Democratizing Data in Government Agencies

In a digital era of rapid technological advances, it is imperative that governments adopt data-driven approaches in order to provide value for citizens and become more efficient. In a recent webinar, Alex Scriven, Senior Data Scientist from the New South Wales Government in Sydney, Australia, shared his insights around data democratization in a government environment, inspiring use cases where government data was utilized to deliver positive impact, as well as actionable best practices for developing a data-driven culture in government agencies.

What Is Data Democratization?

Data democratization can be described as the situation where everyone in the organization (not only data professionals) has skills, confidence, and access to produce data-informed decisions. While this definition is highly generalizable, it has two specific dimensions in the government context:

• Democratizing data access and skills internally amongst government staff (i.e. within and across government agencies).
• Providing transparency for citizens by making government data available for all—which is often referred to as open data.

These two dimensions are related because a positive data culture within the government where skills and access are democratized, will motivate and inspire towards the concept of open data for its citizens as a flow-on effect.

Why is Data Democratization Important in Government Agencies?

There are multiple reasons to support the case for data democratization in the government. Let’s have a look at the top five reasons:

1. Shortage of talent

It is well known that data talents are in short supply globally, with Quanthub reporting that there are three times more data science job postings than job searches. To make matters worse, this scarcity is felt especially keenly in the government sector. The ‘State of Data Science’ report by Stitch revealed that only 1.2% of data scientists ply their trade in the government sector, a proportion much lower compared to numerous other industries.

This means that there will not be enough data practitioners to support all the units within government agencies. The solution is to work towards data democratization by building data skills amongst the broader non-technical employee base, such that they are able to initiate and undertake data-driven projects with confidence and competence.

2. Better deployment of existing talent

Given the data talent shortage in the government sector, it is also crucial that existing talents are well managed and well deployed. In an ideal setting, data practitioners are using their expertise to deliver the greatest value possible for agencies and citizens alike. In agencies where data is not democratized, data talent is often pigeonholed into creating dashboards and delivering value in the form of descriptive analytics (Singapore Smart Nation & Digital Government Office).

If agencies focus on democratizing data skills and access, non-technical stakeholders will be empowered to draw their own descriptive analytics, and data practitioners’ focus will shift from descriptive analytics to predictive analytics and higher impact projects.

Moreover, it is vital that these talents do not fall into the data support trap, where they face burgeoning workloads from all kinds of unfiltered requests. By providing them with the optimal working environment, it is also more likely that the government can retain these highly sought-after talents.

3. Deeper Insights

The non-data-savvy government staff are subject matter experts in their respective fields, meaning that there is a wealth of domain expertise waiting to be tapped into. Instead of imparting domain knowledge to data talents, it is wiser to develop the data skills of domain experts such that they can readily merge their expertise with technical abilities. By leveraging the existing domain knowledge, the agency can gather deeper insights by asking and answering business questions through data-driven approaches.

4. Overcome Speed Barriers

Government agencies are traditionally associated with slow processes stemming from bureaucracy and red tape. When engaging with data practitioners, this is exacerbated by time-consuming workflows involving iterative interactions between data talents and subject matter experts.

With domain experts empowered, they will instead be able to immediately explore the data by themselves, and discover answers to fundamental business questions without the need to involve data practitioners all the time.

5. Trust and open data

As discussed earlier, data democratization in a government setting can go beyond empowering staff with data skills and access, but also by opening government data up to citizens. The setup of an open data portal delivers many benefits to both the government and the citizens. For example, the citizens will value the higher level of transparency offered by the government, while the government can gain from the innovation fostered by the public and from inter-agency collaboration.

Use Cases of Data Democratization in Government Agencies

The use cases of data democratization in government agencies are broad and many. In general, use cases can be categorized under three broad buckets: faster crisis response, improved operational efficiency, and improved citizen engagement and transparency.

(1) Improved Crisis response

Equipping government agencies with the right data access and tools will allow them to mount rapid and optimal responses to major crises. In the webinar, Alex dives deep into a slew of use-cases, but a few stand out. For example, the World Bank and donor governments launched one of the largest disaster relief efforts in history for the Haiti 2010 earthquake. This was driven by access to high-quality satellite imagery under open mapping schemes that facilitate public distribution and use. The Red Cross, for example, analyzed social media data to identify where storm victims are seeking refuge to optimize disaster response (The Economist).

The West Africa 2014 Ebola crisis also involved numerous public data innovations. For instance, Flowminder used cell phone data to predict potential disease transmission routes, while the Ebola Open Data Initiative allowed health officials to develop an open-source model for Ebola intervention simulations.

(2) Improved operational efficiency

Like in any organization, leveraging data at scale can provide massive gains in operational efficiency, and this is similarly seen in government settings. For example, with better deployment of data talent, the was able to apply machine learning on its data to automatically identify fraudulent health bills of its civil servants, reducing tax payer money waste.

The UK Government was able to save £1.24 billion by leveraging descriptive analytics to analyze technology vendor spend and avoid excessive procurement costs. Moreover, in the US, the city of Kansas was able to utilize data to prioritize housing inspections, leading to an 11-fold decrease in inspection times, and improved outcomes for citizens. (Harvard Ash Center)

(3) Improved citizen engagement and transparency

There are numerous data-driven solutions to help promote engagement between the government and its citizens. For instance, the Tunisian government made use of social media data as a proxy to gauge citizens’ perceptions of the government. This is better compared to traditional household surveys as it is cheaper and can be done more regularly.

Another example can be found in Mexico, where the online Mejora Tu Escuela (improve your school) platform provides citizens with data on school performance — empowering parents to make educated decisions around where to place their children and get involved in improving the school. This is only one of the many open data use-cases covered throughout the webinar.

Best practices of democratizing data in government agencies

Data democratization in government agencies requires a concerted effort in order to make it become a reality. Alex shared several valuable tips on how to get started and keep it going. In summary, it’s about empowering people with confidence, learning, tools, skills, and access to data to galvanize a data culture that drives improved outcomes.

Concretely, galvanizing a data culture in a government setting requires steps across a variety of levers from infrastructure, people, tools, organization, and processes. At the end of the day, the people in the government agencies are the most important piece as they are the most powerful force for groundswell democratization. Data democratization can be accelerated by consistently supporting and building up their skills and confidence, and reaching out to the passionate members to lead the transformation.

If you want to learn more about the steps government agencies can take to democratize data science, make sure to watch the full recording of the webinar.