Thread by @TheGovLab

NOW: Stefaan Verhulst’s panel on “Responsible re-use of data for development through data collaboration” for #UNCTADeWeek. We’ll be livetweeting as he, @PMarteyAddo, @EleanorCarey10 & @dr_babs discuss how dev practitioners can generate new insights and overcome data inequalities.

He begins with an explanation of The GovLab's focus on #data and people and the ways data can help change how we go about development.

He reminds the attendees of the policy paper that The GovLab and AFD developed on data and other emerging tech: afd.fr/en/ressources/…

Stefaan also describes some insights from his paper w/ the OECD on “Reusing data responsibly to achieve development goals.” He notes how data collaboration can be cost-effective and innovative way of multiplying dev impact of data.

Read the piece here: oecd-ilibrary.org/sites/b34be3e0…

Goal of this panel will be to discuss ways to establish and operationalize new structures and frameworks for responsible use and reuse of data, particularly for the benefit of those operating in development contexts.

He gives each panelist an opportunity to introduce themselves.

First Q for @PMarteyAddo: Based on your experiences as head of DataLab, could you talk about the value that data reuse and data collaboration can provide development? What are some of the illustrative examples of the ways they can inform decision-making and improve lives?

Peter answers: We need data to address the key changes we face and help us invest wisely.

He describes how #data can create knowledge, enable better decisionmaking, and create new forces of services & opportunities.

"The data that we observe that we need to make informed decisions & to also address developmental challenges is privately held. We also know stakeholders can help translate data into action. It is important there is some form of collaboration btwn the data holders and others."

He draws on the model of the recently concluded #Data4COVID19 Africa Challenge, which sought to enable innovative reuse of data for pandemic response. Read more here: datachallenge.africa

Identifies priorities to enable reuse:
1. Spurring data literacy
2. Establishing shared repositories where people
3. Need for data stewards in organizations who can promote responsible reuse of data for public good.
4. Data holders collaborate with data holders

Peter: "We need a mechanism for transparency and accountability for sharing of such data. I think that will address some of the key concerns on privacy that we see here and combat the misuse of data."

The next Q is for @dr_babs: Our research suggests that #data reuse and collaboration can be cost-effective, innovative, and inclusive. Do you agree with this assessment? What do you see as the potential from your position in UNDP?

"Following on Peter's comments, the reuse of data is very important. A big part of the investment in development is in collecting data. [...] What we've seen is you spend a lot of time collecting data and there's not enough time to analyze and use it."

He goes on, "A big part of what we've been doing at UNDP is seeing how we can bring all that info together in the first place so they know that data exists & developing tools they can understand."

Is not sufficient to collect one type of information but to maximize globally.

He speaks on one challenge: Monitoring & Impact Evaluation

"That's one area that I know a lot of dev agencies put in so much money. We feel like we need new data every time we do that. We need to identify ways to do this in a better way. We need to leverage data that exists."

Stefaan asks @EleanorCarey10 on trust: Are we currently moving in the direction of gov frameworks that can support reuse of data for sustainable development in an inclusive way?

He references the OECD’s report on “Shaping a Just Digital Future” here: oecd-ilibrary.org/development/de…

"I suppose to sum it up very quickly, in the field of data and digital governance, there's a lot of momentum but there's a big task ahead in establishing a harmonized approach to digital governance."

She describes how COVID showed that governance was not just a niche issue but an essential one that cut across various fields and issues.

There are ways that data governance can be used for good or weaponized. OECD has tried to promote good use & digital cooperation.

"There's this overarching framework for the international community to work within. The rationale is incontrovertible and the trend is toward more, not less."

Now, she argues there are 3 ways of thinking about digital gov:
1. Build a gov model that can max rewards while reducing risks
2. Attempting to contain digital gov as a vertical is futile. Need to be cross-sectoral
3. Recognize and prioritize those voices being left behind

She emphasizes this work requires more resources. There's a great disparity in the starting point for countries.

Developing countries are having to adopt frameworks that are not compatible with their economies or match their capacity to implement.

Stefaan now asks everyone: Do we need a more proactive approach to data governance? How much do we still need to convince people to take governance seriously as a positive? What's been the evolution?

Peter: "For the past few years, we've seen more adoption of the use of data. I've started seeing top management start leveraging on data to try to study feasibility of their projects. Babatunde talked about impact and evaluation which has been a big issue."

He goes on, "What I find a lot less is funding. It comes with a lot of convincing to make people understand. One of the big areas is it has to do with education and understanding the value proposition this can bring to development."

The other issue is GDPR and responsible use. People are seeking to better understand how is the data collected and how we can better handle it.

There's growing cognizance of the sensitivities.

Peter asks @dr_babs: What's your experience with non-traditional data?

Babatunde says it's an area that's grown a lot but there are struggles on how to scale it up. There's a lot of interest in non-traditional use and data but challenges with using it.

He notes that some insights are difficult to generate because it requires data that is highly granular and often only held by private sector.

He calls for education and demystifying non-traditional data. There's a great need for resources that can explain this issue. In his own capacity, Babatunde talks about how his org hosts meetings on this issue.

Stefaan notes how this aligns with Peter's call for #data literacy.

SV: How do you go about due diligence on the data that is being made available, especially as that relates to privacy concerns?

Eleanor: There are different approaches that countries are taking. In some, national stat offices are validating data that is coming out of them, curating.

States legal frameworks alone can be insufficient for regulating big data in both advanced econs and developing ones.

There's a need to reinforce the importance of traditional datasets alongside nontraditional. It is not an "either or" proposition but complimentary.

Peter describes the challenges of biases encoded in data and algorithms used by big companies. One effort to improve data responsibility would be to fund independent research that can bring to bear these issues in an open platform.

He cites this paper: doi.org/10.1145/344218…

Another solution would be testing whether data resembles what we see on the ground. We can have people on the ground validate what we see from insights is actually what we see on the ground.

Calls for collective intelligence & other approaches.

He says that a third approach is collaboration with other orgs and companies. Can influence how data is managed in an ethical and responsible way.

SV emphasizes the challenge of finance and cost. He notes how data collaboration can be used to reduce these challenges and scale solutions but we also do need to better finance reuse.

As we begin to close, SV asks: What do we see as the major priorities? What would be your priority if there were no financial restraints?

.@EleanorCarey10: "We really need to think about strong but appropriate governance frameworks. [...] We need to be thinking about where the balance of powers lie. We need to think about balance btwn risk & reward."

Speaks briefly on development cooperation.

.@dr_babs: "I think that coordination and making sure the data is visible. If people cannot find data, they cannot use it."
@PMarteyAddo: "Improving data capacity through education and training. If we have that, we'll have trust, data stewardship."

And that's all! Thanks all who could attend the panel live as well as those who followed along via Twitter.

If this conversation was of interest to you, you may also be interested in GovLab's forthcoming Summit of Open Data Policy. Sign ups open today: eventbrite.com/e/the-state-of…