Subscribe to @gov

At gov graphic

Register an account and opt-in for Government & Public Sector insights.

Citizens like the idea of digital public services - but they are not as keen on sharing the personal data that governments need to deliver them. A 2017 study by the UK Information Commissioner's Office1 found that only 49% of Britons trusted national government departments and organizations to store their personal data. Other surveys - in Australia2 and the US3- have identified similar levels of distrust.

Headlines about the loss, misuse and inaccuracy of data have undermined public trust. The growing use of artificial intelligence (AI) and machine-learning has also fueled concerns about complex algorithms and big data producing unfair or biased outcomes. Fundamentally, the idea of 'trust' between citizen and state is much more complex than, for example, that between customer and retailer.

As governments seek to introduce more tech-enabled services, some are doing more to reassure the public about their management of data & analytics (D&A). New controls, processes and standards- and greater transparency about the way data is used- will all help build trust.

The most effective step governments can take, says Viral Chawda, Principal- Innovation & Enterprise Solutions, KPMG in the US, is to prove to citizens why their data matters. “There has to be a process of give and take. We have seen from the rise of social media that people will share their data when they can see a clear value in it. So governments need to ensure that citizens understand how they can use their data to help improve their lives- which can be anything from proactively managing traffic flows to safer street lighting and smarter energy use.”

Chawda's analysis is echoed by the Organization for Economic Co-operation and Development (OECD) in its 2017 report Embracing Innovation in Government4, which stressed the need for governments to be “transparent about the data they collect, and clearly demonstrate the resulting value of the resulting products”. 

KPMG International's 2018 report, Guardians of trust5, identified that trust in data and analytics is founded on four key anchors:

  1. Quality. Are the data and analytic models good enough?
  2. Effectiveness. Do the analytics deliver the desired results?
  3. Integrity. Is the use of D&A ethical - and legal?
  4. Resilience. How well are governance, security and accuracy of data managed for the long term?

The trust deficit

The Guardians of trust report, which surveyed global IT and business decision makers6, found that just 31 percent of those in the government sector have a high degree of trust in their own organization's use of analytics- and only 40 percent believed that their organization is completely transparent with customers about the data they hold and how they use it.

Transparency becomes especially critical as governments apply AI in analytics, machine-learning models and automated decision-making. Headlines such as “The rise of the racist robots” underline the importance of ensuring that outcomes are fair and align with social and cultural values. For example, one algorithm in Florida used by courts developed a bias - predicting that black criminals were twice as likely to re-offend as white criminals7.

“Organizations must be transparent not just about outcomes,” says Chawda, “but also about what went into the algorithm that produced the outcome.” The OECD's Embracing Innovation in Government report recommended that algorithms used in the public sector to automate decision-making should be opened up. Keen to make France's taxation regulations as transparent as possible, the government has recently opened up the source code it uses to calculate tax.

Ethics boards can also help. Technology companies in conjunction with organizations such as the Centre for Internet & Society India, Digital Asia Hub and UNICEF have created the Partnership on Artificial Intelligence to Benefit People and Society. The UK's Government Digital Service has also published an ethics framework for public sector data science8.

Regulations and expectations

Public sector organizations must also cope with a rapidly evolving regulatory framework. The European Union's General Data Protection Regulation (GDPR) requires organizations to be transparent about how they collect data, what they do with it and how they process it. Yet many citizens have not grasped what such a high-profile change means - a recent Civica survey found that 88 percent of Britons didn't even know what GDPR is9.

The real issue for governments, Chawda says, is to look beyond regulatory compliance to understand customer expectations and then meet or beat them. To do that, responsibility for D&A must be shared throughout the organization. “Technology is a key element in the public sector's use of data and analytics but it is not the only element. Organizations need to ensure that what they're doing is aligned to their strategic goals, so everything cannot be left to the CIO.”

Governments can construct a more inspirational media narrative by publicizing their success stories. The Wellbeing Project, in Santa Monica, California, has drawn on multiple sources of data to construct a Wellbeing Index to understand community needs and make decisions that can improve citizens' lives10. The city council of Adelaide, based in the driest state (South Australia) in the world's driest country (Australia), is trialing the use of smart sensors to collect the data to manage its water more effectively11.

The human factor

If government organizations are to win the public's trust to use D&A more widely, they need, Chawda says, to start winning the war for tech talent. “There is no magic bullet for this. The sheer volume of demand for data scientists and data engineers means that you can't hire your way out of the problem but you can, through a combination of up-training and the application of technology, especially AI, free your staff to focus on more high value tasks.”

So, for example, he says: “The human mind can only perceive a relationship with another party - say a supplier - in five or six ways. With AI, you can analyze contracts in thousands of ways and help, for example, avoid improper payments and prevent fraud. It can even help an agency which has different contracting options with a single supplier ensure it takes the best option.” Delegating such tasks to AI can enable organizations to use their human expertise where it can make the greatest difference.

The public sector may not be able to match Silicon Valley for salaries or perks but it does, Chawda says, have one undeniable advantage. “Millennials will soon dominate the workforce and we know that one of the traits that distinguishes the best and the brightest of them is their desire to make a difference to society.

“They want to work for organizations that give them that sense of a higher purpose. Governments are using data science to tackle real life problems and to better support society's most vulnerable people. Many millennials will be inspired and motivated by that,” he adds.

Persuading citizens to trust governments to use their data to deliver digital public services is no simple task. Yet it could become easier as organizations get better at governance and process, and begin to make a tangible difference to people's lives. The good news, Chawda says, is that the public sector has only just started on this journey. “Right now, even with existing technology, we have only achieved a fraction of what can be done. That leaves a lot of room for opportunity.”

Visit www.kpmg.com/guardiansoftrust to learn more about KPMG's Guardians of trust survey.

Register now to receive Government & Public Sector updates.

Footnotes

1ICO survey shows most UK citizens don't trust organisations with their data, ICO, 6 November 2017.

2Australians don't trust government or telcos to protect their data: survey, Sydney Morning Herald, 1 July 2015.

3Americans and Cybersecurity, Pew Research Center, 26 January 2017.

4Embracing Innovation in Government (PDF 1.4 MB), OECD, February 2017.

5Guardians of trust, KPMG International, February 2018.

6Survey source: A commissioned study conducted by Forrester Consulting on behalf of KPMG International, July 2017.

7Even algorithms are biased against black men, The Guardian, 26 June 2016.

8Government's big data dilemma - building public trust during a data science skills crisis, Diginomica, 22 February 217.

9Do citizens trust the government to handle their personal data?, Digital By Default News, 17 January 2018.

10Embracing Innovation in Government (PDF 1.2 MB), OECD, February 2017.

11Managing water supply with smart technology, Utility, 24 April 2017.

Connect with us