Session 8A - Mobile phone data as a source for official statistics
Using mobile positioning data for official statistics: daydream nation or promised land?
NTTS 2015 9 – 13 March 2015, Brussels
Christophe Demunter, Fernando Reis
EUROSTAT – Unit G-3 "Short-term business statistics and tourism" & TF Big Data
Recent Eurostat project
Feasibility study on the use of mobile positioning data for tourism statistics
Origins
Changing geo-political environment
Quickly evolving technology and large-scale adoption of tools/devices
Changing working environment of official statisticians
New technologies, new techniques, new sources and a new 'Zeitgeist'
boost and stimulate a paradigm shift in official statistics
Recent Eurostat project
Feasibility study on the use of mobile positioning data for tourism statistics
December 2012 – June 2014
Carried out by a multidisciplinary, international consortium of 6 partners from 4 countries (EE, FI, DE, FR)
National statistical institutes
Tourism/mobility reseachers
Academics
Data scientists
Recent Eurostat project
Feasibility study on the use of mobile positioning data for tourism statistics
All reports are on the Eurostat website
Stock-taking
Feasibility of access
Feasibility of use (methodological issues)
Feasibility of use (coherence)
Opportunities and benefits
Consolidated report (34 pages)
Unanswered call ? – barriers to access
Protection of personal data
Interpretation of concepts such as 'personal data', 'anonymised', etc.
Fear of public opinion
Strong need for a less fuzzy legal environment at national &
international level !
Technical challenges
Treatment of very large datasets
Choice between a decentralised
or centralised system
Complex but not impossible;
not considered a hard barrier
Unanswered call ? – barriers to access
Financial and business related barriers
Business secrets for Mobile Network Operators (MNOs)
Cost and burden for MNOs
Need for a mutually beneficial relationship to motivate or incite MNOs
Improving access to mobile positioning data is
THE main short term challenge in order to pave the way
for a more generalised use of this source of big data!
Wrong number ? – methodological issues
Shortcomings that are inherent to mobile phone data
Overcoverage and undercoverage issues
Not more significant than similar shortcomings of 'traditional' sources
Shortcomings that are inherent to new technologies
Continuity of data (what if MNOs drop out?)
Consumer behaviour and preferences are not stable
Need for constant innovation and search for complementary sources
Wrong number ? – methodological issues
Reproduce existing statistical indicators
Not always easy to reconstruct the existing scope and definitions
Absence of socio-demographic breakdowns and domain-specific info
(e.g. purpose, transport, expenditure in the case of tourism statistics)
But additional indicators previously not available, e.g. granularity of
regional breakdowns
Improved quality (better timeliness, less respondent burden, less
recall bias)
Promising results in
terms of coherence
Trade-off to be made
by 'open-minded' users
and producers
0
50 000
100 000
150 000
200 000
250 000
300 000
350 000
400 000
450 000
500 000
Q1-
09
Q2-
09
Q3-
09
Q4-
09
Q1-
10
Q2-
10
Q3-
10
Q4-
10
Q1-
11
Q2-
11
Q3-
11
Q4-
11
Q1-
12
Q2-
12
Q3-
12
Q4-
12
MOB_OUT(EU-27)_OVERNIGHT DEMAND_EE(EU-27)_OVERNIGHT
Wrong number ? – methodological issues
Official statisticians have to think out of the box,
out of the comfort zone :
Origin and relevance of current statistics
primarily driven by relevance & user needs?
Or by available sources & methods 'at the time' ?
We shall not repeat, but do better !
Use of big data necessitates a revolution of the
mindset rather than a simple evolution !
Rethinking indicators, zero-base user need
analysis, not only incremental changes in the
existing frame.
First caller wins ! – the war on data
Mobile phone data is not likely to entirely replace the existing production methodology of official statistics
Explore mixed-mode solutions (e.g. large samples based on big data
+ smaller follow-up surveys to collect additional information)
But … to remain an efficient and competitive player, official statisticians will sooner or later have to rely on big data for part of their business
Join forces across statistical domains that can use the same big data
source
Launch production of experimental statistics before someone else
dominates 'the market'
A next Eurostat/ESS project ?
Multi-country and multi-domain project in the pipeline
Given that getting access is a critical factor, the number of domains
analysed and assessed should be maximised: e.g. population, balance
of payments (travel), transport and urban mobility, tourism
Involve several countries (ESSnet-style), possibly two-speed approach
Use of data stored by Mobile Network Operators
Call detail records and data detail records
Expected output
Partnerships with MNOs
Studying data structures and defining data access standards
Testing data compilation and assessing quality
Thanks for your attention