By Toby McIntosh
UNESCO is preparing a new survey that may shed more light on whether government record-keeping systems are adequate to assess the implementation of access to information laws.
The lack of uniform statistics was highlighted in a recent study in which researchers looked at 265 transparency evaluations conducted in Latin America between 2003 and 2017, reporting a “striking diversity” of methodologies.
“The task of forging consensus on definitions and measurements to gauge compliance with transparency remains urgent and should represent a priority for policymakers, activists, and scholars,” they recommended.
The international body with the potential to encourage better record-keeping is UNESCO, which has responsibility for measuring the Sustainable Development Goal on access to information (SDG 16.10.2), but can’t dictate national record-keeping standards.
UNESCO is planning its second survey on SDG 16.10.2 and has taken some steps to improve the low response rate for its 2020 survey, according to a UNESCO official.
Core Question Concerns Request/Response Rate
One of the core questions in surveys about access to information involves how many requests were received and how they were answered, and if they were answered.
This deceptively simple question has been the subject of much debate, resulting in many variations on the theme. A three-part approach – denied, fulfilled, or partially fulfilled – is perhaps the most common way to analyze responses.
UNESCO, in its forthcoming survey will ask whether requests were denied, fulfilled, or partially fulfilled. In addition, UNESCO will ask how many matters were pending or were dismissed as ineligible, a UNESCO official said. The survey also asks for the number of cases appealed.
(Full details of the planned survey were not available, but the previous one covered the processing of appeals and many other topics.)
Partial Fulfillment
The “partial fulfillment” category can cover many levels of fulfillment, from an almost complete response, to an almost incomplete response.
The variety was illustrated by the National Security Archive, a Washington nongovernmental organization, with a graphic example (see graphic).
More broadly, the Archive challenged as exaggerated the data in US Justice Department’s annual report on government-wide FOI performance, where the official release rate for fiscal year 2019 was 94.4 percent. The Archive recalculated it as much lower.
“A more accurate release rate calculated by the Archive and others hovers between 50 and 60 percent,” the Archive said in its Unredacted blog.
The government gets to “that overly-generous figure by counting nearly entirely redacted documents as successful partial releases (see above for an example), and excluding more than 270,700 requests denied (often improperly) over fees, referrals, “no records” responses, and requests “improper for other reasons.”
In India, a close reading of government statistics was recently provided by the Commonwealth Human Rights Initiative, which questions the government’s annual report claims about a falling rate of rejection of Right to Information requests. CHRI says “a deeper probing” of the data tables “presents a more worrisome picture.” The CHRI report also delves deeply into data on what exemptions were used to deny requests, another common element in surveys.
Lack of Basic Record-Keeping Seen
Although it would seem to be fairly simple for governments to account for responses, many governments may not keep such data.
This gap is evidenced by UNESCO’s experience and by the comprehensive study in Latin America.
UNESCO got answers to its request/response question from only 40 countries.
The researchers in Latin American found only 11 government reports with such data.
(More on these findings follows._
Because of the lack of official information, civil society groups in many countries have done their own evaluations, sometimes making test requests and recording the answers to evaluate transparency.
One such methodology was created by the Freedom of Information Advocates Network.
An example of a national survey is one conducted in Pakistan by the Centre for Law and Democracy using an assessment tool similar to that of FOIANet.
(Full disclosure: this author serves on the FOIANet Steering Committee.)
Latin American Research Finds Lots of Variety
The researchers from Brazil looked at 265 transparency evaluations conducted in Latin America conducted between 2003 and 2017, reporting a “striking diversity” of methodologies.
That diversity presented a challenge to the three researchers, led by Gregory Michener, Associate Professor of Government at the Brazilian School of Public and Business Administration, Getulio Vargas Foundation, Rio de Janeiro (FGV-EBAPE). Their article was published in Government Information Quarterly. (Subscription or payment, $41.95, is required, but try libraries).
“Compliance rates are critical for gauging how well governments are honoring commitments to transparency in practice,” wrote Michener and colleagues Jonas Coelho and Davi Moreira.
Grappling with the data, they generously interpreted the various approaches, counting partial responses as positive, as they waded through the data to create comparable compliance scores.
They calculated an overall transparency compliance rate of 46.2 percent for “passive transparency” (information provided after a request).
The authors stressed that the definition of compliance means different things depending on the evaluation, most done by governments. “This dilemma is most vividly illustrated by passive transparency evaluations,” they noted. “While some evaluations count FOI response rates as any response at all to a request, others calculate response rates as answers that fulfill pre-specified criteria.”
“The most worrying compliance gaps” were found at the municipal level, they found. “In the case of passive transparency, response rates are nearly 40% lower for municipal versus national level governments (20.4% versus 66.1%). In short, the numbers clearly show that compliance with transparency across the region is wanting.”
Passive transparency was assessed in 84 evaluations, which covered 39,305 requests for information over the 14 year period.
“An important observation derived from passive transparency evaluations is the comparative scarcity of governmental evaluations,” according to the article.
Governments authored just 28.5 percent of all passive transparency evaluations (11), the rest being done by nongovernmental organizations.
They also studied data on “active transparency” (information voluntarily disclosed by governments).
The authors express the hope “that this article may help galvanize a ‘Transparency Evaluation Network’, and that evaluators will contribute their neglected compliance evaluations to this ongoing initiative.”
“Only through dialogue can proponents of transparency establish solid grounding for universal definitions, measurements, increased evaluation, greater compliance and, ultimately, better informational outcomes for governments and citizens,” the study concludes.
UNESCO Seeks to Fill Gaps
UNESCO’s 2020 report “revealed gaps in record-keeping.”
To boost the return rate on its next survey, a UNESCO official said the agency has engaged in or is planning:
- Testing of questionnaire with information commissioners (feasibility of obtaining information and pointing to information resources)
- Expansion of the glossary and usage of the manual
- Clinics with information commissioners in filling in the questionnaire
UNESCO in February of 2020 sent out a survey to 209 countries and associated territories. The responses were summarized in a November report. The response rate may have been affected by the pandemic and some confusion over where it as sent.
The report says:
Out of the 62 responding countries and territories with ATI legislation, 40 responded to the question on the number of requests filed, granted and denied. However, not all of them provided data for all elements of this question, for instance, the number of requests denied. Several countries and territories also reported that data is not collected or available centrally. This mirrors findings from 2019, and also raises a question about data informing annual reports where these are published.
Only seven countries (Australia, Brazil, Cayman Islands, Croatia, Ireland, Israel, Republic of Korea) provided all the data UNESCO sought about the disclosure and refusal of information.
UNESCO’s report says. “In some cases, some countries also reported that all requests received were disclosed, which raises questions about the accuracy of data, as there is usually a number of requests which are dismissed as they are incomplete or do not fall into an ATI request, or there are even a few requests carried over from the previous year or into the next year.”
Question 11 of the 2020 survey asked for the total number of requests and for the instances of full disclosure, partial disclosure, and denial.
Question 12 asked for the reasons for denial, offering four categories: national security, privacy concerns, commercial confidentiality and other.
See EYE article on the November 2020 report, with links to earlier stories.
UNESCO is the “custodian agency for reporting on global progress” toward fulfilling Sustainable Development Goal 16.10.2, which calls for the adoption and implementation “of constitutional, statutory and/or policy guarantees for public access to information.”
UNESCO Consulted ICIC on New Survey
The forthcoming UNESCO questionnaire has been developed in consultation with members of the International Conference of Information Commissioner, with which it signed a memorandum of understanding with UNESCO in December. (See announcement.)
UNESCO held sessions with commissioners in three languages — English, French and Spanish – asking especially about the “feasibility of obtaining information,” a UNESCO official said.
The ICIC website does not show information about the consultations, which were confidential. EYE’s requested for an interviews with the current ICIC chairperson, Elizabeth Denham, the UK’s information commissioner, but has not received a reply.
An early draft of the 2020 survey was discussed by a group of 18 invited nongovernmental experts who met virtually for two hours on Nov. 30, 2020.
