Most people have a sense of justice and preferences about the punishments for crimes. Sometimes, these preferences are strongly felt and widely agreed across a population. Almost everyone wants to see serious offenders imprisoned, and this consensus is unlikely to change anytime soon. More often, however, these preferences are not consistent enough to offer any guidance for the justice system. For example, there is little consensus among the public about how long a particular offender should spend behind bars.
This is the conclusion of a study by Cambridge PhD student Andrzej Uhl and US criminologist Justin Pickett, who collaborated to research punishment preferences in Europe. Their study, titled ‘The (In)Stability of Punishment Preferences: Implications for Empirical Desert’, was recently published in the British Journal of Criminology.
Uhl and Pickett decided to investigate punishment preferences in response to a lack of longitudinal research on this subject. This meant that no one had studied whether the individual preferences changed over time. Repeat surveys of public attitudes gathered new samples each year. However, they made no commitment to survey the same people, despite this providing a more accurate assessment of changing preferences.
“You have studies saying that on average, people recommend 10 years in prison for a given crime. And this is usually how results from surveys are presented,” Uhl explains. “But different types of variation that might affect how seriously we take this 10-year average. If 90% of people said 10 years, and some people said 9 years and some said 11 years, you take these results differently than if you found out that half of the people didn't want any punishment and half wanted 20 years in prison. The larger the variation, the more carefully we should approach these averages.”
“On top of that, there is individual inconsistency. The same person can recommend completely different sentences now and a year later – not because their attitudes have changed but because they lack a strong sense of what exactly the just punishment should be," he adds. “You would give less weight to these preferences if they turned out to change all the time.”
This study set out to investigate whether individuals were consistent in their punishment preferences over time. Uhl and Pickett used data from the Central European Social Survey, collected from Germany, Austria, Poland, Hungary, the Czech Republic, and Slovakia. All of these countries have different penal systems, which is most visible in their sentencing lengths. Some of them, such as Germany, have relied on legal scholars and experts when designing their penal laws. Others, such as Uhl’s home country of Poland, have prioritised public opinion on punishment- or at least what politicians perceive to be public opinion. Uhl warns that this can lead to ‘penal populism', wherein the policies implemented are not necessarily the most effective at reducing crime."
The survey interviewed people about their preferred punishment for a variety of crimes, from child support evasion to stranger rape. They were asked to choose from a set of potential punishments, from a fine to an unsuspended prison sentence. Those who chose the latter were asked to provide the appropriate sentence length. The same people were then asked these same questions one year later.
The study found that people very rarely clung to their punishment preferences. For one crime, some first chose a fine and then a community service. If they were consistent enough to choose imprisonment twice, they often still suggested a very different prison term. Sometimes their preference for how long the hypothetical offender should remain imprisoned changed by several months or years.
However, these volatile individual preferences, when averaged across the population, do not change much from one year to another. Pickett attributed this to the so-called 'miracle of aggregation'.
“Individual people change a lot, but it somehow averages out," Uhl explains. “Yours or my opinion might be very unstable, but unless something important happens in society, the average mood tends to stay the same.”
The study also argues that people might have a broad range of punishments they consider acceptable for each crime. For example, somebody might recommend a sentence of 10 years for assault, but their sense of justice would not be offended if that assault was sentenced to 8 or 12 years. Sometimes it might be better to study what is acceptable for the public instead of searching for the ‘ideal’ sentence, especially if people quickly change their minds about what it is.
Pickett suggested that these specific punishment recommendations can change based on trivial issues in people’s lives, such as the weather, a recently watched film, or a general bad mood. These small things were already known to influence judges, leading to different sentences for identical crimes. In their next project, Uhl and Pickett will look into what happens when hundreds of judges receive an identical court file to sentence. They will research whether judges are more consistent than the public in choosing punishments.
Uhl and Pickett also observed the other factors that correlate with changing punishment preferences. One of the most notable is the correlation between level of education and consistent attitudes. Those with less education were much more likely to change their punishment recommendations after a year. Uhl points to criminological studies around public perceptions in the UK, which show how these perceptions change with access to knowledge.
“They showed that if you really give people all the details about some cases, they sentence these cases very similarly to the British judges,” he says. “It's only with very great questions like ‘should criminals be punished more severely?’ that people always say yes. But it's because they have no idea how criminals are punished right now."
The frequent changes in individual attitudes led Uhl and Pickett to conclude that there is not enough clarity on attitudes in general. Because of this, he argued against translating public preferences straight into criminal policy. Instead, he proposes that lawmakers should use the public attitudes only as a starting point. How much weight is given to that starting point could depend on whether the opinions are consistent and stable. But ultimately, decisions should be informed by the best available evidence.