Fighting crime with DNA

The opportunities available and organizations that are working to seize them

May 26, 2026

Inspired by Nan Ransohoff’s recent post on the “Third Wave of Philanthropy”, as well as Greg Berman’s phenomenal book, The Nonprofit Crisis, I thought y’all might be interested in which groups and efforts my team (Criminal Justice at Arnold Ventures) is funding, and why.

The traditional nonprofit space leaves a lot to be desired. There is plenty of good intent, but far too little focus on impact and ROI. In a world where most of our good ideas don’t work, and some even backfire, this a recipe for disappointment and cynicism. This is certainly true in the public safety space. My team and I are trying to change this, one grant at a time.

So, I’ll start sharing examples of nonprofits I think are doing amazing work, and deserve more support. If you are someone with the financial means to help, consider this your invitation to do so.

First up: Expanding DNA databases and the use of DNA evidence to solve more crimes, faster.

To change behavior, focus on increasing the probability of consequences

In the US we’ve primarily tried to reduce crime by making sentences longer, in the hopes of deterring crime and keeping dangerous people off the street. But this has turned out to have a pretty low ROI, for two reasons. Most people we’re trying to deter aren’t thinking very far ahead, so the threat of a long sentence doesn’t register when they’re making decisions today. And most people age out of crime as they grow up and their brains fully develop (most of us become less reckless and impulsive by our late 20s); this reduces the public safety value of keeping people behind bars for a long time. (Of course there are exceptions, but those prove the rule.)

We now know that a far better way to deter and prevent crime is to increase the probability of consequences, rather than the severity of those consequences. We’re learning! That’s good news!

The bad news: The probability of getting caught if you commit a crime is pretty low. In the US, clearance rates (a useful though imperfect proxy for the likelihood a reported crime results in an arrest) hover around 55% for homicide. That means it’s essentially a coin flip whether you’ll be arrested if you commit murder. Not good! For other offenses, you’re even more likely to get away with it.

Based on recent FBI data, clearance rates average 44% for all violent crime. Only 31% of reported robberies are solved. Only 25% of reported rapes result in an arrest (and we know the vast majority of these crimes are not reported in the first place). Clearance rates for property crime are even more dismaying: 16% on average. Only 9% of reported motor vehicle thefts result in an arrest. As I often quip, given that low number it’s amazing we all aren’t stealing cars, all the time.

This means that to reduce crime rates, we should focus on solving more crimes, faster. It turns out this solution has broad, bipartisan appeal — my team is seeing lots of action on this front in a variety of states across the US. (More on this in a future newsletter.)

But of course, solving crime is easier said than done. What are practical, scalable approaches we can use to increase the likelihood that criminal offenders face consequences?

DNA provides one answer. I have research from Denmark that uses a strong natural experiment to show that adding people charged with felonies (not waiting for a conviction) reduced future offending by 42%. The effects were even larger for younger groups. This is a huge impact, coming primarily through deterrence (the crime never happened, sparing would-be victims and putting those offenders on a better path as well — they were more likely be working or enrolled in school after being added to the database).

Other research highlights that DNA is useful in property crimes, not just violent crimes. One study randomized a subset of burglary cases to have DNA evidence analyzed and uploaded to CODIS (the federal DNA database in the US). Cases in that treatment group were 18 percentage points (138%) more likely to have a suspect identified and 12 percentage points (120%) more likely to result in an arrest.

From a public safety perspective, these interventions were a huge win! Collecting DNA from more people at risk of committing crime, and from more crime scenes, will solve more crimes that do happen and deter many people from offending in the first place. Because of the big deterrent effect, I genuinely believe this is our ticket out of mass incarceration.

The opportunity

In the US, all states require DNA be taken for felony convictions. The policy frontier is adding people arrested or charged with a felony (not waiting for the conviction, though samples can be expunged if the case is dropped), or adding people convicted of misdemeanors. A challenge to such expansions is huge backlogs in forensic crime labs — analyzing these additional samples would take more resources these labs don’t have. (You might have heard of backlogs of rape kits that that haven’t undergone testing — it turns out this is a problem for other offense categories as well.) Recent updates to (very conservative) FBI guidance make it easier to use Rapid DNA machines that can create DNA profiles in 90 minutes — much better than the previous timeline (typically several weeks). Police can use these machines at their own booking stations, creating DNA profiles that identify individuals alongside the mugshots and fingerprints law enforcement has been collecting for many decades.

This brings us to privacy. What is a DNA profile, you ask? Importantly, this isn’t 23andMe, with all the sensitive information you might be used to seeing there. A forensic DNA profile is a string of numbers that is only useful for matching to other DNA. Like a Social Security number, the individual numbers don’t mean anything, but together they uniquely identify you. Indeed, many people working in this space put their DNA profile on their business cards, to demonstrate that the information is not sensitive.

Of course, we need strict regulations governing what data are collected and how they are used, especially as the science advances and new techniques become possible. (I value the Policing Project’s thoughtful advice on how to balance public safety with civil liberties.) But in the case of CODIS and the use of Rapid machines, these regulations are already in place — the FBI is quite conservative here. DNA databases have been around since the 1990s and there has never been a data breach. (This is perhaps because they don’t contain information of any value to hackers.)

I could go on about the opportunity here, even with these guardrails in place. But suffice it to say that my team has determined that expanding DNA databases is one of the highest-ROI opportunities available in the US public safety space.

If we expand DNA databases to include people arrested for felonies in all US states (this is already current policy in 19 states), then (based on conservative assumptions) we expect at least $7.8B in social benefits during the next 10 years, from crime prevention alone. Other strategies — such as increasing the use of Rapid DNA machines — would likely complement this and have even broader crime-reduction benefits, as they help crime labs speed through forensic sample backlogs.

The path forward

Who is doing the work necessary to seize this opportunity? It turns out there haven't been many organizations focused on this, which is part of the reason we see so much potential for impact. (It fits Rugter Bregman’s Triple-S framework — this issue is “sizable, solvable, and sorely neglected.”)

Here are two organizations that Arnold Ventures supports in this space; I’ll highlight others in future missives.

The newly-launched Center for Forensic DNA Policy and Practice, at the National Criminal Justice Association, will be the cornerstone of this policy work, assisting states and local agencies in responsibly expanding their use of DNA. They will be a national hub for resources, technical assistance, and policy development. Overseen by Chris Asplen, who has deep expertise in the forensic DNA space, this new Center will ground our and others’ work in the years ahead. I am so excited about this.
The DNA Justice Project, led by Ashley Spence, educates policymakers about the value of DNA and DNA databases. Ashley is a force of nature, combining a powerful personal story that highlights the value of DNA databases to solve crime, with a deep understanding of research and policy in this space. This makes her a smart partner to lawmakers who want to find a path forward on this issue. (Check out her book, Silent Witness.)

And of course, we need more research to rigorously measure the impacts of new DNA policies and practices! More on some ongoing studies I’m excited about, another time.