In conversation with Dr Sofia Collignon: local elections, political violence, and quantitative methods

Welcome to our Dods Research interview series. This is where we talk with leading academics in the UK about their area of expertise and what they can tell us about UK politics. As part of our Dods Research collaboration with Queen Mary University of London, our second interview features Dr Sofia Collignon, Senior Lecturer in Comparative Politics and Director of Mile End Institute, and it covers topics including local elections, political violence, and how data can be used to inform public policy.

Joshua Wells: We’re speaking on 2nd May, with some of the local election results already in. Can you share your initial reaction to what we’re seeing?

Sofia Collignon: One interesting thing about local elections is that political scientists often refer to them as second-order elections. This means they typically don’t weigh as heavily on voters’ minds as first-order elections, such as general elections.

Because of that, voters often treat local elections as a sort of test run. They feel freer to support parties they might not back in a general election. It’s an opportunity to experiment a bit—to express alternative political preferences without the perceived higher stakes of a national vote.

So, when we see strong performances from parties like Reform, we should be cautious about projecting those results onto a future general election. For one, the general election is still some time away, and a lot can change. But more importantly, the context is different—voters might be more willing to give smaller or non-traditional parties a chance locally than they would nationally.

Joshua Wells: Given that, how should the Labour and Conservative parties interpret the outcome of these local elections?

Sofia Collignon: They can use these results as a diagnostic tool. Second-order elections offer insight into what issues are resonating with voters right now—what’s working and what isn’t. The Conservative Party, for example, suffered significant losses. That points to clear dissatisfaction. It suggests they haven’t yet managed to redirect public perception or restore trust in a way that convinces voters they’re offering meaningful change.

For Labour, it’s a chance to reflect on how their current messaging and policies are landing with the public. Even though these elections are lower stakes, they still reveal patterns. Are people responding positively to Labour’s economic message? Is there a clear break from the austerity years under the Conservatives? Or is the public still waiting for policies that feel truly different?

So, both parties should see this as an opportunity to listen closely—what are the current public concerns? What narratives are cutting through? What are the challenges they’ll face going into the next general election? It’s a moment for strategic reflection as much as political assessment.

Joshua Wells: One of your main areas of expertise is political violence. Looking at the UK, there’s a growing sense that incidents have increased in recent years. Is it fair to say we’re living through a period of heightened political violence?

Sofia Collignon: There are two parallel things happening. On the one hand, we’re becoming far more aware of what political violence in the UK actually looks like—what forms it takes, who is affected, how widespread it is, and what the specific challenges are. Since the tragic murder of MP Jo Cox, we’ve seen more politicians openly discussing their experiences with abuse and threats. That’s helped us define and document these issues more clearly.

In 2017, we launched a candidate survey during the general election, where we contacted every candidate standing. It included questions about their background, campaign strategy, and—for the first time—a detailed section on political violence and intimidation. When we repeated the survey in 2019, the number of candidates reporting abuse, harassment, or intimidation had increased significantly. Notably, this rise was steeper for women candidates than for men.

Now in 2024, we’re still collecting data, so it’s too early to provide definitive figures. But what we can say is that violence—particularly on social media—is widespread. We’re also seeing more in-person incidents: candidates being followed, shouted at, and increasingly feeling unsafe.

That’s important because we risk normalising this environment. Candidates are beginning to expect abuse as part of the job. And because women, ethnic minorities, and other underrepresented groups tend to be targeted more, their experience of running for office is often more hostile. That undermines the principle that everyone should have the right to participate in politics peacefully and equally. Despite various initiatives, the proportion of candidates experiencing abuse does not appear to be decreasing. That’s deeply concerning.

Joshua Wells: You’ve spoken about how violence disproportionately affects minorities, particularly women and ethnic minorities. But is it also skewed by political party—do some parties experience more of it than others?

Sofia Collignon: At the moment, we’re seeing that candidates from all parties report experiencing violence. What differs is the level of support they receive from within their party. One worrying pattern is that many candidates report the abuse they suffer comes from supporters of other parties. That’s significant—because political parties should be able to hold their own supporters accountable and take action when those supporters engage in abusive behaviour.

Anecdotally, some candidates say that experiencing violence puts them off standing again. But when we look at the data, we don’t find a strong correlation between experiencing violence and choosing not to stand in future elections. What does seem to matter is a candidate’s sense of safety. If they feel less safe—regardless of whether they’ve been directly targeted—they’re more likely to step away from politics. So the effect isn’t just personal, it’s systemic. It creates a toxic political environment that deters participation more broadly.

Joshua Wells: So, political violence narrows the pool of people willing to stand. You’ve mentioned that much of the abuse comes from party supporters. Has there been any cross-party effort to address this issue?

Sofia Collignon: Yes, there’s been some very important work. The Jo Cox Foundation and the Local Government Association have been instrumental in encouraging parties to adopt codes of conduct—to formally commit to respectful behaviour and to develop internal mechanisms for dealing with abuse. We’re also seeing movement from the Speaker’s Office. In February this year, a Speaker’s Conference was held, bringing together experts to examine the scale and nature of political violence. That’s a strong sign that the issue is now firmly on the political agenda.

I think the UK is moving in the right direction. NGOs are pushing for reform, political actors are increasingly on board, and institutional responses are starting to take shape. That said, we still face major challenges. One of the most pressing is the lack of a centralised system for recording abuse. While we have surveys—like the candidate survey and those directed at MPs—not all abusive behaviour is criminal, even if it’s deeply damaging. Much of it goes unreported. We need a proper national database to track this kind of behaviour. Without it, we can’t evaluate whether the measures being taken are actually working. That’s a gap we urgently need to fill.

Joshua Wells: This invites a broader question, is political violence against elected officials simply inevitable?

Sofia Collignon: No, I don’t think it’s inevitable. We do need to distinguish between passionate, even heated political disagreement—which is part of a healthy democracy—and harassment or intimidation, which are not. Some people argue that freedom of expression gives them the right to say whatever they like, but there are legal limits. The UK’s Protection from Harassment Act defines harassment as any behaviour that is repetitive, offensive, and threatening to another person’s wellbeing. Once you cross that line, it’s no longer just robust debate—it’s abuse.

What’s worrying is that when political discourse turns personal—when the focus shifts from policies to individuals, with the intent to insult or threaten—we degrade the quality of our political debate. Social media compounds this problem by amplifying harmful rhetoric and spreading misinformation. Political leaders and institutions have a huge role to play in reversing this trend. They need to communicate clearly that democratic engagement should be respected, not attacked.

Joshua Wells: That really resonates—but let me push a little. You’ve emphasised focusing on policy, not personal attacks. But what about situations like Partygate? People weren’t outraged by a particular policy, but by the perceived hypocrisy of the Prime Minister at the time. What should we do if voters are concerned about individual behaviour as opposed to policy?

Sofia Collignon: Absolutely—when scandals like Partygate emerge, public frustration is natural and justified. Holding politicians accountable is crucial, and it’s fair for people to express anger when leaders fail to live up to the standards they set for others. But accountability doesn’t require abuse. We can demand answers, call for resignations, or seek legal and institutional redress—without resorting to threats or insults.

There are proper mechanisms for dealing with misconduct: codes of conduct, parliamentary oversight, legal procedures. Citizens should be empowered to pressure institutions to act, and institutions should respond in a way that rebuilds public trust. So yes—criticise behaviour, demand accountability, engage vigorously. But let’s do that in a way that strengthens democracy, rather than poisoning it.

Joshua Wells: You’re an expert in quantitative methods, so let’s imagine I’m trying to run a public policy campaign. What role could quantitative methods play in helping me gather information to influence policy?

Sofia Collignon: If you’re planning a campaign, I’d always begin by encouraging you to define the problem clearly. It’s important to ask how the issue is currently being framed and how you would define it differently or more precisely.

Joshua Wells: Alright—let’s imagine my campaign. I want every school to have a chessboard. The problem is, not every school has access to one.

Sofia Collignon: That immediately invites two questions. First, why do you believe every school should have a chessboard? Second, what’s the deeper issue this campaign is trying to address? For instance, if your aim is to promote strategic thinking among schoolchildren, then the chessboard is not the end goal—it’s the policy tool. The real outcome you’re pursuing is the development of strategic thinking.

Once that’s established, we need to think carefully about how to measure this outcome. Does strategic thinking manifest through mathematical ability, problem-solving skills, comprehension, or systems-level thinking? Defining this is crucial because one of the main strengths of quantitative methods is their ability to take complex or abstract ideas and translate them into something observable and measurable. That lets you build a study to track whether the intervention—placing chessboards in schools—actually makes a difference. You could examine data across different schools or regions, comparing those that implemented the programme with those that didn’t. In doing so, you start to uncover whether the chessboard policy has any tangible effect on the outcome you’re most interested in.

Quantitative research allows you to take a structured approach. It helps you define your goals, determine what is measurable, and think logically about cause and effect. It allows you to trace changes over time and to assess the effectiveness of a policy. All of this makes your campaign far more persuasive to policymakers, who are often looking for evidence of impact and feasibility.

Joshua Wells: That’s really helpful. I want to draw attention to something you’ve said that I think is crucial—if you’re trying to persuade a policymaker, you need to show them what success looks like. And if you can quantify that success, you’re probably making your proposal more persuasive. Do you have any examples of unusual or interesting things we can measure? It’s easy to imagine measuring educational outcomes or economic growth, but beyond these headline metrics, it’s much harder to know what’s actually measurable.

Sofia Collignon: It’s a great question, and one that’s central to the philosophy of quantitative research. Many researchers believe that if a question is properly formulated, then almost anything can be measured. Of course, that’s a big claim—and it comes with a host of challenges. Take something like happiness. How do you measure that? Or how do you assess satisfaction with a public policy? There’s a lot of debate around these issues, but the field has developed tools like large-scale surveys, batteries of related questions, and composite indexes that help approximate these abstract concepts.

Inequality, for example, is notoriously difficult to pin down, but economists and social scientists have developed standardised tools to do so. The Gini index is now a widely accepted way of capturing income inequality within a population. Similarly, measuring development is not straightforward. Some argue it’s purely about GDP, while others insist it should include wellbeing or life opportunities. The Human Development Index is an attempt to capture a more holistic view of development, using indicators like life expectancy, education, and income. It isn’t perfect—no measure ever is—but it’s solid, it’s been validated, and it gives us a common benchmark for comparing progress.

So even with abstract or contested concepts, there are thoughtful and evidence-based ways to measure them. The key is to work through a process of careful definition, standardisation, and validation so that what you’re measuring is both meaningful and comparable.

Joshua Wells: So if I brought my chessboard campaign to you, you’d begin by asking me to define the problem, not just the solution?

Sofia Collignon: Exactly. I would encourage you to start by thinking about the actual problem you’re trying to solve. If your concern is a lack of opportunities for children to develop strategic thinking, then that’s what you need to focus on first. The chessboard is a potential mechanism—but it only makes sense as a policy intervention if we can link it to a clear and measurable outcome.

That’s where quantitative methods come into play. They allow us to translate that strategic thinking into something observable, so we can evaluate whether your policy is effective. It might seem like a simple idea—every school should have a chessboard—but once you start thinking in terms of measurement, outcomes, and impact, you realise there’s a more sophisticated process involved. And that’s precisely where the value of quantitative research lies.