An Introduction To “Nudging” In The Criminal Legal System
Nearly everyone agrees that prosecutors should use more evidence-based practices and engage in more evaluations to build that evidence base. Unlike our health and education systems, the criminal legal system lacks research integration — with almost no modern policy or practice evaluations. And, even where there are evaluations of prosecutions, those evaluations tend to focus on extremely limited metrics like crime rates, conviction rates, and recidivism that would be considered outdated in other professions. Justice Innovation Lab collaborates with partners to develop robust evaluations that include both traditional and more meaningful modern metrics, all aimed at improving prosecutorial outcomes. This series’ first post explores the potential of using one form of innovation in prosecution — behavioral “nudges” — to promote public safety and better serve victims and defendants, how to base nudges on evidence-based best practices or — if there are no best practices because the evidence base is thin — how to evaluate those nudges to build an evidence base.
“Nudging” the Government
Nudging for most people is somewhere between a friend’s encouragement and a parent’s annoying reminder to do something. For governments, researchers, and private industry nudging has taken on a much more significant meaning as a practice to use lessons from behavioral science and economics in order to shape human behavior. In recent years, mostly since the publication of Nudge in 2008, governments recognized the potential to include behavioral science learning and evaluation in designing and implementing policy changes. Since then, two dominant strains of intentional nudging by government emerged — incorporating behavioral science into agency-level policies and creating small, research-focused branches within government that evaluate adjustments to operations and policy.¹
There are many examples of government agencies using both approaches in the United States. Likely the most well-known example of large scale policy change with a conscious behavioral component in the United States was the Affordable Care Act. The law includes clear lessons from behavioral science through its careful setting of defaults and inclusion of penalties to influence individuals toward a certain behavior. As discussed later, there are more recent examples of such large, sweeping changes in local criminal jurisdictions by prosecutors, even if those prosecutors did not view these changes as behavioral nudges.
Across the different levels of the U.S. government, there are also many examples of agencies beginning to evaluate incremental policy changes. Some local governments have even funded internal research groups to work across agencies to develop and evaluate behavioral nudges.² This more incremental approach to change often comes in the form of evaluating small changes to how an office or agency works — changes in defaults on government forms, targeted government advertising, or whether forms are mailed directly to some persons are all examples of these smaller nudges.
While governments using nudges do not have to include measurement and evaluation when instituting changes based upon behavioral sciences, one benefit of approaching change through this lens is that it moves an agency towards measurement and evaluation. This often comes in the form of the government relying upon social scientists when drawing up changes to design rollout in such a way as to measure impact. Furthermore, this process leads agencies to another important, though often neglected step — defining the desired outcome and the metrics to measure it by. To some extent, this framework and nudging is beginning to influence the criminal legal system, though its reach is far more limited, especially among prosecutors, than other public sectors.
Current Nudging in the Criminal Legal System
Many nudge interventions in the criminal legal system currently target behavior by police and frequent “consumers” of the criminal legal system. Police are understandably the subject of more nudge interventions as they have larger budgets that can pay for research, are prominent public servants, and, as frontline workers, affect many people. Interventions targeting users of the criminal legal system are often driven by non-law enforcement agencies and thus benefit from those agencies’ experience with behavior-based interventions.
Some of the most well-known nudges — even if they aren’t discussed as such — within policing are the use of body-cams, various hot-spot policing techniques, and the rise in implicit bias and de-escalation trainings. While departments implementing these measures may not always employ researchers to measure the impact of these interventions, each of these is based on ideas from behavioral science research. For instance, researchers theorized that the use of body-cams would “encourage positive behavioral change” among police officers, presumably by unconsciously deterring officers from abuse. Though many departments adopted body-cams without instituting a formal study, most based the decision, at least in part, on the belief that it might curb police misconduct.
In addition to police nudges, there have been many interventions that target the behavior of those entering the criminal legal system. In recent years, there is a growing push by public health researchers to work within the criminal legal system to try to prevent people from entering the system in the first place. Once an individual is “involved” in the criminal legal system there have also been efforts to reduce involvement through intensive behavioral programs such as Focused Deterrence or minimal touches such as text reminders of court dates.
In contrast, evaluations of behavioral interventions to influence prosecutors are few and far between. This is likely due to a number of factors unique to prosecutorial work: its historical nature, weak connections to non-law enforcement agencies, underfunding of prosecutor offices, and a lack of interest in research and evaluation of prosecutor work.
Hurdles to Nudging in Prosecution
First, there ARE efforts to bring behavioral interventions into prosecutors offices and some of those DO involve research and evaluation. The next post in this series will outline some of those efforts and discuss future avenues for interventions. But, it is also important to understand some of the hurdles with designing and implementing interventions in prosecutors offices.
- Prosecutors and criminal law generally are aimed at assessing an individual’s guilt for a past criminal act. The work is naturally and historically backward-looking. While prosecution may also have a deterrent (forward-looking) effect, a prosecutor’s work is mostly aimed at assessing what happened and often recommending a punishment. Prosecutors are not trained in social science theory to consider how the punishment they seek will affect a defendant’s future — though most prosecutors do at least consider this. As such, the historical nature of prosecution does not lend itself to behavioral interventions that tend to measure success by what happens in the future.
- There are often limited to non-existent links from public health and education agencies to prosecutor offices which hinders prosecutors in finding alternatives to traditional prosecution. It is not uncommon to find prosecutors who are unaware of diversionary programs in their own jurisdictions or who do not know the requirements for referring someone to a diversionary program. Furthermore, prosecutors are often provided with limited information regarding a defendant’s arrest and the underlying facts. By design, our adversarial system places the burden on the defense and the defendant to provide evidence of ‘mitigating factors.’ It is hardly surprising then that prosecutors are rarely the subject of interventions that might lead them to treat defendants differently when their access to resources and information to do so is limited.
- Prosecutor offices tend to be underfunded and understaffed such that funding for research and evaluation rarely exists. Most local prosecutors are publicly elected and are not a part of the local executive branch (e.g. the mayor’s office). With this independence comes the consequence of needing to advocate for an office’s budget from local city, county, and state legislatures without a guaranteed ally.
- Public prosecutors, as popularly elected officials, are evaluated by metrics that are most salient to the public — even if those metrics are flawed and weakly connected to most prosecutor work. While researchers are very interested in understanding how prosecutors work and their effect on defendants, victims, and their communities, the public tends to evaluate prosecutors along two metrics: performance in high-profile cases and crime rates. These measures the public relies upon are only weakly correlated with prosecutor performance — high profile cases that often go to trial are an extremely small sample of prosecutor work and crime rates are nearly non-responsive to prosecution efforts. But, between these poor metrics, constraints on prosecutor time, and the risks involved with deviating from the status quo, prosecutors are pushed away from investing in research and evaluation of alternatives.
Despite these challenges, there have been recent efforts to incorporate behavioral science into prosecutor offices. Those offices, whether consciously adopting nudge strategies or not, recognize that prosecution needs to change and improve in order to better serve the public and take advantage of multi-disciplinary learning.
Why Nudging is Important
Prosecutor offices should consider adopting behavioral interventions and investing money in research and evaluation because doing so can improve outcomes across the prosecutorial landscape. As of now, many established, effective diversion programs that reduce recidivism and offer victims an alternative method of reconciliation are underused, but we don’t know the role of prosecutors in that underuse. Many prosecutors are inundated with large caseloads — mostly drug possession and petty thefts related to addiction and mental health — and most prosecutors feel pressure to prosecute such cases because they are proveable. Victims are generally included in the prosecution process to the extent that they are willing to testify and aid in the incarceration of an individual. Perhaps were prosecutor offices to begin adopting incremental changes to practice that utilize behavioral science, we could begin addressing some of these issues. By addressing these issues, we would be improving public safety and the outcomes for victims and defendants while building a body of evidence of best practices in prosecution.
¹ See Nudging by government: Progress, impact and lessons learnt, David Halpern & Michael Sanders (2016).
² See The Lab @ DC; NYC Mayor’s Office for Economic Opportunity.
By: Rory Pulvino, Justice Innovation Lab Director of Analytics
For more information about Justice Innovation Lab, visit www.JusticeInnovationLab.org.
