Measuring digital trade in bop - A supplementary estimate of private household's digital purchases using internet data

Measuring digital trade in bop
A supplementary estimate of private household’s digital purchases using
internet data
Annette Meinusch and Benny Hessel, Deutsche Bundesbank

I. Motivation

 Internet and digitalization fundamentally change the way people, businesses and
 governments interact and open new windows to trade services internationally

 Examples: video streaming, online games, the purchase of apps and cloud computing

 Especially business to consumer (B2C) micro-transactions challenge compilers:
 measurement often falls outside of the traditional collection systems due to
 high exemption thresholds

 On the business side these thresholds are usually of minor relevance. However, in
 case of services consumed by households they lead to data gaps

 To close these gaps the Federal Statistical Office (Destatis) and the Deutsche
 Bundesbank initiated a project in 2018 to improve the measurement of digital trade

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 2

II. Definitions

 As a starting point, the new OECD-WTO handbook on "Measuring Digital Trade" was
 used to identify digital trade involved in cross-border transactions

The focus was laid on micro transactions that fall below the €12,500 exemption
threshold applied in Germany
 What: Services
 How: Digitally ordered, Platform enabled or digitally delivered
 Who: Business to Consumer
Annette Meinusch, Deutsche Bundesbank
February 2020
Page 3

III. Estimation Approach

 Bottom-up approach

 The online market was broken down into categories and the work between
 the institutions was assigned as follows:

 a. Video on demand
 b. Music on demand Deutsche Bundesbank
 c. Buying and using software
 d. Gambling
 Destatis
 e. Cloud Services

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 4

III. Estimation Approach
 a) Video on Demand (VoD)

 Estimates are based on a 2018 consumer panel, which was carried out at the behalf of
 Germany’s national film funding institution (FFA)

 Aggregate sales for VoD for Germany in 2017 account for €0.77 billion

 The share of video on demand that is relevant for the balance of payments (non-
 resident providers) is set as being 38% (Goldmedia) of total sales €290 million

 Sales in € bn, Sales
 Year according to in € bn,
 consumer panel non-resident
 2013 0.120 0.046
 2014 0.150 0.057
 2015 0.423 0.161
 2016 0.545 0.207
 2017 0.768 0.292

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 5

III. Estimation Approach
 b) Music on Demand

  Source: Federal Music Industry Association (BVMI)

  German digital music sales account for €0,741 billion in 2017

  Included:

 − Music downloads
 − Audio streaming
 − Real tones
 − Digital audio books

  Roughly 90% of sales from music on demand is generated by non-
 resident providers (Statista)

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 6

III. Estimation Approach
 b) Music on Demand - Adjustments

 Sales in € bn, according to
 Year Sales in € bn, non-resident
 consumer panel

 2008 0.140 0.126
 2009 0.163 0.147
 2010 0.203 0.183
 2011 0.247 0.222
 2012 0.294 0.265
 2013 0.328 0.295
 2014 0.371 0.334
 2015 0.486 0.437
 2016 0.604 0.544
 2017 0.741 0.667

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 7

III. Estimation Approach
 c) Buying and using software

Buying and using software can be broken down into the following two sub-
categories

 c1. Mobile applications (incl. games) for smartphones and tablets

 c2. Video games for PC/games consoles as well as online or browser
 games

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 8

III. Estimation Approach
 c1) Mobile applications (incl. games) for smartphones and tablets

 The German app market is estimated to be worth €1.17 billion

 Publishers of third-party apps use Android stores as sales platforms

 We assumed that the global market shares of Apple and Android stores are equivalent
 to their market shares in Germany

 Apple Store 41 bn US Dollar share of total app market sales of 50.6%
 Android Stores 40 bn US Dollar share of total app market sales of 49.4%

 − Apple Store sales in Germany: €0.592 billion
 − Android Store sales in Germany: €0.578 billion

 For the Android stores, using different sources, the assumption was made that 10% of
 sales (€0.0578 billion) are generated by German publishers and are not relevant to the
 balance of payments

 Preliminary result: The total amount of sales was estimated at € 1.11 billion

Source: APPANNIE https://www.appannie.com/de/ is the leading provider of app market data

 Annette Meinusch, Deutsche Bundesbank
 February 2020
 Page 9

III. Estimation Approach
 c1) Mobile applications (incl. games) for smartphones and tablets

 Adjustment for sales billed via one of the two app stores (Android Stores and iTunes)
 for transactions which are included in the “video on demand” and “music on demand”
 segment

 Sales Of which: Of which: Sales in € bn
 Year Growth acc. to Bitkom
 in € bn Video on demand Music on demand excl. VoD and MoD

 2009 732.1% 0.017 0.006 0.011

 2010 316.3% 0.072 0.007 0.065

 2011 149.5% 0.180 0.009 0.171

 2012 77.7% 0.319 0.011 0.308

 2013 27.2% 0.406 0.001 0.013 0.392

 2014 66.2% 0.675 0.001 0.015 0.659

 2015 41.6% 0.955 0.009 0.017 0.929

 2016 13.1% 1.081 0.011 0.022 1.048

 2017 3.6% 1.110 0.016 0.027 1.067

Source: Study provided by Bitkom, Germany’s digital association

 Annette Meinusch, Deutsche Bundesbank
 February 2020
 Page 10

III. Estimation Approach
 c2) Video games for PC / games consoles / online or browser games

 Sales can be broken down into seven categories
 Games for PC /Mac

 Adjustments × Share of
 Games for TV consoles downloads
 − Correction for physical games
 − Elimination of category “Games for mobile
 Games for handhelds
 phones / smartphones / tables“
 − Correction for In-App Purchases Games for mobile phones / smartphones /
 − Correction for German publisher tables

 Subscriptions / monthly fees for online
 games

  Share of sales from German games Monthly fee for online networks

 developments - In-App
 Spending on virtual goods in games purchases
 2014 2015 2016 2017 =
 6.9% 6.5% 6.4% 5.4% Sales in Germany × 1 − 

 =
Source: Game - the association of the games industry Games BOP

 Annette Meinusch, Deutsche Bundesbank
 February 2020
 Page 11

III. Estimation Approach
 d) Gambling

 The German Interstate Treaty on Gambling came into force on 1 July 2012

 The state governments’ gambling supervisory authorities have been drawing
 up annual reports since then (2014-18)

 Those include information on

 − private sport and horse racing betting
 − online casino games
 − online poker
 − online secondary lotteries

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 12

III. Estimation Approach
 d) Gambling

 Gross gambling revenue = gross stakes - winnings paid out
 Gross gambling revenue also includes a portion of sales that is generated not online but at a
 gambling outlet

 Gross gambling revenue
 Gross gambling revenue
 € million Share at
 Year € million
 Gambling outlets gambling outlets
 online only
 and online

 2012 418 21% 330

 2013 1,673 21% 1,322

 2014 1,746 23% 1,344

 2015 2,270 19% 1,839

 2016 2,558 20% 2,046

 2017 3,284 21% 2,594

Source: State governments’ gambling supervisory authorities, annual reports

 Annette Meinusch, Deutsche Bundesbank
 February 2020
 Page 13

III. Estimation Approach
 e) Cloud Services

 Although households get a certain amount of free storage space free of charge from their
 internet provider some of them buy extra space

 To calculate purchases of extra storage space by households a model was developed which
 uses information from

 1. ICT survey which contains information on cloud usage
 2. Internet research on storage space provide free of charge by different providers prices
 for cloud purchases Household purchases of storage space
 3. Share of German providers Year abroad
 € million

 2010 37
 Assumption: around 5% of cloud users do 2011 51
 indeed purchase storage space from abroad 2012 64
 2013 87
 2014 123
 2015 155
 2016 184
 2017 218

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 14

IV. Results I

 In 2019, estimated digital trade of German households amounted to € 7,426 million

 €m 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

 Apps 11 65 171 308 392 659 929 1,048 1,067 1,139 1,174

 Games 258 344 390 339 370 589 690 812 995 1,219 1,443

 Video 46 57 161 207 292 412 531

 Music 146 182 222 264 295 334 437 544 667 779 923

 Gamb 330 1,322 1,344 1,839 2,046 2,594 2,824 3,054

 Cloud 0 37 51 64 87 123 155 184 218 258 298

 Total 415 628 0.8 1,305 2,512 3,106 4,211 4,841 5,873 6,672 7,426

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 15

IV. Results II

 Data for digital household purchases is already included in German bop data (debits)

 in € m 2014 2015 2016 2017 2018 2019

 Total
 Services 254,428 272,197 285,599 303,693 311,333 293,894
 (debits)

 Digital
 Purchases 3,106 4,211 4,841 5,873 6,672 7,426
 abroad
 Share of
 digital
 purchases in 1,2% 1,5% 1,7% 1,9% 2,1% 2,5%
 total service
 expenditures

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 16

IV. Results III

 Effect of digital purchases on the net service balance (-44,2% in 2019)

 in € m Effect of digital purchases on the net service balance
 0
 2014 2015 2016 2017 2018 2019

 (5’000)

 (10’000)

 (14’014)
 (15’000) (15’031)
 (16’065)
 (16’973) (16’802)

 (19’242)
 (20’000)
 (20’686)
 (21’814) (21’938)
 (22’767)
 (24’228)
 (25’000)
 (25’873)

 (30’000)

 Net Service Balance excl. Digital Purchases Net Service Balance

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 17

V. Conclusions

 Bottom-up approach delivers a proxy for service debits of households provided via the internet

 Estimation model is flexible and expandable
 − implementation of other categories / subcategories (e.g. e - learning)
 − outdated categories can be excluded when appropriate

 Main data sources and data providers are known, which simplifies data updates in the future

 Use of internet data allows for a timely provision of information without the implementation of costly
 data collection methods such as household surveys

 Country breakdown is still to be calculated

 By using internet data compilers face the risk of losing access to data sources

 Quality issues: utmost care regarding sample sizes and the reputation of the data providers

 Project has to be seen as a first step to better measure internet induced micro-transactions in bop and
 provide data on digital trade according to the definitions of the handbook

Annette Meinusch, Deutsche Bundesbank
February 2020
Page 18