The more oppositional the tech, the harder it is to gain govt approval: mySociety CEO Louise Crow

Louise also serves as a non-executive director at Connected by Data, a non-profit organisation that campaigns for communities to have a powerful voice in decisions about data.

Before joining mySociety, Louise worked in digital service design and development across a range of organisations, including digital and biotech startups, academia, and the UK Parliament.

She holds a PhD in Artificial Intelligence, a master’s degree in IT, and a bachelor’s degree in psychology, all from the University of Nottingham.

In an interview with indianexpress.com. Louise spoke about mySociety’s work, the challenges of building civic tech products, success stories, and new trends in civic tech products. Edited excerpts:

Venkatesh Kannaiah: Tell us about mySociety and the work you do.

Louise Crow: mySociety was founded more than 20 years ago to build web applications to give people simple, tangible benefits in the same way that e-commerce sites were starting to do, but in the civic and democratic parts of their lives.

This subsequently became known as civic tech. We use digital expertise and human-centred design approaches to build open, usable services that help people inform themselves, take part in decision-making, and work to improve their lives and the communities where they live.

We are based in the UK, and our services reach millions of people each year. We found that activists and civic tech enthusiasts around the world had been inspired by us and had set up similar sites in around 87 countries, using our reusable code.

We use open-source code by default, plan for customisation, and write documentation so that, in practice, other people can run services built with our code in countries around the world.

We are a pioneer in civic tech, and as a result, we’ve also tried to take on a convening role. TICTeC, our flagship event, brings together a global community of technologists, researchers, funders, and government practitioners to research and build civic and pro-democracy tech.

Venkatesh Kannaiah: Give us details of your products and an idea of your impact.

Louise Crow: The tools and applications that we have built focus on practical issues that introduce people to civic life. They cover everything from removing garbage in your cities to housing issues, and all the different aspects of life managed by public institutions and covered by right to information laws. Our tools cover themes like voting, representation in Parliament, what MPs and other representatives are saying, and how they’re voting.

FixMyStreet.com is an online platform that gives people a simple way to report issues like garbage collection delay or potholes in their local area, without them needing to know who’s responsible. They simply mark the location on a map-based interface and describe the problem. The application then figures out who’s responsible for fixing it and sends them the complaint. We are now working with the city and local governments in the UK to make FixMyStreet their official reporting tool.

WhatDoTheyKnow.com is an online platform that helps people make right to information requests to public authorities, which then automatically publishes the responses online so that everyone can benefit from the information. The code that runs WhatDoTheyKnow also supports more than 20 similar initiatives around the world.

Not only citizens, but journalists and civil society organisations use our tools to bring out information, and these have led to investigative reports in leading publications.

TheyWorkForYou.com was one of our very first civic tech sites. It helped define what we now call the parliamentary monitoring category and was one of the earliest parliamentary monitoring tools.

These tools help people find out what their MPs in the UK are saying in Parliament and what they’re actually voting on.

For building this platform, we had to challenge parliamentary copyright to produce a navigable, accessible version of the record, rather than relying solely on the official one. Over time, that contributed to the creation of the Open Government Licence, one of the first licences for the open reuse of government content. It’s now widely used for government and parliamentary data.

We also have a service called WriteToThem, which helps people identify their representatives at different levels of government. Users enter their postcode — the unique identifier for where they live — and the service tells them who their representatives are and what they’re responsible for.

We’ve also worked over the last few years on a collaborative project with a partner called Climate Emergency UK. Together, we’ve created what is essentially a scorecard of local government climate action. This looks into what the councils have committed to in terms of climate action and what they have actually done every 18 months to meet those commitments.

Venkatesh Kannaiah: How are your products integrated into government redressal systems?

Louise Crow: When we began, all complaints were sent by email. But as we’ve started working with local governments, we’ve integrated with their back-end systems. So for the local governments we’re actively working with, reports can go directly into their asset management and other systems, allowing them to send people out quickly to fix the issue. There are more than 30 cities, including London, that use these tools.

Venkatesh Kannaiah: Tell us about the challenges in building civic tech projects?

Louise Crow: If you’re building services that are essentially civic infrastructure, the challenge is to fund them sustainably. That’s why I’m really interested to see the growing recognition of digital public goods and Digital Public Infrastructure (DPI).

DPI is a major topic in India. The idea is that these are digital services that are available to everyone and don’t diminish in value as more people gain access to them. In some ways, they become more valuable the more widely they’re used. I’d like to see more civic applications and civic technology recognised and supported as public goods.

We as technologists like the idea that open-source code can be reused in different places. But you have to understand the local context and adapt it accordingly. You might build a parliamentary monitoring site that’s centered on debates, only to discover that most of the real work happens in committees. Or you might build an access-to-information site, but find there’s such a strong culture of institutional silence that no one responds to information requests. Really understanding how things work in practice is essential, and it can be quite challenging.

Venkatesh Kannaiah: Why has civic tech as a category not taken off? Or am I wrong?

Louise Crow: People are often thinking of that optimistic period in the early 2000s, when everyone was excited about the potential of the internet, and civic tech was part of that energy — it was about empowering people in relation to the government and seeing technology as a positive, democratising force.

That sense of optimism has been tempered now. There’s a greater understanding that technology on its own can’t bring about significant change. You need other enabling conditions as well.

I also think that, on the global stage, in the civic and democratic arena, it’s clear that anti-democratic actors aren’t standing still. They’re using technology in ways that reinforce power.

We have seen the rise of the social media ecosystem, which has been incredibly powerful. But it hasn’t really solved the problem of helping people come together in a constructive, democratic way.

The question is how we make that a positive impact, and how we design for that positive impact. I don’t think it’s because of a lack of ideas. The fact that civic tech remains a niche reflects the priority societies give to democratic health and democratic participation.

Venkatesh Kannaiah: Tell us about successful global civic tech projects.

Louise Crow: There is Polis. It’s a platform for gathering, analysing, and understanding what large groups of people think, in their own words. One of the reasons it’s been successful is that it incorporates an innovative, mathematical way of identifying bridging statements that gain agreement from groups that otherwise disagree. In other words, people who are quite opposed to each other on a particular topic. But what do they actually agree on? I think that’s a fundamental democratic question. Polis started as part of a Computational Democracy Project in the US and was used extensively in Taiwan as part of the vTaiwan process. The vTaiwan process is an open, technology-enabled public consultation model developed in Taiwan to help citizens, experts, businesses, and government agencies deliberate on complex policy issues.

There is the Tree Company in Belgium. They work on voting advice applications — tools that ask people a series of questions and then match their positions with those of political parties to help them make informed choices. They’ve been very successful in Belgium.

There is Ushahidi, a crowdsourcing platform. One of the reasons it’s been successful is due to its flexibility. It was initially used for election monitoring, but has been adapted for flood response and for gathering community needs during the COVID pandemic.

Venkatesh Kannaiah: Tell us about civic tech projects that have failed.

Louise Crow: In my experience, the common failure is when somebody has an idea they think is brilliant, but there isn’t a real need for it. Either there just isn’t an audience, or the need they are trying to address is much more complex than they are anticipating with a technology-first approach.

You have to talk to people about what they want, what they need, how they work, and how they solve problems to build a project that actually meets real needs.

The other relates to the challenge of finding a sustainable funding model. If a service is helping a large number of individuals, but each person receives only a relatively small benefit, how do you capture enough value to keep improving and maintaining that service? I think those are real challenges.

Venkatesh Kannaiah: Tell us about civic tech startups that you are impressed with.

Louise Crow: Working with startups is not part of our day-to-day work, and we come across them mostly through our TICTeC events. I’m impressed by people who start projects in challenging democratic environments.

There is We Visualize Data for Democracy (WeVis), a Thai project operating in a difficult democratic environment.

They have a really systematic approach to learning from projects around the world. They’ve built a first-class parliamentary monitoring tool by looking at similar projects elsewhere.

Venkatesh Kannaiah: From Civic tech to Govtech. Why is the adoption so slow?

Louise Crow: The more oppositional a technology is, the harder it is for that technology to gain institutional support.

One is that governments are inherently slow to change. As institutions, they are designed to be the operating system of a country, and that is supposed to be stable rather than innovative. So they’re perhaps not natural early adopters of new technologies.

Often, the civic technologies that governments adopt are related to participation. In many places, participation in early-stage decision-making is still seen as something of an optional extra — a nice-to-have rather than a core function.

Where we’ve seen significant adoption of different approaches, such as in Taiwan through the vTaiwan process, it’s often been in conjunction with a broader crisis. So I think that if you want rapid adoption of something fundamentally different, there sometimes needs to be a broader sense of crisis, a feeling that what the government has been doing is no longer working.

Venkatesh Kannaiah: Where does AI fit in the civic tech mix?

Louise Crow: We are looking at the use of AI because many of our services contain large amounts of text. Our focus is more on analysis, clustering, and understanding than on content generation. For example, it would be very valuable in the context of access to information to be able to identify that, over the last week, people asked questions on a particular topic but didn’t receive answers. Because AI can cluster related pieces of text based on their meaning, I think that’s a particularly interesting application for us.

I think navigation and search are inherently easier applications. The generative side is more challenging because it’s harder to check and verify the output.