29.11.2017 | Global Policy Watch

Data is the new gold

Development players mine a new seam

Global Policy Watch Briefing #19

by Barbara Adams and Karen Judd

Download this briefing (pdf version)

“Data is the new Gold” headlined a 2014 article in the business press on the marketing power it offers. “Each click, like, and share creates new data in the world, much of which can be used to deliver relevant marketing information and bring increased value to consumer audiences.” Picking up on the potential of so-called Big Data to measure national and global progress on development goals agreed in the 2030 Agenda for Sustainable Development,  the 2030 agenda has driven a variety of new initiatives, bringing together a vast array of global corporations, foundations, and CSOs ready to mine this new seam.

Three of these new data initiatives are the Global Partnership for Sustainable Development Data (GPSDD), Data 2X and the Digital Impact Alliance, all of which are housed at the UN Foundation and which therefore claim only to advance UN goals and priorities, not the UN itself. Most of them are financed by a few major donors, public and private.

Perhaps the most high-profile of these partnerships is the GPSDD, launched in September 2015. Despite the recommendation of the Independent Experts Advisory Group (IEAG) appointed by then Secretary General Ban Ki-moon that the global partnership be UN-led, it was decided to go outside. However, rather than create an entirely new entity, which would be time consuming and involve decisions by Member States, it was agreed that it would be hosted by an existing entity, and in November 2015 UNF was selected to serve as “institutional home for the secretariat”. With its honorary chair Deputy Secretary-General, Amina Mohammed, GPSSD now includes a growing list of   ‘champions’ all working to show governments how their activities – and in many cases their products – can enable  national statistical offices to collect and analyse the data needed to measure achievements towards the SDGs.

Data Champions include a wide variety of governments, corporations, civil society organizations, UN and other international organizations, academic institutions, foundations as well as official statistics and data communities. Being a champion has many meanings and involves diverse relationships, such as advocates and conveners, funders and providers–along with recipients– of various forms of expertise. These include data giants such as MasterCard, IBM, and Facebook, civil society organizations like Civicus, an array of UN development entities, Global Pulse-the Secretary-General’s innovative initiative on big data, the IMF and World Bank Group and a select group of developing countries, such as Colombia, Sierra Leone and Tanzania.

A separate entity but active member of the GPSDD is Data2X, a global gender data collaborative  led by the UN Foundation with support from the William and Flora Hewlett Foundation and the Bill & Melinda Gates Foundation and, according to the UN Foundation, ongoing collaboration with the Office of Hillary Clinton”.

As a data champion in the GPSDD, DATA2X commits to “bridge the expertise and track record of Data2X and leverage the convening power of the Global Partnership to ensure improved gender data is at the heart of our efforts to drive the data revolution for sustainable development.” It states: “Our work will have a particular focus on private sector engagement and innovations for data collection, analysis, and use to fill persistent gender data gaps”….

The Digital Impact Alliance (DIAL) funded by its “Founding Partners” UNF, Bill & Melinda Gates Foundation, the Government of Sweden and USAID, was set up to “to bring the public and private sectors together to realize an inclusive digital society that connects everyone to life enhancing and life-enabling technology.” Staffed by a team of tech researchers, developers, investors, negotiators, and policymakers it claims that “by channeling resources through a neutral entity such as the UN Foundation, and working collaboratively across partners, geographies and verticals, [it] can address the key bottlenecks preventing platforms and services and data for development from scaling”. In June 2017 DIAL announced its own partnership with the World Bank to support its “identification for development (ID4D) efforts”, a World Bank programme intended to help countries increase the number of people with official identification and the development impact of the overall identification system.

Open and accessible data, but who owns it?

The multi-stakeholder data initiatives claim that they make the data they produce from a host of new sources – like cell phones, satellite imagery, bank accounts— totally open and available to National Statistics Offices. In countries that lack the capacity to measure progress through expensive and time-consuming conventional means, such as surveys and questionnaires, the partnership approach is clearly attractive.

To be sure, there could be much to gain for NSOs, struggling to find a way to measure progress on 17 targets and upwards of 340 indicators in the 2030 Agenda. Indicators that gauge public perceptions for example can much more efficiently be measured by professional polling firms, such as Gallup, while indicators of financial inclusion are more efficiently measured by cell phone deposits or credit card subscriptions. But different stakeholders have different rights and responsibilities. Governments are not just one of many stakeholders in these global partnerships but have responsibilities to implement policy, informed by statistics, to achieve the SDGs and will be held accountable for their commitments and performance.

The extent to which governments have enduring rights to analyse data generated through these initiatives largely is influenced by how skillfully they are able to negotiate the terms of the partnership, including the length of its duration.  Licensing of phone companies and regulation of credit card services, for example, is done by government. GPSDD, Data 2X and DIAL serve as platforms to put governments and corporations together and they negotiate a deal – which means that NSOs can offer some useful analysis to companies which in turn continue to give them data access. But in reality the terms of such contracts are also subject to trade and investment rules and intellectual property rights, while Freedom of Information obligations apply only to the public sector, not the private. Additionally, the length of commitment ultimately depends on success of these companies in realizing return on investment – for example, on cell phone users continuing to periodically upgrade and continue to be able to pay for service. Should its business no longer be viable presumably the firm would pull out.

While the global data initiatives all include public and private participants, the primary focus of the Open Data movement has been on the extent to which governments make their statistical data open and available to their citizens – not on whether corporate data gatherers/providers make data open and available to other users.  “Open data exists in a wide variety of fields and domains,” a background paper posted on the Open Data Watch website explains (p.3), adding that the bulk of big data is produced by governments, scientists and corporations. While acknowledging the importance of scientific and corporate data which, like crowdsourced data, “are also often mashed up with open government data”, their own interest is in explaining why open government data matters to developing economies”

“It is now widely recognized that data as a new kind of asset or knowledge is a form of wealth.” (p.7)

Let the Buyer Beware

These concerns and inconsistencies are at play in a new collaboration between GPSDD and the World Bank’s Development Data Group (DECDG) on data innovation in developing countries, Collaborative Data Innovations for Sustainable Development, supported by the Bank’s Trust Fund for Statistical Capacity Building (TFSCB).

Through this innovation window, the DECDG and the GPSDD have issued a call for proposals – with a priority for funding work in low- and lower-middle-income countries.  The awards are made in the form of vendor contracts (not grants) and are subject to compliance with the World Bank Group General Terms and Conditions for Consulting Services. These terms and conditions between the awardee (contractor) and the World Bank Group (purchaser) include the following copyright provisions:

21.01 The deliverable report(s) and other creative work developed by the Contractor specifically and exclusively for the Purchaser as provided under the Contract, including all written, graphic, audio, visual and any other materials, contributions, applicable work product and production elements contained therein whether on paper, disk, tape, digital file or any other media, (the “Deliverable Service(s)”) are considered work made for hire in accordance with the copyright laws of the United States. Purchaser is the sole proprietor of the Deliverable Service(s) from the time of their creation and owns all rights, titles and interests therein throughout the world including, without limitation, the copyright and all related rights.

21.03 Under no circumstances shall Contractor use, disclose, reproduce, publish, distribute or display copies to the public, modify, or prepare deliverables produced as a result of or in connection with, the Deliverable Service(s) including derivative works, in whole or in part, without Purchaser’s prior written consent. …

21.04 All right, title and interest (including, without limitation, rights in patents, trademarks, copyright, and related rights) in Contractor’s pre-existing proprietary intellectual property included in the Deliverable Service(s) (the “Pre-Existing Intellectual Property”) shall remain with the Contractor. The Contractor hereby grants the Purchaser and the World Bank Group an irrevocable, royalty-free, worldwide license to use, disclose, reproduce, publish, distribute or display copies to the public, or modify or prepare derivative works of such Pre- Existing Intellectual Property, in whole or in part, without the prior written consent of the Contractor.

The new pilot window, introduced in late 2016, approved 15 projects all of which are either regional in scope or, while being conducted at country level, aim at a wider applicability of results.

Among these projects, is one headed by a UN entity – ESCWA. The Activity [5] is to Advance Civil Registration and Vital Statistics Systems in the Service of Syrian Refugees. Its collaborating organizations include WHO, UNICEF and a Norwegian research foundation with technical support from the Catholic University of Louvain in Belgium.

What are implications of a UN entity entering into a research partnership in which its results will be wholly owned by the World Bank under US Copyright law? Who is held accountable for increasing civil registrations?

How to disentangle the blended or obscured governance and reporting responsibilities that result from bilateral agreements between the secretariat of partnerships (UNF) of over 100 data champions and the World Bank?

Where does the responsibility rest for the due diligence needed to protect the mission as well as the credibility of the United Nations, both as a participant and endorser of the GPSDD?

Going beyond access to the question of re-use rights the situation gets more complicated. The World Bank for example has its own internal requirements regarding the openness of its data, but does not entirely follow them: countries and UN agencies that partner with the Bank on reports on different goals – – do not always have right to publish the results of this work, as the World Bank may retain those rights.

Who is accountable for SDG progress?

Similar questions arise with regard to other data collection partnerships involving UN agencies.  ILO, IOM and FAO, for example, have teamed up with Gallup –a US-based opinion polling firm—at times in association with the World Bank. An ILO-Gallup partnership examined how women feel about workplace opportunities while an IOM-International Data Analysis Centre report produced in collaboration with the Gallup World Poll examines public attitudes towards migration. Ownership rights to the results vary. In the case of the ILO-Gallup report on women and work the copyright is owned jointly by ILO and Gallup, while the IOM-Gallup apparently has no copyright restrictions.

More problematic however, is the FAO- Gallup partnership, which involves two SDG indicators that will be used for monitoring and accountability of Member States in implementing the SDGs. A Gallup press release states:

Gallup, on behalf of its clients at the World Bank and the U.N.’s Food and Agriculture Organization (FAO), collects data through its World Poll on the topics of financial inclusion and food insecurity that the two organizations use to inform SDG indicators 8.10.2 and 2.1.2, respectively. …Through the “Voices of the Hungry,” a collaborative project between Gallup and the FAO, Gallup supports FAO’s monitoring of global progress toward the goal to “end hunger (and) achieve food security.” [http://news.gallup.com/opinion/gallup/193892/gallup-helps-track-progress-hunger-financial-inclusion.aspx]

Although Gallup makes clear that “national institutions” can use the project’s data base, called the Food Insecurity Experience Scale, ultimately the Gallup organization owns the data. However, the official data used to report on progress on the SDGs is meant to be owned by NSOs, a point frequently repeated by all players, from the Statistical Commission to UN agencies, from the World Bank to mobile phone companies.

A similar contradiction is likely to arise regarding a partnership launched in February 2017 between the UN Foundation and Groupe Spéciale Mobile Association (GSMA)–“Big Data for Social Good” – “which will leverage mobile operators’ big data capabilities to address humanitarian crises, including epidemics and natural disasters. The programme is being launched with 16 of the world’s leading mobile operators, which collectively account for over 2 billion connections across more than 100 countries” according to the UN Foundation.

Thus while data partnerships of all kinds have provided resources and capacity that could assist NSOs in measuring progress on the SDGs, they have apparently been subject to no common rules or guidelines, with each contract different and subject to different restrictions. With different experiences and capacities, NSOs do not have a common approach about whether or how to engage in private sector partnerships and where the responsibility rests to set and ensure related standards. How do NSOs combine being implementers as well as adjudicators?

The use or private sector firms to collect data can divert badly needed funds from the NSOs, despite the expressed concern about the need for capacity building for these offices. It also uses taxpayer money and should make the data available to the public. This concern has been recognized in Mexico, where INEGI is legally required to make all data publically available, but few countries have similar legislation.

What is the UN advocating? How are UN agencies putting in place standards and safeguards or are they avoiding this critical question? How can participants in public-private partnerships ensure public sector access to data generated by or in partnership with the commercial sector?

The same concerns apply to financial data.  An example is the Global Findex Database, operated by the World Bank. The database, collected in partnership with the Gallup World Poll and funded by the Gates Foundation, is based on interviews with individuals in over 140 countries and “provides in depth data on how individuals save, borrow, make payments and manage risks.”

Account ownership is a first step towards financial inclusion. But what really matters is whether people actually use their account – and the data are promising.  More than 65 percent of account users in developing countries report having used their accounts at least three times a month, to save, or to make or receive electronic payments directly from their account. Yet, 1.3 billion adults with an account  in developing countries pay their trash, water, and electric bills in cash, and over half a billion adults with an account in developing countries pay school fees in cash.’ [http://www.worldbank.org/en/programs/globalfindex/overview]

Although countries now seem willing to regard ‘financial inclusion’ as an unquestioned good, one which is included in the global indicator framework for the SDGs, is important that it be more than a more sophisticated packaging vehicle for the once widely-embraced microfinance/microcredit programmes, the track record of which is at best mixed. In addition, although the Findex Database is open to NSOs in every country, the data itself is not collected by these offices and so they cannot stand by the results. As with other statistics compiled by international agencies, such as the World Bank on poverty or the ILO on employment, they often challenge the figures through the UN Statistical Commission, requiring agencies to reconcile any discrepancies.

Donor dependency: the new face of tied aid?

TFSCB is a global grant-making facility administered by the World Bank’s Development Data Group “on behalf of contributing donors”—which today include only one, DfID.  As the Trust Fund’s Advisory Panel 2017 report states:

The opening of two new windows has generated new demands for TFSCB grants but presently, its funding situation must be considered precarious. …” As the sole donor DFID “can largely determine the remit and funding priorities of the TFSCB. As a case in point, the recent DFID contribution of $ 20 million was earmarked for data production in household surveys. While this purpose was certainly in alignment with World Bank concerns about the lack of respective data from low capacity countries, a situation could arise where DFID goals in this respect would not be completely in line with TFSCB funding objectives. Attached contributions by a major donor can carry the risk of changing the remit of the TFSCB from its original or core goals if this major donor is responsible for almost the entire TFSCB budget.

Furthermore, financial dependence on one donor makes the TFSCB quite vulnerable to policy changes affecting IDA in the major donor country. A change of government priorities or budget cutbacks could quickly result in a critical funding situation of the TFSCB. Generally speaking, the dependence on one major donor can negatively affect the stability, the topical flexibility and the demand orientation of the TFSCB.

Transparency and accountability

A number of different global initiatives have been agreed in an attempt to move beyond the lack of transparency of information in such areas as aid and investment as well as increase monitoring and accountability.  While again the focus in on governments, particularly developing country governments, the obligation extends in some cases to corporations as well.

A specific initiative for official development assistance is IATI, used by many governments and highlighted in proposals of the Secretary-General  “Strengthening accountability to guide the United Nations development system’s support for implementing the 2030 Agenda” [Chapter V under B. Increasing transparency on system-wide results]

“In strengthening internal accountability to deliver on collective mandates, we will work with the United Nations development system to accomplish the following initiatives:

(b) Reinforced transparency on entity-specific expenditures and results through system-wide enrollment into the International Aid Transparency Initiative, so as to ensure that States and citizens have real-time visibility into our expenditures. Entities will also build upon significant progress in results reporting systems to make the Organization’s contribution to sustainable development more visible and concrete (para: 108).

An initiative that addresses the fraught area of natural resource extraction and public and private actors is EITI, established in 2007, to strengthen the reporting systems of governments and corporations with regard to the extraction of natural resources.  Details about oil, gas and mining contracts and revenues, as well as, most recently, beneficial owners should be transparent in order to inform public debate. But participation requires laws mandating such disclosures, as done in EU and Canada for example, and until recently the US. US legislation was held up because two of its largest oil companies – Chevron and Exxon Mobil- refuse to release this information, including amounts paid in taxes, and the Trump administration simply pulled out of the initiative entirely.

An independent assessment of the impact of this initiative, while applauding the goals of the initiative, found that measuring impact is contentious and ineffective, as it is largely based on the views of different stakeholders engaged in implementation. While relevant to the global accountability agenda, promoted by SDG 17, the initiative lacks any overall metrics for measuring impact, and largely amounts to measurement of perceptions. [https://eiti.org/sites/default/files/documents/2011-EITI-evaluation-report.pdf]

The 2030 Agenda has been successful in shining light on the importance of measurement, data, the data-policy nexus and the need for national ownership, but this has also spawned market place dynamics without uniform standards and processes to determine what is a genuine—and lasting– contribution to these goals and what is just marketing and positioning for funding.

SDGs how you measure matters

by Roberto Bissio

Poverty eradication (“leave no one behind”) is at the heart of the 2030 Agenda and, in that regard, it requires to “enhance capacity-building support to developing countries” in order ” to increase significantly the availability of high-quality, timely and reliable data disaggregated by income, gender, age, race, ethnicity, migratory status, disability, geographic location and other characteristics relevant in national contexts” (SDG 17.18). The majority of the world’s poor are rural or live in places of difficult access. They are usually excluded from opinion polls and at the edge or outside the reach of communication technologies. While the private sector may not have much data to contribute to the work of the census takers or household surveyors of NSOs, they have all to win from access to the data generated by statistical offices in planning their market outreach.

On the other end, the very rich are usually left out of household surveys, as they are unlikely to be randomly selected to be interviewed or to answer the pollster ringing their doors if they do. This is why, for example, inequality indexes of wealth based on surveys tend to report less disparity then those based on tax declarations (as underreported as those might be). The private banks and credit companies could provide valuable information here, but are unlikely to do so.

The potential conflicts of interests and different definitions of privacy are already leading to major conflicts between European regulators and big data corporations. Which privacy and confidentiality rules will prevail when developing country NSOs and big data holders come together in a partnership?