Classification of institutional sectors follows the recommendations (without the split in national, private and foreign controlled sub-sectors). The activity classification is based on the Swedish Standard Industrial Classification, which is based on the EU classification NACE Rev. 2 and is at aggregated level identical with ISIC Rev. 4. The product classification into 400 product groups is based on EU's central product classification CPA. Final consumption expenditure is classified by purpose. General government by COFOG and households by COICOP. Household is also classified by durability. Gross fixed capital formation is cross-classified by industry and type of capital good.
Scope of the data
- Transactions coverage: All activities production boundary are included except for illegal transactions that are not properly covered.
- Geographic coverage: The Swedish economic territory includes the mainland, Swedish embassies abroad less foreign embassies in Sweden and Swedish ships and aircrafts in international transportation.
- Geographic detail: Annual regional breakdown is available on counties for value added, wages and salaries, employment and household disposable income. Gross fixed capital formation is distributed on 8 regions.
Output is valued at basic prices, intermediate consumption at purchasers prices and value added at basic prices. Final use is valued at purchasers prices. Exports of goods is valued f.o.b. and imports of goods c.i.f. On total level imports of goods a f.o.b. value is calculated. Owner-occupied housing is valued at comparable rental value. Non-market production is valued at cost.
Source data collection programs
- Production approach: Annual calculations are more detailed and we can get value added from output less intermediate consumption. Quarterly calculation is less detailed and we have to use indicators in many cases. Change in value added for market producers and producers for own final use are calculated from volumes and we get current prices by using price indices. For agriculture, forestry, fishing, energy, construction and some service activities special information is used. Manufacturing is calculated by using production volume indices and most services by using information from survey on turn-over in the service sector. For non-market producers, we get surveys for local government and accounting record data for central government direct in current prices.
- Expenditure approach: For household final consumption, most indicators are from the survey on turn-over in the service sector. The final consumption in government is calculated from the same information as for the production side. Information for exports and imports are from the balance of payments statistics published by the central bank. Data on gross capital formation comes from quarterly surveys on gross fixed capital formation and changes in inventories.
Source data statistical techniques
- Annual data on production and expenditure is balanced in an integrated supply and use table framework. This is done on 400 product groups. Quarterly data are then adjusted to the annual levels. Most quarterly data are derived by volume-indicators and shortage in coverage will indirectly by minimized. An aggregated model of the annual supply and use system has recently been implemented to improve the quality on quarterly data.
- Production approach: To get current price data on the production side, producer price indices are used for manufacturing. For services, new indices are constructed to improve our calculations. These are continuously introduced and replace earlier indices based on salaries or consumer price indices.
- Production side in government is calculated in constant prices by using information on hours worked in the sub-sectors for compensation of employees and indices from gross fixed capital formation for consumption of fixed capital.
- Expenditure approach: On the expenditure side we are using consumer price indices for household consumption; export and import price indices for exports and imports; and price indices for construction of buildings and producer price indices for gross fixed capital formation. These indices are used to get from volume to current prices and from current to constant prices.
- Government final consumption consists of intermediate consumption, compensation of employees and consumption of fixed capital. Intermediate consumption is deflated by using a mix from consumer prices indices and producer price indices.
- Base year and update cycle: SNA/ESA recommendations are adopted for calculation of values at constant prices, where the price level of the preceding year is used. The advantage of this method compared to the method used earlier - with a fixed base year - is that the different components are weighted together using latest possible information and no part of the system will have wrong weights due to changed price relations. For presentation of the constant prices which are calculated at different price levels, the figures are chained to the price level of a common year. The Swedish national accounts are now presented at constant prices in 2000 reference year prices. With reference year prices, the different components and aggregates are not additive. Each different time series must be calculated separately.
Other statistical procedures
Reconciliation: The quarterly accounts are revised during the ongoing year and when data for the 1st quarter next year is published. Earlier years are normaly revised when new annual accounts are released in November. At that time full annual national accounts are published for the year two years back in time and also preliminary accounts for the previous year. At certain times, annual accounts for a longer period are revised which means that also the quarterly accounts are revised for that period. The differences between annual and quarterly data is spread to the quarters with unchanged seasonally pattern if no new information is available.
Seasonal adjustment: Calendar adjustments and seasonal adjustments are integrated and made in Tramo/Seats, a computer software recommended by Eurostat. Calendar adjustment of value added and employment is made in two steps. First number of working days quarterly are calculated. Different industrial sectors are assumed to produce at a weighted combination of five or six days a week and continuous production. Often only one type of production week is assumed. The share may differ between value added and hours worked. A ratio between the weighted number of working days an average quarter and the selected quarter is calculated. The proportional effect calculated from the calendar is used in the regression model for the variable to adjust for seasonality. The calendar adjustment is then estimated as a part of the model. For most variables this treatment gives a smaller calendar effect than if it was calculated proportionally. Each variable has a dedicated model for seasonal adjustment and all models are revised, based on new quarterly information available after the calculation of annual accounts in November. In the tables showing GDP from the expenditure side only GDP is calendar adjusted in this way. For some of the other components, the software tests for calendar effects.
Seasonal adjustment is a method to make analysis of time series less difficult. The results may differ depending on the chosen method and input parameters to the model. Also the results for latest period have a tendency to be considerably adjusted in the accounts when succeeding quarters are calculated. Calendar adjustment is not made on the annual figures so quarterly data cannot be expected to sum to the annual figures. The whole period is adjusted every quarter and new seasonally adjusted figures for the whole period are published. Calendar-adjusted figures are treated differently and new figures are published only for quarters with new published actual figures. The values are then calculated from the percentage change for those quarters. Due to the way of calculating constant prices the series are not additive.
Further information is available through Statistics Sweden: