8
Joint meeting Working Groups on Environmental Accounts & Environmental Expenditure Statistics Luxembourg, 10 March 2015 Confidential data (point 7 of the agenda) Eurostat – Unit E2 Stephan Moll [email protected]

Joint meeting Working Groups on Environmental Accounts & Environmental Expenditure Statistics Luxembourg, 10 March 2015 Confidential data (point 7 of the

Embed Size (px)

Citation preview

Joint meeting Working Groups on Environmental Accounts & Environmental Expenditure Statistics

Luxembourg, 10 March 2015

Confidential data(point 7 of the agenda)

Eurostat – Unit E2

Stephan [email protected]

Eurostat

• Eurostat's goal: => publish EU aggregates

• Bottom-up approach used for EU aggregates

• Country data flagged as confidential => problem

• => Eurostat cannot publish EU aggregates!

• => limited use of environmental accounts!

• => AEA and EW-MFA data become almost useless

• Confidentiality is guaranteed by Regulation (EC) 223/2009

• Dilemma:

• Usefulness of environmental accounts Confidentiality clause

Introduction

Eurostat

• Multidimensional data cubes (AEA, EW-MFA)

• Hierarchically classified dimensions = most important!:

• AEA: NACE with three levels (A*64, A*51, A*21)

• EW-MFA: MATERIAL four levels (1 to 4 digits)

• If one cell is flagged confidential at detailed level:

• => next superior level also confidential (done by country)

• => EU aggregate cannot be shown for cell and superior levels

Problem description (1)

Eurostat

• AEA:

• 23% / 35% confidential in 2012 / 2008

• even NACE A*21 cannot be shown completely for EU !

• AEA becomes useless; cannot be used for input-output modelling (e.g. to calculate carbon footprint)

• EW-MFA:

• ca. 30% confidential

• only MF-1-digit-level for EU

• Policy requests indicators broken down by most detailed materials (ca. 50)(e.g. raw materials initiative of DG GROWTH)

Problem description (2)

Eurostat

1. Countries remove all confidentiality flags

+ perfect option for EU aggregates (all details can be displayed)

- may allow identification of individual units at the level of single countries

2. Countries reduce number of confidentiality flags (unreliable data, check again with experts, changing method)

+ maybe better than present

- still the full detail cannot be shown for EU

Possible solutions enabling EU aggregates (1/3)

Eurostat

3. Countries set 'secondary' confidentiality flags at most detailed level in order to disclose superior cells. Eurostat could provide a tool

+ better than present – EU can be shown at least for NACE A*21 and MF level 1

- still the full detail cannot be shown for EU – more iterations needed

4. Eurostat does controlled rounding of EU aggregates (e.g. to the next thousands, PRODCOM method)

+ EU aggregates can be shown whilst confidential country data stay closed

- the rounded values may become unrealistic or unreasonable

Possible solutions enabling EU aggregates (2/3)

Eurostat

5. Countries do not report confidential cells at the most detailed level; instead mark those as 'not available'

+ Eurostat could estimate the missing/confidential cells and publish all EU aggregates at most detailed level

- in certain cases it could happen that sub-totals need to be estimated too

6. Publication of EU aggregates is given priority. In case of confidential cells all 28 countries are automatically disclosed

+ Countries can report as usual and Eurostat can publish EU aggregates

- country data may be published nationally theoretically enabling identification

Possible solutions enabling EU aggregates (3/3)

Eurostat

• To ensure proper use of environmental accounts it is necessary to reduce to a minimum the number of cells that cannot be published – in particular at EU level

• The Working Group is invited to discuss the problem and move towards a common solution

Conclusions