Summary
Five Just Stop Oil (JSO) supporters recently received four and five-year prison sentences for planning nonviolent protests on the M25 in November 2022. We conducted a nationally representative online poll of 944 adults via Prolific to gauge public reaction to the sentencing. We were interested in what the UK public thought about the sentences, with options ranging from "Far too harsh" to "Far too lenient".
Key Findings:
-
General Sentiment: The majority of the British public (59%) found the sentences too harsh. Only 14% felt the sentences were too lenient. The remainder found them proportionate.
-
Political Affiliation: The more right-leaning someone said they were, the more they tended to view the sentences as lenient compared to those who were more left-leaning.
-
Education Level: Those with higher levels of education had a greater probability of viewing the sentences as harsh.
-
Regional Differences: Opinion varied regionally across the UK, with London, Scotland, and Wales considering the sentences particularly harsh on average.
Contents
The British public generally finds the sentences too harsh
People’s political affiliation has a strong effect on how harsh vs. lenient they found the sentences
People’s education level further affects how harsh vs. lenient they found the sentences
Regional differences​
Introduction
​
On 18 July 2024, Southwark Crown Court sentenced five JSO supporters to four and five years in prison for their involvement in planning the M25 protests that disrupted traffic several days in a row in November 2022. These sentences are considered the longest ever handed out in the UK for non-violent civil disobedience.
These sentences were given in the context of an increasingly restrictive environment for activists in the UK. Last year, the Civicus Monitor, which tracks the state of civil society freedoms in 197 countries, downgraded the UK to ‘obstructed’. It argued that the UK was becoming an increasingly “hostile environment” for activists and civil society groups.
New legislation - the Policing Act and Public Order Act - has given police new powers to shut down protests and criminalise people for taking action. Climate activists have also been forbidden from presenting their reasons for protesting as part of evidence in court hearings; climate activists have not been allowed to mention the climate crisis or even historical examples of protest such as the US Civil Rights movement in court. Protestors who defied the ban on mentioning the climate crisis were jailed for seven weeks. These record prison sentences represent another step in the UK’s increasingly strict treatment of activists.
The sentencing was met with criticism from a range of sources. Michel Forst, UN Special Rapporteur on environmental defenders, described the judgements as a “dark day for peaceful environmental protest, the protection of environmental defenders and indeed anyone concerned with the exercise of their fundamental freedoms in the United Kingdom”. Human rights groups, including Human Rights Watch, have voiced similar concerns over the rulings. But how does the UK public feel about what is happening?
Responses by leading news outlets suggest that opinions on the sentences vary greatly. Progressive voices have highlighted the fact that sentences rival, and in some cases exceed, punishments for violent crimes, including robbery, and have raised concerns about an increasingly authoritarian approach to the fundamental democratic right to protest. At the other end of the political spectrum, conservative outlets described the JSO protestors as “fanatics” , arguing that the sentences were justified due to the scale of public disruption involved. The Daily Mail headline after the ruling (see their title page of July 19, 2024) proposed that Britons were united in supporting the sentences, declaring that the judge “spoke for all of us”.
The present poll sought to find out where the UK public stands on the record prison sentences for JSO campaigners - and, by extension, to get a sense of the public view on the broader role of disruptive protest in society.
Method
We conducted a nationally representative online poll (using the survey software Qualtrics) with 945 adults via Prolific (https://www.prolific.com/). We used Prolific’s nationally representative sample function to collect the data, making the sample representative of the UK population in terms of age, sex (as assigned at birth), and political affiliation (who people indicate they would vote for if there was an election tomorrow). We collected the following additional demographic data: region of residence, ethnicity, and education level. These were used to further adjust the results to be representative (for education, region of residence, and ethnicity) using the R package anesrake.
The wording that respondents read was:
On 18 July, Just Stop Oil activists received prison sentences of four and five years for coordinating non-violent direct action protests on the M25 motorway that disrupted traffic several days in a row.
Do you feel these prison sentences are:
-
Far too harsh
-
Harsh
-
Slightly harsh
-
In proportion to the offence
-
Slightly lenient
-
Lenient
-
Far too lenient
To provide an overview of opinions for the analysis, we collapsed the factors “Far too harsh”, “Harsh”, and “Slightly harsh” into one category labelled “Too harsh”. Similarly, “Far too lenient”, “Lenient”, and “Slightly lenient” were collapsed into one category labelled “Too lenient”. “In proportion to the offence” was renamed “Proportionate” for brevity.
In addition to assessing how the British public felt about the prison sentences, we investigated the extent to which people’s political affiliation and their level of education were linked with their opinions of the sentencing. This was done with Bayesian ordinal regression analyses using the R package brms. For both analyses, we ran four Markov chains with 10.000 samples (2000 for warm-up) and used mildly informative priors for the predictor variables, namely, a normal distribution around zero with a standard deviation of 5. We report the model estimates and 95% credible intervals, the Bayesian equivalent of 95% confidence intervals. The 95% credible interval is the range in which the true value lies with a probability of 95%, based on the observed data. Finally, we describe tentative regional differences across the UK and visualise them on a UK map.
​
​Results
The British public generally finds the sentences too harsh
Our initial results, published as a press release, were nationally representative for age, sex, and political affiliation, as provided by Prolific. Following up on these initial analyses, we used additional information on respondents’ ethnicity, education level, and region, to adjust the sample to be representative of these variables. We report the updated results here. They are very similar to the initial ones, with deviations of a maximum of three points (the percentage for “far too harsh” went down from 22% to 19%).
The weighted percentages suggest that 59% of the UK public found the sentences too harsh, 28% found them proportionate, and 14% found them too lenient.
Fig. 1. Overview of people’s opinion on the prison sentences. “Too harsh” encompasses respondents who selected “Far too harsh”, “Harsh”, and “Slightly harsh”. “Too lenient” encompasses respondents who selected “Far too lenient”, “Lenient”, and “Slightly lenient”. Rounded numbers are presented for clarity.
Below (in Fig. 2), we show the breakdown for all seven possible response levels. It reveals that people are quite evenly distributed over the three response options making up the “too harsh” and “too lenient” categories in the overview. The three response options reflecting that people considered the sentences too lenient were by far the least frequent.
Fig 2. A breakdown of weighted percentages showing the proportion of people who selected each option.
People’s political affiliation has a strong effect on how harsh vs. lenient they found the sentences
​
We assessed the extent to which someone’s position on the political spectrum (left to right) accounted for their opinion on the prison sentences. To do this, we had three UK-based team members and ChatGPT assign numbers from 1-7 to the different political parties, according to where each party lies on the political spectrum (1 = furthest to the left, 7 = furthest to the right, etc.). We used the average response across the four coders and ran a Bayesian ordinal regression to analyse the extent to which this left-to-right index predicts how harsh or lenient a given person considered the sentences. As expected, the results indicated that people more to the left side of the political spectrum tended to find the sentences harsher than people on the right. Specifically, the model estimates that each 1-unit increase in the left-to-right index is associated with 1.72 times greater odds of finding the sentences more lenient. This means that, if a person’s left-to-right index is one unit more conservative than another person’s, their odds of finding the sentences more lenient are 72% higher (95% Bayesian credible interval: [63% - 82%]).
Table 1. Average Likert scores and standard deviations (SD) per political affiliation. More left-leaning people tended to view the sentences as more harsh. People who said they would vote for the Conservative Party or for Reform UK were the only group that on average was on the ‘lenient’ half of the Likert scale, although the average is rather close to 4, representing sentences being in proportion to the offence.
People’s education level further affects how harsh vs. lenient they found the sentences
Next, we investigated whether people’s education level affected their responses, additionally to their political leanings. The corresponding survey question was:
What is the highest level of education you have completed?
-
Primary school
-
Secondary school up to 16 years
-
Higher or secondary or further education (A-levels, BTEC, etc.)
-
College or university
-
Post-graduate degree
Following YouGov’s methodology, the first two levels were categorised as “low”, the last two as “high”, and the middle option as “mid”. This categorical variable was then added to the Bayesian ordinal regression model. A first notable result was that the estimated effect of the political left-to-right index remained virtually unchanged, indicating that its effect is largely independent of education level. Second, education had an additional effect such that people with higher levels of education tended to find the sentences harsher than those with lower levels of education. Specifically, the chance of a person finding the sentences harsh are estimated as 55% (95% credible interval: [19% - 90%]) higher if they have high vs. low levels of education, and 34% higher [5% - 63%] if they have high vs. mid levels of education.
Regional differences
​
We investigated the extent to which opinions differed across regions. To summarise overall responses per region, we converted the response options to numbers (1-7; 1=”Far too harsh”, 7=”Far too lenient”) and calculated simple averages. The results should be viewed with the caveat that the survey was not optimised for regional representation and the sample sizes are rather small for some of the regions (especially Northern Ireland where we had just 7 respondents, see the table embedded in Fig. 3). The regions’ averages are between 2.43 and 3.36 where 2 corresponds to “harsh”, 3 to “slightly harsh”, and 4 to “proportionate”. Thus, despite regional variation, the results indicate that Britons across all regions tended to consider the sentences rather harsh on average. The data indicate that people in London, Scotland, and Wales tended to find the sentences harsher than people in other regions. People in England’s North West and East tended to find the sentences the most lenient. Northern Ireland was also among the regions where people tended to find the sentences relatively lenient but we only had 7 respondents from Northern Ireland, which is not enough to gauge overall opinions.
Fig 3. Average Likert scores per region. Darker/colder colours mean people in the region tended to find the sentences relatively harsh, whereas brighter/warmer colours mean people tended to find the sentences relatively less harsh. Note that these are simple averages for the entire region. It is possible that there is substantial variation within regions. Note also that for some regions the number of respondents was low (especially Northern Ireland).
Discussion
The recent record sentencing of JSO supporters involved in the planning of the M25 protests has sparked significant national and international debate. While media responses were very mixed, our poll reveals that a clear majority of the British public views the prison sentences as too harsh. This general sentiment indicates concern about the proportionality of harsh punishments for non-violent civil disobedience.
Given the unpopularity of JSO, we might have expected opinion to be largely in favour of their long prison sentences. This was not what we found. Less than a third of respondents found the sentences to be proportionate, while just 14% considered them too lenient, compared to 59% who considered them too harsh.
These findings contrast markedly with the Daily Mail’s suggestion that, when handing out the record sentences, the judge “spoke for all of us.” It highlights how far intuitions and media claims about people’s beliefs can diverge from many people’s actual beliefs. The degree to which many news outlets are wrong about the UK public’s opinion is striking and we speculate that this is an extreme example of the echo chamber effect, whereby people tend to surround themselves so much with like minded individuals that they develop the erroneous belief that everyone shares their opinions. Recent research on political communication on social media suggests that this effect might be particularly strong for those with right-leaning opinions.
Our analysis showed a strong correlation between political leanings and perceptions of the sentences. Right-leaning individuals tended to view the sentences as more lenient compared to left-leaning individuals, who mostly considered them too harsh. This finding fits well with previous evidence that conservatives are generally more in favour of stricter punishments than liberals.
Education level also plays a significant role in shaping opinions. Individuals with higher education levels tend to view the sentences as harsher. This correlation may reflect a greater acceptance of the necessity of civil disobedience in driving social and environmental change among more educated individuals.
Finally, tentative evidence on regional differences suggests that London, Scotland, and Wales showed particularly high levels of disapproval for the sentences. This can likely be attributed to these regions being more progressive, which is linked to general punitive attitudes, as well as attitudes towards protest, as the effect of people’s political leanings indicated.
The harsh sentences, combined with recent restrictive legislation, paint a concerning picture of the state of civil liberties in the UK. These developments raise questions about the balance between maintaining public order and protecting the fundamental right to protest. The findings of this poll suggest a significant portion of the UK public is uneasy about the severity of the sentences handed down to the JSO campaigners. As the UK navigates its response to climate change, climate activism and civil disobedience, these results suggest that the British public is generally supportive of protest as a viable and necessary form of political expression.
About Social Change Lab​
Social Change Lab conducts empirical research on disruptive protest and people-powered movements. Through research reports, workshops, and trainings, we provide actionable insights to help movements and funders be more effective. You can find all of our research projects and resources on our website. You can contact us at info@socialchange.lab.org
Appendix
​​
We averaged the Likert scores per region to assess regional differences. Generally speaking, looking at average Likert scores is not recommended because of two main reasons: 1) The same average score can arise from very different underlying distributions; an average of 4 can come from everybody choosing the fourth response option or from half of the people choosing the first and half choosing the seventh, etc. 2) The ordinal nature of Likert scales makes it such average numbers cannot be interpreted in the same way that interval data can, where 4.5 would be halfway between 4 and 5; with ordinal data, such as in the present study, 4.5 cannot be translated to saying people were halfway between “proportionate” and “slightly lenient”. We chose to show averages nevertheless to give an overview of regional differences. Below we show which distributions the averages arose from to address the two points above.
Fig. 4. Regional differences in the proportion of people who considered the sentencing too harsh, too lenient, or proportionate.