Does Shared Identity Between Defendants and Jurors Influence Court Decisions?

This study investigates whether shared identity between defendants and jurors influences court decisions. Using data from the Stockholm District Court (2000–2004), researchers analyze demographic and socioeconomic similarities between defendants and randomly assigned jurors. Findings indicate that defendants are 4–6% less likely to receive a prison sentence when at least one juror shares their identity-forming group. Additionally, shared identity reduces prison sentence length by 10%. Both socioeconomic (education, income) and demographic (gender, ethnicity) similarities contribute to these biases, particularly in longer trials and when defendants are present, suggesting in-group favoritism impacts high-stakes judicial decision-making.

Full Citation and Link to Article

Ahrsjö, Ulrika, Susan Niknami, and Mårten Palme. 2024. "Identity in Court Decision-Making." American Economic Journal: Economic Policy 16(4): 142–164. https://doi.org/10.1257/pol.20220802

Extended Summary

Central Research Question

The research article "Identity in Court Decision-Making" by Ulrika Ahrsjö, Susan Niknami, and Mårten Palme explores the role of identity in high-stakes judicial decisions. Specifically, the study investigates whether shared identity between defendants and jurors influences trial outcomes in Sweden. The central question is: Does a defendant receive different legal outcomes—such as conviction likelihood and sentencing severity—when they share demographic or socioeconomic characteristics with their jurors?

The study contributes to a growing body of literature examining social biases in economic and legal decision-making. By leveraging the random assignment of jurors to cases in Sweden’s judicial system, the researchers aim to determine whether in-group favoritism—bias towards individuals who share certain identity markers—affects verdicts and sentencing.

Previous Literature

This research builds on several strands of existing literature. Social identity theory, as first conceptualized by Akerlof and Kranton (2000), suggests that individuals categorize themselves and others into identity groups based on social, demographic, and economic factors. These identities can influence economic behaviors, workplace dynamics, and even political decision-making.

Prior studies on judicial decision-making have largely focused on race and gender biases. For example, studies in the United States have documented racial disparities in sentencing, showing that Black defendants often receive harsher sentences, particularly when judged by predominantly White juries (Anwar, Bayer, and Hjalmarsson, 2012; Lee, 2017). Other research has examined gender bias in legal rulings, with mixed results. Some studies find that male judges sentence female defendants more leniently, while others suggest no significant gender-based sentencing disparities.

However, most prior studies have focused primarily on observable demographic traits such as race and gender while neglecting the role of socioeconomic status (education and income). The authors argue that socioeconomic identity is equally important in shaping decision-making but has been underexplored in legal contexts. By considering a broader range of identity dimensions—including education and income—the study seeks to provide a more comprehensive understanding of how identity influences judicial outcomes.

Data

The study examines trial data from the Stockholm District Court between 2000 and 2004. The dataset consists of 13,217 case-defendant observations, though the final analysis sample is limited to 12,320 cases that include three-juror panels. The key advantage of this dataset is that jurors are randomly assigned to cases, making it possible to isolate the causal impact of shared identity on legal outcomes.

The researchers compile data from multiple sources, including:

1. Court Transcripts – Contain details on defendants, judges, and jurors, as well as trial outcomes, including conviction and sentencing decisions.

2. Swedish Population Registers – Provide detailed demographic and socioeconomic information on both jurors and defendants, including gender, age, nationality, education level, and income.

3. Criminal Records – Offer historical data on defendants’ past convictions, allowing for controls based on prior criminal behavior.

Using these sources, the researchers construct measures of identity based on four key characteristics:

• Gender

• Ethnic background (native vs. nonnative)

• Education level (less than high school, high school, postsecondary)

• Income rank (relative position in Sweden’s income distribution)

These identity traits are used to examine whether a defendant receives different trial outcomes when their jurors share these characteristics.

Methods

The study employs an empirical strategy based on the random assignment of jurors. The key assumption is that, because jurors are randomly selected from a pool, any observed identity effects on trial outcomes are not driven by selection bias.

The authors estimate a series of regression models to measure the impact of identity similarity on trial outcomes. The primary equation models the probability of conviction, likelihood of a prison sentence, and sentence length as functions of identity similarity between the defendant and jurors. The main independent variables capture the proportion of jurors in a three-person panel who share each identity characteristic with the defendant.

The regression model controls for:

• Defendant characteristics (age, gender, prior convictions, education, income)

• Crime severity and type (e.g., violent crime, property crime, drug offenses)

• Juror characteristics (age, gender, education, income)

• Fixed effects for court division and trial period

The study also examines whether the effects differ by trial characteristics, such as whether the defendant was present in court and the duration of the trial.

Findings/Size Effects

The key findings suggest that shared identity between defendants and jurors significantly affects sentencing outcomes, particularly for incarceration rates and sentence length.

1. Reduced Likelihood of Incarceration

   • When one additional juror shares a defendant’s gender, the likelihood of a prison sentence decreases by 1.5 percentage points (7% reduction from the mean).

   • When one additional juror shares a defendant’s ethnic background, the probability of a prison sentence decreases by 1.2 percentage points (5% reduction).

   • When one additional juror shares a defendant’s education level, the probability of a prison sentence decreases by 0.8 percentage points (4% reduction).

   • When one additional juror has a similar income level, the likelihood of a prison sentence decreases by 1 percentage point (4% reduction).

2. Shorter Prison Sentences

   • When jurors share demographic characteristics (gender, ethnicity), defendants receive sentences that are 10% shorter on average.

   • When jurors share socioeconomic characteristics (education, income), sentences are also 10% shorter.

3. Effects Stronger in Long Trials and When Defendants Are Present

   • Identity effects are stronger in trials that last longer than the median duration (45 minutes).

   • The effects are most pronounced when the defendant is physically present in court, suggesting that in-group bias is heightened when jurors directly interact with defendants.

4. Demographic vs. Socioeconomic Identity

   • When combining the identity measures into two composite indices (demographic and socioeconomic identity), the magnitude of their effects is similar, indicating that socioeconomic identity is just as influential as gender or ethnicity in judicial decisions.

     
     

Conclusion

The findings provide strong evidence that identity-based biases influence high-stakes decision-making in the courtroom. Even in a system where jurors are randomly assigned, defendants are significantly more likely to receive lenient sentences when they share demographic or socioeconomic traits with their jurors.

The study has broad implications for legal systems worldwide. It highlights that in-group favoritism is not limited to race and gender but extends to socioeconomic status. The results suggest that increasing juror diversity—across both demographic and economic lines—could help reduce biases in legal decision-making.

Moreover, the findings raise concerns about fairness and equality in judicial systems, even in a country like Sweden, where the judiciary is designed to be impartial. Given that lay judges are often from higher-income backgrounds, their decisions may systematically disadvantage lower-income defendants, reinforcing existing inequalities.

Finally, the study suggests that bias is most pronounced when jurors have prolonged exposure to the defendant, reinforcing the idea that personal interactions increase identity salience. Future research could explore whether interventions—such as structured deliberations or bias training—could mitigate these effects.

Overall, the study contributes to a richer understanding of identity-driven biases in legal decision-making and emphasizes the need for greater representation in jury systems to ensure fairness in the courts.