Everything we learned from Ministry of Data

The complexity of the challenges we deal with – whether it is conflict, climate change or governance – combined with the high value and cost of technological skills and exponential growth of the volume of data requires new approaches, which tap into the potential, talents and skills of different parts of society. It is with this in mind that UNDP, in partnership with the Slovak Ministry of Finance, launched Ministry of Data in 2015 – a challenge to open and reuse data for the public good. 

Three years on we’ve engaged more than 1,000 young people from 13 countries to use government data for greater civic engagement and increased institutional accountability, soliciting over 150 ideas, and working with over 40 mentors. In 2018, we paused to think, and take stock of our experience, and here’s what we’ve learned so far.

Lesson #1: Listen, Adjust, Adapt (or don’t marry a design) 

We’ve made a number of deliberate design choices for the Ministry of Data challenge, but also adjusted the design over time. For instance, we introduced a two-step process in selecting winners to limit the time investment of participants. We also improved on the expert mentoring of participants through bootcamps and online consultations; and we sought to engage in long-term community building across countries by partnering with local organizations and by organizing several meet-ups before and after the challenges. However, not all these attempts were successful. We now know that we were trying to do too many things at once from idea generation to incubation to community building. We’ve learned that it is important (and sufficient) to focus on one aspect of it – but more on that below. 

Lesson #2: Don’t start with the data, but know data access and quality are key 

Many data challenges, competitions or hackathons, in particular those that rely on government data, grapple with issues pertaining to data accessibility, quality and usability. Despite increased investments in recent years, openly available high-quality government data is the exception rather than the norm in the countries we work in. For the Ministry of Data challenge to work, we had to invest significantly more time and energy than we expected in partnering with governments to source, clean and standardize datasets before we were able to make them accessible to challenge participants. Even then, we had to acknowledge limitations in the data and recognize significant differences across countries when in 2017 we ran a challenge in five countries simultaneously. We regularly encouraged the participants to go out and generate their own data, however providing a few basic datasets that allow for the right scale is key in terms of delivering useful and usable software applications.

Lesson #3: Framing determines participation  

Over the years, we’ve learned that the specific framing of the Ministry of Data challenge, including its problem statement, the prize money and our outreach strategy, would determine the kinds of participants we were able to get into the room. For instance, we saw that higher monetary prices were more likely to attract more experienced individual participants as well as start-ups or larger civil society organizations, and that non-monetary awards attracted younger participants and more non-formalized groups. However, the level of interest and number of submissions also seemed to be influenced by external factors that we were not able to influence, such as the growing number of other data or tech competitions in the region.

It wasn’t only the prize money and the kind of mentoring we provided, but less obvious factors such as licensing arrangements, that had an impact on who participated. While students and civil society groups were less interested in who would own the final outputs and under what conditions, it was important for small firms and developers to have the permission to market their software applications and other outputs. What’s more, while we consistently had submissions from women, the majority of submissions still came from young men – forcing us to rethink what are the types of skills and networks we need to build to be able to stimulate and facilitate women’s engagement. We’ve learned that for many women, it was the targeted expert mentoring and training that was perceived to be of much greater value than the price money itself.  

Lesson #4: Design for impact

Challenges can be a useful tool to demonstrate the value of open data, in particular, to policy-makers and government officials. Ministry of Data has allowed us to move rapidly by crowdsourcing ideas from people with very different sets of problem-solving skills and by developing solutions in an agile way. While through Ministry of Data we were able to quickly engage with different communities, we’ve learned that other interventions are needed to sustain the momentum beyond the initial engagement. And whereas challenges are good at generating ideas, bespoke incubation support is needed to get these ideas off the ground. The same applies to the provision of government data where hackathons or challenges can help drive the initial release of data, but where longer-term investments in data quality, institutional capabilities and technical frameworks are needed for long-term change and sustainability. Therefore, when designing an open data challenge, it is key to define clearly the value-added of reaching out to different communities and building in the right type of partnerships – both in terms of leveraging the talent outside and ensuring that the solutions are ingrained within bigger frameworks allowing us to address the issues we deal with at a larger scale.

Now that the first wave of open data competitions, challenges and hackathons has passed, with the accumulated experience and learning, we are ready to design the next generation of community engagement activities.