Guidance note 19 - Open data policies and disclosures
Guidance note 19 - Requirement 7.2
Open data can help make governments more transparent and accountable. It can provide the evidence that public money is being managed well and policies are being implemented. Countries that implement the EITI Standard are required to publish data in open formats to enable a broader use and analysis of the information. To disclose information in open data format means that companies and governments disclose information in a structured manner, that anyone can access, use, and share at any time, for free. This data can be published on reporting entity pages and on cross-governmental platforms, such as open data portals.
A lot of figures, covering topics such as revenues, production, exports, etc, are published as a result of EITI implementation. Users will want to draw on that data to respond to their questions or build their understanding of their sector. Disclosing data in open formats ensure users have the best access to that information. It also means information can be linked with other relevant sources. Publicly available open data is a stepping-stone towards mainstreamed EITI implementation or systematic disclosure of information.
This note provides step-by-step guidance on developing open data policies. It paves the way for creating and publishing open data. It also contains boxes explaining key open data concepts as these are encountered during the reading. The annexes include examples of how governments and companies are publishing data in an open format in EITI implementing countries and provide references for further reading.
Open data is “[…] data and content [that] can be freely used […] and shared by anyone for any purpose [for free].” Examples of open data files include Excel files and comma-separated-value (CSV) files. PDF files are not considered open data, as they are not machine-readable.
More than available, open data is accessible:
1. Legal openness: Ensuring users can legally access the data, build on it and share it. Open data must be licensed. Its licence must permit people to use the data in any way they want, including transforming, combining and sharing it with others, even commercially.
2. Technical openness: Ensuring no technical barriers to data use exist. For example, it is not possible to access underlying data of tables and charts contained in PDF reports. These are only accessible if published alongside Excel files or other open data files containing the underlying data.
The EITI recognises that national and international legislation must be observed, with regards to intellectual property and sensitive information.
Requirement 7.2 Data accessibility and open data
Implementing countries should ensure that EITI disclosures are made publicly accessible. The multi-stakeholder group should:
Source: EITI Standard 2019, https://eiti.org/document/standard
In the EITI the terms ‘systematic disclosure’ and ‘mainstreaming’ are used interchangeably. They refer to the desired end‑state, where the EITI’s disclosure requirements are met through routine and publicly available company and government reporting. This could include public financial reporting, annual reports, information portals and other open data initiatives. Systematic disclosure is the default expectation, with EITI Reports used to provide additional context, collating sources where systematic disclosures can be found, and address any gaps and concerns about data quality. EITI disclosure requirements can be met by referencing publicly available information and/or data.
Less resource-intense data collection will also allow EITI implementation to be more effective and focus more on use and the analysis of data, and support of future policy-development in the sector. However, an emerging risk is that “EITI open data” portals are being developed independently of core administrative information management systems, leading to duplication and decreasing focus on systems as the primary source and guarantor of reliable data. Open data policies in EITI countries have often focused on how to make data in EITI Reports more accessible, while it is often more effective to review the open data policies and practices of the companies and government agencies that are the primary sources of data. Improving reporting entities’ open data policies is likely to increase lasting public access to data.
This section suggests steps that multi-stakeholder groups (MSGs) could consider when developing an open data policy, drawing on examples from EITI implementing countries.
An open data policy should ideally state:whether countries and reporting entities provide information under open licenses (OL), clarifying usability and release;
which level and degree of openness will be used (RE versus OF); and
which degree of standardisation and availability is sought (RDF versus LOD).
It is important that open data policies, where possible, address the comprehensiveness, timeliness, and retention of historical data covered by EITI reporting. Data retention means the continued storage of historical information alongside new data disclosures.
Degrees of openness
Even if EITI data published in Reports have entered the public domain, the accessability of data from reports remains limited, as data contained in large reports are “locked” in the file.
There are five different levels of openness, from “PDF” (not open), to Excel (minimum openness) to “LOD”, each tied to different formats or types of file. These are set out on the scale developed by 5-Star Open Data, as presented below (see also Annex A).
Before considering options for an open data policy, MSGs are encouraged to assess the current situation regarding the access, release, and re-use of the data required by the EITI Standard. MSGs have often found it helpful to consult typical users of EITI data on their needs and expectations.
The first aspect to consider is which information is routinely available in open data formats through government and corporate reporting systems. The second is to address the accessibility of additional data that is created through the EITI process.
The first aspect is linked to the concept of systematic disclosures as highlighted above. MSGs are strongly encouraged to map existing disclosures or undertake a feasibility study to inform this work. The move towards systematic disclosure means that increasingly, MSGs and national secretariats will need to verify whether disclosures are in open data format at source..
There is “no need to reinvent the wheel”. The MSG should examine national policies and standards on open government and open data, and their alignment with international best practices. This helps ensure that the MSG’s work reinforces existing efforts, not duplicating existing ones.
Relevant national policies may include relevant constitutional provisions, government legislation or policies on open government and open data, related legislation such as Freedom of Information Acts (FOIAs), and commitments made through other initiatives, such as the Open Government Partnership (OGP) and the Joint Organisations Data Initiative (JODI). These provide well-established platforms for pursuing open data policies and disclosures. The OGP’s Open Government Declaration, for example, explicitly references a commitment to pro-active disclosure by government, and also specifies that disclosure of information is to be conducted in open data or machine-readable formats (see Figure 1 below).
Figure 1: Screenshot of the Open Government Declaration
With regards to the second aspect, when EITI is the primary disclosure mechanism, it is necessary to consider whether information collected through EITI reporting is sufficiently accessible and develop a policy to improve on public access. At the national level, a survey by the EITI International Secretariat noted that most EITI data continues to be “locked” in pdf reports that are difficult to use. At the international level, since the introduction of EITI Summary data, open data coverage has steadiliy improved and reached a coverage of 90% in 2019. These are available through the International Secretariat’s API or website.
Addressing regularity, timeliness, and methods of release are all important aspects of open data policies, and could e.g. include objectives for timeliness by setting minimum deadlines for the release of data. Annex B includes an extensive list of open data examples, policies and tools.
MSGs can draw from existing practices, such as the guidelines for open data of the Filipino government (see Figure 1). The German EITI and former USEITI provided the building blocks of their websites, the source-code, open for reuse on GitHub., The EITI International Secretariat also provides the source-code used for EITI.org. GitHub allows whoever is interested to look into the source-code providing the framework for EITI’s data portal, and to use the code if they wish.
Figure 1: Guidelines on Open Data Implementation
The open data policy should clarify the procedures for the release of data, including provisions on the terms of data use. The table below shows the different types of open licenses commonly associated with open data. It shows the various sharing-levels associated with different licenses.
Table: Open licenses
Sharing-level of Licence
Creative Commons License
Open Data Commons License
Attribution & Share-Alike
Public domain licenses mean that the data in question is free of any copyright and the publisher(s) waive any rights towards the data. In this case, the users are free to: (i) copy and distribute the data; (ii) produce new works using the data; and (iii) modify, adapt and build upon the data.
Users can, in this instance opt out of copyright and data protection-licenses if they choose to re-use or re-publish the data. This means users are not obligated to cite the source of the data. Attribution licenses, includes all the characteristics of public domain-licenses, with the only additional restriction that the source of the data must be cited, and that any notices accompanying the publication must be kept intact. Lastly, attribution & share-alike licenses include the same rights and restrictions as the two preceeding types, but new works must be published under the same license as the source, and can only be published in ‘locked’ formats if a copy is published alongside in an open format. An alternative to these standard open licenses is to create a custom license, like the United Kingdom.
The above alternatives should, alongside assessments of current national priorities and policies, be brought to the attention of and discussed by the MSG for subsequent decision. Open licenses do not require registration in order to be used, only a statement on the website or publications, including the name of the license, with a link to the relevant explanation. An example is provided below from the Creative Commons’ website. 
MSGs should agree on the most effective way to publish data based on the profile, resources, and technology of its data users. When feasible, MSGs may opt to craft policies that are specific to certain types of users, e.g. media, academia, parliament, local communities, etc. The USEITI, for example, undertook extensive consultations with likely users of EITI data as a key first step in developing their EITI data portal. While the United States no longer implements the EITI, the platform is still being used by the Department of the Interior.
Once the MSG has determined the users needs and capacity, it may chose which degree(s) of openness would be relevant for the disclosures of the data. The EITI Standard requires publication of EITI data in Excel format at minimum. EITI data refers to all tables, charts and figures associated with EITI reporting. Submission of summary data according to the Board-approved template is also required, increasingly making use of data standardisation.
Further, the EITI Standard expects that government agencies and companies to publish EITI data under open licenses, and encourages national secretariats and MSGs to systematically disclose machine-readable and interoperable data. By using the 5-star figure introduced previously, the requirements of the Standard can be identified as below (for descriptions of the various levels, see Annex A):
Demand driven EITI data in the United States
"If we focused just on the requirements, and not what the users wanted to see from it there would be a disconnect. Therefore, we spent a lot of time with users and included them in the design process. This helped us to answer the questions that people actually wanted to know."
The USEITI undertook extensive stakeholder consultation before developing their EITI data portal. This enabled USEITI to identify the data-points and types of visualisations that were in the highest demand. They also considered definitions and explanations of terms that would be useful for less experienced users.
This is commonly referred to as user-, or demand-driven data. It focusses on identifying likely user groups and mapping their needs. USEITI also allows for continuous feedback.
This led to a broader push for using the website as a source for USEITI’s reporting. As a result, the EITI Report was a short document including information of methodology and key facts, and outcomes of the reconciliation process. The detailed data was available through a portal.
For more information or to visit the US Department of the Interior’s data portal, visit https://revenuedata.doi.gov/explore/.
Using data standards means that data is published using a standard methodology. For numbers, this means a consistent use of calculations or categorisation.
To categorise revenue streams, a data standard has been developed in collaboration with the International Monetary Fund (IMF). This classification uses the Government Finance Statistics (GFS) Manual 2014, which is already used by most finance ministries for international reporting. Guidance on how to use the framework is available as Technical notes, which have been drafted by the IMF in consultation with the EITI International Secretariat.
For text-based data, other forms of standardisation may be needed, like codes: for country names, the EITI International Secretariat uses ISO 3166, a widely used country codes.
For more details on how the EITI International Secretariat uses data standards for summary data, please see the links and descritions provided via https://eiti.org/data
The degree of openness of data largely hinges on its interoperability with other information or datasets. EITI Requirements require open data disclosures in the form of Excel or csv-files, meeting the minimum of open data definitions. Using the five-star framework of Annex A, RDF- and LOD-files are examples of more flexible open data formats, in which data-points may be embedded in websites and are linkable to other datasets. More flexible options enable users of data to constantly have the most updated data available.
MSGs and national secretariats should therefore ensure they are aware of the relevant data standards used by various reporting entities in the country, such as national budget codes and classficiations, economic activity codes for companies and sectors, HS Commodity Codes for production and export data, System of National Accounts for macroeconomic statistics, Global Legal Entity Identifiers for government and corporate entities, as well as other standards. MSGs are free to explore different standards than described in this guidance. Additionally, although the leap towards RDF or LOD-files may not be immediately realistic, MSGs should keep these in mind as they greatly improve data access and reduce time spent on data collection. Such data formats are highly relevant for the future of mainstreaming EITI disclosures, and the EITI Board expects that MSGs develop workplans for systematically disclosing EITI data within a five-year period.
According to the steps outlined above, MSGs should agree a written statement outlining their policy on access, release and re-use of EITI data. In the event of adopting a wider national policy, an MSG statement of endorsement may be used. Examples of existing policies are available here. Below is model language drawn from the EITI open data policy which can guide MSGs in the drafting of their own policies:
“The multi-stakeholder group of [country] commits to publication of information that is provided in accessible formats such as Excel and CSV files [or more sophisticated solutions], to be published through [identify platforms] on a [semi-annual/quarterly] basis, which may be freely used or re-used according to [identify license].”
The MSG may also wish to consider additional points: “The multi-stakeholder group commits to …”
“… systematic publication of open data by embedding open data policies and strategies in reporting entities involved in EITI reporting to ensure timely and quality data, accessibility and cost effectiveness of data delivery.”
“… working with users to identify priority data sets and the form of data delivery.”
“… consider user needs and access of different gender, ethnic and geographic users.”
“… ensure that data are provided in granular, machine-readable formats and fully described, so users have sufficient information of strengths, weaknesses, analytical limitations and security requirements, as well as how to process the data.”
“… release data as early as possible, allowing users to provide feedback, and continue to make revisions to ensure the highest standards of open data quality.”
“… share technical expertise and experience with other countries to maximise the potential of open data in a socially inclusive manner.”
“… work to increase open data literacy and encourage people, such as developers of applications and civil society organisations, to unlock the value of open data.”
“… ensure that data is interoperable with national and international standards, including adopting data standards such as [list of data standards to be used in publication] and additional guidance provided by the EITI International Secretariat.”
“… where possible support the cross-referencing of data with other datasets by using unique, persistent and public identifiers for commercial and government entities.”
“… consider the technical infrastructure to deliver and use the open data.”
“… consider the governance and sustainability of open data policies as to ensure that reporting entities have a data steward, data is retained, and security standards are in place.”
Once plans and policies are in place, a data champion and/or data coordinator should be appointed with the mandate to review and update open data disclosures regularly (quarterly or semi-annually).
Step 1 – Accessibility of data required by the EITI
Step 2 – Legal openness
Step 3 – Degree of openness
Step 4 – Drafting the open data policy
Monitoring and implementation
This section provides guidance on what MSGs could consider when implementing their open data policies drawing on examples from EITI countries.
EITI implementing countries publish tables, charts and figures (“data”) as well as narrative text on countries’ extractive industries. The 2019 EITI Standard emphasises that data should be publicly accessible, and expects that countries take steps to integrate publication of open data in company and government systems. The sections below describe three levels of sophistication in publishing extractive industries data, ranging from the most basic – Excel and standardised data files – to automated and mainstreamed examples.
The EITI Standard already requires the MSGs to:
Make the data available in an open data format online and publicise its availability. Open data format means that data can be made accessible in CSV or Excel format and could contain all tables, charts and figures from EITI Reports. (Requirement 7.2.b)
Complete summary data files for each fiscal year covered by the EITI in accordance with the template approved by the EITI Board. (Requirement 7.2.c)
In cases where the MSG hires an Independent Administrator to undertake parts of the reporting process, phase 5.3 and 5.4 of the “Standard Terms of Reference for Independent Administrators” requires the creation and delivery of the same open data products.
For the Dominican Republic, the EITI website is the EITI Report: Instead of producing large physical copies of EITI Reports, all the necessary disclosures are made online as the default, while a reconciliation report is published as an attachment.
The 2019 EITI Standard requires production of machine-readable files that can be published alongside any print/pdf version of EITI reporting. In practice this means that the MSG must ensure that CSV or Microsoft Excel files are created, containing all tables and charts used in EITI reporting. National secretariats and MSGs should ensure that these files are made available for download through their own websites, alongside any non-open data publications. In practice, this could be as simple as publishing tables from reports in an excel file. See annex B for specific examples.
The next stage of data publication aims to ensure that data is interoperable and standardised. As described in Part I, it is important to consider data standards as one develop reporting templates for publication, to ensure that a standard methodology is adopted already at the data creation stage. This helps improve data relevance and use, as data can be combined with other datasets using similar or related standards.
As an example, the EITI recently launched a Board-approved Summary data template, a tool to publish interoperable and standardised EITI data. National secretariats and MSGs are required to submit one file for each fiscal year to the International Secretariat alongside EITI reports.The template can be found at https://eiti.org/summary-data-template, and is used to generate various sets of standardised data. As an example, below is a figure presenting the revenues of the Democratic Republic of Congo, using the GFS classification of national revenue streams.
Figure 2: Revenue profile of Democratic Republic of Congo using GFS classification
Source: EITI Total revenues by GFS, using Summary data. Available at: https://eiti.org/files/190313_-_api_query_-_total_revenues_by_gfs.xlsx
Figure 3: Extractive sector contribution to the economy per capita (2015)
Source: EITI Economic contribution per capita, based on Summary data. Available at: https://eiti.org/files/191114_-_economic_contribution_percapita.xlsx
Additionally, Figure 4 combines EITI data with World Bank demographic data. It presents the contribution of the extractive sector per citizen in government revenues, exports and GDP..
Summary data files are used to facilitate analysis by extractive data users and function as an international reporting template for EITI’s global database. It is accessible through an API, or Application Programming Interface, using a JSON programming language. An API enables other users to link their data with realtime updates. The EITI also provides guidance for advanced users on how to access the database.
However, so far EITI’s API is not linked with other useful data sources such as the World Bank’s Databank, IMF’s datapages, nor UN Statistics Database directly. Therefore, the EITI has not yet reached the final level of openness described in Annex A; linked open data.
The 2019 EITI Standard includes several provisions that encourage and enable mainstreaming, including:
The broad definition of “EITI Report” in the 2019 Standard: “The data [required by the EITI] can be disclosed in the form of an EITI Report, or constitute publicly available information and data gathered or cross-referenced as part of the EITI process.”
The default expectation is that government and corporate entities systematically disclose data in open formats, and several provisions explicitly encourage routine disclosure of data, e.g., Requirement 4.1 on Comprehensive financial reporting: “The expectation is that implementing countries will disclose the requisite information through routine government and corporate reporting (websites, annual reports, etc.), with EITI Reports used to collate this information and address any concerns about gaps and data quality.”
While most implementing countries have embraced some aspects of mainstreaming, especially with respect to contextual information relating to licensing, contract transparency, production and export data, many countries face challenges with “full systematic disclosure”, i.e., routine disclosure of all the data required by the EITI Standard in requisite detail. A common challenge in the Latin American and Caribbean region, for example, lies in confidentiality or privacy provisions that require a company waiver before government agencies can disclose company-specific data. These challenges may necessitate legal review or reforms. In Indonesia, there are numerous government platforms. Each agency seems to have at least one outwards facing data portal for public access to information, such as the Minerba One Data (MODI)  of the extractice sector regulator versus their internal reporting system Minerba Online Monitoring System (MOMS). While these two systems might speak to oneanother, the priority is to ensure the databases are linked to other government agencies’ systems, and Indonesia EITI’s own portal..
All implementing countries are expected to develop a costed work plan for 2019 and onwards that includes steps to mainstream EITI implementation in company and government systems within 3-5 years. The MSG is encouraged to ensure that systematic disclosures adhere to the open data policy of the MSG, or wider government policies. The MSG should also ensure that systematic disclosures adhere to the publication of the required data and at the required level of disaggregation. For examples of systematically disclosed data, see Annex B or check https://eiti.org/systematic-disclosure.
 EITI (2017), ‘Systematic disclosures’. https://eiti.org/systematic-disclosure
 EITI (2017), ‘Systematic disclosure toolkit’. https://eiti.org/document/systematic-disclosure-toolkit
 EITI (2018), ‘Terms of Reference: EITI systematic disclosure feasibility study’. https://eiti.org/document/terms-of-reference-eiti-systematic-disclosure-feasibility-study
 See further guidance on systematic disclosures: https://eiti.org/systematic-disclosure
 Open Government Partnership (2011), ‘Open Government Declaration’. http://www.opengovpartnership.org/about/open-government-declaration
 EITI (2017), ‘Survey on open data in EITI implementing countries’. https://eiti.org/blog/survey-on-open-data-in-eiti-implementing-countries
 Republic of the Philippines (2015), ‘Guidelines on Open Data Implementation: JMC no. 2015-01’. https://www.gov.ph/data/story/guidelines-open-data-implementation-jmc-no-2015-01
 United Kingdom, ‘Open Government Licence for public sector information’. https://www.nationalarchives.gov.uk/doc/open-government-licence/version/3/
 United Nations (2017), ‘Harmonized Commodity Description and Coding Systems (HS)’. https://unstats.un.org/unsd/tradekb/Knowledgebase/50018/Harmonized-Commodity-Description-and-Coding-Systems-HS
 United Nations (2008), ‘The System of National Accounts’. https://unstats.un.org/unsd/nationalaccount/sna.asp
 Global Legal Identifier Foundation (n.d.), ‘ISO17442: The LEI Code Structure’. https://www.gleif.org/en/about-lei/iso-17442-the-lei-code-structure#
 EITI (2019), ‘Economic contribution of the extractive sector per capita’. https://eiti.org/files/191114_-_economic_contribution_percapita.xlsx
 World Bank (n.d.), ‘World Bank Databank’.