NIJ Is Paying $23.8M to Find Out Whether Red Flag Laws and Crisis Courts Work. The Real Constraint Is the Program's September 2029 Sunset

On October 9, 2026, the National Institute of Justice (NIJ) posted one of the largest single research solicitations it has issued this fiscal year: the FY 2026 Evaluation of the Effectiveness of the Bureau of Justice Assistance (BJA) Byrne State Crisis Intervention Program (SCIP), opportunity number O-NIJ-2026-172798.

The money is large for a criminal justice evaluation. NIJ expects to award $23,773,974 across four cooperative agreements, each worth up to $5,943,493. That ceiling is exactly one quarter of the total, which means NIJ plans to fund four projects at or near full size rather than spread the pool thin.

The subject is politically charged. Byrne SCIP is the formula program the Bipartisan Safer Communities Act of 2022 created to fund state crisis intervention programs, and those explicitly include extreme risk protection order (ERPO) programs, often called "red flag" laws. NIJ now wants an independent answer to whether that money prevented violence and suicide.

The deadlines are close:

• SAM.gov registration or renewal: NIJ says to start immediately and finish no later than November 3, 2026
• Step 1, Grants.gov (SF-424): November 16, 2026, 11:59 p.m. ET
• Step 2, full application in JustGrants: November 23, 2026, 8:59 p.m. ET

A pre-application webinar is planned, but NIJ had not set a date when the solicitation was posted.

What Byrne SCIP is, and why the evaluation is happening now

The Bipartisan Safer Communities Act (Pub. L. No. 117-159), passed after the Uvalde school shooting, set aside $750 million over five fiscal years (FY 2022 through FY 2026) for Byrne SCIP. States receive formula funds to create or expand "state crisis intervention court proceedings and related programs or initiatives." The NOFO names four kinds of eligible programs:

1. Law enforcement crisis interventions
2. Court-based programming, including mental health courts, drug courts and veterans treatment courts
3. Referrals to treatment for people in crisis
4. ERPO programs

FY 2026 is the last year of the statute's five-year funding window, which explains the timing. The money for this evaluation comes from the same BSCA appropriation, and the solicitation cites the BSCA as its statutory authority. NIJ is in effect using the last of the program's dedicated funds to find out what the first four years produced.

The NOFO puts a hard date on the window: "Data collection activities and project timelines should include planning for the possible end of BJA Byrne SCIP program activities on or before September 30, 2029."

This sentence matters more than anything else in the solicitation, and the strategy section below comes back to it.

How the award is built: two phases, eight sites, 48 months

All four awards share one structure. Applications must propose both phases. NIJ states that an application missing either phase "would not be considered responsive."

Phase I: Evaluability assessment (up to 15 months)

Phase I asks you to evaluate up to eight site-level programs in one or more of the four program areas. The NOFO is strict about the unit of analysis. An application that proposes to assess "statewide SCIP programs, all BJA SCIP grantees, or an individual BJA SCIP grantee" will be found non-responsive. NIJ wants local programs, such as a county ERPO petition unit, a city's co-responder team or a particular mental health court, not a state's whole portfolio.

You do not have to name your sites in the application. NIJ and BJA will pick the final sites later, and the number may change. If you have secured sites, attach letters of support.

The Phase I plan must cover ten tasks:

• Describing program activities at each site, "with a particular focus on gun violence"
• Assessing program maturity and implementation fidelity
• Building a logic model for each site
• Auditing the availability and quality of operational and administrative data
• Developing primary data collection tools (attach drafts if you have them)
• Judging each site's capacity to take part in a rigorous, multi-site, multi-year evaluation
• Specifying a counterfactual, including sizes for treatment and comparison groups
• Running a power analysis, with both pooled and site-specific calculations for multi-site designs
• Identifying sources of contamination and threats to internal validity
• Specifying outcome measures for violence and crime, "particularly gun violence"

Phase I deliverables come on a fixed schedule: a report outline five months before Phase I ends, a draft report and executive summary three months before the end, a final report with a profile for each site, and a memo recommending changes to the Phase II design.

Phase II: Multi-site impact evaluation (up to 33 months)

Phase II evaluates up to eight sites that Phase I found ready, assessing both implementation and outcomes. NIJ states its outcome of interest directly: the effectiveness of SCIP programs "in preventing violence and suicide." Every proposal must include cost-benefit and cost-effectiveness analyses.

Phase II funds are released at the NIJ Director's discretion. Release depends on clearing human subjects and privacy conditions and on NIJ approving the Phase I deliverables.

Funding controls you need to plan around

• Ten percent drawdown caps on both phases. You can obligate or draw down no more than 10% of the Phase I budget until NIJ approves the revised evaluability assessment plan. The same 10% cap applies to Phase II until NIJ approves the revised evaluation plan.
• Fixed budgets. The Phase I budget "will be final." NIJ says no additional money will be provided for either phase.
• About three months to clear award conditions, during which, in NIJ's words, "research may be prohibited or significantly constrained." Your timeline has to absorb this for both phases.
• Evaluation-only costs. Funds cannot pay for program services. They can cover reasonable costs incurred by program staff for work required by the evaluation, such as manual extraction of administrative records.

NIJ expects to make awards by January 1, 2027, with a performance period of up to 48 months.

The timeline problem in this NOFO

Lay the dates side by side:

On realistic assumptions, an impact evaluation that starts collecting data in mid-2028 can observe programs that are still running for about 15 months before their federal funding may stop. A program that defunds in September 2029 may also wind down earlier, as staff leave once renewal looks unlikely.

For outcomes like gun violence and suicide, which are rare events at the site level, 15 months of prospective data is thin. Reviewers who know the field will see this, and the strongest applications will address it directly instead of hoping nobody notices.

Three design choices hold up under that constraint:

1. Lean heavily on retrospective administrative data. Programs have been operating since 2023. Court dockets, ERPO petition and order records, firearm surrender logs, crisis call records, mortality data and hospital records already exist for the program's early years. A design that rebuilds a treatment cohort from historical records, rather than enrolling participants going forward, is not limited by the 2029 date.
2. Choose quasi-experimental methods that fit the data you can get. For ERPOs specifically, the published literature on red flag laws already uses case-level matched comparisons and synthetic control methods at the state and county level. Propose a method you can defend for the data available in your area.
3. Make the power analysis do real work. NIJ asks for both pooled and site-specific power calculations. Use them to show reviewers that, with the observation window available, your design can detect an effect of plausible size on suicide or firearm injury. If it cannot, say which intermediate outcomes, such as firearm removals, treatment engagement or re-petition rates, you can measure reliably.

How the application is scored

The proposal narrative is limited to 30 double-spaced pages. Weighting Potential Impact at 25% is unusually high for an NIJ evaluation. NIJ also says grantees should put "equal effort" into practitioner-facing dissemination products as into scholarly papers. A dissemination plan aimed at state court administrators, ERPO petitioners and crisis response leads will score better than a list of journals.

Under Capabilities, NIJ explicitly asks applicants to show past experience with evaluability assessments and multi-site impact evaluations, and invites past reports or data collection tools as appendices. If your team has done an evaluability assessment for NIJ, BJA or SAMHSA, attach it.

Required attachments include the NSPM-33 Biographical Sketch and Current and Pending (Other) Support common forms for all key personnel, and the solicitation also lists a Gold Standard Science component. A data management, sharing and archiving plan, a privacy certificate and a human subjects protection plan are standard for NIJ research and should be ready before you start the narrative.

Eligibility and the priority section

Eligibility is broad: state, county, city, township and special district governments; federally recognized and non-federally recognized tribal governments; public and private universities; nonprofits with or without 501(c)(3) status; and for-profit firms, including small businesses. Foreign entities are not eligible. No match is required.

In practice, this competition will be fought among university research centers and established evaluation firms. A small team can still compete as a subrecipient. A firm with deep relationships in one state's court system, for example, could be very valuable to a university applicant proposing a court-based programming evaluation.

The solicitation includes the standard set of FY 2026 OJP priority considerations: directly supporting law enforcement operations, including immigration enforcement; combating violent crime; supporting services to American citizens; protecting American children; and supporting American victims of trafficking and sexual assault. For a research evaluation, the honest fit is combating violent crime, and that is the priority to claim. Claiming priorities your project does not actually advance will not help, and the narrative must explain how the project advances each one you select.

The standard unallowable-activities section also applies, including the 8 U.S.C. 1373 immigration-cooperation scope language that now appears across OJP's FY 2026 solicitations. We covered how that language plays out in our analysis of BJA's Integrity in American Elections program. Universities should have their sponsored programs office review the certifications before the JustGrants submission.

Which program area to choose

NIJ expects to fund at least one award in each of the four areas, and applicants may propose more than one area. The four areas are not equally hard to win.

• ERPO programs will probably draw the most applicants, because the research community around red flag laws is active and well published. Expect strong competition.
• Court-based programming is the area with the longest history of evaluation. There is extensive literature on drug courts, mental health courts and veterans treatment courts. A competitive application needs to explain what SCIP funding changed compared with what these courts were already doing.
• Law enforcement crisis interventions cover co-responder models, crisis intervention team training and alternative response units. Data quality varies widely from site to site, so the Phase I data audit decides whether this area can be evaluated at all.
• Referrals to treatment is the least defined of the four, and it may be where a well-scoped proposal has the least competition.

Proposing more than one area lets you compete for more than one award, but it also spreads your eight-site budget across more program models. Unless you already have site access across several areas, one deeply scoped area usually scores better than two thin ones.

Practical steps for the next three weeks

1. Check SAM.gov now. The November 3 target leaves almost no margin if your registration has lapsed.
2. Call state administering agencies. Each state's Byrne SCIP subgrant list is public information. Identify five to ten local subgrantees in your chosen area and ask about data access before you write the design section.
3. Draft the counterfactual first. The 40% design score depends on whether your comparison group is believable. Build the rest of the proposal around it.
4. Address the September 2029 date explicitly in the project timeline and in the limitations discussion.
5. Budget Phase I as if it were the whole award. No additional money will come, and only 10% is available until NIJ approves your revised plan.

For a related NIJ and BJA competition with a December deadline, see our breakdown of BJA's FY 2026 MUHR program. For the bigger picture of where DOJ's FY 2026 grantmaking went, see our analysis of the FY 2026 federal notice recomposition.

If you are building an evaluation team for this solicitation, Granted can help you draft the narrative, check it against the NOFO's scoring criteria and assemble the JustGrants attachments before the November 16 Grants.gov deadline.