Issuu

The value of data JUNE 3, 2021 James Tebrake Assistant Director IMF Statistics Department IMF | Statistics

Disclaimer: The views expressed in this presentation are those of the authors and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.

The value of data to Forest Elephants in Africa

A team of Cornell researchers placed 50 custom audio recorders in a Central African Rainforest and collected hundreds of thousands of hours of jungle sounds. The team of researchers build a neural network to identify and digitize the elephant sounds. In addition to the elephant sounds the network also isolated the sound of gunshots which was also digitized.

The team of researchers built a database that included, among other things, the time and location of the elephant sounds and the gunshots.

This information was analyzed and used to identify where the elephants were being killed and to develop tactics to stop poachers before they fired.

The digitization of elephants sounds, the storage of this information in a database, and the insight drawn out of the data is helping to conserve the elephant population.

 The use of data is not new, what is changing is that the production and use of data is now widespread. Almost every business is capturing “events”, “observable phenomena” in digital format and using this data in some form to drive production, sales, logistics, and innovation. Temple-Raston, D., “Elephants under Attack have an unlikely ally: Artificial Intelligence”, NPR October 25, 2019. https://www.npr.org/2019/10/25/760487476/elephants-underattack-have-an-unlikely-ally-artificial-intelligence IMF | Statistics

It may be best to view this activity as a chain...

Events

Actions, events, observable phenomena, … stuff happens every day, whether it is someone getting a COVID-19 test, clicking a link on a website, driving to visit a friend, logging on to their computer…. or a tree falling in the forest. IMF | Statistics

Digitized

Sometimes, someone or some entity decides to record this and digitize this “stuff” (i.e. convert it into digital format or produce a datum).

Databases

These datum (individual digitized events or observable phenomena) are organized and stored in databases.

Knowledge

Knowledge is acquired by analyzing the information in the database and this knowledge is used to devise new applications.

What does the System of National Accounts (SNA) say about “data”?

Events, observable phenomena, stuff happening is not production i.e. it is not “an activity, carried out under the responsibility, control and management of an institutional unit, that uses inputs of labour, capital, and goods and services to produce outputs of goods and services.” SNA 6.2 IMF | Statistics

“The cost of preparing

“Databases consist of

“Research and

data in the appropriate format is included in the cost of the database but not the cost of

files of data organized in such a way as to permit resourceeffective access and use of the data. Databases may be developed exclusively for own use or for sale as an entity or for sale by means of a licence to access the information contained.” SNA 10.112

development consists of the value of expenditures on creative work undertaken on a systematic basis in order to increase the stock of knowledge, including knowledge of man, culture and society, and use of this stock of knowledge to devise new applications.” SNA 10.103

acquiring or producing the data.” SNA 10.113

What does the System of National Accounts (SNA) say about valuing data? SNA Recommended Treatment vs Practice SNA Recommended Treatment

Practice

IMF | Statistics

No value as per 2008 SNA

Market prices or sum of costs if produced for own use.

Some of this may be included in the value of databases as the cost of preparing data.

This is mostly likely being undervalued given the proliferation of firms developing databases. It is also often combined with estimates of software.

This may be undervalued given the proliferation of firms deriving knowledge from their data and using it to devise new applications.

If data is an investment flow, what is the stock of data?  The SNA has various methods to estimate the stock of data such as accumulating the value of investment flows through time. This requires information about the investment flows, prices and asset lives.  To complicate things… the value of the stock can change over time (i.e. through depreciation, revaluations, obsolescence etc..).

 Some of this may be captured in Goodwill – but… Goodwill is usually a “catchall” and is only recorded when its value is evidenced by a market transaction. “For reasons of reliability, goodwill is only recorded in the SNA when its value is evidenced by a market transaction, usually the sale of the whole corporation. Exceptionally, identified marketing assets may be sold individually and separately from the whole corporation, in which case their sale should also be recorded under this item.” (SNA para 10.199) IMF | Statistics

A possible way forward for the SNA?  Create a new intellectual property product asset class “Data” or expand the “boundary” of databases to include the costs of acquiring or producing data.  Encourage countries to delineate between databases and software (and provide appropriate guidance) and develop research and development asset classes – with one related to data science.  Work with the business accounting community to delineate data / databases from goodwill, record databases on balance sheets when created and develop appropriate rates of depreciation / consumption of fixed capital.

IMF | Statistics

The Value of Data some experimentation Labor Costs Estimates for Data-Related Activities

“Net Stock” of Data, Data Science and Databases

 A few statistical agencies have developed experimental estimates of data. Early indications are that data is economically significant and growing.  The update of the SNA is an opportune time to ensure all the activity related to the production and use of data is included in the production and asset boundary and the estimates are “highly” visible given the implication for policy, including consumer protection policy. IMF | Statistics

References  Temple-Raston, D., “Elephants under Attack have an unlikely ally: Artificial Intelligence”, NPR October 25, 2019. https://www.npr.org/2019/10/25/760487476/elephants-under-attack-have-anunlikely-ally-artificial-intelligence

 Reinsdorf M., Ribarsky J. (2019); “Measuring the Digital Economy in Macroeconomic Statistics: The role of data”. Presented at the 2020 ASSA Meetings.  Quirós G., Reinsdorf M., “Measuring the Digital Economy”. Informal Briefing to the IMF Executive Board Paper. Available at: file:///C:/Users/JTebrake/Downloads/022818MeasuringDigitalEconomy%20(1).pdf  Bugge, B., Smith, P., Tebrake, J., “The value of data in Canada: Experimental Estimates”. Available at: https://www150.statcan.gc.ca/n1/en/catalogue/13-605-X201900100009

IMF | Statistics

References  ISWGNA Task Team on Digitalization. “An update on recording and measuring data in the system of national accounts”. Presented at the 14th meeting of the Advisory Expert Group on National Accounts.  Ahmad N., van de Ven P. (2018); “Recording and measuring data in the System of National Accounts”. Available at https://unstats.un.org/unsd/nationalaccount/aeg/2018/M12_3c1_Data_SNA_asset_boundary.pdf  United Nations, European Commission, International Monetary Fund, Organisation for Economic Co-operation and Development, World Bank, System of National Accounts (SNA) 2008, New York, 2009. Available at https://unstats.un.org/unsd/nationalaccount/docs/sna2008.pdf .

IMF | Statistics

Thank you!

James Tebrake (jtebrake@imf.org) IMF Statistics Department

IMF | Statistics