Page 267 - Special Topic Session (STS) - Volume 2

STS493 Irene S.
2.2 Product development
In addition to using administrative data in official statistics, these data also
prove to be very valuable in combination with data from sample surveys and
registrations. The administrative registers often contain far more detailed
information than sample surveys can provide, generating new information.
Combining data from a variety of registers also creates new possibilities. Hence
new, more detailed and up to date statistical information is made available. For
example, in recent years in the Netherlands, real estate drew a lot of attention,
especially empty shops and offices. The National Monitor on Disused premises
gives a complete view on all real estate (houses, offices and shops) at
municipality level and for houses even on a district and local area level. Based
on several registers like the Addresses of buildings, Valuation for tax purposes,
Resident’s registration and the Trade Register it was determined whether or
not a property is disused. After a first national publication in 2017, the method
was further customized by determining whether there was a matter of energy
use in empty buildings as indicated by data from energy companies and reality
checks were carried out to determine the relation between actual and
administrative disuse. Our users consider the figures appropriate to monitor
policy concerning property disuse. Another example is the combining of data
on ownership of vehicles, characteristics of vehicles, drivers’ licenses and
travelled distances with characteristics of persons and households, to satisfy
the needs on information on traffic and mobility related to trends in society, as
well as to develop more regional data and information on specific population
groups. On the international level, to get a better view on cross boundary
payments CBS is investigating which data on payment transactions could be
useful and is available at banks.

2.3 Adaptive survey design
The nature of primary data collection for official statistics is bound to
change. In response to budget pressure due to gradual but persistent declines
of response rates, designs like adaptive and responsive survey design have
received a lot of interest over the last decade e.g. Chun (2018). During the past
years, many surveys at CBS were redesigned to reduce cost and to increase or
maintain response rates where also alternative approaches like adaptive survey
design is investigated. Adaptive survey design assumes that differentiation of
effort over relevant population subgroups is either effective in improving
survey quality or efficient in reducing survey costs. Currently, adaptive survey
design is a standard option in redesigns of persons and household surveys at
CBS and was implemented in the Dutch Health Survey, e.g. Berkel et al. (2018).
How does adaptive survey design relate to the use of administrative data?
Adaptive survey designs have four main elements: quality and cost objectives
and metrics, stratification of the target population, design features, and an

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