Using behavioural insights to build a data culture

image of imen's essay 'using behavioural insights to build a data culture' from a pdf of the proof

The digital transformation of the public sector in Tunisia is at the top of the national agenda. The pandemic highlighted a large digital divide between citizens which meant many households faced challenges when receiving social aid, pension payments and cash transfers as well as accessing public health care. The discrepancy in levels of digital literacy also left many people vulnerable to fraud as fake social media pages appearing to offer social aid tricked citizens into sharing personal data such as national identity card numbers.

Introducing decrees

To accelerate digitalisation, the Tunisian government introduced a series of measures including several decrees dedicated to tackling the resistance to change and the administrative inertia, and encouraging interoperability.

In theory, introducing decrees to help facilitate change when it is urgently needed sounds like a good idea. But in practice, Tunisia found that mandated change meant the focus was not on citizen or public sector teams’ needs, and consequently, this would not translate to implementing proper public policy tools.

We quickly learnt that decrees are not the panacea and that organisational culture is the biggest barrier to change. To accomplish its goals, Tunisia needed to build a culture of trust, autonomy, and collaboration across organisational borders. Tunisia needed a new approach to address these.

Advancing interoperability through behavioural insights

Interoperability of information systems was one of the top priorities for (former) Minister Mohammed Fadhel Kraiem. The social protection experience through the AMEN Social programme highlighted the importance of data sharing and data as evidence to build and implement a fair and efficient social protection model.

Behavioural insights helped find approaches to help the organisation responsibly share and integrate data as part of our efforts to establish interoperability.

We decided to use behavioural insights because we felt that although traditional, top-down approaches like decrees don’t work by themselves, the 2 approaches used together might be effective.

As a behavioural researcher at the Ministry of Technology, my role is to help understand the challenges civil servants face when digital transformation programmes are launched. Research findings can then be used to inform the implementation of these programmes.

When the government issued a decree on interoperability, I organised a series of workshops with colleagues from the Ministries. I soon learnt that the main challenge was change resistance: stakeholders did not want to set up interoperability processes and share data because of a combination of psychological and social barriers.

Insights into data sharing between organisations

At the Ministry of Technology, we’ve been using the Organisation for Economic Co-operation and Development’s (OECD) ABCD framework to guide our research. It focuses on 4 drivers of behavioural problems:

1. Attention.

2. Belief formation.

3. Choice.

4. Determination.

Here are some of our observations from the workshops we organised to understand the barriers to data interoperability.

1. Attention (how easily people are distracted) People did not pay much attention to the data interoperability programme because of its lack of clarity and the complexity of the enrollment process. They had limited visibility on how the data archiving system would work, whether it would mean there’d be additional costs, what type of technical assistance they could receive or how the once-only policy could benefit then. The once-only policy is a principle which states that public administrations should collect information from citizens and businesses only once, and then share and reuse this information.

2. Belief formation (how first judgments often turn out to be inaccurate) We realised people were reluctant to share data because of a fear of losing control and unveiling poor data encoding and protection practices, for example, classifying data, anonymising data, developing consistent metadata.

3. Choice (how people are influenced by the framing of a choice, as well as the social and situational context) Data interoperability was available as an on-demand service. But because people tend to stick with what they’re used to, there was a lack of appetite to drive change and proactively ask for, and embrace, data interoperability.

4. Determination (how people’s willpower is limited and subject to psychological biases) It took a significant amount of time for people to feel committed to the programme, so the ones who bought into it earlier on had difficulties staying motivated.

Encouraging data interoperability

Insights from these workshops helped inform how we should design the implementation of the data interoperability programme. We knew there was work to do before agents from both business services and sectorial decision makers felt comfortable sharing data.

To help, we developed a series of offline meetings around social protection services to understand user needs and encourage partners to meet and interact. This offline meeting was an opportunity to discover data use and data needs from each stakeholder and unveil their concerns about anonymisation as well as level of abstraction of their data.

Together, we co-designed a framework for sharing and managing data. We also introduced a ‘data interoperability simulator’ to help users visualise what an open data provider/data consumer process might look like. We felt the best way for business sectors to build trust in a more transparent way of working was by giving them the opportunity to experiment safely with their data, and be involved in considering what could be shared with by third parties.

The process helped deconstruct long-ingrained beliefs that prevent some sectors from sharing their data. Using behavioural insight methodology and experimenting with a data interoperability simulator have gone a long way to supporting the organisational transformation of the National Center for Computing into a National Operator for Interoperability that will help steer the interoperability deployment with diverse stakeholders.

We’re still learning. But we believe that using behavioural insights to improve culture and trust, help us build sustainable foundations for digitisation in the public sector.

Public Digital