Presentation
Download
Report
Transcript Presentation
INDEX OF SERVICE
PRODUCTION (ISP)
Presentation
2009-09-14/15
Daniel Lennartsson
Statistics Sweden
[email protected]
2017-04-06
Growth of service sector
(Share of GDP for different sectors)
60
50
Agriculture (SIC 01-05)
40
Priv ate serv ices (SIC 5093)
30
Manufacturing (SIC 01-45)
20
Public serv ices
10
0
1950
2017-04-06
1960
1970
1980
1990
2000
2010
Service sector and GDP
16
14
12
10
8
GDP
6
01-45
4
50-95
2
0
-2
-4
2017-04-06
Swedish investigation about the economic
statistics 2002
-The users of statistics want statistics for the growing
private service sector
-The users want short term indicators, similiar
indicators to those in the manufacturing sector.
Production (volume) index, new orders, prices,
turnover and capacity)
-The department of Econcomic Statistics was given
about 600 thousand Euro / year to produce new
indicators. Most of this amonth are used to produce a
index of service production (ISP) and turnover statistics
on a monthly basis.
2017-04-06
Project plan with purpose to develop a
production (volume) index for the service
sector on a monthly basis
Starting points and Conditions
-OECD manual
-Great Britain’s index of services
-National Accounts
-Discussions with producers and users of economic statistics
Goals
-Set a design to produce an index of service production
-Publish the first result spring 2008.
-Timeliness 35 days after the end of the reference month
2017-04-06
Volume index in the service sector
The theoritical form of an index of production is a
volume index of Laspeyres type
Where q is produced quantity
N
Qt
p
i ,0
i 1
N
p
i 1
i ,0
Mt
qi ,t ai , 0 i ,t
j 1
M0
qi , 0 ai , 0 i , 0
j 1
p is price of produced quantity
δ is quantity for consumption
a is price of quantity for consumption
i is one of the produced quantities
j is one of the quantities for
consumption
0 is the base period
t is the present period
2017-04-06
Production index in the service sector
Qt is just a theoretical index and can not be calculated in
practice. Instead consider it as a value added index.
According to the theory of value added index the index
should be calculated as deflated production value minus
deflated consumption value.
To calculate (deflated) value added on monthly basis is
very difficult. Instead use some other output as a proxy
for value added.
We use deflated turnover for each activity as a proxy.
To produce index of production we also need to weight the
activities together up to aggregates
2017-04-06
How to calculate the production index in
practice
•
Weight (wb) is
wb
VA
VAb
bB
•
When applicate base years weights, the index of
production by time t can then be given by
QBAlt,t
(Turnb,t ) /( Pb,t / Pb,0 )
wb
Turn
bB
b
,
0
100
wb
bB
2017-04-06
The main components to
construct an index of production
• Output (Turnover)
• Deflators
• Weights
----------------------------------------------------• Trading Day effects
• Seasonal adjustment
2017-04-06
Turnover
• Sampled based survey on a monthly basis
• Coverage NACE 45 -96 excl. 64-66
• Low cut off threshold (> 20 000 Euro / year)
• The sampling frame is stratified by activity
and size (128 *6)
• Total sample size is 7 000 enterprises on
monthly basis and 6 000 on quarterly basis
2017-04-06
Deflators
Best deflator:
Service producers price indicies (SPPI)
- Only produced on quarterly basis
- Forecasting of SPPI is done for the present month by a
forecasting model, exponential smoothing
Other deflators:
Consumer price index (CPI), wage index (WI), producer
price index in industry (IPPI), import price index (IMPI)
– all indicies are produced on monthly basis
2017-04-06
List of deflators – per NACE Rev 1.1
ISP for quarter 20071 compared to 20061, constant prices in 20061
Nace
Def
Prel
Diff (%) Priceindex
Periodicity
50
51
52
55
60
61
62
63
64
70
71
72
73
74
80
85
90
91
92
93
Total
2017-04-06
104,2
104,3
106,3
105,9
103,5
93,8
101,4
106,8
108,0
99,1
111,0
108,5
86,0
111,0
104,1
110,1
110,0
115,0
100,3
101,5
104,2
104,3
106,3
106,7
103,5
92,3
103,8
106,6
108,0
99,1
111,0
109,1
86,0
111,5
104,1
110,1
110,0
115,0
100,3
101,5
105,4 105,6
0,0
0,0
0,0
-0,8
0,1
1,6
-2,4
0,2
0,1
0,0
0,0
-0,6
0,0
-0,5
0,0
0,0
0,0
0,0
0,0
0,0
-0,2
CPI
IPPI
CPI
SPPI, CPI
SPPI, CPI
SPPI
SPPI
SPPI, CPI
SPPI
SPPI, CPI
SPPI, IPPI
SPPI
SPPI, IMPI
SPPI
CPI, WI
CPI, WI
SPPI
WI
CPI, WI
CPI
Month
Month
Month
Qrt, Month
Qrt, Month
Qrt
Qrt
Qrt, Month
Qrt
Qrt, Month
Qrt, Month
Qrt
Month
Qrt, Month
Month
Month
Qrt
Month
Month
Month
Weights
• Value added weights
– Aggregates from national accounts
– Details from SBS survey
• Yearly weights
• The weights updates every year (small
changes though)
2017-04-06
Trading day effects and seasonal
adjustment
• Estimation of trading day effects
- Regression model, based on time series data (Major part of activities)
- Expert group estimating trading day effects (Retail Trade)
- It’s very difficult to adjust for trading day effects in an accurate way!!
• Seasonal adjustment of the index serie
- No seasonal adjustment at the moment
- Seasonal adjustment will start in a near future
2017-04-06
Results, total ISP
Index present month compared to same month previous year
15
10
5
0
-5
-10
-15
2017-04-06
Results, ISP compared to IIP
Index present month compared to same month previous year
10
5
0
-5
ISP
IIP
-10
-15
-20
-25
2017-04-06
Dissemination
• Monthly publishing
- Timeliness, 35 days after reference month
- Press release (webpage)
- Statistical databases
• Monthly index revised each quarter due to
– Forecasted SPPI replaced by definite figures
– Estimates of turnover based on a larger sample
2017-04-06
Comparison with Index of
industrial production (IIP)
• Differences in the output
(IIP: 70 percent turnover, 25 percent volumes, 5 percent nr of
working hours. ISP: 100 percent turnover)
• IIP has high cut-off threshold (>10 employees)
• IIP has best deflators (priceindex on monthly basis)
• Same source for calculating yearly weights (National accounts
plus SBS)
• Same method to calculate index series, chain index, base year
changes every year, reference year = 2005.
• IIP has seasonal adjusted series
• About the same timeliness (35-40 days after reference month)
2017-04-06
• In the future an index of production for the whole private sector