,
25 tweets,
5 min read
Read on Twitter
LONG THREAD: With today's announcement of substantial investment in data-driven pretrial criminal legal system reform, we are long-overdue for a broad conversation on addressing community needs, the role of philanthropy, & the ethics of certain models of social science research.
We hope this will be a starting point, though it appears the train has already left the station:
Good quality research. Evidence-based. Data-driven.
These buzzwords are quite popular in #cjreform circles, among nonprofits, researchers, foundations, & elected officials.
Good quality research. Evidence-based. Data-driven.
These buzzwords are quite popular in #cjreform circles, among nonprofits, researchers, foundations, & elected officials.
Yet no one required such rigorous investigation to incarcerate millions of people, including hundreds of thousands pretrial, which has caused generations of harm.
Research did not create our present criminal punishment system.
Racism & fear did.
Research did not create our present criminal punishment system.
Racism & fear did.
Decades of legislative reforms designed to remove leniency, reduce discretion, and increase punishment were not rooted in data showing the breadth of a particular social ill. Retribution is not research-backed.
Instead, bills were named for individual crime victims; the story of a single crime committed during a prison furlough catapulted into the national consciousness, weaponized through campaign advertising; policy was implemented based on anecdote.
But to reform this system built on specific cases obscured and magnified with smoke and mirrors?
Researchers, nonprofits, and foundations call for data-backed solutions to deconstruct the punitive, carceral structures that comprise the criminal legal system.
Researchers, nonprofits, and foundations call for data-backed solutions to deconstruct the punitive, carceral structures that comprise the criminal legal system.
We aim to name & confront the impulse to require voluminous research to convince policymakers that non-carceral approaches to safety & healing are worthy of consideration.
This is not the standard to which we held policies that vastly expanded policing, jails, & prisons.
This is not the standard to which we held policies that vastly expanded policing, jails, & prisons.
There is an undeniable appeal and value in crafting public policy based on something provable and replicable. Of course examining data can advise effective policy, can unpack and disrupt assumptions, can help us identify truth.
As an institute housed at a law school within a large research university, we especially understand this appeal for academics at the intersection of law and policy. It is logical to value and require tested, verifiable evidence as proof in order to bolster a conclusion.
But this push for data-driven solutions to a crisis created by racism and fear seems to rest on an assumption that “fact” and “truth” must derive from independently tested research.
We have to allow for the possibility that other sources of truth are equally valid & compelling.
We have to allow for the possibility that other sources of truth are equally valid & compelling.
Fact and truth also derive from the lived experiences of people directly affected by decades of failed policies. Their voices and solutions should be centered in dialogue about how to undo the effects of those policies—past and ongoing. #communityjustice
To that end, we hope we can serve as a convener for a larger conversation about possible alternative roles for philanthropy in addressing the inequities that pervade our justice system. This must be an ongoing dialogue, and it is not just related to justice system reform.
This kind of philanthropy leads to concentrations of large amounts of money in large institutions, to the detriment or even starvation of smaller scale, community efforts that may not engage in evidence-based evaluation, but do have solutions that work for their communities.
In the specific field of justice reform, we are especially wary of technocratic endeavors that exclude directly affected communities: people who are presently or formerly incarcerated & their families, people who are otherwise justice-involved, activists, grassroots groups.
Being housed in a large research university but committed to supporting community work, we find that this kind of philanthropy is damaging for our model. It creates a baseline of distrust with communities. When we approach community groups, they assume we aim to conduct research.
Their immediate response: we don't need research to confirm what we already know.
The research community should not scoff at this reaction. We must hear & absorb the pain in being the frequent object of study but never the author of policy.
The research community should not scoff at this reaction. We must hear & absorb the pain in being the frequent object of study but never the author of policy.
That pain has specific resonance when it comes to Randomized Control Trials ("RCTs"), often framed as the "gold standard" of empirical research.
To our community partners, it's fool's gold.
They raise a moral objection to randomizing access to justice system improvements.
To our community partners, it's fool's gold.
They raise a moral objection to randomizing access to justice system improvements.
For one thing, especially in a criminal legal setting, such studies convert communities of color & poor people into research subjects of randomized public policy. We fear this idea of randomization normalizes public policy by chance, like a housing or school lottery.
We are concerned that public policy by chance is fundamentally inconsistent with the project of equity. And even though RCTs are trial runs, they are not just experiments. They accept and enshrine randomization, even in a pilot setting, operating as policy in the real world.
For example, one RCT funded by this new endeavor is a study in Allegheny County, Pennsylvania, on the impact of having a public defender at a criminal defendant’s preliminary arraignment, with the overall goal of improving the criminal justice system.
When research already exists (see, e.g., below) to confirm the efficacy of an intervention, it disrupts a necessary precondition for RCTs: equipoise—true uncertainty as to whether the intervention or the status quo is more effective. How does an IRB approve such a study? 
RCTs do not just make an existing random distribution of a scare resource more explicit.
1) The resource is only scarce b/c of mutable public policy
2) In systematizing who is freed & who is bailed out, who has a lawyer & who doesn't, we become complicit in the system's harm
1) The resource is only scarce b/c of mutable public policy
2) In systematizing who is freed & who is bailed out, who has a lawyer & who doesn't, we become complicit in the system's harm
No matter the underlying question, keeping someone in jail for research purposes intentionally subjects them to jail's harms. We raise a deontological concern for randomized human experiment on predominantly poor people/people of color to determine who should be freed from cages.
Policymakers in jurisdictions engaged in these studies are equally to blame. Your constituents didn't elect you to harm people in the name of research. Why must we devote public partnerships to studies instead of implementing policies that begin to undo our racist system for all?
The #reform community has an opportunity to embrace new kinds of evidence to shift punitive policy. We know the name Kalief Browder as well as we know Willie Horton. We allow anecdote to help us identify harm. Why don't we invest in directly affected people identifying solutions?
