"Public digital infrastructure like India's stack drives inclusion better than private platforms"

"101 Interoperability of data and systems can be supported by adopting harmonized standards— ideally, open standards. Open standards are often determined by sectoral or international standard setting organizations (SSOs) in order to support the interoperability of data and systems within a partic- ular market or sector. They are therefore designed collaboratively based on user needs.102 Public intent data should also be published under an open license and at no charge or at a marginal price to cover the costs of dissemination or reproduc- tion.103 Nearly 48 percent of the surveyed countries have adopted some form of open licensing regime for public intent data. All the high-income countries covered in the survey have done so, compared with about 40 percent of middle-income countries. Other countries, such as Jordan and Mauritius,104 have adopted Creative Commons Attribution 4.0 Inter national Licenses for government datasets released as open data. In Jordan, datasets published by the government are open to all and licensed under a Jordanian Open Government Data License, which allows the use, reuse, and sharing of data, in compat- ibility with the Creative Commons (CC-BY) license.105 To ensure that data prioritized for publication meet the needs of nongovernmental actors in the private Data policies, laws, and regulations: Creating a trust environment | 203 sector and civil society, these decisions should be guided by consultations with multiple stakeholders (see chapter 8). Enabling access to and the seamless transfer of public sector data between different entities within the public sector and with end users (including individuals and businesses) requires ensuring the interoperability of information technology (IT) systems (including platforms) and data (syntactic and semantic interoperability). As defined by ISO, syntactic interoperability enables “the formats of the exchanged information [to] be understood by the participating systems,” while semantic interopera- bility enables the “meaning of the data model within the context of a subject area to be understood by the participating systems.”106 Effective data and systems interoperability requires the implementation of sev- eral technical protocols and a government interoper- ability platform. In addition to technical enablers for interoperabil- ity across the whole of government, an enabling legal and regulatory framework is often required. This framework mandates the use of the government’s interoperability platform and data exchange proto- cols, ensuring that all government entities connect to and use the platform as a vehicle for exchanging data. Very few countries surveyed have adopted a full range of common technical standards (such as the FAIR principles) that enable the interoperability of systems, registries, and databases (figure 6.6). Estonia is among the few countries surveyed that has established standards for open APIs for govern- ment to government (G2G), government to business (G2B), and government to consumer (G2C) services; standardized communications protocols for access- ing metadata; and developed semantic catalogues for data and metadata. A distinct advantage of implementing interop- erability is the possibility of applying the once-only principle to access to data, which reduces the admin- istrative burden. Citizens and businesses are asked to provide their data only once, thereby requiring pub- lic sector entities to internally share and reuse data— with the appropriate safeguards—in the provision of administrative services. Because the risk of data breaches and misuse increases when data are stored in centralized or decentralized but linked reposito- ries, the once-only principle should be complemented with robust legal and technical data protection as well as cybersecurity and cybercrime safeguards, implemented in a citizen"

"That said, because APIs can expose data to unauthorized access during data transfers, they may prove challenging to use effec- tively in lower-middle-income countries that do not have sufficient technical capacity to respond to cyber- security risks. Fully leveraging APIs to enable effective interop- erability and portability requires ensuring that they are developed according to common standards. These standards can be developed through regulation or by industry, based on a multistakeholder approach. Examples of successful initiatives include the Berlin Group, which has developed common API standards for the European banking industry.145 Cases from the financial services sector (such as the UK Open Bank- ing Initiative and implementation of the European Union’s Second Payment Service Directive) may pro- vide helpful lessons for the effective implementation of these mechanisms as enablers for data reuse and sharing. Forging data partnerships. An alternative modality for private sector data sharing is through data public- private partnerships (PPPs) entered into on mutually agreed contractual terms between private sector entities or between government and businesses. For example, the traffic application Waze has partnered 206 | World Development Report 2021 with more than 1,000 cities and other local public sector entities through its Connected Citizens Pro- gram146 to exchange traffic data and derive insights to inform mobility projects, manage traffic and con- gestion, support emergency response, and share data with citizens through a cloud-based platform.147 Data partnerships pose several challenges. Part- nerships between large companies and small and medium enterprises may raise concerns about fairness because of asymmetries in information or market power. Partnerships between public and pri- vate entities may lead to conflicts of interest because of the government’s dual role as data user and data regulator.148 In either case, partnerships may create uncertainties around the allocation of the legal lia- bility associated with the use of the data, as well as potential compliance costs due to lack of harmoniza- tion of legal frameworks applicable to both parties.149 Some of these risks can be mitigated by developing contract guidelines or standard contractual terms to harmonize provisions and rectify information asymmetries. Some public sector initiatives have attempted to develop such standard terms to promote data sharing.150 Not all data sharing partnerships are designed for profit. Some businesses provide their data and digital tools at no charge to governments, academia, and nongovernmental organizations for “social good.” Data philanthropy,151 particularly in the area of big data, has enabled the World Bank,152 together with UN agencies—the World Health Organization (WHO), United Nations Development Programme (UNDP), World Food Programme (WFP), and United Nations Children’s Fund (UNICEF)—and others, to leverage companies’ data stock and digital capabilities to fully exploit the value of data for development, while bene- fiting the private sector through positive externalities. Limiting intermediary liability. One of the great enablers of the flow of data across the internet are rules limiting the liability of intermediaries for content that flows over their platforms. The intermediary liability concept has roots in US telecommunications law dat- ing back to the 1930s,153 and it has been informed by subsequent US case law.154 Crucially, this exemption from liability was extended to “interactive computer services” (internet service providers) in Section 230 of the 1996 amendments to the Communications Act of 1934155 and in the Digital Millennium Copyright Act.156 The advent of data platform business models has led to growing requests from users for the “take-down” of their personal information and has triggered an ongoing debate between privacy advocates and"