Data holds the potential to drive innovation and help us make evidence-based policies and decisions. Data innovation is also crucial to supporting data collection and analysis to monitor the achievement of the 2030 Agenda for Sustainable Development. Innovations in data enable policymakers to identify and integrate insights into planning processes, course-correct strategies, and get real-time feedback on how well policies are resonating with people. This creates opportunities for more targeted and cost-effective interventions, improvement of public services, as well as empowerment of citizens – even more critical as communities seek to recover from the impact of COVID-19 and work towards more sustainable and inclusive societies by 2030.
One of the biggest challenges facing many of the SDGs, and SDG 16 on promoting peaceful, just and inclusive societies in particular, is the lack of available data on many of the indicators. In 2021, only two indicators under the goal had a coverage above 75%, five indicators had a coverage between 50% and 75%, two indicators had a coverage between 25% and 50%, and thirteen had a coverage below 25%. Time, cost, political will and the relative newness of many of the indicators (e.g. SDG 16.3.3 on access to dispute resolution mechanisms was added to the indicator framework in 2020) all pose additional strain on many data producers and especially National Statistics Offices responsible for collecting this data.
SDG 16 Data Innovation
A wide range of official and unofficial data actors are stepping up to fill the gap on SDG 16 data by building new partnerships, identifying technological innovations for real-time data, moving towards full digitalization of their data collection methods, identifying innovative data sources to complement official statistics as well as innovative data collection tools and methodologies. This is now even more relevant in the new reality of COVID-19 where many traditional, paper-based methods is no longer possible. To capture some of these examples, UNDP OGC is launching a collection of stories on SDG 16 Data Practices featuring a wide range of innovative SDG 16 data initiatives led by National Statistics Offices, national data producers as well as independent data actors.
Two examples from UNDP support in this area come from Armenia and Tunisia. UNDP has been working with national actors to test out how using big data can complement our understanding of the different dimensions of SDG 16 in the two countries.
Data Innovation for measuring inclusive and responsive decision-making– an example from Armenia
Alternative data sources, innovative data collection methodologies and tools, have been extensively used in UNDP Armenia’s National SDG Innovation Lab. In one of its first projects, SDG 16, and particularly the indicator measuring the proportion of the population who believe decision-making is inclusive and responsive (SDG 16.7.2 indicator) was selected for the SDG Monitor pilot for which a SDGmonitor.live platform was initiated.
Facebook/META — the most popular social media platform and the main channel of informal communication between the citizens and public officials in Armenia, and E-draft — the Government’s official website on the publication of draft legal acts - were identified as the primary non-conventional data sources.
Combining official data sources with state-of-the-art machine learning algorithms, data was analyzed from 8 public institutions, 30 Facebook profiles of current and former public officials, 45000 FB posts, and 469264 comments, as well as public feedback on 2226 draft legal acts. Based on the collected data, the analysis reveals trends in government and public activity levels. As a result, the SDG Monitor pilot, a user-friendly, AI-powered, open-access interactive online data analytics tool for real-time SDG 16 progress monitoring in Armenia which couples conventional and non-conventional data sources was developed.
Lessons
From the onset of this journey, a successful model of cross-sectoral partnership among various public institutions and development agencies was established. A Task Force composed of representatives from the Deputy Prime Minister’s office, various prominent public institutions, and UN agencies was set up to provide necessary field expertise and support the SDG Innovation Lab team through each stage of platform development — from idea generation to the validation of the ongoing processes.
Where to next?
Following the development of innovative methodologies of in-depth country-level data collection and success of implementation of SDG Monitor pilot, the SDG Innovation Lab team started working with Data4Now global initiative led by UNSD, to support the National Administrative Department of Statistics of Colombia with integration of unconventional data sources in their national statistics monitoring efforts. The knowledge gained through this process and multi-stakeholder collaborations will be utilized to scale-up the SDG Monitor as well as expand engagement in mutual learning and solution sharing for more rapid and sustainable development progress in Armenia and beyond.
Analysing sentiments for corruption-related offences – an example from Tunisia
In Tunisia, as part of contextualizing governance indicators and identifying proxies to complement SDG 16 indicators, the Tunisian National Statistics Institute (INS) experimented with the use of social media analysis to monitor corruption. The aim was to explore whether non-traditional sources of SDG 16 data, including social media, can be used to assess perceptions of corruption. The exercise also compared the data to the baseline established by the official data on corruption collected by the INS. The objective was to find a way in which social media and non-conventional data can complement traditional statistics to monitor citizens’ perceptions and attitudes.
UNDP Tunisia worked with the INS and UN Global Pulse on “sentiments” on social media platforms to measure SDG 16.5.1 on the proportion of persons having been in contact with bribery. A network analysis on web and social media was conducted, which searched for content that included pertinent corruption-related keywords in Arabic, French and English. A content analysis was subsequently done to determine the relevance of the posts and whether these had a positive or negative annotation.
The results were encouraging, showing fairly strong convergence between results obtained through social media analysis and survey data generated by the Governance, Peace and Democracy household survey conducted by the INS of Tunisia.
Lessons
Sentiment analysis, often referred to as opinion mining and emotion AI make use of natural language processing to assess affective states of particular topics. Social media analyses can be regarded as a sort of heart rate monitor providing a diagnosis based on parameters like variability and activity over time. From this perspective it can provide a significant value-added to complement traditional data sources. However, there are still challenges to confront including concerns over privacy issues of data harvesting. In addition, the reliability of data is also questioned, given the limited technical rigor compared to traditional household surveys (e.g. lack of representative sample of those who have access to the internet or predominance of negative opinions, etc.).
Where to next?
UNDP Tunisia will continue to work with the INS to institutionalize the use of social media analysis for official statistics and SDG monitoring. To this end, it will support the development of capacities and tools, including a methodology, to conduct social media sentiment analysis on trust as part of the Technology for Democracy initiative, with the objective to complement official measurement of SDG 16 progress.
This blog has been developed by Ulrika Johnsson, UNDP Oslo Governance Centre, Eduardo Lopez-Mancisidor, UNDP Tunisia, and Elen Sahradyan, UNDP Armenia
This blog was first featured on SDG 16 Hub.
Cover photo: Unsplash.com