Chapter Five

Data Policy

As a complement to the section "data policies and legislation" outlined in Chapter 3, this chapter explains what a data policy is, why it is important, explores the foundation of data policies and provides a few examples taken from Africa.

What is a data policy in the context of the SDI?

To explain a data policy, one has to go back to the fact that information is a meaningful conclusion derived from the processing of some data. Information policies are common: For example (1) TV channels are not allowed to show chocking images of injured people or dead bodies while reporting on natural disasters or terrorist attacks in the US; (2) In the context of a corporate body, an information policy will define the type of management information, the communication media to use, and the rules to be followed by the staff for that information to be shared within the various departments, sections and units of the institution, and with the outside world.

A data policy is more specific. In the SDI context, it constitutes the premises of a legal framework defining some basic principles specific to data, to be observed by individuals and institutions when generating, collecting, transforming, disseminating, and making use of data. Data in this case could be geospatial data, socio-economic data, policy data, etc.

While a distinction is to be made between information policy and data policy, it should be kept in mind that, during the SDI process, a piece of information may be used as an input to a further process, and therefore become an input data for that process. Consequently, it is sometimes difficult to draw a line between data and information, and therefore to clearly distinguish information policy from data policy in some cases.

Why data policies are important in the SDI context ?

In chapter 3 it was mentioned that some institutions refuse to share their data, either because of a restriction on these particular data, or because of the absence of a policy relating to provision of data altogether.

Even in the situation where a data policy exists, it may not have been ideally elaborated, i.e. in such a way that people apply it by consent and not by constraint. The data policy should be initiated with a clear understanding of why for example, some people and institutions are reluctant to share their data.

By nature, human beings tend to be suspicious about the quality of what comes from their neighbours, and, in the opposite, have confidence in the quality of their own produce. This feeling is not limited to individuals. Institutions too, develop the same feeling with regards to the outside world. It is understandable to be cautious about every relationship or partnership that could present a risk for one own image, be it individual or institutional. This is particularly true when data exchange is involved.

From an institutional point of view, risks, virtual or real, are generally associated with the principle of sharing data :

1. as mentioned above, an a priori suspicion of the quality of third party data is common¹. This generates a cautious attitude due to the false risk of deriving questionable information from the third party data;

2. another virtual risk arises from an a priori presumption that the institutions' own data (generally deemed of high quality by the latter) may be "wrongly" used if shared with a third party, or even that ownership thereof may be lost. This ends up in a paternalistic attitude characterised by attempts to find out in advance what the data will be used for by the third party;

3. some institutions may fear that other users discover the poor quality of their data by sharing them. Such a fear is likely to dictate a protectionist attitude with regards to sharing data, in an attempt to avoid the risk of deteriorating the image of the data owner institution.

The purpose of developing an SDI being merely to provide easy access to development information through data sharing, a sound data policy should look carefully at ways to remove the potential risks summarized above, so that the data producers be happy and confident in sharing their data.

All this shows how important the data policy is in achieving full participation of all stakeholders and their involvement and commitment to the SDI process.

Foundation of data policies

An insight in the historical evolution of information and data policies shows that they stem from the basic human right to freedom of opinion and expression thereof. Data policy is enshrined in the Universal Declaration of Human Rights which forms the basis of public access to information : Article 19 thereof stipulates that "Everyone has the right to freedom of opinion and expression ; this right includes freedom to hold opinions without interference and to seek, receive, and impact information and ideas through any media and regardless of frontiers". The World Resources Institute, in the framework of a study on environmental governance, compiled a document on existing data policy sources giving an overview of information policies in general, and environmental information in particular. (WRI presentation on Environmental Governance for Equitable Natural Resources Management at the USAID/WRI's Information Working Group Meeting in April 2001 in Washington DC)

This document shows that the human right to freedom of opinion and expression, was the basis and the reference for the drafting of constitutions and environmental information policies in various countries and institutions around the world.

Another important element that favoured the development of many data policies is Chapter 40 of Agenda 21 on "Information for Decision making" Two program areas are proposed in this Chapter for implementation, to ensure that decisions are based increasingly on sound information : (1) bridging the data gap, and (2) improving information availability.

Examples of data policies in Africa

These main sources mentioned above, backbone of information/data policy initiatives worldwide, influenced a series of information policies. Following are a few examples of such policies:

· the Executive order of President Clinton on "Coordinating Geographic Data Acquisition And Access: The National Spatial Data Infrastructure" (1994)

· the Aarhus Convention on European public access to information, participation to decision making and access to justice in Environmental maters (June 1998)

· the Netherlands Government's Information Act (WOB) concerning public access to information

· the Global Terrestrial Observing System - GTOS Data and Information Plan

· the Biodiversity and Conservation Information System - BCIS Data Policy

Africa is not absent from the global « data-policy » picture. The following summarizes the best knowledge about data policy initiatives to-date in Africa :

South Africa : under the National Spatial Information Framework - NSIF initiative, a Spatial Information bill was developed (see Annex)

Nigeria : under the National Geo-spatial Data Infrastructure - NGDI project, a Nigerian National Policy on Geo-spatial Information was developed (see Annex)

Africover Under the FAO Africover Eastern Africa project, guidelines for custodianship were adopted. Data distribution and management is based on these Guidelines, whereas specific data access policies were developed in agreement with the National Focal Point Institutions (NFPI) for the different types of data sets. (see annex)

Benin In the framework of the SISEI (Internet -based Environmental monitoring and information system) project, an Information Charter was adopted to define the rights and obligations of the data providers and data users part taking in the SISEI network (see annex)

Burkina : The institutions members of the PNGIM (network of environmental data producers and users) drafted bylaws to regulate the process of data sharing within the network (see annex)

The table below shows how of few basic principles of data sharing, namely data ownership/custodianship, data discovery mechanism, and data maintenance are addressed in these examples of data policy in Africa.:

Data ownership/custodianship

South Africa

Data ownership is not explicit, but implied through clauses protecting "the copyright of the State and any other interested party". However, data custodianship is defined as follows : "a data custodian is an organ of state which is officially responsible for the capture, maintenance, management, integration, distribution or utilization of spatial information on behalf of the State and the public"

Nigeria

· "Ownership implies intellectual property right over a dataset by a body or individual."

· "A custodian is a body or person designated as having a certain right and responsibility for development and or management of spatial data. A custodian may have the right on behalf of the community to determine the condition for use, accessibility and distribution of data."

Africover

· Custodianship is seen by the AFRICOVER Steering Committee as being at the core of efficient and effective management and access to the Multipurpose Africover Database on Environmental Resources (MADE). The principle of custodianship assigns to each National Focal Point Institution certain rights and responsibilities for the management of MADE on behalf of the relevant national country and its agencies.

· A custodian is a recognized contact point for the distribution, transfer and sharing of the information and has responsibilities regarding the maintenance and quality of the information. The custodian ensure accessibility to the information, and has the right to apply market conditions provided that this does not significantly disrupt accessibility

SISEI Bénin

· Data generated using public funds are by principle of public nature and should be made accessible, except where national security or institutional interest may be affected by such public access.

· Copyright, and intellectual property rights in general, are guaranteed for all institutional members. Implicitly data custodianship is recognized as well as data ownership.

Burkina Faso

Same as Benin above

Data discovery mechanism

South Africa

· "an electronic metadata catalogue which enables users to search for and gain access to spatial information".

· "An organ of state must capture and maintain metadata for any spatial information held by it.

· Metadata must conform to the standards and prescriptions referred to in section 17.

· An organ of state must ensure that metadata are available to users by -

o including them in a manual on functions as described in section 14 of the Promotion of Access to Information Act, 2000; and either

o establishing an electronic data base containing metadata for this purpose at the electronic metadata catalogue; or

o making its metadata records available to the Department, in the prescribed manner, for inclusion in the electronic metadata catalogue"

Nigeria

· Every geospatial data producer shall provide metadata for each of its data holdings.

· Government, through the lead agency and in consultation with the NGDI Committee, shall establish electronic geospatial metadata catalogue and Clearinghouses in NGDI node agencies in partnership with those agencies

Africover

National Focal Point Institutions (NFPI) designated as custodians are responsible and accountable for ...ensuring data documentation (conforming to ISO standards) is available so that datasets can be discovered through a metadata clearinghouse mechanism (and also website)

SISEI Bénin

· The Institutional Profile of the Environmental in Benin: a document detailing among others, environmental actors, their mandates, their products, the international agreements in which Benin is part taking, the national strategies, policy statements, etc.

· The SISEI website : particularly its search engine allowing the use of key words to retrieve information, the institutional data source giving, by field of operation, quick access to the WebPages of the contributing institutions, the thematic data source, giving access to classified environmental information, the Information and Products Data Source including metadata.

BurkinaFaso

A metadatabase fed and maintained according to the following clause : "All stakeholders involved or concerned with environmental information production are committed to inform the network of new activities undertaken in their respective fields, the data involved, and the findings resulting from their efforts,"

Data maintenance

South Africa

· "The Minister must, in consultation with the Committee, from money appropriated by Parliament for this purpose, establish and maintain an electronic metadata catalogue as a component of the SASII"

· "An organ of state must capture and maintain metadata for any spatial information held by it"

· "If an organ of state, other than a data custodian, captures or updates spatial information, it must only update the spatial information fields maintained by the data custodian and, within 30 days thereof, provide an electronic copy of those fields or records of the captured or updated information by the custodian, at no cost."

· "A data custodian of a base data set must, within 30 days of updating, furnish all updates of the base data set to the data custodian of a derivative dataset, in order to ensure synchronous maintenance of the two datasets."

· "A data custodian of a derivative dataset must update the dataset with reasonable promptness after receiving an update of the base data set, to avoid possible ambiguity in the spatial information presented."

· "A user or data vendor must, in terms of sub-sections (2) and (3), report any problem or shortcoming which in his or her opinion affects the quality of spatial information, within 30 days after discovering the problem or shortcoming, to the data custodian or data vendor who supplied the information."

· "The data custodian or data vendor must, subject to sub-section (5), respond, in the prescribed manner, to the user or data vendor within 30 days of receiving the report and such response must either convey the corrected spatial information or, if more appropriate, a reasonable explanation regarding the problem or shortcoming perceived by the user or data vendor."

Nigeria

· "Updating of these datasets [fundamental datasets ] shall be done on a continuous basis but not later than five years after production."

· "All GI projects should contain in-built programme of data updating in line with policy item 5 above."

· "A custodian of a fundamental dataset must, not later than 30 days after updating, furnish all updates of the base dataset to the clearinghouse; the clearinghouse shall in turn inform the custodian(s) of the derivative dataset(s) within 7 days, in order to ensure synchronous maintenance of the fundamental and derivative datasets."

· "The owners of the datasets have responsibility to update their datasets when considered old in line with policy item 5."

· Efforts should be intensified on the operationalisation of Nigerian Satellites to provide stable primary data source for the production and updating of the relevant fundamental datasets.

· The producer of each dataset [Thematic dataset}shall ensure updating of the dataset on a regular basis as appropriate.

· A producer of thematic data who used a fundamental dataset as input shall only update the geospatial data fields produced and maintained by that particular producer.

· A custodian of a derivative dataset must update the dataset with reasonable promptness after receiving an update of the base dataset, to avoid possible ambiguity in the geospatial data presented.

· The custodian of a base dataset shall render all reasonable assistance to the data custodian of a derivative dataset to perform the updating contemplated in 6 above.

· The metadata of any dataset shall be updated whenever the dataset is updated.

· A user shall report any problem or shortcoming, which in his/her opinion affects the quality of a geospatial data, within 30 days after discovering the problem or shortcoming, to the clearinghouse.

· The clearinghouse shall convey the information to the relevant data custodian while denying access to the affected data until the producer has rectified the identified problem.

Africover

· "Custodian NFPIs must maintain plans for information collection, conversion and maintenance in conformity with the needs of users. Consequently they must liaise with Africover Steering Committee or an equivalent regional body that helps coordinate standards implementation at the regional level, clients and other affected parties when making any significant information management or dataset changes, so that the impact upon the user community and its clients can be assessed. The custodian NFPI is also responsible for negotiating the terms and conditions under which other agencies collect and maintain the MADE information on its behalf"

· "To achieve the purposes behind custodianship, NFPIs designated as custodians are responsible and accountable for:

o maintaining the quality of the MADE information assigned to them e.g. accuracy, integrity, currency, and completeness

o publicly declaring, through different facilities the status of MADE information concerning coverage, source and compliance with national and international standards."

· "User agencies receiving MADE information from a custodian should advise the custodian of any errors or omissions detected in the information received"

· "Where a user agency collects specific information on behalf of a custodian, it should do this according to the standard set by the custodian. User agencies are also obliged to pass the information back to the custodian for maintenance or storage free of charge."

· "User agencies producing information products from information which is the responsibility of another custodian agency should consider the passing back of the information product to the custodian as part of their agreement for the use of the information."

· "The user agency shall also pass on to the custodian information that has been improved or upgraded as part of this process. The custodian shall in turn ensure that that the improved or upgraded information is made available to any other users."

SISEI-Benin

"Every partner must ensure the regular updating of the information provided through its Web pages."

Burkina Faso

· "Every partner within the Network shall ensure the reliability of the data and information provided and is accountable thereof" [ maintenance implied]

· "Every partner within the Network must undertake regular data collection, processing, storing and dissemination, in the limit of the field defined by its mandate". [maintenance included]

ANNEX 1: South Africa Spatial Information Bill

ANNEX 2: Nigeria National Policy on Geoinformation

ANNEX 3: FAO Africover Eastern Africa Guidelines for custodianship

ANNEX 4: SISEI Bénin : La charte Informationnelle

ANNEX 5: Burkina Faso : Règlement Intérieur du PNGIM

In the SDI context standards guide data collection and processing, while existing data (and new datasets) are documented, which removes the risk of poor quality data being accidentally used in decision making.