Page 131 - Special Topic Session (STS) - Volume 4

STS571 Jaanus Kroon
population. The data are readily available, which makes data collection faster
and more cost-effective. As mobile phones are widely used, the resulting data
set is comprehensive; it minimises the human factor affecting interviewer
interpretation in surveys and ensures homogeneity. So, the data are more
accurate and of better quality than those collected by traditional data
collection methods. Although each transaction with a payment card would
give a location-based fact similar to the roaming transaction of a mobile
owner, ten years ago it was obvious that the facts from card data were not as
numerous as those detected by mobile positioning. That is why it was decided
to continue payment card data collection on an aggregated basis for
estimating travel expenditures and to use it as one of the main alternative data
sources that can help to validate the border-crossing aggregates derived from
the mobile positioning data (MPD).

2. Mobile positioning-based statistics of border crossing:
methodological aspects
The jointly developed methodology is based on the use of readily available
log files from Mobile Network Operators (MNOs), registering the information
needed for billing incoming and outgoing roaming activities like voice calls,
SMSs and MMSs, mobile-data usage, and mobile supported GPS usage. These
activity events are called Call Detail Records (CDR). The parameters that the
methodology needs for each call activity are:

• SIM card ID, replaced by a randomly generated pseudonymous ID for
statistical use;
• date and time;
• antenna ID with location data;
• country ID.
In line with the Balance of Payments (BoP) methodology, mobile
positioning determines the residence of a traveller using the permanent
residence criterion, regardless of the resident’s citizenship or nationality. As a
rule people sign a contract with a mobile company in the country where the
phone will be used most frequently, and so the phone owners are presumed
to reside in the country where their SIM card is registered. This approach is
supposed to give even more precise results under the residency concept of
BoP statistics. The amount, length and nature of the visits of Estonian residents
and non-residents are determined by the location-based anonymised use
patterns of mobile phones derived from the roaming activities in the reporting
resident operator network, and operator clients’ roaming activities in networks
abroad. The statistics on inbound and outbound travel reflect both same-day
and overnight visits:
• The number of visitors is determined from the encoded number IDs.

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