Page 268 - Special Topic Session (STS) - Volume 2

STS493 Irene S.
optimization and implementation strategy. See Schouten et al. (2017) and
Tourangeau et al. (2017). It is within the domain of the stratification of the
target population that administrative data prove their value added in
improving survey designs. Determining target groups, also called
segmentation or clustering of the target population, is done with a
classification tree. People are divided into groups based on personal
characteristics. Within the Dutch Health Survey, demographic and regional
characteristics have been used that are known to have a different response
distribution than the population. Examples are ethnicity, ethnicity of parents,
age, income, urban character of the neighbourhood or municipality, education,
household-type and size, marital status, wealth, gender, and home ownership.
The strata were based on administrative variables that are used in post-survey
adjustments.

2.4 Improving Statistical Business Register backbone function
Next to improving survey design for social statistics, administrative data
also prove their value in the production of business demography and other
statistics. By linking administrative data to administrative and statistical units
stored in the Statistical Business Register (SBR) data sets can be enriched,
enterprises can be characterised and sub populations can be determined. For
example, Family Businesses (FB) are recognized to play an important role in
economies of the member states of the European Union (EU). FB make up
between 65 to 80% of all European companies, they make a significant
contribution to Europe's GNP and employment (40 to 50% of all jobs), and tend
to be great innovators, with a longer-term vision and specific commitment to
local communities (http://www.europeanfamilybusinesses.eu). In order to
measure the importance of FB’s their performance and characteristics
separating them from other kind of businesses they need to be characterized
within a SBR. As described by Konen (2017) CBS identified FB’s without
sampling and surveying but by using information from the SBR and
administrative registrations (Trade register, Payroll Tax register, management
of relations of tax authorities, satellite of Self-employed Entrepreneurs,
household register, alliance register and Child parent register). The research on
detecting Family Businesses in the SBR fits into a broader field of research on
‘profiling’ enterprises thereby differentiating businesses based on certain
characteristics. In addition, policy makers show interest in different typologies
for Small and Medium Enterprises. Besides Family Businesses and the Self
Employed, there is interest for Hidden Champions, Almost Failed Firms,
Ambitious Entrepreneurs, (Un)-Consciously Constraint Entrepreneurs and
Corporate Social Responsibility. These “sub-populations” can only be derived
by combining the SBR with a multitude of various data sources (registers,
administrative data, internet data etc.). At the Register Department of CBS

257 | I S I W S C 2 0 1 9