How To Rethink Data Culture in Government
Rethinking Data Culture in Government: Why People Matter More Than Platforms
Brent Hoberman, Co-Founder and Chairman of Founders Forum Group, recently posed a pointed question on LinkedIn: 'How good is the data that governments use to make choices?' In a world where data drives nearly every sector, his query lands at the heart of policymaking—and exposes a much bigger issue.
The problem isn’t just data quality. It’s what happens after the data is gathered: how it’s interpreted, challenged, applied, and communicated. This is where people and culture play an outsized role. And it’s why leaders—in government, investment, and business—must stop thinking about data as a technical asset alone and start treating it as a strategic one that depends on human judgment and institutional design.
I want to unpack these issues from the perspective of data quality, which is important, but the people interpreting it and the culture of the environment in which they have to question the data and make assumptions.
In my view, GenAI is not there to deliver shortcuts but to present different perspectives that our critical thinking needs to consider to unlock the outcomes of improving productivity and growth, which AI can deliver.
Good Data, Bad Decisions: Where Things Go Wrong
We often treat data as inherently objective, but how it is interpreted—who interprets it, through what lens, and under what constraints—matters as much as the data itself.
Take the UK’s experience during COVID-19. Despite access to extensive health and economic datasets, inconsistent interpretations led to wavering policy responses. The Office for Budget Responsibility later admitted significant forecasting errors, shaking public confidence in data-driven decisions.
Similar issues surfaced in significant infrastructure projects like HS2. Initial economic modelling drastically underestimated costs, leading to public mistrust and reactive policy changes. These aren’t failures of data—they’re failures of how it was applied.
Even Stripe CEO Patrick Collison, a voice from the private sector, noted the danger of false confidence in large datasets—his point: insufficient data isn’t the only problem. Misapplied data—interpreted without critical thinking or contextual understanding—can be just as damaging.
In the eight years I have worked as a specialist within the UK Government, in Digital Data and Technology, Policy, Trade and Internal Audit professions, I’ve had the pleasure of working with some great people and civil servants. However, what I have noticed in my time is that the culture is what has held true innovation from taking place.
The Human Layer: Risk Aversion, Bias, and Bureaucracy
Looking at the examples that Brent highlighted, let’s look at why these policy-making issues and outcomes keep repeating themselves. In my view, it comes down to three interlocking human and cultural challenges:
1. Risk Aversion Is Rational—But Limiting
Civil servants operate in a high-stakes environment. Their decisions are under constant scrutiny by media, politicians, and the public. In this context, risk-taking is often viewed as not innovation but liability.
Sarah Munby, Permanent Secretary at the UK Department for Science, Innovation & Technology, has acknowledged that it’s often rational for civil servants to avoid risk. But logical or not, it breeds inertia.
When failure is penalised more than success is rewarded, the safest decision is to do nothing new.
2. Cognitive Bias Distorts Data Use
Confirmation bias, anchoring, and availability heuristics aren’t abstract psychological concepts. They shape how policies are made.
A policymaker invested in a particular narrative may unconsciously seek data confirming their view and discount contradictory evidence.
Over-interpretation is also a significant issue. Data stretched to fit political needs loses its integrity—and can lead to flawed, even dangerous, decisions.
3. Bureaucratic Silos Kill Momentum
Government departments often operate in silos. Data is hoarded, not shared. Systems don’t talk to each other. This report by the UK National Audit Office confirms that departments need to ‘work together more effectively on industrial strategy.’
And insights that could drive better outcomes get lost in translation—or trapped in incompatible formats.
The UK's approach remains fragmented and outdated compared to digitally integrated governments like South Korea, Singapore or Japan - the latter two I have got to know quite well.
Global Lessons: What Innovative Nations Get Right
While the UK wrestles with entrenched bureaucracy, other countries show what’s possible when data, leadership, and culture align.
Singapore: Through its 'Smart Nation' initiative, Singapore integrates real-time data platforms to make public services seamless. Government, academia, and private industry collaborate deeply, removing institutional silos.
Japan's ‘Society 5.0’ vision blends AI and big data to plan more intelligent, sustainable urban environments. In fact, this strategy is central to the Osaka (Kansai) 2025 Global Expo.
South Korea: Its ‘Digital New Deal’ enabled swift data-led responses during COVID-19, powered by strong partnerships between government and tech companies.
United Arab Emirates: The UAE—particularly Abu Dhabi—has taken a bold, strategic lead in AI adoption by integrating it into national policy, investing billions through sovereign wealth funds like ADIA, Mubadala, and MGX, and forming global partnerships to position itself as a hub for innovation and digital governance. Their national strategy is called the UAE national Strategy for Artificial Intellegence 2031.
The U.S.: While federal agencies vary widely, collaborations with private firms like Palantir and Microsoft have produced some of the world’s most advanced public data systems.
In all these examples, one thing is clear: technology alone doesn’t drive transformation. It’s the willingness to take calculated risks, to experiment, and to bring in diverse expertise that makes the difference.
Lessons from Business: Why the Private Sector Moves Faster
The cultural divide between government and business regarding data is stark. That said, the ability of the public to innovate at pace can be remarkable, but in the private sector, data decisions are often tied directly to customer feedback, revenue impact, and competitive pressure—that forces action and rewards adaptation.
Silicon Valley startups and tech companies in Shenzhen and Tokyo iterate constantly. They make small bets, test them, learn fast, and scale what works. It is in the culture that risk is learnt from, which enables the unlocking of innovation.
This mindset—rapid experimentation over exhaustive analysis—is still rare in government, where long policy cycles and political accountability inhibit quick movement. Yet, this agile approach is exactly what data requires.
A Smarter Path Forward: Strategic Recommendations
Governments must recalibrate how they think about data to close the gap between aspiration and impact.
This means shifting from a tech-first mindset to a people-and-culture-first strategy is one that McKinset, one of my clients, promotes with confidence in the ‘Never Just Tech’ way of working and communications campaign.
Here’s how government needs to rewire itself:
1. Build a Culture of Informed Risk-Taking
Encourage pilot projects or 'policy sandboxes' that allow for low-risk testing of new approaches.
Create internal protections for innovative civil servants so failures are treated as learning opportunities, not career risks.
2. Strengthen Cross-Departmental Collaboration
Mandate interoperability standards for data systems across departments.
Fund cross-agency task forces to address complex, multi-dimensional challenges like climate, health, and housing.
3. Invest in Data Literacy and Critical Thinking
Embed data literacy into civil service training—not just technical skills but also bias awareness, ethical interpretation, and critical evaluation.
Include diverse expertise on policymaking teams: behavioral scientists, data analysts, domain experts, and communicators.
4. Prioritise Ethical Data Use and Public Trust
Develop and publicise transparent guidelines for collecting, storing, and applying data.
Engage citizens in how their data is used—building understanding and trust through plain-language communication.
5. Benchmark Globally, Act Locally
Use international models not as copy-paste solutions but as inspiration tailored to local political and institutional realities.
Create a structured approach to learning from countries with more substantial digital infrastructure and integrated policy systems.
6. Communicate with Honesty and Clarity
Communications and positioning are critical. Be upfront about uncertainties and trade-offs in data-led policy. Voters are more likely to support change when they understand the rationale.
Use storytelling to humanise data—show how real people benefit when better insights inform policies.
The Bottom Line: It’s Not Just About Data
If the UK and other governments want to deliver better services, smarter spending, and more substantial outcomes, they need more than dashboards and datasets. They need cultural change. They need institutions that reward learning, not just control. They need to empower people who can ask hard questions about the data—not just accept it at face value.
This doesn’t mean we abandon analytics or modeling. It means we ground them in a human-centered strategy supported by ethics, collaboration, and a willingness to evolve.
As Brent Hoberman’s question rightly implied, 'How good is the data?' isn’t the only thing we should be asking. We also need to ask:
Who is interpreting the data?
What assumptions are they bringing?
And are we creating the right conditions for the best insights to surface—and stick?
The future of effective policy isn’t just data-driven. It’s people-powered and not just about the technology!
