NR-PM: 2010:16
Quarterly N ational Accounts
Inventory
Sources and m ethod s for the Sw ed ish
N ational Accou nts
Septem ber 2010
Quarterly N ational A ccounts Inventory
Contacts for Quarterly N ational Accounts:
Jessica Engd ahl
E-m ail: jessica.engd [email protected]
Telephone: +46 8 506 94117
And ers Jäd er
E-Mail: and ers.jad [email protected]
Telephone: +46 8 506 94728
1
Quarterly N ational A ccounts Inventory
Table of contents
Chapter 1 Overview of the system of quarterly accounts ................................... 3
1.1 Organisation and institutional arrangements ....................................................... 3
1.2 Publication timetable, revisions policy and dissemination of QNA ...................... 3
1.3 QNA compilation approach .................................................................................. 4
1.4 Balancing, benchmarking and other reconciliation procedures ........................... 4
1.5 Volume estimates ................................................................................................ 5
1.6 Seasonal adjustment and working day correction ............................................... 5
1.7 Additional information .......................................................................................... 5
Chapter 2 Publication timetable, revisions policy and dissemination of
QNA ........................................................................................................................... 6
2.1 Release policy...................................................................................................... 6
2.2 Contents published .............................................................................................. 7
2.3 Special transmissions .......................................................................................... 9
2.4 Policy for metadata .............................................................................................. 9
Chapter 3: Overall QNA compilation approach .................................................. 10
3.1 General architecture of the QNA system ........................................................... 10
3.2 Balancing, benchmarking and other reconciliation ............................................ 11
3.3 Volume estimates .............................................................................................. 15
3.4 Seasonal adjustment and working day correction ............................................. 16
Chapter 4 GDP and components: the production approach ............................. 18
4.1 Gross value added, including industry breakdowns .......................................... 18
4.2 FISIM ................................................................................................................. 23
4.3 Taxes less subsidies on products...................................................................... 24
Chapter 5 GDP components: the expenditure approach .................................. 25
5.1 Household final consumption ............................................................................ 25
5.2 Government final consumption, including split into individual/collective
consumption ............................................................................................................ 26
5.3 NPISH final consumption ................................................................................... 28
5. 4 Gross fixed capital formation ............................................................................ 28
5.5 Changes in inventories and valuables ............................................................... 30
5.6 Imports, exports ................................................................................................. 30
Chapter 6 GDP components: the income approach .......................................... 32
6.1 Compensation of employees, including components (wages and salaries) ...... 32
6.2 Taxes less subsidies on production ................................................................... 33
6.3 Gross operating surplus & mixed income .......................................................... 33
Chapter 7 Population and employment ............................................................... 34
7.1 Population .......................................................................................................... 34
7.2 Employment: persons ........................................................................................ 34
7.3 Employment: hours worked ............................................................................... 35
Chapter 8 From GDP to net lending/borrowing .................................................. 37
8.1 Primary income from/to the ROW (D.1 to D.4), gross national income ............. 37
8.2 Consumption of fixed capital (K.1), net national income, acq. less disp. of
non-financial and non-produced assets (K.2) .......................................................... 40
8.3 Current transfers from/to the ROW (D.5 to D.7), net national disposable
income (B.6n) .......................................................................................................... 40
8.4 Adjustment for the change in net equity (D.8), net saving (B.8) ........................ 41
8.5 Capital transfers (D.9), net lending/borrowing (B.9) .......................................... 41
Chapter 9 Flash estimates .................................................................................... 42
9.1 Flash GDP and employment estimate ............................................................... 42
Chapter 10 Main data sources used .................................................................... 44
2
Quarterly N ational A ccounts Inventory
Chapter 1 Overview of the system of
quarterly accounts
1.1 Organisation and institutional arrangements
The prod uction of statistics and responsibility for the various statistical
areas is organised in such a w ay that Statistics Sw ed en (SCB) has overall
responsibility for the coord ination and supervision of official statistics and
for the d evelopm ent of statistical nom enclature and classifications. In
ad d ition Statistics Sw ed en is responsible for coord inating international
statistical reporting and contributes actively to international cooperation.
Statistics Sw ed en has d irect responsibility for official statistics in cert ain
general areas of society. This applies, for exam ple, to the labour m arket, the
econom y, ind ustry and prices, the population and w elfare as w ell as to
housing and construction. In a num ber of other areas of society, som e 25
other agencies bear the responsibility for official statistics. The prod uction
of these statistics is effected partly by Statistics Sw ed en and partly by other
prod ucers of statistics.
The 1st of June 2008 Statistics Sw ed en w as reorganised and a separate
d epartm ent for N ational Accounts w as form ed . The N ational Accounts
Departm ent consists of six units:
Prod uct Accounts
Public Finance
Financial and Sector Accounts
Coord ination of Econom ic statistics
Econom ic Micro sim ulations
Econom ic Analysis
The quarterly as w ell as the annual GDP calculations are m ainly prod uced
at the Prod uct Accounts unit together w ith the Public Finance unit. A
separate unit are responsible for the Financial and Sector Accounts
includ ing the d ata that refer to sector accounts in the QN A d elivery
accord ing to ESA 95 transm ission program m e, i.e. the net
lend ing/ borrow ing as w ell as the household d isposal incom e.
1.2 Publication timetable, revisions policy and
dissemination of QNA
The quarterly accounts are published w ithin 60 d ays after the end of the
reference quarter. For the second quarter a flash estim ate is published
w ithin 35 d ays after the end of the quarter w ith alm ost the sam e content as
an ord inary release. Most requirem ents of the ESA95 Transm ission
program m e are fulfilled . The d om estic release is generally on a m ore
3
Quarterly N ational A ccounts Inventory
d etailed level than the ESA95 transm ission. The press release, the quarterly
publication (BN P-kvartal) and upd ated figures in the d atabase are
available at the w ebsite at 09.30 on the d ay of publication .
From 2009 the non-financial sector accou nts are published at the sam e tim e
as GDP. H ow ever, at this point no d ivision is d one betw een non -financial
and financial corporation, this split is d one in the full release of the sector
accounts about 90 d ays after the reference quarter.
Using new inform ation available, revisions take place w ith every new
publication of the quarterly figures. When releasing second , third and
fourth quarter the other quarters of the reference year can be revised w hile
figures in the previous years are fixed . When the first quarter is released all
quarters of the previous year m ay be revised . The quarterly accounts are
benchm arked to the final annual accounts w hen they are prod uced in
N ovem ber tw o years after the reference year . The benchm arked quarters
are published at the sam e tim e as the annual accounts to keep consistency
betw een annual and quarterly figures at every release.
1.3 QNA compilation approach
GDP are calculated accord ing to the prod uction approach as w ell as the
expend iture approach. The m ost com m on m ethod used in the calculations
is to extrapolate the N ational Accounts value from the sam e quarter
previous year w ith the grow th rate accord ing to an ind icator, e.g. value
ad d ed in a certain m anufacturing ind ustry is extrapolated w ith the grow th
rate estim ated in the Ind ustrial p rod uction ind ex for that ind ustry. For
som e parts of the expend iture approach values from the sources are used
d irectly. This is possible w hen the sam e source is used in the quarterly
accounts as in the annual accounts and it m ainly con cerns changes in
inventories and exports and im ports. The calculations are based on non ad justed values. GN I is calculated by ad justing GDP for prim ary incom e to
and from the rest of the w orld .
1.4 Balancing, benchmarking and other reconciliation
procedures
The balancing betw een the GDP by expend iture approach and GDP by
prod uction approach is d one w ith the help of supply and use tables. GDP
by incom e approach is balanced through a resid ual calculation of the item
operating surplus and m ixed incom e. Before balancing in the supply and
use table checks and valid ation of the source d ata are m ad e in the process
of calculation by the expert responsible for a certain area. Relations like
labour prod uctivity and hourly w ages are also analysed in the
reconciliation process.
When the annual calculations are com piled the quarterly accounts are
aligned to the annual accounts for the year at issue through the least square
m ethod , MinD4. Value ad d ed at a d etailed level is benchm arked through
MinD4 and the balancing of the quarterly supply and use tables and GDP
are m ad e after the benchm arking.
4
Quarterly N ational A ccounts Inventory
1.5 Volume estimates
The Sw ed ish N ational Accounts introd uced the m ethod of chain linking
constant prices in 1999 in coord ination w ith the im plem entation of ESA 95.
Calculations are m ad e in current prices and constant prices. The current
prices for the previous years, converted into average prices, form the base
for the calculations in constant prices. Using the average prices of year 2000
the tim e series are published in constant prices w ith reference year 2000.
The volum e changes are calculated through Laspeyres ind ices and price
changes are consequently calculated as Paasche ind ices. Constant prices
are also calculated for GN I and other incom e variables accord ing to table
0109 in ESA 95 transm ission program m e.
1.6 Seasonal adjustment and working day correction
Working-d ay and seasonal ad justm ents are integrated and m ad e in
Tram o/ Seats, the com puter softw are recom m end ed by Eurostat.
Seasonally ad justm ent is m ad e using a d irect approach, this m eans that
each series are ad justed separately.
Working-d ay ad justm ent is m ad e for value ad d ed in constant prices and
hours w orked . Seasonal ad justed figures are presented for value ad d ed in
constant prices and hours w orked at the sam e levels as w orking d ay
corrected figures. Seasonal ad justed figures are also presented for the
expend iture sid e in constant prices at an aggregate level. The w hole period
back to 1993 is ad justed every quarter and new w orking -d ay and
seasonally ad justed figures for the w hole period are published . For now no
ad justm ents are m ad e for d ata in current prices, w hich m eans that the
requirem ents of ad justed d ata in table 0102 in ESA 95 transm ission
program d ata are not fulfilled .
1.7 Additional information
Home page Statistics Sweden and the page for Quarterly
National Accounts
Startpage - w w w .scb.se
http:/ / w w w .scb.se/ Pages/ Prod uct____22922.aspx
Press releases of the quarterly National Accounts
http:/ / w w w .scb.se/ Pages/ List____250611.aspx
Statistical database
http:/ / w w w .ssd .scb.se/ d atabaser/ m akro/ MainTable.asp?yp=bergm an&x
u=scb&om rad ekod =N R&om rad etext=N ational+accounts&lang=2&langd b
=
Description of the annual and quarterly calculations (Swedish
only)
http:/ / w w w .scb.se/ statistik/ N R/ N R0103/ _d okum ent/ SOU2002.pd f
5
Quarterly N ational A ccounts Inventory
Chapter 2 Publication timetable,
revisions policy and dissemination of
QNA
2.1 Release policy
Quarterly estim ates are released w ithin 60 d ays of the reference period
follow ing the targets of the Action Plan on EMU statistical requirem ents
and of the Principal European Econom ic Ind icators. The press release, the
quarterly publication (BN P-kvartal) and upd ated figures in the d atabase
are available at the w ebsite at 09.30 on the d ay of publication. The release
d ates are published in Septem ber t-1 at the SCB w ebsite
http:/ / w w w .scb.se/ Pages/ PublishingCalend arStartPage____259922.aspx
At the sam e d ay as the release on the w ebsite d ata are sent to Eurostat
accord ing to the ESA95 transm ission program m e. For the second quarter a
flash estim ate is released w ithin 35 d ays after the end of the reference
quarter follow ed by an ord inary publication after about 70 d ays.
Using new inform ation available, revisions take place w ith every new
publication of the quarterly figures. The policy for revising the Sw ed ish
N ational Accounts is show n in table 2.1. When releasing second , th ird and
fourth quarter the other quarters of the reference year can be revised w hile
figures in the previous years are fixed . When the first quarter is released all
quarters the previous year m ay be revised .
The d etailed annual accounts t-2 is released at the sam e tim e as the QN A
for the third quarter at the end of N ovem ber. At that tim e the quarterly
estim ates for t-2 are benchm arked to the new annual figures using the
MinD4 m ethod . In this benchm ark process quarterly figures back to year t 5 are revised . The calculation of Quarterly N on-financial Institutional
Sector Accounts alw ays uses the QN A as the benchm ark, in ord er to be
fully consistent w ith QN A.
Table 2.1 Revision p olicy of the Sw ed ish N ational Accou nts (GDP calcu lations)
Reference
Year
T
T
T
T
Time of
publishing
End May
Mid September
End November
End February
T+1
T+1
T+1
T+1
End May
Revised
Mid September
End November Revised
End February
T+2
T+2
T+2
T+2
End May
Mid September
End November Final
End February
Q1 Year T Q2 Year T Q3 Year T Q4 Year T Year T
First
Revised
Revised
Revised
First
Revised
Revised
First
Revised
First
First sum of quarters
Revised
Revised
Revised
Revised sum of quarters
Revised
Revised
Revised
Prel. annual accounts
Final
Final
Final
Final annual accounts
6
Quarterly N ational A ccounts Inventory
2.2 Contents published
The release of the regular Quarterly N ational Accounts (QN A) as w ell as
the flash estim ate covers GDP by expend iture and GDP prod uction
approach w hereas no com plete com pilation of the incom e approach is
m ad e as the operating surplus is d erived as a resid ual.
2.2.1 ESA95 transmission programme
Tables accord ing to the ESA95 transm ission program m e for QN A are
d elivered to Eurostat the sam e d ay as the d om estic release of the results,
i.e. w ithin 60 d ays after the reference quarter. The requirem ents of ESA 95
transm issions are fulfilled to a large extent but som e com ponents are still
not com piled . All tim e series are available from 1st quarter 1993 and the
follow ing content is d elivered :
Table 0101: Gross value ad d ed at basic prices and gross d om estic
prod uct at m arket prices
- Com plete tables
Table 0102: GDP id entity from the expend iture sid e
- N ot com plete coverage of seasonally ad justed figures, all d etails is
not d elivered . Also no d ata on w orking d ays ad justed expend itures
are d elivered .
Table 0103: GDP id entity from the incom e sid e
- Com plete table
Table 0107: Disposable incom e, saving, net lend ing/ borrow ing
- Com plete table
Table 0109: Real d isposable incom e
- Com plete tables
Table 0110: Population and em ploym ent
- Com plete table
Table 0111: Employm ent by ind ustry
- N ot com plete coverage of w orking d ay ad justed and seasonally
ad justed figures, all d etails are not d elivered .
Table 0117: Final consum ption expend iture of household s by
d urability
- Com plete tables
Table 0120: Exports of good s and services by Mem ber states of the
EU/ third countries
- Com plete tables but the d ata that is on voluntary basis are not
d elivered .
Table 0121: Im ports of good s and services by Mem ber states of the
EU/ third countries
- Com plete tables but the d ata that is on voluntary basis are not
d elivered .
7
Quarterly N ational A ccounts Inventory
2.2.2 The release on the domestic website
In the d om estic release on the w ebsite the results of the quarterly
accounts are published on a m ore d etailed level than d elivered in the
ESA 95 transm ission for m any of the com ponents.
The tim e series run from 1993 on a d etailed level w hile d ata back to 1980
are published on an aggregated level covering the prod uction approach
and expend iture approach in constant prices as w ell as hours w orked .
The full set of N on-financial Quarterly Institutional Sector Accounts
(QSA) is released w ithin t+90 d ays after the reference quarter . Starting
2009 a first release of the sector accounts, w here no split is d one betw een
non-financial corporations and financial corporations w as m ad e
together w ith the QN A release, w ithin 60 d ays after the en d of the
quarter. Since before the prelim inary annual figures for the sector
accounts have been released together w ith QN A at the fourth quarter .
Data on net lend ing/ borrow ing have also been sent to Eurostat every
quarter together w ith the QN A d ata, accord ing to th e ESA95
transm ission program .
Production approach
Value ad d ed for m arket prod ucers and prod ucers for ow n final use are
published both in current and constant prices. Current prices are
presented in the national release at a breakd ow n of 15 ind ustries w hile
actual values in constant prices (reference year 2009) as w ell as grow th
rates are presented for 40 ind ustries. N on-m arket prod uction is
presented for Central governm ent, Local governm ents and N on -Profit
Institutions Serving H ousehold s (N PISH )
Working d ay ad justed and seasonally ad justed figures for value ad d ed
(reference year 2009) are presented w ith a breakd ow n of 15 ind ustries as
w ell as for central governm ent, local governm ent and N PISH .
Expenditure approach
The com ponents of the expend iture sid e are presented at the sam e levels
both in current prices and constant prices (reference year 2009). Grow th
rates are also show n for constant prices.
Final consum ption expend iture of household s (incl. N PISH 1) is
presented in the national release w ith a break d ow n by purpose, 14
subgroups, and d urability, 9 subgroups. Final consum ption expend iture
of general governm ent is split into central governm ent and local
authorities. Gross fixed capital form ation is p ublished by ind ustry, 15
subgroups, and by type of investm ent, 6 subgroups. Inventories are only
published as an aggregate although acquisitions less d isposals of
valuables are show n separately. For exports and im ports a d ivision is
m ad e betw een good s and services.
Seasonally ad justed figures are published on the m ain variables, w ith a
breakd ow n on 15 series, for the expend iture approach (reference year
2009).
Income approach and employment
Com pensation of em ployees, num ber of em ployed persons and hours
w orked are published w ith the sam e breakd ow n as value ad d ed , i.e. 40
8
Quarterly N ational A ccounts Inventory
ind ustries for m arket prod ucers and prod ucers for ow n final use. And
for non-m arket prod ucers the presentation is m ad e for central
governm ent, local governm ent and N PISH . Labour costs are split into
the tw o com ponents - w ages and salaries - and em ployers´ social
contributions and payroll taxes.
Working d ay ad justed and seasonally ad justed figures for hours w orked
are presented at the sam e level as value ad d ed , w ith a breakd ow n of 15
ind ustries as w ell as for central governm ent, local governm ent and
N PISH .
2.3 Special transmissions
The N ational Institute of Econom ic Research gets a special transm ission on
a m ore d etailed level than published , a d elivery that the Ministry of
Finance also get access to. A separate transm ission is also carried out for to
the Central Bank (Riksbank). A few other special transm issions are also
m ad e on subscription.
2.4 Policy for metadata
When publishing the quarterly accounts com m ents on the balancing
proced ures as w ell as com m ents on revised estim ates ar e posted on the
w ebsite. Brief general com m ents of sources used and som e concepts of
N ational Accounts are also includ ed . In these com m ents the d iscrepancy
betw een the prod uction approach and expend iture approach a re show as
w ell as the steps taken to balance the system . The causes of revision on
earlier quarters are also briefly explained .
Docum entation of the statistics accord ing to the stand ard of Statistics
Sw ed en (BAS), are available at the SCB w ebsite but only in Sw ed ish.
Metad ata are also available accord ing to the SDDS Special Data
Dissem ination Stand ard (SDDS) at the IMF w ebsite:
http:/ / d sbb.im f.org/ Applications/ w eb/ sd d sctycatbaselist/ ?strcod e=SWE
&strcat=N
9
Quarterly N ational A ccounts Inventory
Chapter 3: Overall QNA compilation
approach
3.1 General architecture of the QNA system
In term s of organisation the annual and quarterly accounts are m anaged by
the sam e organisational unit and in general the sam e persons responsible
for a certain area in the quarterly accounts are also m anaging these
calculations in the annual accounts.
The Gross Dom estic Prod uct is ind epend ently estim ated by the prod uction
approach as w ell as by the expend iture approach. The calculations are
carried out on a d etailed level and GDP is sum m ated from the d etails. For
GDP by incom e approach operating surplus and m ixed incom e is
calculated as a resid ual. H ow ever com pensation of em ployees is calculated
on a d etailed level as w ell as taxes less subsid ies.
The statistics that the quarterly accounts are based on are for m ost areas
not as com prehensive as for the Annual N ational Accounts. The d ata
sources are to a larger extent based on surveys. Com pared to the annual
accounts, the single largest d ifference in the d ata available is that no
inform ation of the com panies‟ interm ed iate consum ption is collected .
The m ethod s used in the quarterly accounts could be classified into the
follow ing three categories.
-
The value in the N ational Accounts are extrapolated buy using an
ind icator
The d ata from the source are used d irectly in the N ational Accounts
Mod els
For m ost areas of the quarterly accounts the short-term statistics used in
the calculation w ill be replaced by other sources w hen th e annual accounts
are com piled . The estim ates in the short-term statistics and the annual
statistics are rarely totally com parable. Som etim es there are d ifferences in
the d efinition of the variables an d even if the d efinition is the sam e the
estim ates w ill d iffer d ue to the fact that annual statistics in general have
better coverage. Thus the level estim ates in the short-term statistics are not
used d irectly rather the change betw een the reference quarter and the sam e
quarter previous year are used for extrapolation of the national accounts
value. For som e areas ind icators both in current and constant prices are
d elivered from the sources statistics, for other are variables d eflation into
constant prices (or reflation into current prices) are m ad e w ithin the
national accounts d epartm ent. Either w ay the price ind exes that are used
are analysed by the expert responsible for a certain ind ustry or other area
at the N ational Accounts d epartm ent to ensure the relation betw een
current prices and constant prices.
For som e variables the source is the sam e in the quarterly and annual
accounts, such as exports and im ports and inventories, so values from the
sources are used d irectly. For a few areas there is no inform ation available
10
Quarterly N ational A ccounts Inventory
on quarterly basis, e.g. gross fixed capital form ation in the agriculture
ind ustry, and for these areas suitable m od els are used .
The basis for the calculations is non -ad justed figures. The d etailed
calculations of the prod uction approach are d one at basic prices and
estim ates on taxes and su bsid ies are ad d ed at an aggregate level. The
expend iture approach is calculated at m arket prices on d etailed levels.
Although not integrated in the w hole calculating process a system of
supply and use tables are used in the balancing process. Seasonally
ad justm ent and calend ar ad justm ent are carried out in a separate process
and the release of the figures contains both non -ad justed and ad justed
figures.
Beginning from the end of the reference quarter the follow ing steps of the
calculation process can be id entified :
– 51 d ays
Acquisition of source d ata from surveys and ad m inistrative
sources
45 – 51 d ays
Valid ation of source d ata and calculation of initially
estim ates in the N ational Accounts
52 d ays
Reconciliation m eetings, scrutinizing the N ational Accou nts
estim ates
53 d ays
Prelim inary calend ar ad justed and seasonally ad justed d ata
available for analysis
53 – 58 d ays
Balancing in supply and use tables. Writing of articles and
the press release.
58 d ays
Chain-linking, calend ar ad justm ent and seasonally
ad justm ent
58 d ays
GDP non-ad justed and ad justed values d eterm ined
59 d ays
Preparation of release, w riting of press release etc
60 d ays
Release at 09.30
3.2 Balancing, benchmarking and other
reconciliation procedures
3.2.1 Quarterly GDP balancing procedure
It is very unusual for the first com pilation of GDP from the prod uction sid e
and the expend iture sid e to give a unanim ous estim ate of the GDP grow th.
N orm ally the d ifference in GDP-grow th accord ing to the tw o calculations
w ould be around 1 percentage poin t.
After the initial estim ates are calculated the calculations are scrutinized in
m eetings betw een the persons responsible for the balancing proced ure and
the persons responsible for the separate calculation . These d iscussions
could lead to noting that som e variables need further investigation and
contact is then taken w ith the source statistics. Along w ith this process the
balancing proced ure starts. About three d ays are spent on balancing the
11
Quarterly N ational A ccounts Inventory
supply and use in the accounts and rechecking and confirm ing p rim ary
d ata.
The balancing is d one in tw o parallel processes w here one group looks at
the m acro perspective analyzing the d evelopm ent of the d ifferent
com ponents of GDP. At the sam e tim e an analysis in supply and use tables
are carried out by another group . This analysis of supply and use is d one
for 77 prod ucts and ind ustries. The find ings from the tw o groups are
d iscussed in regular m eetings and d ecisions are m ad e as to w hat steps that
should be taken to balance the system . H ow ever, the tim e for investiga ting
problem s is quite lim ited . Therefore, w hen for exam ple no obvious
explanation is found in the contacts w ith source statistics, d ecisions as to
w hat variable should be ad justed often have to be based on general
know led ge about the prim ary statistics an d the relations betw een
prelim inary and final statistics, rather than on d irect inform ation about the
und erlying problem causing the d iscrepancies.
The quarterly supply and use tables
The structures in the supply and use tables are m ainly based on the
stru ctures accord ing to the latest d etailed annual accounts, concerning year
t-2. The input d ata is the existing QN A inform ation on outp ut and
expend itures. Using this d ata, the m od el generates the ad d itional d etail
required to com plete the supply and use tables for each quarter of the
estim ation period .
This m eans that it is assum ed that the supply and use relationships in the
annual tables also are relevant for the estim ation period of the quarterly
supply and use tables. For exam ple, it is assum ed that the share of
interm ed iate consum ption in relation to output is the sam e as it w as in year
t-2.
The tables are used to id entify d iscrepancies betw een the supply and use of
prod ucts that can then be investigated and resolved as far as possible prior
to publication. It is prim arily a tool for d etecting larger inconsistencies in
the accounts and for the m om ent the tables are not fully balanced at each
quarterly release. The full balancing of the tables is d one w hen d etailed
annual accounts are com piled for the reference period .
3.2.2 Benchmarking of QNA and ANA
SCB has used d ifferent m ethod s of benchm arking (BM). Up to 1985, the so
called Bassie m ethod w as used . After that both the MinQ m ethod and
MinD4 m ethod has been used . All m ethod s of benchm arking ad just
Quarterly N ational Accounts (QN A) to final Annual N ational Accounts
(AN A) in ord er to secure consistency in every tim e series and at the sam e
tim e m inim ize the revision of the quarterly tim e series. These and other
m ethod s of benchm arking have been investiga ted from several points of
view , theoretical, practical, and num erical and IT 1. The benchm arking
problem could be consid ered as a bias ad justm ent problem . Prelim inary
statistics includ e a certain am ount of bias. H ow this bias is d istributed on
See Öhlén (2006), „Benchm arking and Seasonal Ad justm ent – A Stud y of Sw ed ish GDP‟.
http:/ / epp.eurostat.ec.europa.eu/ pls/ portal/ d ocs/ PAGE/ PGP_DS_EUROIN D/ PGE_DS_E
UROIN D_WSA/ TAB58876947/ OH LEN %20AB.PDF
12
1
Quarterly N ational A ccounts Inventory
quarters in the sources of QN A is im portant and any benchm arking
proced ure should con sid er this issue. For exam ple, Ind ustrial prod uction
ind ex (IPI) is used as ind icators for the prod uction sid e of QN A. The
survey d esign for IPI (one sam ple a year) causes an und er coverage in
upsw ings. As a consequence, the bias is increasing d uring the year.
Due to the experience that the algorithm for Min Q d id not alw ays
prod uced reliable results, a research w as m ad e d uring 2007 to further
analyze the Min Q m ethod and if necessary, provid e a better proced ure. 2
This research suggested that Min Q should not be used . The old Denton‟s
m ethod as given by m inim izing the m easure D4 below seem ed to provid e
an acceptable solution.
q n
D4
(Z q / X q
Zq 1 / X q 1)2 ,
q 1
Where n is the num ber of quarters used . The D4 algorithm is further
d iscussed in Öhlén (2006). It has been program m ed in the SAS –softw are
and has been used for benchm arking of QN A 1993-2006 and w as
introd uced in 2007.
The find ings in Öhlén (2006) show that MinD4 (Dentons m ethod ) is qu ite
robust in term s of d ifferent structures of bias. MinD4 is also a linear
proced ure. This property is very practical because bench m arking could be
perform ed at the low est level. All aggregates of MinD4 benchm arked
series could be calculated as sum s of benchm arked series at low er levels.
For instance, if total exports of good s and total exports of services are
separately benchm arked , you d o not need to benchm ark total exports. It is
calculated as the sum of the tw o com ponents. The „im plicit‟ benchm arked
sum shares the optim ality properties from the com ponents.
The seasonal pattern of QN A from the prod uction sid e is consid ered m ore
reliable than the seasonal pattern of the d em and sid e. This is referring to
the initial estim ates. Therefore benchm arking is initially m ad e for the
prod uction sid e. This m eans that a balancing proced ure has to be m ad e
after the benchm arking of the quarterly figures. The benchm arking is
carried out for chain linked values and current price values and covers the
quarters of the year for the final annual accounts (year t-2) and the quarters
for the tw o preced ing years (t-3, t-4). A restriction for the benchm ark is
therefore that the fourth quarter of year t-5 should be unchanged .
3.2.3 Other reconciliation(s) of QNA different from balancing and
benchmarking
The person responsible for a calculation m akes the first analysis of the d ata
used in the calculations. Ad justm ents of source d ata to m eet the d efinitions
and coverage of N ational Accounts are m ad e. Explanations of larger
d eviations from a “reasonable” d evelopm ent of the variables accord ing to
the tim e series are sought. If the d ata d elivered from the source statistics
for som e reason seem s questionable contacts are taken w ith persons at the
2
See Öhlén (2006) op.cit.
13
Quarterly N ational A ccounts Inventory
source statistics to d iscuss the resu lts. This could lead to ad justm ents of the
d ata for the N A calculations. Even if no d irect faults are d etected in the
source statistics ad justm ents can be m ad e at the N ational Accounts if
original d ata seem s unreasonable, e.g. com plem entary inform ation point in
another d irection or the d ata d on‟t m atch other inform ation in the N A.
H ow ever, the m ain principle is that ad justm ent like this shall be d one
rarely and only if they are w ell ground ed .
In the calculation process the results of value ad d ed are also confronted
w ith the calculations of hours w orked , analyzing labour prod uctivity. This
analysis is m ad e to d etect apparent problem s in source d ata. H ow ever,
ad justm ent on basis of labour prod uctivity is d one w ith great caution since
there is no separate estim ate to confront the N A estim ate against and the
relation betw een hours w orked and value ad d ed is w eak in m any
ind ustries. Also hourly w ages are analysed and confronted w ith
inform ation on hourly w ages from the statistics on w ages and salaries in
the private and the public sectors.
3.2.4 Amount of estimation in various releases
N ational Accounts are only published once a quarter, except for the second
quarter w hen a flash estim ate is calculated . Betw een the releases of the
quarters w ithin a year m ost source statistics are revised to som e extent.
This is m ainly d ue to com plem enting reporting from the com panies. N ew
sources for the calculations are m ainly introd uced w hen the d etailed
annual calculations are com piled about 23 m onths after the reference year
(N ovem ber t+2). The m ost substantial im provem ent in source d ata w ould
be that inform ation on interm ed iate consum ption is available w hen the
d etailed annual are com piled .
Table 3.2 show s the revisions betw een the first tim e the quarter is
calculated com pared to the revised sum of quarters released in May, about
5 m onth after the reference year (May t+1), as w ell as the revision betw een
the first calculation of the quarter and the release of the d etailed annual
calculation in N ovem ber t+2, for the period 1999-2005. The results are
presented both as a norm al average for the period (colum n Average) as
w ell as the average in absolute values (colum n Absolute). The table show s
that revisions m ainly take place w hen the final annual accounts are
com piled for the reference period in N ovem ber t+2.
Table 3.2 Revision of qu arters in May year t+1 and N ovem ber year t+2
GDP
Household consumption
Government consumption
Gross fixed capital formation
Inventories (contribution to GDP)
Exports
Imports
May t+1
Average
May t+1
Absolute
Novmber t+2
Average
Novmber t+2
Absolute
0,0
0,2
0,4
0,6
0,0
0,3
0,3
0,3
0,1
0,6
0,1
0,1
-0,1
0,9
0,8
0,8
0,0
0,2
0,0
0,0
0,4
0,6
1,1
1,1
0,3
0,6
0,8
0,8
14
Quarterly N ational A ccounts Inventory
3.3 Volume estimates
Most d ata used in the N ational Accounts are collected in current prices and
then d eflated into constant prices using relevant price ind ices like prod ucer
price ind ices, service price ind ices and consum er price ind ices. H ow ever
for a few series quantities are used for calculating constant prices d irectly,
then using price ind ices to reflate into current prices. This is fore exam ple
the case for parts of the calculation on prod uction and consum ption of
energy. A third m ethod , m ainly used for non -m arket prod uction, is to
calculate constant prices and current prices separately, m aking prices
changes a resid ual. For m any parts of the calculations d eflation into
constant prices is carried out in the source statistics. This is for exam ple the
case for foreign trad e statistics, changes in inventories, ind ustrial
prod uction ind ex and service prod uction ind ex. For other parts of the
calculations d eflation is m ad e at the N ational Accounts unit, e.g. for
household consum ption, foreign trad e in services and gross fixed capital
form ation. Either w ay the calculations are m ad e at a very d etailed level
and the aggregated to the levels of publication. The calculations are based
on chain ind ices and the constant prices are therby based on the price level
the previous year (t-1). In the N ational Accounts as w ell as in the price
statistics Paasche price ind ices are calculated and thereby the volum e
changes in N ational Accounts and in source d ata are Laspeyres ind ices.
3.3.2 Chain-linking and benchmarking.
To be able to com pare d ata based on chain ind ices the constant prices need
to be chain-linked into a tim e series expressed in a price level for a certain
year, so called reference year. In the d om estic release the reference year is
3
the previous year . When changing to chain -linking in 1999 all three
m ethod s of chaining w ere consid ered , i.e. „Annual Overlap ‟ (AO), „Overthe-Year‟ (OY), and „Quarterly Overlap‟ (QO). In general, chaining is used
to justify tim e com parisons in tim es of structural changes of the econom y.
Until May 2010 the chaining w as m ad e w ith the OY m ethod . In May 2010
the chaining m ethod w as changed into the AO m ethod . The d ecision to
change to the AO m ethod w ere m ainly based on results from tests
show ing that the AO m ethod is m ore com patible w ith the m ethod s used
for benchm arking and seasonally ad justm ent.
When chaining accord ing to the Annual Overlap m ethod the form ula
below is used . In the exam ple the 2nd quarter of 2002 is calculated in
reference prices 2000, w here LQ stand s for Laspeyres volum e ind ex. The
base is the current price value for year 2000.
LQ(ao)2000,(2002,2)= LQ 2000,2001*LQ 2001,(2002,2)
w here
LQ 2000,2001 = the change betw een year 2000 and year 2001.
and
LQ 2001,(2002,2) = the change betw een year 2000 and the second quarter 2002.
3
In the ESA transm ission the reference year are 2000.
15
Quarterly N ational A ccounts Inventory
The form ula is used on transactions and aggregates w ith constant positive
or negative signs. For item s that change sign over tim e, other m ethod s
have to be used . In the Sw ed ish QN A changes in inventories in reference
year prices, that could be both positive and negative, are presented w ith
the sam e share of GDP as in t-1 prices. For external balance of good s and
services the net exports in referen ce year prices is subtracted from exports
and im ports in reference year prices.
The benchm arking is carried out in the chain -linked series and in current
prices. The benchm arked chain -linked series in reference prices are then
recalculated to constant prices (t-1). After cond ucting this benchm arking
proced ure the accounts w ill not be balanced . Therefore a new
reconciliation/ balancing is have to be carried out for all price levels i.e.
current prices, constant prices (t-1) and average prices.
Below , w e show the prod uction of QN A for a year t-2 to year t-4 for
d ifferent quarters w hen new AN A are available.
1) Chain-linking
2) Benchm arking of three years t-4, t-3 and t-2
a) Benchm arking,
b) Recalculation to current prices and constant prices (t -1
prices)
3) Reconciliation in current prices and constant prices (t-1 prices) for
the w hole period , year t-4, t-3 and t-2.
4) Chain-linking
5) Seasonal ad justm ent
3.3.3 Chain-linking and seasonal adjustment
As d escribed in 3.3.2 seasonal ad justm ent of volum e m easures is
perform ed after chain -linking and benchm arking. A d esirable property
from the user point of view is that seasonally ad justed annual totals equals
the non-ad justed annual totals, i.e. tim e consistency, even if this
requirem ent could how ever d eteriorate the quality of the season ally
ad justm ent som ew hat. Therefore the objective has been that the
requirem ent of tim e consistency shall be fulfilled for all series w here it is
technically possible.
3.4 Seasonal adjustment and working day correction
Until May 2010 a direct approach of seasonally adjustment were used. This
meant that each series was adjusted separately with an optimal model for the
specific series and no balancing was carried out and therefore the series
were non-additive. Based on user demand it was decided in 2009 that the
seasonally adjustment should be changed in a way so that the subseries
should be able to aggregate to totals. A project were launched to look into
what would be the best method to use for making the different series
16
Quarterly N ational A ccounts Inventory
additive . The result is a model where all series are direct seasonally
adjusted at first, then the subseries are reconciled so that they sum to larger
aggregates as well as to total GDP. The residual is split on the sub series in
proportion to the level of the series and the uncertainty of the seasonally
adjusted estimate where the uncertainty is measured as the variant in the
irregular component. The new model was introduced in the release of the
first quarter of 2010, in May 2010.
About 150 series of QN A are seasonal ad justed every quarter and includ es
all series of QN A from the prod uction sid e in constant prices as w ell as
hours w orked and the aggregates on the expend iture sid e in constant
prices. H ow ever the release of the prod uction sid e and hours w orked are
not m ad e at that d etailed level and as d escribed in chapter 2.2 all
requirem ents of the ESA95 transm ission program m e are not fulfilled as
series in current prices are presently not seasonally ad justed .
The DOS-program s of TRAMO/ SEATS is used together w ith an interface
w ritten in SAS. These program s have been used since 1998. Every series is
ind ivid ually m od elled based on the recom m end ations as given by Eurostat
since 1998. In Öhlén (2006) an overview of these issues is given 4.
Choice of ARIMA-model
Up till now , about 50 ARIMA-m od els have been investigated for every
series before the final choice is m ad e. The criteria used are based on
statistical principles and variability issues as w ell as graphical output on
the ad justed series and the resid uals.
Working day/Trading-day adjustment
All prod uction series and the num ber of hours w orked are w orking d ay
ad justed based on the Sw ed ish calend ar. The num ber of w orking d ays in a
quarter (norm alized ) is used as an external variable in the regression m od el
in TRAMO.
4
See Öhlén (2006) op.cit.
17
Quarterly N ational A ccounts Inventory
Chapter 4 GDP and components: the
production approach5
4.1 Gross value added, including industry breakdowns
In the first step of the calculations gross value ad d ed is d erived d irectly
through using ind icators for output, like prod uction or turnover, to
extrapolate the value ad d ed from the correspond ing quarter of the
previous year. For constant prices the base is the current prices previous
year converted into average prices. Average prices (t prices) for the
quarters are calculated from constant prices (t-1 prices) by u sing the
current prices (CUP) over the year d ivid ed w ith t-1 prices over the year.
This is expressed in the follow ing form ula:
00Q1t =00Q1t-1*Sum 00Qn CUP / Sum Qn t-1
N o statistical m od el is used to take into account the relationship betw een
the quarterly ind icator and the annual estim ate. Rather the com piler
responsible for a certain ind ustry evaluates how w ell the used ind icator
pred icts in calculating the future annual estim ate. Calculations are carried
out at both current prices and constant prices at the sam e level of d etail.
Interm ed iate consum ption is not part of the initial calculations but are
d erived w hen the supply and use tables are created and are analysed w hen
balancing the system of supply and use tables, as d escribed in section 3.2.2.
For m ost ind ustries both constant prices and current prices are d elivered
from the source statistics w here inform ation is collected in current prices
and then d eflated w ith relevant ind ices. Inform ation about prices m ainly
consists of Prod ucer p rice ind ices (source 24) and Service price ind ices
(source 25) although a variety of d ifferent ind ices are used . The m ain
sources for calculating value ad d ed for m arket prod ucers and prod ucers
for ow n final use is the Ind ustrial prod uction ind ex (source 4), N ew ord ers
and d eliveries (source 5), the Service prod uction ind ex (source 8) and the
Turnover statistics (source 7). H ow ever a lot of other sources are also used
as d escribed below .
The non-m arket prod uction in the governm ent sector is not calculated
d irectly by ind ustry but by COFOG 6. In the d om estic releases central and
local governm ent sectors are presented separately. H ow ever, governm ent
activities by ind ustry can be d erived by the connection betw een COFOG
5
The m ain sou rces u sed in the calcu lations are d escribed in m ore d etail in chap ter
10. The sou rces nu m bers refer to the nu m ber given in chap ter 10.
6
Classification of the functions of the governm ent
18
Quarterly N ational A ccounts Inventory
7
and N ACE (Statistical Classification of Econom ic Activities) and are
presented by ind ustry in the ESA95 transm ission program m e to Eurostat.
4.1.1 Agriculture, hunting and foresting and fishing (NACE
A+B)
The d ata used to calculate value ad d ed in the agricult ure indust ry consists
of an annual forecast on the crop harvest, or statistics on the output of the
crop harvest (source 1) d epend ing on the quarter calculated , and m onthly
statistics on anim al prod uction (source 2). All d ata is d elivered from the
Sw ed ish Board of Agriculture. The annual figure for the harvest is split in
equal shares over the four quarters accord ing to the SN A guid elines.
Calculations are m ad e for five subgroups and are based on inform ation
about quantities and prices. Anim al prod uction is calculated in the sam e
w ay, quantities m ultiplied by prices, and is split in seven subgroups. To
calculate constant prices the values in previous year‟s prices are
extrapolated using the change in quantities.
The calculations of prod uction in the fishing indust ry are based on
m onthly statistics on quantities and values from the N ational Board of
Fisheries (source 3). Inform ation is d elivered on a very d etailed level but in
the calculations the d evelopm ent of the land ing of the five m ain species of
fish are used as ind icators for the w hole ind ustry. Since the statistics only
cover salt w ater fishing ad justm ents are m ad e for cultivated fish and fresh
w ater fishing. Constant prices are calculated by extrapolating the value in
previous year‟s prices using the change in quantities.
Output in the forest ry indust ry is regard ed as being prod uced
continuously over the entire period of prod uction, not just w hen the tim ber
is felled . Grow ing trees are treated as inventories of w ork in progress.
They are transform ed into inventories of finished stocks w hen they are
m ature. Output is com posed of the prod uction of the felling of trees
together w ith the prod uction of stand ing tim ber (net increm ent).
Inform ation about saw m ill tim ber and pulp w ood prod ucts used in the
saw m ills and pulp m ills are used as ind icators for total prod uction for the
felling of trees. Saw m ill tim ber and pulp w ood d elivered from road sid e
inventories as w ell as im ports and exports of these prod ucts are ad justed
for. For exports and im ports d ata is available in foreign trad e statistics.
Inform ation about saw m ill tim ber d ata is collected by Statistics Sw ed en on
a m onthly basis and w ood pulp d ata is collected by The Sw ed ish Forest
Ind ustries Fed eration . Data on inventories of w ood pulp is collected on a
quarterly basis by Statistics Sw ed en. Statistics on inventories of saw m ill
tim ber are only available on an annual basis lead ing to m od el based
calculations for the first to third quarters of each year. The calculations are
first carried out in constant prices using d ata on quantities to extrapolate
value ad d ed . Curren t prices are d erived through using inform ation on
prices on saw m ill tim ber and w ood pulp.
Output of stand ing tim ber (net increm ent) in constant prices is d erived
from the total of cultivated forestry resources, w hich grow and are felled
on forestry land . Gross increm ent is d erived from annual d ata on the forest
7
In this d ocum ent N ACE allw ays refers to N ACE rev1.1
19
Quarterly N ational A ccounts Inventory
stand from the Sw ed ish University of Agricultural Sciences. The annual
d ata d ivid ed by four is used per quarter. The price used to calculate the
gross increm ent value in current prices is the d eliv ery prices of felled
tim ber d elivered to forest road sid es. The felling of trees is calculated as
d escribed above using inform ation on the use of saw m ill tim ber and w ood
pulp.
4.1.2 Manufacturing and mining (NACE C–D)
In the m anufacturing ind ustries value ad d ed in constant prices is
calculated by using the Ind ustrial prod uction ind ex (source 4) to
extrapolate the value ad d ed for the sam e quarter previous year. To
calculate value ad d ed in current prices, the m onthly inquiry on N ew
ord ers and d eliveries (source 5) is used for extrapolation of the value
ad d ed for a given ind ustry). Ind ustrial prod uction ind ex is m ainly based
on the survey on d eliveries in the m anufacturing ind ustries w hich is then
d eflated using Prod ucer price ind ices (source 24). The N ational Accounts
calculations are m ad e for 38 ind ustries.
In the N ational Accounts som e ad justm ents are m ad e to the source
statistics to enhance coverage and concepts. Since the source estim ates
d eliveries rather than prod uction ad justm ents are m ad e for changes in
inventories to take into account the prod uction that is not d elivered d uring
the reference quarter but put into inventories as w ell as for the d eliveries
that are m ad e from inventories that are not prod uced d uring the reference
quarter. These ad justm ents are m ad e based on the quarterly survey on
Ind ustrial Inventories and take into account the changes in inventories of
good s in process as w ell as finished prod ucts. The source d ata only covers
the prod uction of good s in the m anufacturing ind ustries but not the
prod uction of services. For som e ind ustries in w hich the prod uction of
services is m ore substantial supplem ents are m ad e for som e prod ucts, e.g.
for m erchanting, using inform ation about the export of services.
4.1.3 Energy (NACE E)
Calculations of electricity, gas, steam and hot w ater supply (N ACE 40) are
m ainly d one using d ata on quantities then reflating into current prices. A
variety of price ind ices are used , e.g. relevant Consum er price ind ices
(source 26) and Prod ucer price ind ices. The short-term inform ation is
d etailed and the m ain source is the Monthly electricity statistics (source 6).
Other inform ation used is Monthly fuel, gas and inventory statistics and
inform ation on im ports of gas.
The Turnover statistics are used to extrapolate value ad d ed in current
prices in N ACE 41, Collection, purification and d istribution of w ater. The
Consum er Price Ind ex for w ater is used for d eflation. In Sw ed en
w aterw orks are extensively integrated w ith sew eage functions. Thus it is
not possible to obtain d ata for w aterw orks separately and hence the
ind ustry covers both w ater supply and sew age d isposal.
20
Quarterly N ational A ccounts Inventory
4.1.4 Construction industries (NACE F)
The approach to calculating prod uction in the construction ind ustry uses
inform ation about constant capital form ation and repairs. The volum e
changes for investm ents and expend itures for purchased repair and
m aintenance services regard ing build ings, structures, and houses are used
as an ind icator for value ad d ed in constant prices. Repairs are b ased on the
assum ption that they d evelop in the sam e w ay they d id in the year t-2, (the
m ost recent d etailed annual calculation). Current prices are obtained by
reflating w ith a w eighted ind ex, an ind ex that is based on the im plicit
ind ex for total gross capital form ation in build ings, structures an d houses
as w ell as a num ber of construction ind ices reflecting the price change for
repairs.
4.1.5 General approach on services (NACE G-O)
The m ain source to calculate services, N ACE G-O, is Turnover statistics
(source 7) in current prices and the related Serv ice prod uction ind ex
(source 8) expressed as volum e change. The basis for the Service
prod uction ind ex is the Turnover statistics. For d eflation, a num ber of
d ifferent ind ices are used like Service price ind ex (source 25) and
Consum er price ind ex (source 26).
The inform ation from the Turnover statistics and Service prod uction ind ex
is used in the N ational Accounts to extrapolate value ad d ed in current
prices and constant prices, respectively. The calculations are carried out on
a d etailed level for about 40 ind ustries. Even if the Turnover statistics and
Service prod uction ind ex form the m ain source other inform ation such as
quarterly financial statem ents for larger com panies are also used as
reference inform ation in som e ind ustries. The follow ing are those
ind ustries w here Turnover statistics and Service prod uction ind ex are not
used as the m ain source for the calculation of value ad d ed w ith attend ant
com m ents.
4.1.7 Financial intermediation (NACE J)
The source of the calculations for financial interm ed iat ion is based on
quarterly surveys on banks and insurance com panies respectively (source 9
and 10) issued by Statistics Sw ed en on behalf of The Sw ed ish Financial
Supervisory. FISIM is d escribed in section 4.2.
For financial services d irectly m easured , com m issions etc (N ACE 65), the
value ad d ed in current prices is extrapolated from the d evelopm ent of
com m issions in banks, cred it m arket enterprises, securities corporations,
m utual fund s, fund corporations and investm ent corporations. The price
ind ex used for d eflation consists of a w eighted ind ex betw een the w age
ind ex for the financial ind ustries (N ACE 65-67) and an ind ex on fund s.
Insurance services (N ACE 66) consist of life insurance, pension fund ing,
non-life insurance, and reinsurance. Value ad d ed for life insurance is
extrapolated based on the d evelopm ent of ad m inistrative costs and the
w age ind ex for N ACE 65-67 is used for d eflation. For non -life insurance in
current prices value ad d ed is extrapolated using the change in output,
21
Quarterly N ational A ccounts Inventory
w here output is m easured in accord ance w ith ESA 95 as prem ium s,
includ ing equalisation provisions, applicable to the period plus prem ium
supplem ents less claim s d ue. Value ad d ed in constant prices is
extrapolated w ith the change in the num ber of insurances.
The value ad d ed for activities auxiliary to financial interm ed iation (N ACE
67) is m od el-based and is calculated as the w eighted average of the
d evelopm ent in N ACE 65 (exclud ing FISIM) and N ACE 66.
4.1.8 One-to-two dwelling houses and leisure houses (part of
NACE 70.2).
The calculation of constant prices is based on a w eighted volum e ind icator
for the d evelopm ent of the stock of one-to-tw o d w elling houses and leisure
houses. To calculate the d evelopm ent for one-to-tw o d w elling houses the
num ber of com pleted d w ellings are used , ad justing for m od ernising
renovations and d em olitions. For leisure houses inform ation on build ing
perm its are used as a volum e ind icator. Current Prices for one-to-tw o
d w elling houses is obtained by reflation using a w eighted ind ex of both
d w elling form s. Th e ind ices are partial ind ices of Consum er Price Ind ex.
4.1.9 Private households with employed persons, NACE P
Inform ation on the com pensation that d isabled persons receive to em ploy
personal assistants are the basis for the calculation of value ad d ed in
current prices. There is no interm ed iate consum ption in the ind ustry, so the
output value is equal to value ad d ed . Constant prices are com piled by
d eflation by an hourly w age ind ex for the self-em ployed persons.
4.1.10 Non-profit institutions serving households (NPISH)
Value ad d ed for N PISH s is m easured , in accord ance w ith the d efinitions in
ESA 95, using the cost m ethod . The extrapolation m ethod is used . Value
ad d ed are first calculated in current prices using inform ation on changes in
w ages and salaries and em ployers‟ social contributions accord ing to the
“Aggregate gross pay, payroll taxes and prelim inary tax statistics from
em ployers‟ m onthly tax returns” (LAPS, source 20) as w ell as inform ation
on capital consum ption and payroll taxes. Consum ption of fixed capital is
calculated using a m od el based approach as d escribed in chapter 8.2.
Deflation is d one using a w age ind ex.
4.1.11 Production in the Government sector
Value ad d ed in current prices is calculated using the cost m ethod . The
central governm ent sector accounting d ata for authorities is collected
quarterly by the Sw ed ish Financial Managem ent Authority (ESV) (Total
activity of central governm ent source 11). Local governm ent d ata is
attained from quarterly inquiries based on a sam ple of about 80 out of 290
m unicipalities and com plete coverage of the 20 county councils (source 12,
13).
Sw ed en introd uced volum e m easures for the calculations of the prod uction
of ind ivid ual services in constant prices in 2007. Accord ing to the
Com m ission Decision N o 2002/ 990 on the principles for m easuring prices
and volum es, volum e m easures are introd uced for the part of governm ent
22
Quarterly N ational A ccounts Inventory
prod uction that constitutes ind ivid ual services. For collective services at
constant prices the cost m ethod is used .
See further chapter 5.2 Governm ent final consum ption.
4.2 FISIM
The m ain sources for the FISIM calculations are the Riksbank‟s m onthly
balance d ata of Monetary Financial Institutions (MFI), com piled by Statistic
Sw ed en, and the Sw ed ish Financial Supervisory Authority‟s quart erly
balance and profit and loss d ata for financial enterprises. Im port and
export of FISIM is calculated using the Riksbank‟s Balance of Paym ents
d ata on inflow and outflow of interest.
FISIM for the d om estic sectors is calculated using m ethod 1 as outlin ed in
Council Regulation (EC) N o 448/ 98 of 16 February 1998 and thus uses the
follow ing three com ponents:
Average stocks of loans and d eposits for MFI except the central bank and
sector 123 (Other financial interm ed iaries, OFI, except insurance
corporation s and pension fund s,) are includ ed .
Accrued interest for MFI (except the central bank and OFI) by user sectors
is includ ed .
The internal reference rate given by the ratio of interest receivable on loans
to the stocks of loans betw een MFI (except the centra l bank and OFI) is
includ ed .
FISIM is calculated as follow s, (d eposit stocks × internal reference rate) interest payable on d eposits + interest receivable on loans - (loan stocks ×
internal reference rate).
To calculate the reference rate, quarterly, Sw ed ish m arket rates are used .
The rates that are used are STIBOR (Stockholm Interbank Offered Rate 3m onth period ), treasury bond s of 5-years d uration and the governm ent
lend ing rate. The am ount of interest, payable and receivable, is calculated
using the Riksbank‟s d eposit and lend ing rates upon the stock d ata.
Inform ation on profit and loss account d ata from the Sw ed ish Financial
Supervisory Authority is used for cross-checking the results of the FISIM
calculation.
Changes in volum e reflect the changes in stocks. For sub-sectors Banks and
H ousing Finance institutions the Riksbank‟s m onthly balance d ata of
Monetary Financial Institutions is used for d ata on stocks of loans and
d eposits by sector. An average of the opening and closing balance of each
m onth in the quarter is used . The correspond ing inform ation for the sub sector Other m onetary cred it m arket corporations is collected from the
Sw ed ish Financial Supervisory Authority‟s quarterly d ata. This
inform ation is also available by sector. An average of op ening and closing
balance is used .
The level of d etail and quality of the com ponents necessary for FISIM
im port and export calculations are not available in Sw ed en. The result of
the external reference rate calculations, the ratio of interest received on
23
Quarterly N ational A ccounts Inventory
loans from non-resid ent cred it institutions plus interest paid on d eposits to
non-resid ent cred it institutions to the correspond ing stocks of loans, are
very volatile and therefore not suitable for use. The volatility probably
stem s from the financial reports of the non-resid ent branches. The level of
d etail of these reports is d ifferent from the d om estic one. The Riksbank‟s
Balance of Paym ents d ata on interest inflow and outflow by sector is used
to calculate im port and export for FISIM. The ratio of FISIM to accrued
interest for each d om estic sector is used to calculate inflow and outflow for
FISIM. Only Monetary Financial Institutions (MFIs) export FISIM. FISIM
betw een resid ent and non -resid ent MFIs are record ed in term s of net
results. Therefore only one third of the ratio of FISIM to accrued interest is
used for this sector.
Municipalities and county councils‟ loans in foreign currency are used as a
key to d ivid e FISIM into the tw o sub-sectors of local governm ent.
On a quarterly basis FISIM is allocated by ind ustry through the supply and
use analysis.
4.3 Taxes less subsidies on products
In calculations of GDP prod uction approach, value ad d ed is valued at basic
prices. Taxes and subsid ies on prod ucts are ad d ed or subtracted , on an
aggregate level in ord er to obtain GDP at m arket prices.
The calculations for taxes (exclud ing VAT) and subsid ies on prod ucts are
based m ainly on the record s from the Financial Managem ent Authority
(Ekonom istyrningsverket, ESV) of the incom e of central governm ent
d epartm ents and agencies und er revenue head ings, w hich are upd ated on
a m onthly basis. Period reallocations m ust be und ertaken in ord er to obtain
the accrued value, since ESV record s are cash -based and paym ents entered
und er the revenue head ings are usually m ad e in arrea rs.
To attain VAT in the prod uction and expend iture sid e calculations of GDP,
theoretical VAT is u sed . This is calculated in the supply and use tables for
the d ifferent uses. The calculations are m ad e both in current and constant
prices.
24
Quarterly N ational A ccounts Inventory
Chapter 5 GDP components: the
expenditure approach
In the quarterly accounts as w ell as in the annual accounts com plete
calculations are m ad e of the expend iture approach, i.e. on an aggregate
level no com ponent is d eterm ined as a resid ual.
5.1 Household final consumption
H ousehold final consum ption is com piled on a d etailed level, in
accord ance w ith 147 consum ption purposes (COICOP 8). The calculations
are m ostly based on extrapolation and m any d ifferent sources are used to
get ind icators for the extrapolation. The m ain source is Turnover Statistics
(source 7) w hich accounts for about 30 percent of total household
consum ption. Besid es Turnover Statistics a lot of other sources are used .
For som e consum ption purposes w here one large com pany is d om inant
specific inquiries are used , or inform ation is collected through the
com pany's financial statem ents. For other consum ption purposes like
energy and cars, inform ation from m onthly and quarterly statistics
com piled by Statistics Sw ed en are used . The m ain m ethod s and sources for
the calculations are d escribed below w ith exam ples on consum ption
purposes w here the m ethod is used . H ow ever, the d escription d oes not
cover in d etail w hat m ethod is used for the d ifferent consum ption
purposes.
Household consumption matrix
For a num ber of consum ption purposes, m ainly w ithin clothing and
footw ear (COICOP 3) and furnishing, household equipm ent, and routine
m aintenance (COICOP 5) a m atrix is used to split the turnover figures from
the Turnover Statistics into prod uct groups and to d eterm ine how m uch of
these prod ucts are used by household s.
The Turnover Statistics m easure total turnover in each sub -ind ustry. Since
the ind ustries sell a large num ber of d ifferent good s and services, and
household consum ption is calculated and record ed for each good and
service ind ivid ually, the trend figures for the d ifferent ind ustries have to be
converted to trend figures for the various good s and services. This is d one
by a m atrix w here prod uction (turnover) by ind ustry is split into those
good s and services that are sold and how m uch of such is ascribed to
household consum ption. Inform ation on prod ucts by ind ustry is attained
from the Retail Trad e Survey. By com paring the d istributed values
betw een tw o years for the sam e quarter, a trend is obtained for the good
and service in question. These trend figures are linked to the d ifferent
purposes for calculating the quarterly levels of household consum ption.
The values calculated through the m atrix in current prices are d eflated
w ith Consum er Price Ind ices (CPI).
8
Classification of Ind ivid ual Consum ption by Purpose
25
Quarterly N ational A ccounts Inventory
Other methods
When d om estic prod uction is alm ost exclusively used for consum ption of
the resid ent household s, like it is for som e services, the sam e ind icator
used for extrapolation of value ad d ed is also used for extrapolation of
household consum ption, assum ing that household consum ption is a fixed
fraction of the prod uction. Som e of the purposes are, how ever, estim ated
this w ay because there are no other good ind icators available. The
Turnover Statistics are used this w ay for som e services, for exam ple for
consum ption groups w ithin transports (COICOP 7) like taxi, railw ay
transportation and public transportation. This m ethod is also used for
electricity, gas and heating (COICOP 045) and financial services (COICOP
126) w here the calculations are based on inform ation from m onthly and
quarterly statistics com piled by Statistics Sw ed en and The Financial
Supervisory Authority respectively. For the consum ption in occupant
ow ned hom es (COICOP 042) w here household consum ption is the only
use for prod uction the output m ethod is used as w ell. This also applies to
other services w ithin housing.
For som e consum ption purposes w here one large com pany d om inates,
specific inquiries are used or inform ation is collected through the
corporation‟s financial statem ents. This for exam ple provid es the basis for
calculating the consum ption of alcoholic beverages and tobacco (COICOP
021, 022) prescription m ed icines (COICOP 06111), parts of com m unication
(COICOP 08), gam ing (COICOP 0943), and new spapers and stationery
(COICOP 0952-54).
H ousehold consum ption of m otor vehicles (COICOP 071) is calculated
based on the ad m inistrative source, Statistics Sw ed en‟s vehicle statistics
(source 14), and inform ation on average prices.
Most COICOP groups d escribed above are calculated based on
extrapolation of current prices and d eflation into constant prices using CPI.
A few COICOP, how ever, are extrapolated in constant prices using
inform ation on quantities and CPI is then used for reflating to current
prices. This m ethod is for exam ple used in the calculations of the
consum ption of electricity, gas and heating (COICOP 045).
The calculations of household consum ption are based on inform ation
about sales in Sw ed en, thus supplem ents are m ad e for d irect purchases by
Sw ed es abroad (COICOP 15) and d ed uctions are m ad e for purchases in
Sw ed en by non-resid ent household s (COICOP 16). These ad justm ents are
m ad e on an aggregate level using inform ation in current prices from
exports and im ports of services. A w eighted consum er price ind ex based
on a nu m ber of d ifferent countries is used for d om estic consum ption
abroad . Consum ption expend iture by Sw ed es abroad is d eflated by an
average of consum er price ind ices for the countries of the m ost frequent
visited tourism d estinations.
5.2 Government final consumption, including split into
individual/collective consumption
Governm ent final consum ption in current prices is calculated from the cost
sid e. The calculation of total consum ption expend iture is broken d ow n into
the com ponents interm ed iate consum ption, w ages and salaries, social
26
Quarterly N ational A ccounts Inventory
contributions, other taxes on prod uction, other subsid ies, consum ption of
fixed assets, sales, other assets prod uced for ow n account, and social
benefits in kind . Central governm ent d ata in current prices is obtained at a
d etailed level. The calculations are based on the record s of the Sw ed ish
N ational Financial Managem ent Authority (ESV) for the expend iture of
d epartm ents and agencies of central governm ent by type of expend iture,
the so-called total activity of central governm ent (source 11). The basis for
these record s is d erived from the central governm ent accounting system
and applies to all the expend itures of these bod ies, even those financed
from sources outsid e the central governm ent bud get. All expend iture is
thus treated as consum ption, investm ent or a transfer and is specified by
type of expend iture and purpose. The calculations of the social security
sector, w here the national pension insurance fund s constitute the m ain part
of the sector, are based on forecasts m ad e by the Social Insurance Office as
w ell as annual financial statem ents.
The statistical sources for county councils and m unicipalities are quarterly
surveys that collect rather d etailed inform ation on costs and incom es. The
survey for m unicipalities is a sam p le survey cond ucted by Statistics
Sw ed en (source 12) and the survey for county councils is a census covering
all units and also cond ucted by Statistics Sw ed en (source 13). In 2008 the
local authorities com prised 290 prim ary m unicipalities, 18 county councils,
2 regions and 103 local fed erations. The calculations of the local
governm ent sector includ e the Sw ed ish Association of Local Authorities,
the Fed eration of Sw ed ish County Councils and non -profit institutions
belonging to the local governm ent sector.
Since the third quarter 2007, volum e m easurem ents for governm ent
prod uction have been introd uced in the Sw ed ish N ational Accounts on an
annual basis as w ell as on a quarterly basis. Accord ing to the Eurostat
d ecision, volum e m easurem ents are only used for the ind ivid ual
prod uction. About 65 percent of the total prod uction in the governm ent
sector is ind ivid ual prod uction in Sw ed en. On an annual basis 7 per cent of
the ind ivid ual prod uction is not covered by volum e m easures. On a
quarterly basis the inform ation is scarce and the volum e m easurem ents are
to a large extent based on forecasts or on estim ates accord ing to the last
annual accounts. One area w here forecast is used is for the consum ption
purpose Ed ucation (COFOG9 9). The annual volum e m easurem ent is based
on num ber of stud ents/ stud ent hours and the quarterly estim ates are
based on forecasts using d em ographic d ata. This is believed to give
relatively good estim ates. The m ost im portant areas to get better
inform ation on a quarterly basis is purposes w ith a large w eight like Old
age (COFOG 1020), Sickness and d isability (COFOG 1012) and parts of
H ealth (COFOG 7). Old age and sickness and d isability are also based on
forecast using d em ographic d ata. For health w here there is no inform ation
available on quarterly basis and the quarterly estim ates are the sam e as in
the last annual accounts. Work is ongoing to evaluate the existing m ethod s
both on annual and quarterly basis and to get better inform ation, especially
on a quarterly basis.
9
Classification of the functions of the governm ent
27
Quarterly N ational A ccounts Inventory
To d eterm ine constant prices for collective prod uction hours w orked is
used for the m ajor part that consists of com pensation of em ployees. For the
other item s of consum ption, such as sales, other assets prod uced for ow n
account, and social benefits in kind , w eighted ind ices based m ainly on CPI
are used .
5.3 NPISH final consumption
N PISH is calculated using the cost m ethod . The calculations of w ages and
salaries are based on “Aggregate gross pay, payroll taxes and prelim inary
tax statistics from em ployers‟ m onthly tax returns” (LAPS, source 20). The
d ata is com prehensive and com prises w ages and salaries paid and pay related benefits. Social contributions and other taxes on pro d uction are
calculated using the inform ation on their share of w ages and salaries for
the sector. Consum ption of fixed capital have a m od el based approach (see
chapter 8.2.1). Subsid ies com prise subsid ies to w ages and salaries and are
obtained from the calculations for central governm ent subsid ies. Sales as
w ell as interm ed iate consum ption is calculated u sing ratios from the
previous year. Deflation is carried out using the CPI total.
5. 4 Gross fixed capital formation
Gross Fixed Capital Form ation is calculated both by ind ustry and by type
of investm ent. Many d ifferent sources are used in the calculations although
the m ajor source is the Investm ent Survey (source 15) cond ucted by
Statistics Sw ed en. Extrapolation m ethod s based on the trend s accord ing to
the sources are used for m ost parts of the calculations. Source values based
on d ata on exports and im ports are used d irectly for a few item s, m ainly
ships and aeroplanes. For m ost parts the source statistics are in current
prices and for d eflation ind ices for d om estic supply (ITPI) as w ell as a
variety of other ind ices are used .
5.4.1 The Business Sector by type of investment
Machinery, equipment and transportation equipment
The investm ent survey prod uced by Statistic Sw ed en is used to account
m ost investm ents in other build ings and structures, m achinery, equipm ent
and rail transportation for the business and household sector and it covers
m ost ind ustries and it is cond ucted three tim es a year: in February, May
and October. The survey has been ad apted to the investm ent d efinition of
the Sw ed ish N ational Accounts but d isposals of m achinery are only
separately given once a year. The survey covers both com pleted and
anticipated investm ent, as w ell as intangible investm ent, in ord er to gain a
com prehensive picture of business investm ents. The source values are not
used d irectly but as a basis for extrapolation of annual benchm arked levels
in current prices. Deflation to constant prices is d one using ind ices for
d om estic supply. The ind ices are w eighted on a d etailed level accord ing to
the m ost recent annual calculation for year t-2.
28
Quarterly N ational A ccounts Inventory
The ind ustries agriculture, forestry and fishing are not covered in the
Investm ent Survey. Calculations for these ind ustries are to large extent
m od el-based using annual inform ation from The Sw ed ish Board of
Agriculture, The N ational Board of Fisheries and The Sw ed ish Forest
Ind ustries Fed eration.
Housing
To calculate investm ents in housing, a m od el is used based on the num ber
of apartm ents/ houses treated in term s of their com m enced construction
expend itures, norm al construction, and com pletion tim e. The basis for the
calculation is volum e changes and current prices are obtained by reflation
using a num ber of ind ices for construction.
Motor vehicle
A separate calculation is carried out for m otor vehicles based on
ad m inistrative source m aterial. The m ain source for calculation of both
investm ent and household consum ption of vehicles are Statistics Sw ed en‟s
vehicle statistics.
Softw are
The calculation of gross fixed capital form ation in softw are is based on
extrapolation w ith the N ACE 72's prod uction value (com puter consultancy
ind ustry SN I72) in current prices as w ell as in constant prices. The output
value is based on the Turnover Statistics and the d eflator used is Service
Prod ucer Price Ind ex for com puter consultants.
Leasing
The capital item leased is record ed in accord ance w ith the ESA as gross
fixed capital form ation in the leasing activity. The leasing charges are
consequently entered in the N ational Accounts w ith a breakd ow n as to
interest and am ortisation costs. The calculations are based on both
ad m inistrative and survey m aterial. H ow ever, the sources used give no
inform ation about users‟ kind of activity. To get this split the investm ent
survey is used .
5.4.2 Government Sector
Total central governm ent activity, com piled by the Financial Managem ent
Authority (ESV), record s investm ent in accord ance w ith the N ational
Accounts d efinitions grouped by build ings and structures, m achinery,
transport equipm ent and other equipm ent, research and d evelopm ent, and
intangible investm ent.
The sam e surveys that are the basis for the calculations of consum ption are
the m ain source for the calculation of investm ent in the prim ary
m unicipalities and county councils. Investm ent is calculated into a
breakd ow n by build ings and structures and m achinery and equipm ent,
softw are purchased and prod uced for ow n account, and investm ent
financed by leasing. The inform ation from the sources are not used d irectly
but are used as a basis for extrapolation of the value the sam e quarter
previous year.
29
Quarterly N ational A ccounts Inventory
5.5 Changes in inventories and valuables
The calculations of chan ges in inventories are based on a num ber of
sources. The m ain part of the stocks of inventories consists of inventories in
the m anufacturing ind ustries and in retail and w holesale trad e. A quarterly
stock inquiry is carried out for the m ining, quarrying an d m anufacturing
ind ustries (source 16). The establishm ent‟s report separately finished
good s, raw m aterials and w ork in progress. A quarterly survey is
cond ucted for retail and w holesale trad e (source 17). Inform ation is also
available on a quarterly basis for inventories of fuel, oil, agriculture, and
inventories in the governm ent sector. The calculations of inventories in
forestry are based on inform ation about input into the pulp and paper
ind ustry and saw m ills. For all types of inventories the source d a ta is the
sam e in the quarterly accounts as in the annual accounts and levels form
the source d ata are used d irectly in the quarterly accounts.
A total value estim ate for assets in the form of valuables is calculated by
using the trend of a w eighted ind icator based on net exports of valuables,
Turnover statistics for d om estic trad e w ith antiques and a forecast on new
prod uction of valuables.
5.6 Imports, exports
Exports and im ports are based on the m onthly Foreign trad e in good s
statistics (Extrastat an d Intrastat) (source 18), and the quarterly Foreign
trad e in services statistics (source 19). Som e ad justm ents are m ad e in the
N ational Accounts in com parison to source d ata.
External trad e statistics are d elivered from the source statistics both in
current prices and constant prices. The calculations of constant prices are in
all essentials based on the export and im port price ind ices (source 24).
Since international trad e in good s statistics only partly cover non cross bord er trad e, a few m inor com plem en tary sources are used in the N ational
Accounts. This m ainly concerns estim ates on procurem ents. Certain CN 10
num bers in the trad e in good s statistics are treated as services in the
N ational Accounts and are therefore reclassified . The types of services in
question are consultancy services for com puter system s and softw are,
architectural and technical consultancy services, m iscellaneous other
business services, film and vid eo services and creative, literary and artistic
services. Another ad justm ent m ad e in th e N ational Accounts is that invoice
values are used in the calculation of exports of good s and not statistical
values (i.e. exports valued free on board ). This approach also affects the
calculation of trad e in freight services, w hich are also expressed in invoice
prices. The reason for shifting to invoice values is the d ifficulties to collect
d ata in a w ay that m eet necessary requirem ents for valuing exports f.o.b.
(free on board ) and im ports at c.i.f. (cost insurance and freight). Shifting to
invoice values also allow s greater consistency w ith d ata on prod uction,
consum ption, etc. For Intrastat export d ata is collected d irectly from
invoice values. For Extrastat d ata, the inform ation provid ed by Sw ed ish
Custom s is transform ed to invoice values. Because of this approach a
10
Com bined N om enclature
30
Quarterly N ational A ccounts Inventory
global f.o.b./ f.o.b.-calculation, accord ing to the ESA 95 m anual, is not
carried out.
For trad e in services the estim ates in the N ational Accounts are based on
the trad e in services survey and the com plem entary calculations that the
Riksbank is responsible for. Data from the trad e in service statistics are
prod uced in current prices and calculations in constant prices are m ad e by
the N ational Accounts d epartm ent. Since no ind ices are available for trad e
in services other ind ices have to be u sed . One third of the prod ucts are
d eflated w ith Service price ind ices (source 25) for the correspond ing
ind ustry and one third are d eflated by labour cost ind ices. Other than
these, ind ices for export and im ports of good s and the Consum er price
ind ices are used .
There are a few d ifferences in the trad e in services accord ing to the
N ational Accounts and the Balance of Paym ents. These d iscrepancies are
d ue to d ifferent valuation of freights, und er coverage in the sources and
m anual d ifferences. As m entioned above, shifting from statistical values to
invoice values results in freight services being calculated at invoice values
in the N ational Accounts w hile the statistical values are record ed in the
Balance of Paym ents. Further m ore som e ad justm ents are m ad e on
transportation services e.g. d ata on exports of port and navigation services
are calculated on the basis of statistics covering ships entering and leaving
ports. This item is not covered in the survey on trad e in services. Travel
services are record ed gross in N ational Accounts accord ing to ESA but net
in the Balance of Paym ents. Financial Services Ind irectly Measured (FISIM)
that is a part of N ational Accounts is not yet introd uced in the balance of
paym ents.
The d istinction betw een EU-m em ber states and non-m em ber states is
based on the inform ation collected in the Foreign trad e in good s statistics
and foreign trad e in services statistics. The split into m em ber states and
non-m em ber of the Monetary Union as w ell as EU institutions is not
subm itted in the QN A ESA95 transm ission (table 0120 ant 0221).
31
Quarterly N ational A ccounts Inventory
Chapter 6 GDP components: the
income approach
Presently it is not possible to calculate GDP w ith regard to the incom e sid e
of the Sw ed ish N ational Accounts since there is no reliable inform ation on
Gross operating surplus and Gross m ixed incom e. These item s are
calculated as a resid ual at an aggregate level. Com pensation of em ployees
is calculated on a d etailed level and published in the sam e breakd ow n as
prod uction and hours w orked .
6.1 Compensation of employees, including
components (wages and salaries)
For com pensation of em ployees estim ates are carried out separately for
total levels as w ell as for the separate ind ustries and sectors.
The source for calculating Com pensation of em ployees for the w hole
econom y as w ell as for the business sector and N PISH is “Aggregate gross
pay, payroll taxes and prelim inary tax statistics from em ployers‟ m onthly
tax returns” (LAPS) (source 20). For central governm ent and local
governm ent the sources used are the sam e as for calculating consum ption.
For central governm ent the source is the record s kept by the Financial
Managem ent Authority (ESV) and for the local governm ents the
inform ation collected in the specific surveys is used as w ell as LAPS. For
the central and local governm ents d irect values are m ainly used . H ow ever
for the total econom y, as w ell as for the business sector and N PISH , the
source is used to extrapolate quarterly values from previous year, w ith the
trend s form ed in accord ance w ith the source. The d ifference betw een the
sector sum s and the separate estim ate of total w ages and salaries are
chiefly allocated to the business sector.
To scrutinize the estim ates on w ages and salaries as w ell as hours w orked
and num ber of em ployees, the im plicit estim ates of hourly w ages and
w ages per em ployee are com pared to estim ates in the short -term statistics
on w ages and salaries. Inexplicable d ifferences betw een the estim ates are a
basis for ad justm ent in the N A estim ates of w ages and salaries, num ber of
em ployees or num ber of hours w orked .
Employers‟ social contributions are split into tw o com ponents, social
contributions com pulsory by law and social contributions regulated by
agreem ent. Calculations are m ad e for the total levels, based on total
paym ents and d eposits w ithin com panies, as w ell as for the separate
ind ustries. The total sum of social contributions regulated by law is
calculated on the basis of the am ount of social contributions that has been
paid to the governm ent and social security sectors from all sectors. For
social contributions by agreem ent, the total level is com piled using
inform ation on paym ents from em ployers to insurance com panies and the
allocations m ad e w ithin the com panies.
Calculations for the separate ind ustries are carried out using w ages and
salaries and the established percentage rates for social contributions by law
32
Quarterly N ational A ccounts Inventory
and by agreem ent w ithin the reference period . The sam e m ethod is used
for N PISH w ith regard to local governm ent. For central governm ent and
m unicipalities, inform ation on actual paym ents is used . The d ifference
betw een the sector sum s and the separate estim ate of total social
contributions is allocated to the business sector.
Payroll taxes are calculated in the sam e m anner as social contributions.
Total payroll taxes are based on the paym ents to the central governm ent.
For the separate ind ustries the payroll taxes are calculated by applying the
established rates for payroll taxes on the values for w ages and salaries. For
central governm ent and m unicipalities actual values are collected in the
d ata sources.
6.2 Taxes less subsidies on production
The Sw ed ish calculations for other taxes on prod uction are based for the
m ost part on the record s kept by the Financial Managem ent Authority
(source 11) (ESV) of the incom e of d epartm ents and agencies of central
governm ent by revenue head ings that are entered m onthly. Since the ESV‟s
record s are cash -based and inw ard paym ents to the revenue head ings
usually take place in arrears, period read justm ents are m ad e by ESV in
ord er to obtain the accrued value. In practice, the incom es are shifted back
in tim e, for exam ple incom e item s for February -January m ay instead be
record ed as incom e for the fourth quarter.
6.3 Gross operating surplus & mixed income
The sum of 'Gross operating surplus' and 'Gross m ixed incom e' is d erived
as the resid ual betw een GDP and the sum of com pensation of em ployees
(D.1) and taxes less subsid ies on prod uction (D.2 – D.3).
33
Quarterly N ational A ccounts Inventory
Chapter 7 Population and
employment
7.1 Population
The population presented in the N ational Accounts are com piled on the
Population statistics, w hich are based on the population registration,
ad m inistrated by the Tax Authority. The population in N ational Accounts
refers to the average betw een the beginning and end of the reference
period (quarter).
7.2 Employment: persons
Employm ent in the Sw ed ish N ational Accounts refers to persons
em ployed , no estim ation of num ber of jobs is com piled . Three sources
are used on a quarterly basis, the Labour Force Survey (LFS) (source 21),
Short-term em ploym ent (KS) (source 22), Wages and salaries in the
private and public sectors (KL)(source 23). Direct calculations of levels
are not d one on a quarterly basis. Instead the m ethod used is to
extrapolate the level from the sam e quarter the previous year w ith the
trend s form ed in accord ance w ith the sources. When calculating
em ploym ent as w ell as hours w orked estim ates are carried out
separately for total levels as w ell as for the separate ind ustries and
sectors.
The Labour force survey is the source for estim ating the quarterly
em ploym ent in N A for the w hole econom y. The trend used for
extrapolation is the grow th rate for em ployed persons 15-74 years
accord ing to the LFS. In the LFS d ata used in the N ational Accounts
Sw ed ish resid ents (one-year rule) w orking abroad are exclud ed .
Supplem ents for non -resid ents w orking in Sw ed en are only d one on an
annual basis. In the quarterly calculations these follow the sam e trend as
the total econom y.
To m ake m ore accurate estim ates for d etailed econom ic activities
enterprise surveys are used to calculate num ber of em ployees. The
second ary sources are Short-term em ploym ent, Wages and salaries in
the private sector, Salaries in the Governm ent sector and Salaries in the
prim ary local authorities and county councils. Emp loyed persons, as
w ell as hours w orked , is calculated separately as em ployees and self
em ployed persons at a breakd ow n of 51 ind ustries. H ow ever in the
d om estic release, on 31 ind ustries, there is no split on em ployees and
self-em ployed persons they are presented together as em ployed
persons. The reason for this is the lack of quality in the estim ates of selfem ployed persons at d etailed level.
Self-em ployed persons are based on the LFS for the total as w ell as for
the ind ustry breakd ow n. There is a conceptual d ifference betw een selfem ployed persons accord ing to the LFS and accord ing to the N ational
Accounts concerning self-em ployed persons w orking in their ow n
corporation. In N ational Accounts persons w orking in their ow n
34
Quarterly N ational A ccounts Inventory
corporation are regard ed as em ployees and not as self-em ployed
persons as they are in the LFS.
7.3 Employment: hours worked
The volum e of hours w orked on annual as w ell as quarterly basis is
com piled together w ith the num ber of em ployed persons. As for
em ployed persons the levels are n ot calculated d irectly on quarterly
basis, rather the extrapolation m ethod is used .
The Labour force survey is the source for estim ating hours w orked for
the w hole econom y. The treatm ent of hours w orked for Sw ed ish
resid ents w orking abroad and non -resid ents w orking in Sw ed en
correspond s is the sam e as for em ployed persons (see 7.2).
To m ake m ore accurate estim ates for d etailed econom ic activities hours
w orked accord ing to the enterprise surveys Wages and salaries in the
private sector, Salaries in the Governm ent sector and Salaries in the
prim ary local authorities and county councils are used .
H ours w orked by self-em ployed persons are calculated using the LFS
for the total econom y as w ell as for the ind ustries. Reclassification are
m ad e for hours w orked by persons em ployed in their ow n com pany
analogues w ith the treatm ent of num ber of self-em ployed persons.
Tw o m ethod s are used to calculate the volum e of hours w orked in the
Sw ed ish econom y:
Direct m ethod , the total hours actually w orked are estim ated
accord ing to the trend s of LFS and the enterprise statistics
Accounts m ethod , d ata on the num ber of em ployed persons are
com bined w ith average hours w orked accord ing to the LFS and
the enterprise survey.
The d irect m ethod is used for the aggregates of m arket prod ucers and
total econom y w here the LFS is used . When calculating ind ustries the
accounts m ethod is used , com bining estim ates on em ployees accord ing
to Short-term em ploym ent w ith actual w orking tim e accord ing to LFS
and “Short-term statistics, w ages and salaries, private sector”. The
accounts m ethod is also used for estim ates of hours w orked in central
and local governm ent and also for N on -profit institutions serving
household s. The num ber of em ployees in the governm ent sector and
N PISH are based on ad m inistrative sources. To estim ate the volum e of
hours w orked the trend for quarterly w orking tim e accord ing to LFS is
used (Volum e of hours w orked = N um ber of em ployees x annual
w orking tim e).
H ours w orked in the hid d en econom y are on quarterly basis assum ed to
have the sam e d evelopm ent as the ind ustry for w hich they are
calculated .
35
Quarterly N ational A ccounts Inventory
Verifying the validity of employees and hours w orked
The valid ity of em ployees and hours w orked is verified by a d irect
m ethod and an ind irect m ethod . The d irect m ethod is to com p are the
estim ates from the d ifferent sources used in the calculations. The
num ber of em ployed persons and hours w orked are m easured both by
LFS and KL and num ber the of em ployees are also estim ated in the
Short-term em ploym ent statistics. The ind irect m eth od is to com pare
estim ates of hourly w ages and w ages per em ployee accord ing to the N A
w ith survey estim ates of w ages, salaries and labour costs. There are
som e conceptual d ifferences betw een these estim ates, for instance
w ages in N A is not on accrual basis, nevertheless the com parison
ind icates lack of consistency betw een em ployees/ hours w orked and
w ages and salaries.
36
Quarterly N ational A ccounts Inventory
Chapter 8 From GDP to net
lending/borrowing
In the Sw ed ish N ational Accounts net lend ing/ borrow ing (B.9) is
calculated through the external balance. The resid ual betw een net
lend ing/ borrow ing accord ing to the external balance and accord ing to the
d om estic sectors is d istributed to the non -financial corporation sector. The
m ain sources used are - Foreign trad e in services (source 19) and other
Balance of Paym ents statistics like, Survey on Foreign Direct Investm ent
(annual), m onthly survey of d irect investm ents, Balance statistics for non financial corporations.
8.1 Primary income from/to the ROW (D.1 to D.4), gross
national income
GN I is calculated by ad justing GDP for prim ary incom e to and from the
rest of the w orld . The d ata for m ost of the transactions are collected from
the balance of paym ents statistics, BoP. Statistics Sw ed en collects the bulk
of the BoP statistics on com m ission by the Riksbanken. To get gross accrual
figures on taxes to and subsid ies from EU d ata from the Sw ed ish Financial
Managem ent Authority (ESV) are used in the N ational Accents instead of
BoP d ata.
8.1.1 Compensation of employees (D.1) from/to the ROW
From 2003 these d ata are part of the quarterly survey on external trad e in
services. Wages and salaries earned and also other rem unerations are
includ ed . Wages for construction w ork is record ed separately. Sw ed ish
build ing sites in the rest of the w orld are consid ered as prod ucer units in
the RoW. Wages paid to Sw ed ish build ing w orkers em ployed on these sites
are therefore consid ered as com ing from the rest of the w orld , even if they
are paid from the Sw ed ish firm . Inform ation concerning em bassy
personnel is obtained from the Ministry of Foreign Affairs. ESA 95 requires
gross record ing, and a calculation on taxes is m ad e at the processing unit of
Statistics Sw ed en. The estim ate how ever covers w ages and salaries earned
in a foreign country regard less of the length of the stay. ESA 95 allow s for a
m axim um of one year. This d ifference is consid ered to be of m inor
significance in practice.
8.1.2 Taxes on production and imports (D.2) to the ROW
Data on taxes only arise on the outflow sid e. They consist of custom s
d uties, VAT, agricultural levies and sugar levies, w hich m ake up parts of
Sw ed en‟s d ues or contributions to the EU. These are record ed as
transactions precisely affecting taxes on prod uction and im ports. The GN I
levy, also part of Sw ed en‟s contributions to the EU, is record ed instead as a
current transfer. Data are obtained d irectly from the ESV (see below ).
37
Quarterly N ational A ccounts Inventory
8.1.3 Subsidies (D.3) from the ROW
The d ata on subsid ies apply to both subsid ies on prod ucts and other
subsid ies on prod uction from the EU and therefore only affect the inflow
sid e. It is m ainly a question of subsid ies to agriculture. These d ata are
obtained from the ESV, w hich com piles the basis for central governm ent
net lend ing. The m aterial also com prises that part of the activity of centr al
governm ent, w hich is not record ed in the national bud get. The national
bud get record s the subsid ies that d epartm ents and agencies of governm ent
pay out but are financed by EU fund s. The d epartm ents and agencies are
obliged to record types of expend iture und er bud get head ings. This
enables the ESV to d eterm ine how paym ents are d istributed as betw een
subsid ies and other expend itures.
8.1.4 Property income (D.4) from/to the ROW
Interest (D 41)
The d ata in the balance of paym ents statistics on financial retur ns are
structured in accord ance w ith the m ain item s in the financial balance.
Thus, three types of interest incom e and expend iture are d istinguished .
Interest flow s linked to d irect investm ent. These represent interest on loans
in a d irect investm ent relationship. Interest paid to affiliated com panies,
such as intra-group banks etc., is includ ed . The interest is record ed on an
accrual basis.
Interest flow s linked to portfolio investm ents exclud ing financial
d erivatives. These com prise interest on bond s and m oney m arket
instrum ents that are record ed in Sw ed ish kronor and foreign currency and
specified by resid ent sectors and rest of the w orld . Interest on Sw ed ish
bond issues in foreign currency abroad is record ed as accrued interest.
Riksbanken carries out separate calculations in ord er to obtain accrued
interest for interest on stocks that arises m ainly through trad e, e.g.
securities issued by the State and housing finance institutions and
securities issued abroad .
Interest on loans etc., includ ing interest on financial d erivatives Record ing
has been m ainly on an accrual basis.
The d ata are collected m ainly by d irect reporting by transactors w ho have
stocks of assets or d ebt vis-à-vis the rest of the w orld involving large
am ounts; a guid e value, but not an absolute lim it, is approx. SEK 200
m illion. All the m ajor banks report inform ation on interests by country.
Estim ates for other m onetary financial institutes are calculated based on
the reports from the banks. N on -financial enterprises are covered by the
survey on Balance Statistics for non -financial com panies (BAST).
D istributed income of corporations (D 42)
The d istribution concept covers, apart from interest, d ivid end s on shares in
portfolio investm ents, repatriated share d ivid end s and reinvested profits
from d irect investm ents.
Returns on portfolio shares com prise d istributions from hold ings
am ounting to less than 10 percent of the share capital or sharehold er voting
rights. The d istribution is recorded on an accrual basis. The data are obtained
38
Quarterly N ational A ccounts Inventory
for d ebt securities by d irect reporting, w hile earnings on equity securities
are based on calculation.
Divid end s on shares in a d irect investm ent enterprise are record ed partly
in respect of the d ate of paym ent and partly w hen the d ivid end s are
payable. Returns on d irect investm ent are calculated as the net am ount of
financial incom e and costs and are record ed after tax in the host country.
Depreciation, capital gains and capital losses are not includ ed . An annual
sam ple survey is cond ucted on d irect investm ent in the rest of the w orld
that m akes use of a register continuously upd ated from Statistics Sw ed en‟s
register of d irect investm ent com panies, new spapers and m agazines etc.
Direct reporting agents for the continuous reporting are selected , inter alia,
w ith the aid of the survey register. Data collected on financial services are
supplem ented by a calculation of brokerage com m issions from d eals in
shares.
Reinvested earnings on foreign direct investments to/from the rest of the
w orld (D 43)
Reinvested earnings con stitute that part of a com pany‟s results that are not
d istributed to the sharehold ers but are retained in the com pany. These
earnings are calculated as the d ifference betw een the com pany‟s total profit
after tax and the d istributed profit. Data on d istribu ted profits are obtained
via annual d irect reporting as d iscussed above. Whereas d ivid end s are
record ed by d ate of paym ent, reinvested earnings are attributed to the year
d uring w hich the com pany d eclared the profit.
Property income attributed to insurance policy holders (D 44)
This incom e com prises the total of prim ary incom e generated by
investm ent in so-called insurance technical reserves. The investm ent return
is attributed to the insurance policyhold ers as property incom e. H ow ever,
this is retained in p ractice by the insurance com panies and pension fund s
and m ust therefore be treated as though the policyhold ers had paid
prem ium s and contributions of a com m ensurate am ount to the com panies
and fund s in question. These prem ium s and contributions are ad d ed t o the
prem ium s and contributions actually paid .
Statistics Sw ed en obtains inform ation through its collection of d ata from
the insurance com panies on how they invested the reserves, w hich they d o
m ainly by w ay of portfolio assets. The returns are also repo rted but are not
record ed separately. The stock of life insurance policies held by the
household sector in the rest of the w orld is calculated and the returns to
shares and other assets are allocated , though these are also not record ed
separately.
Rents on land and on sub-soil assets (D 45)
The rents received by land ow ners constitute a form of capital incom e; but
not the rent on build ings on the land in question, w hich is treated as a
service. The sam e applies to royalties, w hich the ow ner of d eposits of
m inerals and fossil fuels collects by granting licences to other institutional
units for the exploration or exploitation of such d eposits. These
transactions are not applicable in the case of the rest of the w orld . It follow s
from the accounting rule that transactions betw een resid ents and non resid ents for land and sub-soil assets are d eem ed to occur betw een resid ent
39
Quarterly N ational A ccounts Inventory
units. In that case, the non -resid ent party accrues a financial claim on a
notional resid ent unit accord ing to BPM5, 312. It is consid ered that this
property incom e, if it exists at all, is of m inor significance as far as Sw ed en
is concerned .
8.2 Consumption of fixed capital (K.1), net national
income, acq. less disp. of non-financial and nonproduced assets (K.2)
By d efinition, net national incom e is calculated by the subtraction of
consum ption of fixed capital (K.1) from gross national incom e (B.5*g) as
d erived in section 8.1.
8.2.1 Consumption of fixed capital (K.1)
Since the third quarter 2007 capital stocks are calculated on a quarterly
basis. The calculation of consum ption of fixed assets at Statistics Sw ed en is
d one by use of a Perpetual Inventory Method (PIM) w ith geom etric
d epreciation rates. The m ethod is used for d ata as far back as 1993.
The consum ption of fixed assets, K1 is calcu lated in current prices
accord ing to the follow ing form ula:
K1
t N
(1 2 1
) GFCF
w here δ t is the geom etric rate of capital consum ption. This rate is kept
fixed over tim e if no specific inform ation on changes in the rate is
available. GFCFt is the gross fixed capital form ation d uring the year. The
GFCF of the year is assum ed to be in service half of the year on average.
The net stock at the end of the year is also equal to the net stock at the
beginning
of the next year in constant prices(from the previou s year) is calculated
accord ing to the follow ing form ula:
N t+1 = N t + GFCFt – K1
8.2.2 Acquisition less disposals of non-financial and nonproduced assets (K.2)
Acquisition less d isposals of non -financial non -prod uced assets are in the
Sw ed ish N ational Accounts only record ed for the d om estic sectors. N o
transactions w ith other countries are registered d ue to lack of inform ation.
For instance, for intangible non -prod uced assets like patented entities, it is
not possible to d istinguish betw een w hat is paym ent s for patented entities
and w hat is paym ents for the use of a patent. Tod ay no split is d one
betw een the tw o and the collected values are record ed as trad e w ith
services as a w hole.
8.3 Current transfers from/to the ROW (D.5 to D.7), net
national disposable income (B.6n)
By d efinition, net national d isposable incom e is calculated by the
subtraction of net current transfers from / to the rest of the w orld (D.5 to
D.7) from net national (B.5*n) incom e as d erived in section 8.2.
40
Quarterly N ational A ccounts Inventory
8.3.1 Current transfers (D .5 to D .7) from the ROW
In the N ational Accounts external balance the Balance Of Paym ents
statistics are used in the quarterly as w ell as the annual accounts.
8.3.2 Current transfers (D .5 to D .7) to the ROW
In the N ational Accounts external balance the Balance of Paym ents
statistics is the m ain source for current transfers to the rest of the w orld .
For a few item s like the GN I, fee inform ation from ESV is used on quarterly
basis because it better com plies to the accrual principal.
8.4 Adjustment for the change in net equity (D.8), net
saving (B.8)
This item is not yet calculated in the Sw ed ish QN A.
8.5 Capital transfers (D.9), net lending/borrowing (B.9)
Capital transfers (D.9) from / to the rest of the w orld are d erived by
sum m ing up capital transfers of the private and the public sector as
published in the Sw ed ish balance of paym ents. In the Sw ed ish N ational
Accounts net lend ing/ borrow ing (B.9) is calculated through the external
balance. The resid ual betw een net lend ing/ borrow ing accord ing to the
external balan ce and accord ing to the d om estic sectors is d istributed to the
non-financial corporation sector.
41
Quarterly N ational A ccounts Inventory
Chapter 9
Flash estimates11
9.1 Flash GDP and employment estimate
Statistics Sw ed en has published N ational Accounts in a flash report
version for the second quarter of every year since 1996. Both the
prod uction sid e and the expend iture sid e are calculated , and both constant
and current prices are published since 2008. Only fixed prices w ere
calculated and published prior to 2008. Com pared to ord ina ry quarterly
publishing, the expend iture sid e of the flash report version is published at
the sam e level of d etail w hile prod uction and em ploym ent are published
on a som ew hat m ore aggregated level. The publication takes place about 30
d ays after the end of the quarter. The flash report calculations started by
com m ission of the Sw ed ish Ministry of Finance w hen a reorganisation of
the bud getary year conferred that relevant N ational Accounts inform ation
w as not available as the basis of forecasting w ork. There w as not tim e for
ord inary publication of the second quarter to be able to serve as a
forecasting tool in the bud getary proced ure.
The source d ata for the flash report calculations are m ore incom plete
com pared to the ord inary quarterly calculations. The am ount of d ata
available for the calculations varies am ong d ifferent areas, d ue to certain
areas being treated by m onthly statistics and others by quarterly statistics.
For areas covered by m onthly statistics tw o of the quarters' three m onths
are usually available for the N ational Accounts calculations. This applies to
ind ustrial m anufacturing and prod uction in the energy ind ustry, as w ell as
housing investm ents. This also applies to em ploym ent calculations to a
certain extent, but estim ates of total person s em ployed and total hours
w orked are based on inform ation for an entire quarter from the Labour
Force Survey. For the flash report, foreign trad e in good s are based on the
ord inary good s-d istributed foreign trad e statistics for tw o m onths and the
flash statistics, based on a m ore lim ited survey of the quarter's last m onth.
In those cases w here the d ata is m ad e up of quarterly statistics, the d ata
used have m ore im puted values than in an ord inary d elivery. For exam ple
Foreign Trad e in Services is hand led in this w ay.
Certain areas have no or very little inform ation available for the flash
calculations, and then the estim ates are based to a great extent on m od els
and evaluations. Such are the cond itions applying for m unicipal
consum ption in current prices, changes in inventories and prod uction in
forestry and financial services. The gross fixed capital form ation has a large
part of the calculations for ind ustrial investm ents in enterprises investm ent
forecasts as w ell as ord inary second quarter estim ates are based on such
m aterial.
There is a three m onth Consum er Price Ind ex and Prod ucer Price Ind ex for
calculation in constant prices. The Service Prod uction Price Ind ex, a
quarterly survey, uses a forecasting m od el.
11
For a d efinition of flash estim ates see footnote in section 1.2
42
Quarterly N ational A ccounts Inventory
9.1.1 Results of Flash estimates 2001-2008
Table 1 show s the d ifference in volum e changes betw een the Flash Report
Calculation and the ord inary quarterly calculation for t he second quarter
for years 2001-2007. The revisions to GDP have been m od erate d uring this
period , on average am ounting to 0.2 percentage points in absolute
num bers. On the other hand , revisions have been extensive for ind ivid ual
sections.
Table 1 Revisions of the Flash Report
The difference between the Flash Report and subsequent quarterly calculations (actual
volume development according to the Flash Report less actual volume development in
ordinary quarterly calculations) Using positive numbers mean that the Flash Report has
overestimated the development and negative numbers mean that the Flash Report has
underestimate the development .
2001
2002
2003
2004
2005
2006
2007
2008
0,5
-0,3
0,4
-0,1
-0,1
0,2
0,2
0,6
0,0
0,5
0,3
0,1
-3,5
1,0
0,0
-1,7
0,1
-0,1
-1,4
-0,9
-0,4
-0,4
-0,2
-2,1
-2,0
0,8
-0,2
1,0
0,6
-0,6
0,6
1,0
0,4
-2,9
-2,8
-1,5
-1,1
0,9
0,3
1,4
Household consumption
0,9
1,0
0,6
0,3
-0,3
0,0
-0,2
0,0
General government consumption
1,2
-0,2
-0,7
0,1
-0,2
-0,7
0,0
-1,0
-0,3
-4,2
-0,8
-1,1
1,8
0,0
-0,6
-0,3
0,4
-0,2
0,0
0,6
-0,2
0,6
0,3
0,2
-1,4
-1,4
-1,8
-3,2
-1,1
0,5
0,4
1,3
0,4
-0,1
0,2
-0,2
-0,1
0,2
0,1
0,0
Resources
Value added market producers and
producers of own final use
Value added central government and
social authorities
Value added local authorities
Value added NPISH
Imports of goods and services
Uses
Gross fixed capital formation
Changes in inventories
Exports goods and services
GDP
Betw een 1998 and 2009 Statistics Sw ed en publishes a m onthly activity
ind ex for the econom ic activity in the econom y (AI). The ind ex w as based
on a num ber of m onthly ind icators.
43
Quarterly N ational A ccounts Inventory
Chapter 10 Main data sources used
As a background to the in form ation of the m ain sources for QN A a brief
d escription of Statistics Sw ed en‟s Business register (FDB) is given. The FDB
provid es the sam pling fram e for statistics prod uced by Statistics Sw ed en.
This applies in particular to econom ic statistics. All statistics intend ed to
provid e inform ation on the Sw ed ish econom y, regard less of level, call for
coord ination of d efinitions of units to be surveyed , ind ustries, size
categories etc. This in turn requires a register of high quality to serve as an
instrum ent of coord ination. The register covers all Sw ed ish enterprises,
d epartm ents and agencies of governm ent, organisations, their
establishm ents and activity units. With the aid of the FDB, populations are
d em arcated for statistical inquiries as regard s coverage, ind ustries and size
groups. The register serves as a source for nam e and ad d ress d ata for the
enterprises, establishm ents and / or activity units to be covered by various
inquiries.
In January 2008 there w ere about 950 000 enterprise units and about
1 020 000 establishm ents units in the d atabase. Most surveys use the
sam ple fram e in March for the present year.
The production approach
Source 1: Crop production forecast for cereals and oilseed crops
Link to surveys und ertaken at the European level :Type of source: Forecast on harvest.
Population and sam ple: Forecast
Period icity: Annual, the forecast is m ad e in August every year.
Tim e of availability of results: 15 d ays
Main variables used in QN A: Quantities and prices on the m ain crops
Further ad justm ents m ad e to the survey d ata: -
Source 2: Animal products - Annual and Monthly Statistics
Link to surveys und ertaken at the European level : Type of source: Ad m inistrative
Population and sam ple: Covers all slaughter-houses in Sw ed en
Period icity: Monthly
Tim e of availability of results: 45 – 60 d ays
Main variables used in QN A: Quantities and prices on the m ain cattles, m ilk and egg
Further ad justm ents m ad e to the survey d ata: -
44
Quarterly N ational A ccounts Inventory
Source 3: Sw edish sea-fisheries
Link to surveys und ertaken at the European level : Type of source: Ad m inistrative
Population and sam ple: Covers all fisheries in Sw ed en
Period icity: Monthly
Tim e of availability of results: 30 d ays
Main variables used in QN A: Quantities and values on the m ain species
Further ad ju stm ents m ad e to the survey d ata: -
Source 4: Industrial production index
Link to surveys u nd ertaken at the European level: Accord ing to Council Regulation
(EC) no 1165/ 98, concering short-term business statistics
Type of source: Survey
Population and sam ple: Sam e as in N ew ord ers and d eliveries in ind ustry
Period icity: Monthly
Tim e of availability of results: 40 d ays
Main variables used in QN A: Prod uction ind ices, volum e changes
Further ad justm ents m ad e to the survey d ata: Ad justm ent for changes in inventories
to better reflect the prod uction value rather than d eliveries. For som e ind ustries in
w hich the prod uction of services is m ore substantial supplem ents are m ad e
Source 5: N ew orders and deliveries in industry
Link to surveys und ertaken at th e European level :Accord ing to Council Regulation
(EC) no 1165/ 98, concering short-term business statistics
Type of source: Survey
Population and sam ple: The population is all enterprise units w ith at least 10
em ployees in (N ACE 10-40) the sam ple consists of 2 100 enterprise units.
Period icity: Monthly
Tim e of availability of results: 40 d ays
Main variables used in QN A: Prod uction ind ices, value changes by ind ustry (N ACE
10-37)
Further ad justm ents m ad e to the survey d ata: Ad justm ent for changes in inventories
to better reflect the prod uction value rather than d eliveries. For som e ind ustries in
w hich the prod uction of services is m ore substantial supplem ents are m ad e .
45
Quarterly N ational A ccounts Inventory
Source 6: Monthly electricity statistics (also used for H ousehold
consum ption)
Link to surveys und ertaken at the European level :Type of source: Survey
Population and sam ple: The population is all establishm ents in N ACE 10-37, electric
netw ork establishm ents, and railw ay traffic enterprises. Supply of electricity covers
the w hole population, and for the uses of electricity 1 800 establishm ents are
surveyed (all establishm ents w ith an annual use that exceed 2000 MWh).
Period icity: Monthly
Tim e of availability of results: 40 d ays
Main variables used in QN A: Quantities on MWh.
Further ad justm ents m ad e to the survey d ata: -
Source 7: Turnover statistics (also used for H ousehold consum ption)
Link to surveys und ertaken at the European level: Council Regulation N o 1165/ 98
concerning short-term statistics
Type of source: Survey
Population and sam ple: The population is all enterprise units w ithin N ACE 41, 5093, except N ACE 65-67, 70.201 and 91. The sam ple is about 13200 enterprise units.
Period icity: Monthly
Tim e of availability of results: 45 d ays.
Main variables used in QN A: Turnover trend s, value changes by ind ustry
Further ad justm ents m ad e to the survey d ata: -
Source 8: Service production index
Link to surveys und ertaken at the European level: Type of source: Survey
Population and sam ple: Sam e as Turnover Statistics
Period icity: Monthly
Tim e of availability of results: 50 d ays (45 d ays for internal use)
Main variables used in QN A: Turnover trend s, volum e change by ind ustry
Further ad justm ents m ad e to the survey d ata: -
46
Quarterly N ational A ccounts Inventory
Source 9: Financial corporations except insurance companies – quarterly
financial data (also used for H ousehold consum ption)
Link to surveys und ertaken at the European level:Type of source: Survey
Population and sam ple: The sam pling fram e used is the inspection register of the
Financial Supervisory Authority. The statistics cover insurance com panies and
pension institutions, exclud ing pension found ation and benevolent societies. The
inquiry is a full census.
Period icity: Quarterly
Tim e of availability of results: N o separate publication , d ata are sent to N A 45-55
d ays after the quarter.
Main variables used in QN A:
Further ad justm ents m ad e to the survey d ata: -
Source 10: Sw edish insurance companies– quarterly financial data (also
used for H ousehold consum ption)
Link to surveys und ertaken at the European level:Type of source: Survey
Population and sam ple: The sam pling fram e used is the inspection register of the
Financial Supervisory Authority. The statistics cover financial corporations- except
insurance corporations. The inquiry is a full census.
Period icity: Quarterly
Tim e of availability of results: N o separate publication , d ata are sent to N A 45-55
d ays after the quarter.
Main variables used in QN A: Data on ad m inistrative costs, prem ium s, provisions,
claim s and num ber of in surances.
Further ad justm ents m ad e to the survey d ata: -
The expenditure approach
Source 11: Total activity of central government (also used for central
governm ent value ad d ed and w ages and salaries)
Link to surveys und ertaken at the European level: Type of source: Ad m inistrative
Population and sam ple: Covers all the institutional units in the central governm ent
sector.
Period icity: Quarterly
Tim e of availability of results: 45 d ays
Variables used for QN A: Incom e and expend itures
Further ad justm ents m ad e to the d ata: 47
Quarterly N ational A ccounts Inventory
Source 12: Quarterly survey on municipalities (also used for value ad d ed
and w ages and salaries in m unicipalities)
Link to surveys und ertaken at the European level: Type of source: Survey
Population and sam ple: Covers about 80 out of 290 m unicipalities
Period icity: Quarterly
Tim e of availability of results: 40 d ays
Main variables used in QN A: Incom e and expend itures
Further ad justm ents m ad e to the survey d ata: -
Source 13: Quarterly survey on county councils (also used for value
ad d ed and w ages and salaries in county councils)
Link to surveys und ertaken at the European level: Type of source: Survey
Population and sam ple: Covers all the 18 county councils and the 2 regions
Period icity: Quarterly
Tim e of availability of results: 40 d ays
Variables used for QN A: Incom e and expend itures
Further ad justm ents m ad e to the d ata:-
Source 14: Statistical register for vehicles
Link to surveys u nd ertaken at the European level: Type of source: Ad m inistrative based on Central Motor Vehicle Register of the
N ational Road Ad m inistration .
Population and sam ple: The vehicles registered in the Central Motor Vehicle Register
of the N ational Road Ad m inistration
Period icity: Monthly
Tim e of availability of results: 30 d ays
Variables used for QN A: Changes in num ber of vehicles
Further ad justm ents m ad e to the d ata: -
48
Quarterly N ational A ccounts Inventory
Source 15: The investment survey
Link to surveys und ertaken at the European level: Type of source: Survey
Population and sam ple: The population covers all enterprises, except N ACE 1-5, 8085, 91-93. The sam ple includ es all enterprises w ith m ore than 200 em ployees and
enterprises w ith 20 to 199 em ployees are sam pled . The survey has a cut-off for
enterprises w ith less than 20 em ployees.
Period icity: Quarterly (3 quarters, 1st , 2nd , 4th )
Tim e of availability of results: 50 d ays (for internal use 45 d ays)
Main variables used in QN A: Changes in gross capital form ation by ind ustry and
type of investm ent.
Further ad justm ents m ad e to the survey d ata: -
Source 16: Industrial inventories
Link to surveys und ertaken at the European level:Type of source: Survey
Population and sam ple: The population is all enterprise units w ith at least 10
em ployees in (N ACE 10-40). Sam ple of 1300 enterprise units.
Period icity: Quarterly
Tim e of availability of results: 45 d ays
Main variables used in QN A: Stocks of inventories as w ell as changes in inventories
in current prices as w ell as constant prices.
Further ad justm ents m ad e to the survey d ata: -
Source 17: Quarterly survey on inventories in retail trade and w holesale
trade
Link to surveys u nd ertaken at the European level: Type of source: Survey
Population and sam ple: Covers w holesale and retail trad ing enterprises. About 4 500
enterprises are surveyed . Enterprises w ith a turnover exceed ing SEK 100 m illion are
subjected to a full census.
Period icity: Quarterly
Tim e of availability of results: About 45
Main variables used in QN A: Stocks of inventories in current prices as w ell as
constant prices
Further ad justm ents m ad e to the survey d ata: -
49
Quarterly N ational A ccounts Inventory
Source 18: Foreign trade - exports and imports of goods
Link to surveys und ertaken at the European level:
Intrastat
- Council regulation (EEC) no 638/ 2004.
- Com m ission Regulation (EEC) no 1982/ 2004
Extrastat
- Council regulation (EEC) no 1172/ 95
- Com m ission Regulation (EEC) no 1917/ 2000
Type of source: Ad m inistrative and survey (custom d ata for Extrastat and survey for
Intrastat).
Population and sam ple:
Extrastat: For enterprises exporting and im porting good s to and from countries
outsid e the EU (third countries) d ata are used from the export notifications and
im port d eclarations w hich the enterprises supply to Sw ed ish Custom s in conjunction
w ith the export and im port of good s.
Intrastat: Data are collected on a m onthly basis from enterprises w ith total exports of
good s to other EU countries or im ports of good s from other EU countries to a
m inim um value of SEK 2 200 000 or exports to a m inim um value of SEK 4 500 000
Period icity: Monthly in current prices, quarterly in constant prices
Tim e of availability of results: 65 d ays in constant prices (prelim inary d ata for
internal use, 50 d ays)
Main variables used in QN A: Exports and im ports on good s in current and constant
prices.
Further ad justm ents m ad e to the survey d ata: Invoice values are used in N A.
Source 19: Foreign trade in services statistics
Link to surveys u nd ertaken at the European level: The regulation EC-184-2005
regard ing the balance of paym ents.
Type of source: Surveys m ainly
Population and sam ple: For the 2007 statistics, the sam ple fram ew ork w as
approxim ately 47 000 enterprises and the sam ple of approxim ately 4 800 enterprise
units.
Period icity: Quarterly
Tim e of availability of results: 40 d ays after the reference quarter
Main variables used in QN A: Trad e in services in current prices. Prim ary incom e to
and from rest of the w orld , current transfers to and from rest of the w orld .
Further ad justm ents m ad e to the survey d ata: Invoice values are used in N A
50
Quarterly N ational A ccounts Inventory
The income approach
Source 20: Aggregate gross pay, payroll taxes and preliminary tax
statistics from employers’ monthly tax returns (also used for value ad d ed
in N PISH )
Link to surveys u nd ertaken at the European level: Type of source: Ad m inistrative
Population and sam ple: Covers all em ployers that m ake paym ents of w ages and
salaries.
Period icity: Quarterly
Tim e of availability of results: 50 d ays (for internal use 45 d ays)
Variables used for QN A: Wages and salaries
Further ad justm ents m ad e to the d ata: -
Source 21: The labour force survey
Link to surveys und ertaken at the European level: Accord ing the EU Regulation N o
430/ 2005
Type of source: Survey
Population and sam ple: The target population in the Labour force survey is all
persons w ith civil registration in Sw ed en w ho have reached the age of 15 but not 75,
ap prox. 6 800 000 ind ivid uals. The survey is based on a sam ple of about 21 000
persons each m onth.
Period icity: Monthly/ Quarterly
Tim e of availability of results: Quarterly d ata 40 d ays
Main variables used in QN A: N um ber of em ployees and self-em ployed by ind ustry
and by sector. H ours w orked for em ployees and self-em ployed persons by ind ustry
and by sector.
Further ad justm ents m ad e to the survey d ata: Self-em ployed p ersons that w ork in
their ow n corporation are regard ed as em ployees in the N ational Accounts.
51
Quarterly N ational A ccounts Inventory
Source 22: Short-term empolyment statistics
Link to surveys u nd ertaken at the European level: EU regulation (1165/ 98)
concerning short-term em ploym ent statistics.
Type of source: Survey
Population and sam ple: The population includ es all establishm ents in the private
sector and all organizations in the public sector w ith at least one em ployee in
accord ance w ith Statistics Sw ed en‟s Business register (FDB).The sam ple for the
private sector com prises approx. 19 500 establishm ents. The public sector are
covered by a sam ple of 4 100 establishm ents
Period icity: The inquiry is cond ucted every m onth and published quarterly
Tim e of availability of results: 45 d ays
Main variables used in QN A: Changes in the n um ber of em ployees by ind ustry.
Further ad justm ents m ad e to the survey d ata: ´
Source 23: Wages and salaries in the private sector
Link to surveys und ertaken at the European level: Council Regulation N o 1165/ 98
concerning short-term statistics
Type of source: Survey
Population and sam ple: The population consists of enterprises w ith at least five
em ployees. The sam ple in the 2008 survey w as about 7 7800 enterprises.
Period icity: Monthly
Tim e of availability of results: 60 d ays (prelim inary results for internal use 45 d ays)
Main variables used in QN A: N um ber of em ployees, hours w orked , hourly w ages
Further ad justm ents m ad e to the survey d ata: -
52
Quarterly N ational A ccounts Inventory
Prices
Source 24: Price indices in producer and import stages
Link to surveys und ertaken at th e European level :Council Regulation N o 1165/ 98
concerning short-term statistics
Type of source: Survey
Population and sam ple: The p opulation is all transactions concering sales from
prod ucers and purchases from im porters of prod ucts in N ACE A -E. Approximately
1200 producers/importers are surveyed reporting about 4000 quotes (1500 home sales, 1000
exports and 1400 imports).
Period icity: Monthly
Tim e of availability of results: 25 d ays
Main variables used in QN A: Indices for producer prices, home sales; producer prices,
export sales (export price index); producer prices, home sales and exports: import prices, and;
domestic supply prices, home sales and imports.
Further ad justm ents m ad e to the survey d ata: -
Source 25: Service price index
Link to surveys und ertaken at the European level: Council Regulation N o 1165/ 98
concerning short-term statistics
Type of source: Survey
Population and sam ple: Populationen is all transactions concering sales of certain
services in service ind ustries. 1000 enterprise units are surveyed reporting about 4000
quotes
Period icity: Quarterly
Tim e of availability of results: 45 d ays
Main variables used in QN A: Service prices ind ices at d etailed level
Further ad justm ents m ad e to the survey d ata: -
Source 26: Consumer price index
Link to surveys und ertaken at the European level: H arm onized Ind ices of Consum er
Prices (H ICPs) accord ing to Article 121 of the Treaty of Am sterd am (109j of the
Treaty on European Union).
Type of source: Survey
Population and sam ple: The p opulation is all transactions concerning good s and
serviced in the private d om estic consum ption.
Period icity: Monthly
Tim e of availability of results: 15 d ays
Main variables used in QN A: Consum er price ind ices
Further ad justm ents m ad e to the survey d ata: 53
Quarterly N ational A ccounts Inventory
54