3. Aims
• Safebook results will be used for:
• Internal use: SA footnotes, MIP metadata
references, MIP statistics explained
• Ranking/labelling/clustering MIP scoreboard per
each headline/MS
4. Closed
Task Name Duration Resource Initials Start Finish Milestone
% Work
Complete
Preparation of DM document 10 days DB 15 March 2014 27 March 2014 No 100%
DM endorsement 12 days DB 14 April 2014 29 April 2014 Yes 100%
Internal review of MIP Quality
Safebook template questions
30 days DB,IJ,MIP QT 30 April 2014 10 June 2014 No 100%
Presentation to MIP TF 1 day DB,IJ 23 May 2014 23 May 2014 Yes 100%
Peer review of questions by MIP
Steering Group
16 days MIP SG 11 June 2014 02 July 2014 No 100%
Presentation to MIP SG 1 day DB,IJ 24 June 2014 24 June 2014 No 100%
Setting of IPM version 8 days DB 09 July 2014 18 July 2014 Yes 100%
Reading of available inventories and
quality reports
85 days MIP QT 12 May 2014 September 2014 No 100%
Re-run for the 28 EU * 11 headlines 15 days MIP QT 27 August 2014
16 September
2014
No 90%
Preliminary results to be sent to MIP
SG for benchmarking
5 days MIP QT 17 September 30 September NO 100%
Benchmarking exercise 6 days MIP QT,MIP SG 30 September 2014 08 October 2014 No 100%
5. Ongoing and future steps
Task Name Duration Resource Initials Start Finish
Mileston
e
% Work
Complete
Benchmarking exercise 6 days MIP QT,MIP SG 30 September 2014 08 October 2014 No 90%
Director debriefing 1 day DB, JB, IJ 17 October 2014 17 October 2014 No 50%
Preparation of draft text for SA 5 days MIP QT 13 October 2014 20 October 2014 Yes 0%
Identification of Countries to be
visited
5 days AC 1st week of November No 0%
Footnotes for the SA 3 days MIP QT 03 November 2014 05 November 2014 No 0%
Country visits (joint with EDP?) 15 days ?? 01 December 2014 19 December 2014 No 0%
DMES debriefing 1 day PP 10 December 2014 10 December 2014 No 0%
DM debriefing 1 day PP 15 December 2014 15 December 2014 No 0%
6. Main dimensions
• Based on the lecture of all metadata and other
information the analysis focuses on :
• Footnotes + proposed text
• Quality improvements with respect to the previous
year
• Clustering by MIP headlines
• Focus on medium term (10 years)
7. SAFEBOOK
Current
account
balance -
% of GDP
Net
internatio
nal
investmen
t position
Real
Effective
Exchange
Rate
Share of
world
exports
Nominal
unit labour
cost index
House
Price
Index
deflated
Private
credit flow
Private
debt
General
governme
nt gross
debt (EDP)
Unemploy
ment rate
Total
financial
sector
liabilities
Thresholds -4/6% -35% ±5% & ±11% (*) -6% 9% & 12% (*) 6% 14% 133% 60% 10% 16.5%
BE No No No No No No No No No No
BG No No No No No No No No No No
CZ No No No No No No No No No No
DK No No No No No No No No No No
DE No No No No No No No No No No
EE No No No No No No No No No No
IE No No No No No No No No No No
EL No No No No No No No No No No
ES No No No No No No No No No No
FR No No No No Yes No No No yes No
HR Yes Yes No No No No No No No No
IT No No No No No No No No No No
CY No No No No No No No No No No
LV No No No Yes No Yes Yes No No No
LT No No No No No No No No No No
LU No No No No No No No No No Yes
HU No No No No No No No No No No
MT No No No No No No No No No No
NL No No No No No No No No No No
AT No No No No Yes No No No No No
PL No No No No Yes No No No No No
PT No No No No No No No No No No
RO No No No No Yes No No No No No
SI No No No No No No No No No No
SK No No No No No No No No No No
FI No No No No No No No No No No
SE No No No No No No No No No No
UK No No No No No No No No No No
Preliminary results
8. Proposed footnotes
• MIP team has identified so far 11 footnotes for the
Statistical Annex
4 footnotes for the House Price Index
footnotes in 3 domains for Latvia
• Dissemination team will liaise with domain managers to
decide which footnotes to retain and which information
should be used for metadata and which data should be
flagged
9. EDP vs Safebook
EDP Risk Assessment Safebook
Focus Risk Fit for purpose
(safeness = 1 - risk)
Score 0-10 0-100%
Indicators Debt + Deficit (aggregated) 11
Perspective Producer perspective User perspective
Sources Missions, expertise of desk
officers, data, inventories and
other metadata, other information
Knowledge, data, metadata,
quality reports and inventories
Respondents Desk officers MIP team
Benchmarking Management Directorate D Domain managers
ECFIN-ECB ?
Briefing to DG of Eurostat Briefing to Director of D
Usage EC internal use only
(work planning)
Eurostat internal use
(work planning)
10. Fit for purposes assessment
• On the basis of the score assigned to the
different questions and their relative impact
factor a Safebook index [0-100%] was calculated
for all indicators and Member States.
Legend
Safe >75%
Risk under control 75%<x<50%
Risk out of control <50%
Legend
Safe >95%
Risk under control 85%<x<95%
Potential risk 70%<x<85%
Risk out of control <70%
Initial clustering Final clustering (consistent with EDP)
13. GGD Benchmarking
EDP Risk Assessment
Scores
Scores are aggregated to debt
and deficit.
MIP assesses debt only
Linear
combination
Before
benchmarking
After
benchmarking
SAFEBOOK
General
governme
nt gross
debt (EDP)
Thresholds 60%
HR 37%
LU 62%
EL 68%
AT 70%
RO 72%
BE 72%
EE 73%
BG 73%
PT 73%
CZ 74%
IT 74%
CY 74%
LV 74%
SK 74%
ES 76%
PL 76%
HU 77%
DK 78%
IE 79%
LT 82%
FR 87%
SI 90%
UK 90%
DE 92%
MT 92%
NL 92%
FI 92%
SE 92%
SAFEBOOK
General
governme
nt gross
debt (EDP)
Thresholds 60%
HR 31%
BE 36%
ES 44%
RO 48%
IT 50%
AT 52%
PT 54%
LU 56%
IE 57%
LT 60%
BG 65%
EL 65%
LV 65%
DE 68%
CZ 68%
CY 68%
SK 68%
HU 70%
DK 70%
FR 71%
PL 76%
UK 76%
EE 77%
MT 77%
NL 77%
SE 77%
FI 83%
SI 84%
14. SAFEBOOK
Current
account
balance -
% of GDP
Net
internatio
nal
investmen
t position
Real
Effective
Exchange
Rate
Share of
world
exports
Nominal
unit labour
cost index
House
Price
Index
deflated
Private
credit flow
Private
debt
General
governme
nt gross
debt (EDP)
Unemploy
ment rate
Total
financial
sector
liabilities
Thresholds -4/6% -35% ±5% & ±11% (*) -6% 9% & 12% (*) 6% 14% 133% 60% 10% 16.5%
BE No No No No No No No No No No
BG No No No No No No No No No No
CZ No No No No No No No No No No
DK No No No No No No No No No No
DE No No No No No No No No No No
EE No No No No No No No No No No
IE No No No No No No No No No No
EL No No No No No No No No No No
ES No No No No No No No No No No
FR No No No No Yes No No No yes No
HR Yes Yes No No No No No No No No
IT No No No No No No No No No No
CY No No No No No No No No No No
LV No No No Yes No Yes Yes No No No
LT No No No No No No No No No No
LU No No No No No No No No No Yes
HU No No No No No No No No No No
MT No No No No No No No No No No
NL No No No No No No No No No No
AT No No No No Yes No No No No No
PL No No No No Yes No No No No No
PT No No No No No No No No No No
RO No No No No Yes No No No No No
SI No No No No No No No No No No
SK No No No No No No No No No No
FI No No No No No No No No No No
SE No No No No No No No No No No
UK No No No No No No No No No No
The country breakdown was not carried out for REER because there is a single compiler
We are not proposing footnote for GGD because we will take on board the reservations that will be released on the 21 october with the press release on general government deficit and debt (no duplication of work)
3 Footnotes for LV – 2 footnotes for FR and HR
Only HPI with 4 footnotes has more than 1 footnote
Additional information for metadata can also be collected during benchmarking
First step of benchmarking was toward EDP exercise
We read all inventories and metadata. Naïve way of reading -> user perspective
The first ranges correspond to the clustering used by the Court of Auditors for internal quality controls.
The final clustering used applies the ranges of the in the EDP Risk Assessment Exercise
With new clustering it is easier to discriminate
Balance of Payment statistics are relatively safe
During the benchmarking exercise for CA, IIP and EMS the domain managers confirmed our ranking. Indeed there was cooperation during the safebook exercise.
REER there is no country breakdown
Unemployment by far the safest indicator. During benchmarking there was the inclusion of additional information about methodological revisions due to the introduction of the census.
Concerning ULC our proposed ranking was not accepted. This should be followed up as our colleagues were too busy right now to provide a detailed feedback.
Financial Accounts statistics (Financial Sector Liabilites, Private sector debt and Private Sector credit flow) bears some risks. Domain manager suggested a new score for HR Financial Sector Liabilites according to new information (compilation manual from the national compiler).
House Price Index benchmarking is still pending and will be carried out after the 23rd following the adoption of Eurostat backdata estimations by working party
EDP see slide below
EDP assessment has a similar process. Indeed it has inspired our exercise.
However the EDP Risk Assessment Exercise is run on government debt and deficit together while in the MIP scoreboard we have only debt. For this reason we have used 50:50 weights otherwise we would have taken directly the EDP scores
After benchmarking we are able to identify 4 more countries which are not fully fit for purpose (risk is out of control).
Other countries that performed very low in our assessment but for which we do not have identified a footnote
Countries that are exceeding the threshold
We are not proposing footnote for GGD because we will take on board the reservations that will be released on the 21 october with the press release on general government deficit and debt