Open Agricultural Data (BR0115)

What is the public problem that the commitment will address? The various actors in the agricultural value chain have made efforts to promote data availability. From a public policy perspective, a broad regulatory framework is already in place to ensure that government bodies carry out transparency actions aimed at disclosing agricultural data on the Brazilian Open Data Portal. Government bodies that deal with agricultural data, such as the National Supply Company (Conab), are legally competent to collect, systematize, store and make available their data. Education and research institutions, such as the Brazilian Agricultural Research Corporation (Embrapa), are engaged, through the Open Science initiative, in sharing the data generated by publicly-funded agricultural research. The Ministry of Agriculture, Livestock and Supply (Mapa) has a Data Observatory that integrates databases with the objective of strengthening and improving the integration, management, access and monitoring of data and information of strategic interest to the agricultural sector and Brazil. Civil society is also engaged with the theme through the climate, forest and agriculture coalition, the forest code observatory, the climate observatory and even the OGP initiative. Partnerships between government bodies and civil society have established networks and collegiate bodies with functions and duties related to the opening of data, such as the RDA Brasil, which is a neutral and democratic forum for discussing research data management, the Dataverse-Brasil, a discussion group about the development and implementation of data repositories using Dataverse as a tool, and the GoFAIR Brasil Agro Network, which is being established to enhance the dissemination of FAIR principles in the opening of agricultural data and related areas by its members in a coordinated and collaborative manner. Despite the significant number of diverse initiatives aimed at promoting the availability and reuse of data from agricultural value chains and the civil society’s high demand for access to this data, these efforts have not been effective, as they are made in a disconnected manner, which hinders data opening and integration and, consequently, data reuse. Despite the considerable volume of data available, a great part of it cannot be used, since many databases do not conform with open data principles, which prevents information from being findable, accessible, interoperable and reusable, due to the lack of mechanisms for assessing risks and impacts of the opening of agricultural value chains data and the discrepancy between civil society interests and the priorities for opening agricultural value chains databases defined by public bodies.

What is the commitment? The commitment is to engage the government and civil society in the opening and integration of priority databases related to the agricultural value chain, considering the risks, impacts and feasibility of these actions.

How will the commitment contribute to solve the public problem? To fulfill the commitment, government bodies and civil society organizations will participate in a meeting (milestone 1) and carry out an assessment of existing databases (milestone 2). These milestones aim to ensure collaboration between agricultural value chain actors in the opening and integration of databases, understand civil society demands, survey existing initiatives on the theme to avoid duplication of efforts, indicate priority agricultural value chains for opening of data and consolidate, analyze and generate new data and information of strategic interest. These initial actions will serve as preparation for the following actions, which consist of assessing the risks, impacts and feasibility of opening and integrating databases (milestone 3) considering the ethical and responsible reuse of available data, database security and the protection of sensitive data. Milestone 4 involves defining the minimum metadata set needed to ensure data interoperability and transparency. After implementing these actions, priority databases to be opened and made interoperable will be defined (milestone 5).

Why this commitment is relevant to OGP values? The commitment involves significantly strengthening transparency and social participation, as it will promote the opening and integration of databases related to agricultural value chains with an intense participation of various actors in the value chain.

Additional information The commitment is directly related to the 2030 Agenda SDG 2 - Zero Hunger and Sustainable Agriculture and its target 2.4. This goal is intended to end all forms of hunger and malnutrition by 2030, so as to ensure that people – especially children – have access to nutritious and sufficient food all year round. To this end, it is necessary to promote sustainable agricultural practices, through support to family farming and equitable access to land, technology and the market. Its target 2.4 is to ensure, by 2030, sustainable food production systems and implement resilient agricultural practices that increase productivity and production, that help maintain ecosystems, that strengthen capacity for adaptation to climate change, extreme weather, drought, flooding and other disasters and that progressively improve land and soil quality.

Milestones (with verifiable deliverable) Start Date: End Date: Milestone 1 - Holding a meeting between government bodies and civil society to discuss the opening of agricultural value chain databases December 2021 December 2022 Milestone 2 - Assessment of the situation of agricultural value chain databases December 2021 March 2022 Milestone 3 - Assessment of risks, impacts and feasibility of the opening and integration of agricultural value chain databases January 2022 July 2022 Milestone 4 - Development of a metadata protocol for integration and transparency of agricultural value chain data based on web standards and FAIR principles January 2022 July 2022 Milestone 5 - Opening and integration of priority databases July 2022 December 2022

Action Plan Review

Commitment 5. Open Agricultural Data

Commitment cluster #5 and #10: Open Government and Environment (For #5: EMBRAPA, CONAB, MAPA, IPEA, SEMAS/PA, OCF, UFMG, RNP, and C4AI/USP. For #10: IBAMA, MMA, ABRAMPA, and Fiquem Sabendo) [23]

For a complete description of the commitment, see Commitments 5 and 10 in the action plan.

Context and objectives:

Brazil is a major player in the global agribusiness circuit. In 2021, this sector’s exports reached a record value of over 120 billion USD, mostly formed by soy-based products, meats, forest products, sugar and alcohol-based products, and coffee. ^[24] In fact, the country produces one-third of all coffee and sugar consumed worldwide, and it is the world’s largest producer of soybean and corn, as well as the second most important producer of beef. ^[25] The presence of this key agricultural sector makes this cluster of commitments especially relevant in terms of monitoring the environmental effects of this economic sector.

These commitments were drafted from policy topics prioritized by both civil society (Commitment 5) and the government (Commitment 10). Giving their synergies, the IRM has clustered the evaluation and impact analysis of these commitments. However, the implementation is set to happen independently. Both commitments were drafted in partnership with CSOs, in a process led by the CGU. [26] Nevertheless, the final version of the commitment, according to a civil society representative interviewed by the IRM, [27] was much less ambitious than needed to do bring a substantial change. A main constraint was the budgetary limits in place since before the co-creation process given existing budgetary realities. While these constraints were explicit at the co-creation phase, they still limited the potential of the commitments.. According to the civil society representative, there were concerns on the feasibility of effectively leading commitment 10, as the agencies involved did not have or had not assigned the necessary human capacity or financial resources to deliver milestones such as launching the agreed interface. [28]

In the 2018–2021 plan, 2 out of the 11 commitments were related to environmental issues (water resources and climate change). Commitment 5 seeks to open and integrate the agricultural and livestock data available, whereas Commitment 10 aims to improve data transparency regarding environmental licensing and address the issues of having multiple environmental databases scattered across different governmental agencies and lack of access to data. It aims to do this by centralizing data and opening it in one single online platform.

Commitment 5 is a crucial step in identifying unsustainable practices in the sector. For instance, it is crucial to be able to trace the livestock's origin and determine whether the cattle originates from a farm following labor and environmental regulations or if is associated with areas known for illegal deforestation for grazing, among other concerns.. To this day, it is not possible to verify this as the databases are not opened. This is particularly important in one of the largest exporters of meat products and one of the countries with the largest rainforests in the world.

Commitment 10 is also very important, as environmental licensing is a key regulation for large works, such as roads, hydroelectric plants, and mining dams. Although subnational environmental licensing is not included in this commitment, the centralization of data regarding federal oversight is a crucial transparency instrument for civil society to monitor human intervention on the environment.

Potential for results: Modest

Data on agricultural activities currently published through the national open-data portal is not relevant to monitoring economic activities, as it focuses on implementation indicators for governmental programs and is not updated. ^[29] Meanwhile, as indicated in the commitment, relevant institutional efforts such as the Ministry of Agriculture’s Data Observatory ^[30] do not comply with open-data principles and instead provide dashboards to access statistical data, which prevents its reuse and its interoperability with other data sources to obtain further insight.

Regarding the environmental licensing process, the Environmental Licensing National Portal (PNLA) currently publishes information through its website ^[31]; however, this information is not easily accessible, nor does it follow open-data guidelines.

Despite the limitations flagged by civil society organizations, these commitments could substantially advance open government regarding environmental regulation. Currently, data is scattered and, in many cases, unavailable for the public—which was pointed out by a civil society representative interviewed by the IRM researcher. ^[32] By opening data—with parsimony—and centralizing it in one single platform, this commitment could have the potential to modestly improve transparency regarding environmental issues at the federal level.

Opportunities, challenges, and recommendations during implementation

There are two key aspects of the commitments. The first one is to have sufficient IT capacity to allocate to Commitment 10. This task requires several hours of work to build systems and make data available. Governmental agencies should be able to yield human resources to implement this commitment. The second key aspect is realistic budgets. To build up this system, some investment is needed to set it up, in particular funds for hiring IT consultants.

These two aspects, in turn, present two challenges that might risk the implementation of the commitment. To mitigate these challenges, it is important to open up channels with CSOs that might be able to point to open-source IT solutions cheaper than the proprietary software initially considered for this commitment—including some IT solutions developed by the government itself, but not used across different agencies.

The last challenge regards the standardizing of understanding of governmental agencies regarding the new privacy data law ^[33] and the need to open data, in particular to Commitment 5. Many agencies are now denying access or closing access after this law was passed ^[34] in 2018, even though it clearly accepts the opening of data that is of public interest. ^[35]

Results Report

Commitment 5. Open agricultural data

Commitment 5 sought to open and integrate data from agricultural value chains. Access to agricultural data is an ongoing top priority for civil society, as the sector represents a significant portion of Brazil’s GDP [64] and has significant environmental repercussions. There is evidence that actions were conducted for all milestones. For example, a diagnosis was conducted on the characteristics and properties of 19 databases managed by the government institutions participating in the commitment (Milestones 1 and 2). [65] An initial assessment of the risks, impact, and feasibility of opening the databases was also done (Milestone 3), [66] with a preliminary metadata protocol developed by mapping common fields among three main producers of agricultural and livestock commercial data (Milestone 4). [67] Nevertheless, the metadata protocol was not implemented, although the commitment lead mentioned that the University of Sao Paolo, a key researcher in the agricultural sphere, took the initial steps to adopt it. [68] In terms of actually opening and integrating databases (Milestone 5), there is evidence that new datasets were added to Brazil’s open data portal [69] but not integrated. [70]

One main result was the opening of geolocation limits of rural properties of the Environmental Rural Registry. [71] The commitment coordinator expressed that after the commitment’s conclusion, key insights from stakeholders' conversations informally influenced the Ministry of Agriculture. For instance, the Brazilian agricultural observatory online tool [72] now includes rural registry data and incorporates open data features and a metadata section. [73] The rural credit database has also been included in the observatory but fell outside of the IRM evaluation period for early results. [74] Both of these databases were top civil society priorities. [75]

Despite these developments, the commitment fell short of accomplishing what civil society viewed would be its most important contribution. For instance, civil society emphasized that the ownership details of properties were not disclosed in the Environmental Rural Registry, reducing the data's utility and failing to meet their expressed requests. [76] Furthermore, a primary goal for civil society within the commitment was the opening of the Animal Transport Guide (GTA) database, essential for potential monitoring of environmental infringements in livestock transportation. [77] The Ministry of Agriculture, as an administrator of the database, responded that there was no possibility of opening the GTA due to risks to producers and confidentiality concerns under Brazil’s General Personal Data Protection Law (LGPD). Additionally, they contended that the primary function of the GTA was for sanitation monitoring rather than for social control and oversight of environmental matters. [78]

Civil society cited precedence with similar databases on deforestation, [79] but viewed that the Ministry remained unreceptive. [80] The commitment coordinator clarified that the commitment became a new space to debate on the GTA, as there have already been legal inquiries by civil society organizations requesting to open the database. [81] In their opinion, these open and hard conversations, while sometimes making their job more about conflict resolution, had started changing mindsets, such as the changes made to the Brazilian Agricultural Observatory, the creation of a working group for the integration of data systems of the rural sector, and the establishment of the need to include the Public Prosecutor’s Office and the National Data Protection Office to address the disparity of opinions on LGPD application. [82] Given the political context and the shortened implementation period, the IRM considers Commitment 5 as having achieved moderate early results by advancing discussions and evaluations on opening agricultural data that resulted in specific databases being opened while generating ideas that were implemented after the commitment ended. Moving forward, the relevant federal bodies could better evaluate whether opening datasets violates secrecy provisions should be more actively involved in the national action plan process.