OKPOLICY.ORG
•Watch: via YouTube
•View: Presentation slides (PDF)
The Oklahoma House of Representatives held an interim study on Oct. 21, 2026, to examine maximizing the effectiveness of the County Community Safety Investment Fund (CCSIF), which was created by State Question 781 passed by voters in 2016. SQ 781 was part of a justice reform package approved by voters to reclassify simple drug possession and some minor property crimes from felonies to misdemeanors. SQ 781 directed the savings from those changes would then used by county governments to provide substance abuse and mental health services through the County Community Safety Investment Fund (CCSIF).
However, due to legislative and administrative disagreement about the size of those savings, it took more than seven years before the Legislature appropriated funds to the County Community Safety Investment Fund.
During the October 2026 interim study, OK Policy’s Research Director Anthony Flores and Criminal Justice Policy Analyst Cole Allen presented data on the impact of SQ 780 and SQ 781, including prison cost savings, statewide crime trends, and the effectiveness of diversion programs. Their testimony highlighted how current calculations may underestimate savings from reduced incarceration, while crime rates have declined and diversion programs have shown promising public safety outcomes. They also emphasized the opportunity to reinvest those savings into community-based programs that reduce repeat offenses and strengthen public safety across Oklahoma, particularly in rural communities.
The following is a transcript of their comments. The full video of the interim study — IS-26-098, Maximizing the Effectiveness of the County Community Safety Investment Fund (CCSIF) requested by Reps. Meloyde Blancett, Nicole Miller, and David Hardin — can be viewed at the Oklahoma House website.
# # #
Comments from Anthony Flores, OK Policy’s Research Director
Good afternoon. Thank you for the opportunity to speak today. My name is Anthony Flores and I’m the Research Director for Oklahoma Policy Institute. I’m joined by my colleague Cole Allen, who is our Criminal Justice Policy Analyst. We’re here to talk about data related to State Questions 780 and 781. We’ll cover the savings estimates associated with both state questions, then look at crime trends and recidivism outcomes from diversion programs.
OK Policy has worked with state agencies on these calculations since 2019. When the Department of Corrections disputed the first OMES estimate, we helped figure out why the numbers didn’t match, worked with OMES on a new formula, and provide the estimate of prison days averted for OMES and LOFT.
State Question 781 directs LOFT to calculate the savings and averted costs to the state from SQ 780 reducing how many people go to prison and how long they stay. The key words there are averted costs. These are costs that never happened. You can’t track data for a prison stay that didn’t happen, so the statute tells LOFT to use actual data where it exists and the best available estimates where it doesn’t. Then an amount equal to those savings is paid into the County Community Safety Investment Fund, subject to appropriation. For this presentation, since CCSIF is a mouthful, we’ll call it the Community Safety Fund.
The formula itself is simple. Take the prison days we avoided because of SQ 780, what we formally call estimated prison days averted, and multiply by what it costs to keep one person in prison for one day, the average daily cost per inmate. As you’ll see over the next few slides, every change in the savings estimate over the years comes from a change in one of those two numbers. So I’ll describe them one at a time, starting with the average daily cost per inmate.
We’ll first look at the statutory definition of DOC’s average daily cost per inmate. State law, Title 57, Section 561.1, requires DOC to calculate its actual average daily cost per inmate every year, by security level, and to include all direct and indirect costs. Minimum security cost is the number that matters here, since that’s where people convicted of SQ 780 offenses would be housed. Higher security levels have higher costs.
However, different definitions of cost have been used for the official savings calculations. To understand why, it helps to know how costs are grouped. Direct costs are tied to each person, like food, clothing, and medical care. Indirect costs support the whole system, like facility costs or salaries and benefits for staff. Costs can also be variable or fixed. Variable costs go up and down with each person. Fixed costs, like buildings, upkeep, stay about the same until a facility runs out of room.
DOC’s official number counts all direct and indirect costs, both fixed and variable. That distinction matters for the bottom line of the savings estimate. Savings equal days averted multiplied by daily cost, so the narrower the definition of cost, the smaller the savings estimate.
From FY 2018 through 2022, OMES counted only the cost of one more inmate: medical care, food, clothing, linens, and inmate pay. That is a marginal cost: direct and variable costs only, with no indirect or fixed costs. DOC’s minimum security cost which included those indirect costs, was 3.6 to 4.3 times higher. In FY 2022, for example, DOC’s figure was about $61 a day, and OMES used $16.90, 3.6 times higher. That matters because savings is days averted multiplied by daily cost per inmate. If you keep days averted the same, and your daily cost per inmate is 3.6 times higher, then the savings estimate is 3.6 times higher.
In 2023, the Legislature passed Senate Bill 844, which moved the savings calculation from OMES to LOFT. Starting in FY 2023, LOFT widened the calculation. It added more direct costs from the state’s accounting system and each facility’s share of agencywide costs like medical care. It still left out indirect costs like salaries, benefits, and larger fixed costs. That moved the daily cost from $16.90 to somewhere between $30 and $41 a day. In every year, that is still below DOC’s full minimum security cost, because it leaves out indirect costs.
The question of fixed costs came up before LOFT took over. In 2017, the Legislature created the Agency Performance and Accountability Commission, or APAC, to run independent performance audits of state agencies. The outside auditors who reviewed DOC estimated a daily cost of $40.66 for a medium security facility, which counted some fixed costs. In 2021, LOFT reviewed that work and noted that the $13.20 a day OMES used for FY 2020 left out fixed costs such as staff, depreciation, and maintenance of assets. LOFT recommended that cost estimates include all operating expenses, fixed costs included. Using DOC’s FY 2020 minimum security cost of $57.39 a day, LOFT estimated that year’s savings at $46 million instead of $10.6 million. When LOFT took over the calculation, it did not use DOC’s official figure, as its 2021 report had. It built its own calculation that excluded indirect costs such as salaries, benefits, and larger fixed costs. That figure landed closer to the auditors’ $40.
This is why it matters today. DOC’s prison population is projected to exceed capacity in February if this year’s growth continues. When that happens, even a small increase in the prison population forces decisions about staff, beds, and buildings. It has been nearly 10 years since SQ 780 took effect, so it may be time for the calculation to start accounting for fixed costs in some way. Exactly how much is something for LOFT, the legislature, and DOC to work through.
Now the second number, days averted. The main thing to know is that since FY 2023, this number has been stable. A stable number means a predictable number, and that matters for appropriations. But it took a few agencies working together to get there.
The first OMES estimate, for FY 2018, said about 2 million prison days were averted. DOC disputed it using data they had on receptions to DOC, and when we looked into it, we found OMES assumed 100% of cases with an SQ 780 charge ended in a conviction. We agreed with DOC that actual prison receptions should be used instead. Starting in FY 2020, OMES used a new formula developed with DOC, OK Policy, and OCJR. Since SB 844, LOFT produces the estimate instead of OMES. We’ve provided the days averted figures since FY 2020, first to OMES and now to LOFT.
So again, the takeaway: a stable estimate of days averted means a predictable savings estimate, and that matters for appropriations.
This chart puts the two numbers together: days averted times the daily cost per inmate. Yellow is the OMES savings estimate for FY 2018 through 2022, which uses OMES’s estimate of days averted and its marginal cost per inmate. The FY 2018 and 2019 estimates used early methods that were later replaced. Red is LOFT’s official estimate from FY 2023 on, which uses our estimate of days averted and LOFT’s direct and shared cost per inmate. Those changes in both numbers are why the estimates shift over time. Gray is what was appropriated to the fund, which started in FY 2024 at $12.5 million and was $12.5 million again in FY 2025. In every year, the estimated savings were higher than the appropriation.
Those official estimates use the OMES and LOFT daily costs. If we keep the same days averted but use DOC’s official minimum security cost instead, the savings are 1.5 to 4.3 times higher. To be clear, these are not official estimates. They show what the savings would be if the calculation counted indirect costs. The statute says the calculation “shall be final and shall not be adjusted for any subsequent changes in the underlying data.” So past estimates can’t be changed, but the method can change going forward.
If we set aside FY 2018 and 2019 because of the problems I described, the estimated savings from FY 2020 through 2025 add up to $100 million using the official numbers, or $284 million using DOC’s full cost. Over that same period, $25 million was appropriated to the fund.
Again, this is not about going back and filling in past years. It just shows the size of the gap so far.
To sum up. LOFT’s daily cost counts direct and shared costs. DOC’s official cost also counts indirect costs like staff, which a smaller prison population can help avoid as we near capacity. Since FY 2023, the days averted estimate has been stable and predictable. And since the legislature started funding the Community Safety Fund in FY 2024, the gap between estimated savings and appropriations has averaged $5.1 million a year using LOFT’s cost, or $19.2 million a year using DOC’s cost.
And with that, I’ll hand it off to Cole.
# # #
Comments from Anthony Flores, OK Policy’s Research Director
My name is Cole Allen. I am the Criminal Justice Policy Analyst at the Oklahoma Policy Institute, and I’m going to talk about the crime trends that we’ve seen since State Question 780 has passed, and then some recidivism data related to diversion programs.
The big takeaway is that since SQ 780 passed, those crime trends have gone down across the state. That is seen using the best available data that we have, and I will get a little bit more into what those metrics are.
Probably the most well-known aspect of SQ 780 is its impact on drug crimes, and this chart shows the arrest rates across Oklahoma for the possession of the drugs that SQ 780 impacted. Those include things like methamphetamine, narcotics, opioids, and synthetic narcotics. What we can see is that since SQ 780 took effect, arrests across the state have gone down by about 30%, and it’s stabilized really since fiscal year 2022.
But what I think is really interesting about this is that it shows that the arrest rate for sale and manufacturing of these same drugs, which was not affected by SQ 780, has stayed stable. It’s actually risen about 10%, which to me indicates that law enforcement agencies across the state have still been able to maintain their focus on tracking down the sources of these drugs in our communities, which is again the sale and manufacturing. This arrest data is the best data that we have for the prevalence of drug possession in Oklahoma. This is an indication that drug possession is going down somewhat in our state.
The other big aspect of SQ 780 has to do with larceny, theft, and property crimes, which has, like drug possession, also decreased. This chart shows the rate of reports of crime. This is from the FBI’s data, which is really the gold standard report for this, and we can see that it has dropped significantly since SQ 780 took effect.
But that was actually a continuation of a longer trend. We have the data here from fiscal year 2011, but it’s been going on well before then. Larceny rates have been dropping across the state, and actually Oklahoma has been seeing that drop faster than the U.S. national average, which is, I think, a testament to Oklahoma. What this shows generally is that the two major areas of crime affected by SQ 780 have been going down in our state, according to the best available metrics that we have.
So this is the prevalence of the crime. But what about our state’s response to the crime, which is really where SQ 780 changed how we respond?
When we look at the case filings, this is how DAs — prosecutors — respond to cases of possession or larceny or some of the other crimes affected by SQ 780. Case filings across the state for SQ 780 crimes have gone down significantly.
Earlier today, we heard some of the speakers talk about a spike in misdemeanor cases. That is exactly what we would expect after SQ 780, because it changed a number of felonies or raised a threshold to charge someone with a felony to a misdemeanor. And so we see, along with a spike in misdemeanor cases for these types of crimes, we see a massive drop in felonies. And overall, that has led to an overall decrease in the number of case filings for SQ 780 crimes. This is not just in urban Oklahoma. We’ve had many speakers today from rural Oklahoma, and what we see here is a similar trend that we see across all 77 counties when broken down between urban and rural. This is the USDA classification for that. We see that even though the numbers are slightly different, we do see that overall the shape is going down, which is an important trend for us to notice.
And when we look more closely at the specific subcategories of the SQ 780 crimes — so on top of drug possession or larceny, things like receiving receiving stolen property, fraud and forgery, embezzlement, and false declaration of a pawn — while the shapes may be different, while the individual trends may be different between the felonies and the misdemeanors, the overall number of case filings have gone down across the state. This is in all 77 counties on average.
But case filings are not the end all be all of the criminal legal system. That really has to come down to convictions. And so when we look at the counties where we have conviction data for, these are the 13 OCIS counties, the counties that use the Oklahoma Court Information System. This is eight urban counties and five rural counties, so a good mix. We can see that the convictions for these crimes have also gone down in a similar trend to the case filings, which I think leads to some of the reductions in prison population that we have been seeing.
This has been driven by a number of things. One of the things that it has been driven by is some of the alternatives to incarceration that have worked, like diversionary programs. Because when they work, they help get people out of the system. And part of how they do that is by helping reduce recidivism, which is the last thing I want to talk about today.
So, like crime trends, we’ve also seen recidivism efforts — especially for diversion programs that are funded through CCSIF — are working. It has positive public safety impacts.
First, looking at Diversion Hub, which is in Oklahoma County and the surrounding metro area. We can see that participants in the misdemeanor diversion program have a significantly lower rate of rebooking in the county jail than the general population. It’s about 30% on average, and actually, as time goes on, we see that gap widen where Diversion Hub clients are having a better outcome than the general public.
The same holds true when we look at new bookings. Again, this is the DA’s prerogative to book or refile cases against someone. We see similar trends. It ranges from about a 10% to 30% drop between Diversion Hub clients and the general public. And the trend also continues when we look at Justice Link in Tulsa where we see any client who has had interaction with Justice Link has a much more positive outcome when it comes to rebooking rates when compared to the countywide average.
What’s interesting about this is when we look at successful clients — these are clients who have successfully completed the programming and gotten the treatment that they need — they have the lowest rate of rebooking. But even the unsuccessful clients — clients who did not fully participate and clients who did not receive all the treatment — they still have a lower rebooking rate than the general jail population, which I think goes to show what a lot of speakers have been talking about.
Any participation in a diversion program can have positive impacts on someone’s life. It’s better when they’re successful, but when we can get people in contact with treatment providers, with services, with transportation, this does help reduce some of the things that drive crime.
And so, Diversion Hub and Justice Link, of course, are only found in the urban areas. We’ve heard a lot of discussion today about the need for services in rural areas, and there are a lot of things that drive that. But a big part of that is a lack of funding. So that takes us all the way back to calculating savings from SQ 780.
As I recap, from fiscal year 2020 to 2025, the official estimates of SQ 780 has been about $100.2 million. After the legislature began appropriating in fiscal year 2024 and through 2025, we’ve seen $25 million appropriated, and that’s had positive impacts.
Because we know that crimes affected by SQ 780 are down across the state, and diversion programs work, especially diversion programs that are funded through CCSIF. If we can expand investments into non-urban counties to rural counties and to other communities to get similar programs like we’re seeing in Tulsa and Oklahoma County — again tailored to those counties’ needs, we can’t just slap on Tulsa solutions to Pushmataha County needs — we can see better public safety outcomes and hopefully have better, happier, healthier Oklahoma.
