Business Register Quality Improvement

1
Business Register Quality Improvement

• 18th International Roundtable on Business Survey
  Frames, 17 - 22 October 2004

2
Business Register data collection

• Sources
• surveys
• administrative sources
• Frequency
• annual
• monthly, bimonthly

3
Goals of quality project

• revision of processes
• quality measurement
• documentation

4
Quality project - goals / background

• revision of processes
• economic statistics need more real time BR data
• productivity should rise, resource problems
• quality measurement
• EU demands
• needs of national users
• documentation
• user needs
• transfer and extend of knowledge among BR
  personnel
• organisation of work, productivity

5
Revision of processes survey process

• Old practice
• sending mass of paper questionnaires mainly in
  the beginning of the year
• checking and processing in BR took even 6 months
• plenty of checking as batch programs - hard phase
  in the annual work process
• some merger events were updated in BR with a
  delay of a year or more

6
Revision of processes survey process

• Revised practice
• sending of paper questionnaires more frequently -
  smaller amounts at each time
• all round the year - smoothen workload
• shorter delay between response and its processing
• checking procedures for each group
• monthly phone profiling of largest re-structuring
  enterprises and of starting big enterprises

7
Revision of processes

• survey selection hint of significant changes in
  certain variables causes, that unit will be
  selected
• hint may come via
• media
• feedback from Statistics Finland
• administrative data monthly wages and salaries,
  sales
• mergers from Trade Register or Tax Administration
• Finlands Post address data

8
Annual allocation of tasks, before the revision
9
Change of timing of the annual survey process
10
Main features in revision of processes

• Special practice for largest re-structuring cases
• continuous work on checking and
• on error situations
• some of checking procedures could be started and
  timed from on-line application

11
Gains of the revision of processes

• easier to communicate with respondents
• respondents are more motivated to give
  information
• organisation of work, organisation of deputy
  duties
• better quality of collected information
• more real time data

12
Speeding up of the completion of the annual
statistical file
13
New quality indicators

• coverage measurement
• quality control survey
• monitoring will be systematic and carried out
  annually

14
Coverage measurement

• BR comparison against certain administrative
  sources
• data source missing from BR
• business tax 10,614
• monthly payment control 6,465
• pay as you earn 30,957
• Result, missing from BR coverage
• units 13,7 , limited companies 1,0
• turnover 0,6
• wage bill 0,3

15
Quality indicator of variables

• QUALITY SURVEY
• addressed to small enterprises (not belonging to
  ordinary annual surveys)
• stratified sampling
• measures quality of certain variables
• will be carried out annually

16
Quality indicator of variables

• QUALITY SURVEY, results
• NACE-code (Section level) 95
• NACE-code (5-digit level) 90
• municipality 94
• postal code 88
• address 71
• multi-location or not 99

17
Results of quality survey
18
Documentation

• Variable descriptions, 730 in all
• Process descriptions, 30 in all
• forms a Handbook of Business Register for BR
  people and BR users
• use as training material
• available for everyone in Statistics Finland on
  Intranet
• will be kept up-to-date

19
Structure of variable descriptions

• Title name of the variable
• Content explanation of the variable, possible
  code values, purpose of use
• Source how and from where the value is derived,
  may exist more than one
• Processing explanation how the variable is
  processed
• Processing rule(s) rules of the data system,
  which affect to the variable

20
Example of variable description

• Title oyr_omist_tyyppi
• Content Owner type of legal unit. Owner types
  are
• 1 Private domestic, 2 State
• 3 Municipality 4 Region of Åland
• 5 Foreign-owned 6 Other owner type
• Source Tax Administrations customer database
  (ASITK), Bank of Finland, Register of Enterprise
  Groups
• Processing Inferred by default from the code of
  legal form. In type 5 foreign ownership is at
  least 50 per cent. This is used only for
  corporate units. For legal units which are
  subsidiaries of enterprise groups, ownership
  information is updated according to the parent of
  the group.
• Processing rule A31.1 Processing rule for
  updating of the Tax Administrations customer
  database, B62.4 Inference of owner type by means
  of enterprise group information, A34 Updating
  with the Bank of Finland data.

21
Process descriptions

• 2 pages each process flow chart and one page
  text
• Input processes, 15 key or sub-processes
• example Updating by Finlands Post business
  address data
• Processing, 9 key or sub-processes
• example processing of intra-enterprise
  demographic information
• Output processes, 7 key or sub-processes
• example creation and checking of the statistical
  file

22
THANK YOU !