SOME CONCLUSIONS ON OPPORTUNITIES AND CHALLENGES WITH A POLICY VIEW
The past two decades have witnessed extraordinary developments related to data storage capacity, high-speed networks to move and acquire data, and rapid growth in computational power.
Significant changes in the volume and uses of data have been enabled by improvements in the ability to collect, transfer, store, aggregate, link and analyze this data. Therefore, data are nowadays processed, redeployed, shared and transferred around the clock and across the globe. Whether this data creates economic and social value at an increasingly greater pace is no longer in question. However, measuring the value generated remains a complex task, as data are used in so many situations and contexts for many purposes. In addition, it is clear that the Internet is not simply an excellent tool for acquiring and assembling data of all types, but also a key platform for many emerging innovative applications that transform this data into valuable services for consumers and businesses. Throughout this chapter the context for our analysis has been the increasing importance of understanding developments in the Internet economy that are drivers of innovation and growth.Big data economics appears as a challenging field of research with considerable opportunities to provide value to the academic community and to practitioners. In the following paragraphs, ideas, extracted and summarized from the analysis of previous sections, many of them in incipient stages, are presented as potential avenues for further research.
25.5.1 Knowledge Discovery
Deriving knowledge from very large amounts of data - knowledge discovery - needs tools other than data-based traditional models, in particular when considering unstructured information (Dhar, 2013). Unlike conventional database querying that asks ‘What data satisfy this pattern?’ or econometric modeling that follows hypothesis-driven research, knowledge discovery asks ‘What patterns satisfy this data?’, taking advantage of observations that would not have been made under controlled circumstances.
In general, statistical and econometric models need to be augmented by theoretical arguments and additional tests to overcome difficulties with causality or when scaling conclusions to a different situation.Therefore, there is considerable room for developments combining statistics, econometrics and computer science. For instance, the usual approach in machine learning detects structure in data without strong assumptions about linearity or the parameters of underlying distributions. Apart from the need for high quantities of data, a disadvantage of this approach is that it could also pick up spurious relations in data, thus arriving at meaningless conclusions. Therefore, big data analytical techniques will typically benefit from combining them with more traditional (modeling) approaches.
Another area of interest is forecasting. In conventional statistics and econometrics, errors in prediction are derived from three main sources: the mis-specification of the model (too many variables, too few variables, wrong assumptions about the relationships between variables or about their distribution), use of a limited number of observations for the estimation of parameters, and the random disturbances and measurements errors. Using big data allows reducing the first two types of challenges significantly as the need to specify the model in detail decreases and as availability of large numbers of observations and finely grained data mitigates the problems of sampling. The use of big data hence allows the introduction of new forms and methods of forecasting.
25.5.2 Data-based Platforms
From a narrower economic perspective more research is needed on the value of data and the contribution of data to economic growth and jobs. Several areas look promising including the increasing relevance of metadata and context, the implications of open (big) data and data portability across platforms and applications, data formats and types, and the capabilities of users to control their data (i.e., privacy-enhancing technologies).
Another area of interest is research on data platforms and multi-sided markets. From the brief analysis in the section above several issues can be identified as promising avenues for further research. They include the creation and sustainability of both direct and indirect network effects, as well as the design and structure of the platform business. Moreover, the technological and business relationships, the pricing of - and subsidies to - sides belonging to the platform, as well as strategic decisions on quality, degree of openness and differentiation raise challenging research problems.
25.5.3 Data Governance
From an economic policy perspective ‘data governance’ is also an increasingly relevant subject. According to OECD (2014) data governance is the overall management of the availability, accessibility, usability, integrity and security of the data collected and stored. In addition, data governance includes policies and regulations that form a consistent and effective framework for such data management.
We suggest, additionally, that data governance should not be construed merely as an economic matter but that it should also encompass a social perspective. The issue at stake is an emerging mythology of big data (boyd and Crawford, 2012) that could lead to a blind acceptance of this new paradigm. Areas for careful consideration include acknowledging that conclusions based on data are always subject to a process of interpretation, that bigger data does not necessarily equate to better data, the relevance of the context of data, the ethical consequences of the intensive usage of data, and the new digital divides because of lacking access to data.
Governance is particularly important for data from public sources or in open data schemes. In this regard, the health and security domains are paramount examples, particularly if multiple stakeholders and different countries are required to cooperate. Additional challenging areas to those already mentioned appear to be: financial sustainability of initiatives, framework and incentives for exchange and access to data, coordination and complementarity of initiatives, quality and efficiency of collected data, and capacity building and training.
Deficits in existing skills in the big data domain are particularly relevant as main decisions on acquisition, storage, and analysis of data are being taken today and will condition upcoming developments. These skills are transversal in nature and include prominently knowledge on state-of-the art technology (sensors, networks, and databases), analysis (from econometrics to machine learning as discussed previously), and legal aspects (privacy, ownership, and regulation).25.5.4 Policy and Regulatory Framework
As an emerging area, still rather immature and full of innovation promises, many stakeholders including big companies assert that big data should not be regulated (Hemerly, 2013). However, there are some areas affected by big data where public policies and regulation do exist and that can have a deep impact on the evolution of the field. The most prominent of these areas are (1) the domain of privacy and use of - mostly personal - data by service providers; (2) data ownership; and (3) the extent of open data.
Starting with privacy, regulation of service providers’ usage of personal information represents a true challenge for policy-makers, as difficult trade-offs between innovation and user rights may need to be resolved (Gomez-Barroso and Feijoo, 2013). On the one hand, excessively restrictive access to data can inhibit the growth of potentially valuable services or applications. On the other hand, risks associated with privacy intrusions must be carefully inspected because they affect not only economic interests but also people’s lives.
Several approaches are possible from a very general policy perspective ranging from the use of market forces in a laissez faire framework to reliance on some regulation ‘by design’ or ‘by law’ (or any combination of both). Considering the pure market approach, existing evidence as described in the chapter suggests that interested companies are still in a very early phase in their learning curve on big data, basically gathering all available information on every aspect of their businesses.
Most of them are also trying rather basic approaches as business models on this data: deducting behavioral patterns of users and adjusting whatever is the business model - advertising, subscription, sales, other - to these observed patterns, or even simpler: just selling data to interested third parties, which in turn include the user profiling in their business models. Is the market responding rapidly to concerns of users about the (mis)use of their personal information - supposing they are aware? Most evidence today suggests that this is not the case. Unclear terms in service agreements, significant market position of leader providers - and absence of respectful providers - unbounded usage of information gathered in a particular transaction and, above all, prominent examples of data gathering without users knowing all point toward slow market reaction by its own means, if any.Regulation by design is linked with the fact that the amount of information that users reveal to providers increases with the ability of technology to provide better and more personalized services. In this regard, users - or rather, consumers - lack information and tools concerning the amount of data that it is collected and about the preservation of contextual integrity during the flow of information. On the supply side of the market, privacy-enhancing technologies may make it possible to reach equilibria where data holders can still analyze aggregate and anonymized data while subjects’ individual information stays protected. Open big data or data portability could help to increase transparency in the domain and avoid users’ lock-in effects.
Data portability is obviously linked with data ownership. In this area, two issues appear attract increasing attention in current discussions. The first is the application of intellectual property rights (IPRs) on data. There are different positions on how and to what apply IPRs. The traditional position suggests distinguishing between raw data without IPRs and findings and facts on data - structured datasets included - that would have some form of IPR.
It is unclear if this ‘legacy’ solution will contribute to innovation or deter it.The second issue, related to the previous one, refers to data obtained through public funds. There is an increasing pressure to release in an open manner all publicly funded data. For instance, the National Science Foundation in the USA and the Horizon 2020 framework program in Europe contain provisions for open data management in research projects. The provision of open data by all types of government requires technical resources to ensure that data are readily available in adequate formats or even in some standardized version. It is not just a matter of publishing data in whatever circumstances. The availability of this data - often in the context of cities and/or the Internet of things - could give rise to opportunities for innovation and improvement of public services within so-called ‘data civics’ (Hemerly, 2013) (or ‘data governance’ as discussed in this chapter).
In sum, the big data domain is full of opportunities and challenges for economic research and a more robust understanding of the economics of (big) data is needed. The brief introductory review presented in this chapter is a modest first step in that direction.
NOTES
1. See BBVA, ‘Big data: En que punto estamos?’ [Big data: where are we?] Innovation Edge, June 2013, accessed 6 January 2015 at http://www.centrodeinnovacionbbva.com/en/innovation-edge/big-data/big- data-where-we. Translation from the original Spanish document by the authors.
2. Sometimes referred to as the four (or sometimes three) Vs of big data. See IBM (2014), ‘The four V’s of big data’, accessed 23 March 2015 at http://www.ibmbigdatahub.com/infographic/four-vs-big-data.
3. Attempts to define big data by means of size are a daunting task. A few numbers suffice to illustrate the challenges: in 2013 30 billion pieces of content were shared on Facebook every month, 2.9 million emails were sent every second, 72.9 products were ordered on Amazon per second, and so on, reaching 2.5 quintillion bytes of data exchanged every day. See IBM (n.d.), ‘What is big data?’, accessed 22 January 2015 at http://www.ibm.com/big-data/us/en/.
4. See Glossary of Telecommunications Terms from US National Communication Systems (Federal Standard), accessed 6 January 2015 at http://www.its.bldrdoc.gov/fs-1037/fs-1037c.htm.
5. See SAS (n.d.), ‘What is big data?’, SAS.com, accessed 24 January 2015 at http://www.sas.com/en_us/ insights/big-data/what-is-big-data.html.
6. Open Data Institute accessed 23 March 2015 at http://theodi.org/.
7. European Union Open Data Portal accessed 23 March 2015 at http://open-data.europa.eu/.
8. See Data.Gov: The Home of the US Government’s Open Data, accessed 23 March 2015 at http://www. data.gov/.
9. See Data.Gov.UK: Opening Up Government, accessed 23 March 2015 at http://data.gov.uk/.
10. See OECD, Data-driven Innovation for Growth and Well-being, accessed 23 March 2015 at http://oe.cd/ bigdata.
11. ‘Data, data everywhere’, The Economist, 25 February 2010, accessed 22 January 2015 at http://www. economist.com/node/15557443.
12. See the work of the Institute for Prospective Technology Studies (IPTS) of the European Commission (EC), the EC research program on big data within Horizon 2020, and the OECD work program mentioned above.
13. From a very abstract perspective it would be possible to distinguish between data that exist prior to any measurement from data that only exist they have been obtained. The former are sometimes termed ‘raw data’, although in essence all data are raw, as stated by Gitelman (2013). In this chapter the term is largely avoided except in the last section where differences in intellectual property rights (IPR) before and after data acquisition are discussed.
REFERENCES
Acquisti, A. and H. Varian (2005), ‘Conditioning prices on purchase history’, Marketing Science, 24 (3), 1-15. Akred, J., R. Williamson and S. O’Sullivan (2013), ‘Building a data platform’, paper presented at the Strata Conference + Hadoop World, New York, 28 October 2013, accessed 1 January 2016 at http://conferences. oreilly.com/strata/stratany2014/public/schedule/detail/36512.
Ballon, P. (2009), ‘The platformisation of the European mobile industry’, Communications & Strategies, 3 (75), 15-34.
Belleflamme, P. and M. Peitz (2010), Industrial Organization: Markets and Strategies, Cambridge, UK: Cambridge University Press.
boyd, d. and K. Crawford (2012), ‘Critical questions for big data’, Information, Communication & Society, 15 (5), 662-79.
Buchholtz, S., M. Bukowski and A. Sniegocki (2014), Big and Open Data in Europe: A Growth Engine or a Missed Opportunity, report commissioned by demosEUROPA - Centre for European Strategy Foundation, Warsaw Institute for Economic Studies, accessed 12 January 2015 at http://www.bigopendata.eu/wp-content/ uploads/2014/01/bod_europe_2020_full_report_singlepage.pdf.
Chen, H., R.H.L. Chiang and V.C. Storey (2012), ‘Business intelligence and analytics: From big data to big impact’, MIS Quarterly, 36 (4), 1165-88.
Dhar, V. (2013), ‘Data science and prediction’, Communications of the ACM, 56 (12), 64-73.
Feijoo, C., J.L. Gomez-Barroso and P. Voigt (2014), ‘Exploring the economic value of personal information from firms’ financial statements’, International Journal on Information Management, 34 (2), 248-56.
Feijoo, C., J.L. Gomez-Barroso and D. Potoglou et al. (2013), ‘Analysis of the economic value of personal information’, Seville: Institute for Prospective Technological Studies-European Commission.
Feinleib, D. (2012), ‘Big data landscape’, Forbes, 19 June, accessed 22 January 2015 at http://www.forbes.com/ sites/davefeinleib/2012/06/19/the-big-data-landscape/.
Gawer A. and M.A. Cusumano (2002), Platform Leadership: How Intel, Microsoft, and Cisco Drive Industry Innovation, Boston, MA: Harvard Business School Press.
Gawer, A. and R. Henderson (2007), ‘Platform owner entry and innovation in complementary markets: Evidence from Intel’, Journal of Economics & Management Strategy, 16 (1), 1-34.
Gitelman, L. (ed.) (2013), ‘Raw Data’ is an Oxymoron, Cambridge, MA: MIT Press.
Gomez-Barroso, J.L. and C. Feijoo (2013), ‘Information personal: la nueva moneda de la economia digital’ [Personal information: the new currency of the digital economy], El Profesional de la Informacion, 22 (4), 290-97.
Grossklags, J., S. Hall and A. Acquisti (2007), ‘When 25 cents is too much: An experiment on willingness-to- sell and willingness-to-protect personal information’, in Proceedings of Sixth Workshop on the Economics of Information Security (WEIS2007), Pittsburgh, PA, 7-8 June.
Hemerly, J. (2013), ‘Public policy considerations for data-driven innovation’, Computer, 46 (6), 25-31.
Hey, T., S. Tansley and K. Tolle (eds) (2009), Fourth Paradigm. Data Intensive Science Discovery, Redmond, WA: Microsoft Research, accessed 22 January 2015 at http://research.microsoft.com/en-us/collaboration/ fourthparadigm/.
Hidding, G., J. Williams and J. Svlokla (2011), ‘How platforms leaders win’, Journal of Business Strategy, 32 (2), 29-37.
Jacobs, A. (2009), ‘The pathologies of big data’, Communications of the ACM, 52 (8), 36-44.
Jentzsch, N., S. Preibusch and A. Harasser (2012), ‘Study on monetising privacy. An economic model for pricing personal information’, European Union Agency for Network and Information Security, 28 February, accessed 22 January 2015 at http://www.enisa.europa.eu/activities/identity-and-trust/library/deliverables/ monetising-privacy.
Kelly, K. (2014), ‘Big data vendor revenue and market forecast 2013-2017’, Wikibon, 12 February, accessed 23 January 2015 at http://wikibon.org/wiki/v/Big_Data_Vendor_Revenue_and_Market_Forecast_2013-2017.
Lohr, S. (2012), ‘The age of big data’, The New York Times, 11 February, accessed 23 January 2015 at http:// www.nytimes.com/2012/02/12/sunday-review/big-datas-impact-in-the-world.html?pagewanted=all/&_r=0.
Manyika, J.M. Chui and B. Brown et al. (2011), ‘Big data: The next frontier for innovation, competition, and productivity’, McKinsey.com, May, accessed 23 January 2015 at http://www.mckinsey.com/insights/ business_technology/big_data_the_next_frontier_for_innovation.
Marton, A., M. Avital and T.B. Jensen (2013), ‘Reframing open big data’, paper presented at the 21st European Conference on Information Systems, Utrecht, 5-8 June 2013.
Mayer-Schonberger, V. and K. Cukier (2013), Big Data: A Revolution That Will Transform How We Live, Work, and Think, New York: Houghton, Mifflin Harcourt.
McAfee, A. and E. Brynjolfsson (2012), ‘Big data: The management revolution’, Harvard Business Review, 90 (10), 3-9.
Mulligan, M. and D. Card (2014), Sizing the EU App Economy, European Commission Digital Agenda for Europe, accessed 23 January 2015 at https://ec.europa.eu/digital-agenda/en/news/sizing-eu-app-economy.
Murray-Rust, P. (2008), ‘Open data in science’, Serials Review, 34 (1), 52-64.
Nissenbaum, H. (2010), Privacy in Context. Technology, Policy and the Integrity of Social Life, Stanford, CA: Stanford University Press.
OECD (2013), ‘Exploring data-driven innovation as a new source of growth: Mapping the policy issues raised by big data’, OECD Digital Economy Papers, No. 222, OECD Publishing, accessed 23 January 2015 at http:// dx.doi.org/10.1787/5k47zw3fcp43-en.
OECD (2014), ‘Unleashing the power of big data for Alzheimer’s disease and dementia research: Main points of the OECD expert consultation on unlocking global collaboration to accelerate innovation for Alzheimer’s disease and dementia’, OECD Digital Economy Papers, No. 233, OECD Publishing, accessed 23 January 2015 at http://dx.doi.org/10.1787/5jz73kvmvbwb-en.
Potoglou, D., S. Patil, C. Gijon-Tascon, J. Palacios and C. Feijoo (2013), ‘The value of personal information online: Results from three stated preference discrete choice experiments in the UK’, in Proceedings of the 21st European Conference on Information Systems (ECIS), Utrecht, accessed 23 January 2015 at http://www.staff. science.uu.nl/~Vlaan107/ecis/files/ECIS2013-0899-paper.pdf.
Varian, H. (2009), ‘Hal Varian on how the web challenges managers’, McKinsey & Company Insights and Publications, January 2009, accessed 6 January 2015 at http://www.mckinsey.com/insights/innovation/ hal_varian_on_how_the_web_challenges_managers.
World Economic Forum (2012), Big Data, Big Impact: New Possibilities for International Development, Geneva, accessed 23 January 2015 at http://www3.weforum.org/docs/WEF_TC_MFS_BigDataBigImpact_ Briefing_2012.pdf.
Johannes M. Bauer and Michael Latzer - 9780857939845
Downloaded from Elgar Online at 04/05/2017 01:09:57AM via Hong kong University of Science and Technology