Transcript New Guidelines on Protection of Tabular Data at Statistics Finland Annu Nissinen [email protected] Joint UNECE/Eurostat Work Session on Statistical Data Confidentiality Ottawa, Canada, 28-30 October.

New Guidelines on Protection of
Tabular Data at Statistics Finland
Annu Nissinen
[email protected]
Joint UNECE/Eurostat Work Session on Statistical Data Confidentiality
Ottawa, Canada, 28-30 October 2013
Outline of the presentation
1. Protection of tabular data at Statistics Finland: pre-2012
situation
2. What has been done to improve our procedures on SDC
3. Current situation: where are we now and what’s next
Annu Nissinen
30/10/2013
2
Starting point in the end of 2011
Internal guidelines on the protection of tabular data written
in 2000-2002
 What has happened in ten years?

Legislation has changed, data protection methods and tools
have developed
 SDC practices at different departments of the agency have
developed and adopted their own standards. The suitability of
all those methods was not sure. Outdated, far from optimal
methods existed.

→ Updating of the guidelines was needed
Annu Nissinen
30/10/2013
3
Starting point in the end of 2011

Do we have enough resources?
As a state authority, Statistics Finland is dependent on public
money
→ Setting limits to research and development work
→ Can we have up-to-date and optimal protection methods
available?


How can we improve the situation with this starting point?
Annu Nissinen
30/10/2013
4
Renewal of the guidelines

Two working groups to renew the guidelines


First task was to find out how the old guidelines were working
in practice
Problems with the old guidelines:
Not pragmatic
 Only few examples of very general sensitivity rules and
protection methods
 Partially too open to interpretations
 Not even give a clear answer to which statistics they were to
be applied

Annu Nissinen
30/10/2013
5
Renewal of the guidelines

Challenges met by the working groups:
No obligation for the departments or units to write down their
protection methods
→ almost impossible to evaluate or compare the different
methods in use
 Collection of tacit knowledge isn’t easy
 High expectations towards the renewal: “After renewal, all
problems considering SDC will be solved (in theory and in
practise).”

Annu Nissinen
30/10/2013
6
The new guidelines

How do the new guidelines improve the situation?
1. Solving problems caused by the old guidelines

The guidelines are clearer and more pragmatic and they
should lead departments to check their current SDC
methods and update them if needed
2. Preventing new problems

The SDC instructions need to be written and available

The departments are obliged to draft a formal list of the
statistics in their responsibility needing protection
against statistical disclosure
Annu Nissinen
30/10/2013
7
Current situation

New guidelines exist and the next step is applying them in
statistics production
There’s a window of opportunity for changes in SDC practises
due to other big changes in operating environment
 Change takes time inside a big organisation
 Resources for the change in practises?


Inside the agency the new guidelines can be seen as
unwelcome extra red tape and unnecessary increase in the
workload...
 ... or a possibility to improve the quality of statistics

Annu Nissinen
30/10/2013
8
Has the situation improved?

Adoption of the new guidelines (and possibly new SDC
practises) still in process

Crucial success factors:
 Right “marketing”
 Right allocation of required resources
Annu Nissinen
30/10/2013
9