The overview of real GDP in 2003 under the former and new methodologies shows that the biggest changes occurred in the categories of gross fixed capital formation and government final consumption expenditure. Overall, changes in structure prevailed; the revision of 0.2 percentage point to the total GDP growth rate in 2003 was insignificant.

Tab. 1 (Box)
The estimate of household consumption growth was reduced by the revision, whereas the opposite applied to investment and government consumption (annual percentage changes, percentage points)