A Home for Every Child Initiative eight months in: PIPs and predictive algorithms

Summary

Podcast Summary: Priorities Podcast – "A Home for Every Child Initiative Eight Months In: PIPs and Predictive Algorithms"

Host: Sophia Fox (StateScoop)
Guest: Alex Adams, Assistant Secretary, Administration for Children and Families (ACF)
Date: June 17, 2026

---

Episode Overview

This episode explores the progress and impact of the Administration for Children and Families’ (ACF) "A Home for Every Child" initiative, launched eight months ago to address the shortfall of licensed foster homes in the U.S. The discussion focuses on state participation, performance improvement plans (PIPs), the use of predictive algorithms, and how new data-driven approaches can transform outcomes for children and families.

---

Key Discussion Points and Insights

1. Background and Goals of "A Home for Every Child"

• National Shortfall: U.S. currently averages 57 foster homes per 100 children entering care ([01:31]).

• Initiative Launch & Expansion: Launched under the "Fostering the Future" executive order, ACF hoped for 10 state participants but now has over 25 (half the country) committed ([03:20]–[05:02]).

  • Quote: "You never know quite what to expect when you announce an initiative like A Home for every Child... but we far exceeded any reasonable expectation." — Alex Adams ([03:20])

• Dual Approach: States can either:

  • Recruit and retain more foster homes ("increase the numerator").
  • Reduce the number of children entering care through prevention ("shrink the denominator") ([04:13]).
  • "Everything that you're going to see from our agency is designed to move the needle on either the numerator of homes or the denominator of kids." — Alex Adams ([04:33])

2. Performance Improvement Plans (PIPs)

• Old vs. New System:

  • Previously, PIPs required up to 14 measures and massive 200–300-page reports with limited effectiveness ([05:27]).
  • Now, states set monthly, focused, evidence-based goals and report real-time data on foster home to child ratios, safety, and permanency ([05:27]–[07:07]).
  • "We're getting out of the mindset of writing all of these 100 pages of plans, then checking in at the backside... Instead... looking at the right measures more frequently, monthly in this case." — Alex Adams ([05:53])

• Comparative Data: Enables states to benchmark progress against peers and themselves over time; fosters healthy competition and innovation ([07:07]).

3. Regional Participation and Unique Challenges

• Diverse Interest: Both red and blue states, rural and urban, have joined—the main barrier is county-administered states facing extra bureaucratic steps ([07:58]).
• Data Age Limitations: Most public data is currently two years old, but the initiative aims to create real-time, monthly insights to drive decisions and technical assistance ([09:32]).
  • "Data is greater than opinions." (Sticker given by Indiana’s agency head) — Alex Adams ([11:26])
  • "Right now we offer a lot of technical assistance that's based on gut feel... But I want the data to take us that destination." — Alex Adams ([10:30])

4. The Human Impact of Data-Driven Change

• Beyond Numbers: Both host and guest highlight the importance of remembering the real families and children behind the data ([11:55]–[12:50]).
  • "It's so easy to talk about technology and numbers and data and sometimes forget that we were actually talking about, you know, children and families." — Sophia Fox ([11:55])
• Idaho Success Story: Adams describes transforming Idaho’s system from having too few foster homes—forcing kids into hotels or offices—to surplus capacity and 30% fewer group home placements in just a few years ([12:50]).
  • "We went from 74 homes for every 100 kids up to 104... We got every kid out of those non-traditional settings." — Alex Adams ([13:17])

5. State Autonomy and Innovative Approaches

• Flexible Metrics: States choose improvement paths tailored to local needs—some prioritize better placements (e.g., Oklahoma), others focus on prevention or permanency ([15:10]).

  • "By allowing states to channel their energy and their focus on a few high leverage things, I think we're going to redirect a lot of energy and reclaim hours that were Otherwise squandered, filling out paperwork or checking boxes.” — Alex Adams ([16:28])

• Predictive Analytics & Risk Modeling:

  • Deployed to assist (not replace) human decision-making during intake—bringing more data to triage, risk stratify, and inform proportionate responses ([17:46]–[18:00]).
  • "We think it's an area that could be improved by bringing some additional data... Not to supplant the human decision making, but to supplement." — Alex Adams ([18:00])

6. Grants and Incentives

• Innovation Challenge: $7 million in funding; states must join by June 30 to compete for prizes ($3M for the highest ratio, $2M and $1M for other improvements). Competition runs until September 30, 2027 ([09:32], [23:24]).
  • "We hope to win, but more importantly, we want all kids to win. We know this is going to be a win for all kids." — Paraphrased from a governor ([24:13])
• Predictive Analytics Grants: $6 million available for at least 10 states to pilot predictive analytics for intake and/or placements ([18:00], [20:29]).
  • Idaho's pilot funded by a Casey Foundation grant was so successful it's now state-funded ([20:29]).

7. Capacity for Data-Driven Tools

• Infrastructure Concerns: Some states still face challenges with siloed data, but Adams is optimistic—if Idaho can do it, others can, and vendor support plus federal grants lower barriers ([20:29]).
  • "If we got it done at Idaho, I'm confident anyway..." — Alex Adams ([20:29])

---

Notable Quotes and Memorable Moments

• On State Buy-In:
  "It goes without saying we're trying to get states to increase their ratio of homes to kids. And there's two ways of doing that." ([04:13]) — Alex Adams

• On Data vs. Past Practice:
  "Data is greater than opinions." ([11:26]) — Alex Adams, quoting Indiana child welfare director

• On Human Impact:
  "We went from 74 homes for every 100 kids up to 104... every kid out of those non-traditional settings." ([13:17]) — Alex Adams

• On Predictive Analytics:
  "We think it's an area that could be improved by bringing some additional data... Not to supplant the human decision making, but to supplement." ([18:00]) — Alex Adams

• On Competition and Collaboration:
  "Healthy competition... we want all kids to win. We know this is going to be a win for all kids." ([24:13]) — Paraphrased from a participating governor

---

Timestamps for Important Segments

• State Participation & Initiative Launch: [03:20]–[05:02]
• Performance Improvement Plans (PIPs): [05:27]–[07:07]
• Regional Differences & Data Approach: [07:58]–[11:43]
• Idaho Success Story: [12:50]–[14:03]
• State-Specific Metrics & Predictive Analytics: [15:10]–[18:00]
• Grants and Innovation Challenge Details: [09:32], [23:00]–[23:42]
• Capacity & Data Infrastructure: [19:39]–[20:29]

---

Conclusion

"A Home for Every Child" has rapidly outpaced initial expectations, with over half of U.S. states onboard and a strong emphasis on making evidence-based, data-driven improvements in foster care. The initiative is marked by a new spirit of collaboration and competition among states, the adoption of advanced analytics and automation, and an enduring focus on the kids and families behind every data point. The next phase includes real-time data sharing, state-by-state performance comparisons, and significant grant funding to catalyze best-in-class solutions nationwide.

Transcript

[A]

Hello and welcome to State Scoop's Priorities podcast. I'm Sophia Foxo, a reporter for StateScoop. This week, assistant Secretary Adams joins the podcast to discuss the Administration for Children and Families Home for Every Child initiative and how modernizing child welfare systems with predictive algorithms and automated tools can better support families in need. But first, here are the biggest state IT stories of the week. As Missouri hosts its first set of World cup matches this week, state lawmakers approved new restrictions on drone operations near critical infrastructure facilities as part of a sweeping public safety bill governor Mike Kehoe signed last week. The bill lowers the threshold for protected open air facilities from venues with a capacity of 5,000 people to those accommodating just 500. It also increases penalties for violators using drones to deliver weapons, explosives or other prohibited items into restricted areas. A coalition of state lawmakers is once again warning against a federal proposal to preempt state laws regulating artificial intelligence, submitting a letter to Congress on Tuesday urging the nation's lawmakers to oppose the Great American AI act, which was released in draft form earlier this month. In the letter, they argue that the three year moratorium of state AI laws that regulate the development of certain AI models would tie the hands of lawmakers at a moment of rapid technological transformation. Speaking of artificial intelligence, Anthropic recently launched a new cyber defense program aimed at helping state, local, tribal and territorial governments better protect critical infrastructure like power, water, transit and port authorities. The AI giant is committing to up to $15 million in total Claude credits across the program the cost per usage units Anthropic assigns to its AI services. Last year, the Administration for Children and Families launched a Home for Every Child, a national initiative aimed at improving the ratio of foster homes to the number of children in care through preventative measures and technological solutions like predictive algorithms and automation tools. Currently, the national average is roughly 0.57, meaning that for every 100 children entering the foster care system in the United States, there are only about 57 licensed foster homes available. In the eight months since it launched, 25 states have joined the initiative, pledging to upgrade their child welfare IT systems and modernize how state and local agencies support support families in need through a monthly performance improvement plan. ACF has also announced a $7 million innovation challenge that will start in October, potentially incentivizing more states to join. Though the number of states, roughly half the country, has exceeded ACF Assistant Secretary Alex Adams expectations, he said he hopes that by changing the reporting structure and reducing the number of data points required required, it will allow Child welfare agencies to make more evidence based decisions and not just rely on gut instinct.

[B]

I would say we're just getting started. So you know, when President Trump and first lady Melania Trump signed the Fostering the Future executive order last November, they basically set out a whole of government approach to improving child welfare. And ACF is proud to play a part of it. And A Home for every Child is one of our physical manifestations of executing on that successful executive order. You never know quite what to expect when you announce an initiative like A Home for every Child because states have to commit to certain policies and data reporting requirements and things like that as a condition of participation. But certainly it goes without saying that we far exceeded any reasonable expectation of participation to be here in June of 26 and have more than half the country sign on exceeded any of our wildest expectations. So it's great to see it spread. It's great to see states adopt that mindset of managing supply relative to demand. And you know, it goes without saying we're trying to get states to increase their ratio of homes to kids. And there's two ways of doing that. States can either increase their numerator of the number of homes through diligent recruitment, kin first policies, better retention of existing foster homes, or most importantly, states can shrink the number of kids coming into foster care through effective prevention, getting those in care paths to permanency faster. And everything that you're going to see from our agency is designed to move the needle on either the numerator of homes or the denominator of kids. And uh, you touched on some of the ways that we're doing that through upcoming grants.

[C]

Yeah, I mean it sounds like I remember when we first talked, I think earlier this year about the initiative, I think the expectation or the hope was to get at least 10 states to, to join. And now as you said, you have half the country and there seems to be a lot of enthusiasm for this type of effort and this type of work. What have, what has ACF learned so far? What have the states learned or, or changed since, since joining this effort?

[B]

Well, certainly we did learn that there was a lot of bottled up energy for change. I think there was a recognition that the current practices were not working. And just so your listeners know, when a state signs on to a home for every child, what does that mean? What is the next step? A state then will negotiate a PIP with acf. A PIP as a program improvement plan where they outline their goals, their strategies and their measures to increase the ratio of homes to kids in the past, every state was under a pip with ACF where they had to address up to 14 different measures. They had to write long form reports that averaged between 200 and 300 pages and just objectively, you know, had not worked. And what we're getting states to do is we're getting out of the mindset of writing all of these, you know, hundred pages of plans, then checking in at the backside of it to see if they worked. And we're getting states to instead commit to looking at the right measures more frequently, monthly in this case. So every month we'll get data from these states that show their ratio of homes to kids, as well as some wraparound measures related to safety and permanency. And that will set up a system where states will be able to compare themselves to each other. How is Idaho doing relative to Utah? How is Mississippi doing relative to Louisiana? And then it will also allow them to see how they're doing over time. So both having those cross jurisdictional comparisons as well as time series comparisons, we think will unlock competition, we think will unlock innovation and lead to better outcomes for children and families in a manner we just have not seen in the past.

[C]

I'm really happy to hear that. And especially the enthusiasm that states want to participate in an initiative like this, despite the increased reporting measures that have been put in place. I know in April launched a dashboard or a map, an interactive tool that also displays a visual representation of the comparison that you're speaking about. Have you noticed because about half the country has now joined the initiative, have you noticed that there are certain challenges specific to particular regions of the country? I'm thinking about the Southwest versus the Southeast, the Midwest versus more western states or more northern states. Do you notice that similar challenges plague child welfare offices across the country, or do you think that they are specific to regional issues?

[B]

Well, a couple things I would say first is, you know, if you look at who is participating in a home for every child, in terms of the states that have signed on, it is a diverse group. I mean, we have some that border Canada, red states, we have blue states, we have rural states, we have urban states. So in terms of where the enthusiasm is coming from, we've seen it all over the map. I would say probably the rate limiting step for states signing on has been those that have county administered models, those that are state supervised, county administered have additional layers of sign off that they have to go through to join this initiative. Because when a state signs on, they're provided a Stata monthly, therefore they have to get agreement from those counties to get that data. So probably if you see gaps, the number one reason for gaps today is state supervised county administered states. In terms of differences that we're seeing in terms of outcomes, I would say it's premature announces that they're going to join a home for every child. Usually that commitment is the first step and then it's followed up by us negotiating a pipe of ACF in the state. And then once that PIP is complete, signed and negotiated, then the state goes live and starts reporting about a month after the go live date. So I would say it's premature. You referenced in the intro we announced an innovation challenge where states that achieve the highest ratio or the highest growth rate in the.

[C]

Yes, please tell me about that.

[B]

Yeah, no, it's designed to up the ante and bring some resources to bear to the states that achieve the best results. As part of that, to participate in that innovation challenge, a state has to make the decision to join or not join a home for every child by June 30th of this year. So the reason for that is we're trying to get those states online and reporting to coincide with the start of the new fiscal year. On October 1, say later this year we'll have that state by state data that's reported to us monthly where you can do cross state comparisons and then you can look at the time series trends. We have a lot of great data on our website today. The biggest challenge with it is it's two years out of date. I mean business that makes data driven decisions based on data that's two years out of date. So I mean there's certainly some regional trends based on that data. States in Appalachia, for example, tend to have higher entry rates in foster care, likely correlated to opioid use, things like that. But I've always been hesitant to talk in broad strokes based on that older data. What I want to use this new data that's coming into us for since we'll have near real time data, monthly reports and measures that we think are high leverage. What I want to start doing at ACF is use that to build out our technical assistance and best practices. Right now we offer a lot of technical assistance that's based on gut feel, that's based on stakeholders impressions of what's working and what's not working. But I want the data to take us that destination as part of the enthusiasm of working with states. We've had a lot of governors do events with us announcing they're joining and oftentimes they'll have Stakeholders there, former foster youth, current and former foster parents, and it kind of becomes a rah rah event. And one of the ones we read recently in Indiana, the director in Indiana gave me a sticker for my phone that says data is greater than opinions and.

[C]

Oh, wow.

[B]

I think that's kind of a nice synopsis of what we're trying to achieve with this near real time data and leveraging it into improving outcomes for kids.

[C]

You can't see me right now because we're all off camera, but I'm smiling because that really warmed my heart. I appreciate you sharing that because I think with data driven decisions, it's so easy to concentrate on the numbers, but there is a human face behind every number. Whether that's a foster family, whether that's a foster child, and to be able to advance and improve these child welfare outcomes, whether it's preventative measures to keep children in the home, whether it's increased foster families to be able to care for them when they are in need, I think that's incredibly special. And I, I just don't. It's. It's so easy to talk about technology and numbers and data and sometimes forget that we were actually talking about, you know, children and families and people whose real lives are impacted by the, the work that's, that's being done. Um, so I really appreciate you, you sharing that, without a doubt.

[B]

And I was just going to say, you know, a lot of this was based on what I had observed in Idaho. Prior to this role. I ran Idaho's HHS agency and like many states, we had a shortage of foster homes, which meant when a kid came into the system, particularly older youth or foster sibling groups, many of them got placed in non traditional settings like government offices. They got placed in Airbnbs or hotels, which, oh my goodness, you know, caregiver. So I noticed once we started committing to getting our ratio of homes to kids up through effective prevention time, lawyer, permanency, kinship, licensing, all of those types of things, it wasn't just moving a number, it was transforming outcomes. We went from 74 homes for every 100 kids up to 104 homes for every 100 kids pretty quickly. And in between, we got every kid out of those non traditional settings. We stopped using group homes by about 30%. We got kids that had been placed out of state back in state so that they're closer to their biological parents. And we saw how moving those numbers moved lives and moved outcomes. And that's what excites me about this.

[C]

Yeah, absolutely. That sounds amazing. That Sounds like great work. I very candidly, I have two people in my family who were adopted out of the foster care system here in California and they were two older kids when they were adopted. They're my aunt and uncle now, but they were adopted when they were I believe 10 and 12. So my grandparents really wanted to be able to, you know, provide that, that care and love forward and so very fortunate that they were able to, to move into that. And I know how difficult it is to try and find the right fit, the right family for that. So I appreciate this, this data driven enthusiasm. Beyond proving the foster care ratio, know that that is the, the core focus. Are there any other success metrics that you're looking for in these performance improvement plans? Whether it's, you know, better operational efficiency, more technology that's being used in the offices, whether it's predictive Algori assessment tools. That way we get away from gut driven and instinctual decisions and subjective opinions to that more evidence based decision making that you're talking about.

[B]

Yeah, I mean I think the first thing I would say is we wanted states to determine their best paths to achieving their ratio of homes to kids. So it might look a little bit different in each state. You know, Oklahoma is the first to get their PIP in and one of the measures they're looking at is they want to improve their adverse placement score, which is a mouthful. What they're looking at is they want to improve the number of placements in good settings. So kin first, not with kin. They want to improve the number of placements in the same county so that kids are close to their schools and close to relatives and their friend group and all of that. So you know, some states like Oklahoma are focused on improving placements, others are laser focused on prevention, some are focused on permanency. So what we wanted to do is allow states knowing all 50 states are different and our laboratories of democracy in their own right pick different things that they think would move the needle in their state and then the proof will be in the pudding. Once we start getting reports from multiple states, we'll be able to compare, like I said, how Idaho looks relative to Utah. And if Utah is doing better, nothing will motivate Idaho to change some practices, change some policies or say what is Utah doing that they're achieving better results than us? So I think it will do a couple things. By allowing states to channel their energy and their focus on a few high leverage things, I think we're going to redirect a lot of energy and reclaim hours that were Otherwise squandered, filling out paperwork or checking boxes. We do think there are a couple things that are pretty interesting and could help states improve their ratio. One of those, as you said, is predictive analytics or predictive risk modeling. It's been tested in a couple jurisdictions, mostly at the county level, but more recently at the statewide level in Idaho where people have used it on intake. Usually intake is the first step in a foster care trajectory. It's when somebody calls the hotline alleging neglect or abuse of a child. And I used to go sit in with our intake workers in Idaho and the gravity of their jobs was never lost on me. I mean, the information that they can glean in those calls sets off a trajectory that has the prospects of changing a child's life or a family's life. So it's really important we get it right.

[C]

Absolutely. I can only imagine how difficult those conversations must be and how clear headed that the intake workers have to be and be focused to be able to provide the immediacy of help and assessment in real time.

[B]

Yeah, and there's, you know, sit in on a couple of those calls and they're emotional roller coasters. And certainly it helps explain why turnover is high and burnout profession. But we think it's an area that could be improved by bringing some additional data to help with the human decision making. Not to supplant the human decision making, but to supplement the human decision making. So some states have experimented with predictive analytics that will bring in additional data to help triage cases, help the intake workers better see around corners and see some things that might not have otherwise come out on a verbal phone call from a healthcare worker or a school employee. There's been a little bit of data that suggests in some jurisdictions it's helped better triage cases better risk, stratify them and lead to a better and more proportionate response and a right sized response. So because it's been mostly tested at the county level, not at the state level, we did announce $6 million in grants as grants available to states. We didn't limit the use case to intake. There was potential use cases for improving placements and placement stability. So any state can apply and states can outline a use case that they think will help them either increase their ratio of homes to kids through better placement and recruitment, or better triage and helping reduce the number of kids safely coming into foster care.

[C]

Do you think that states already have the type of data infrastructure to be able to successfully implement these types of predictive risk models? I'm imagining that for a lot of states, because I cover this pretty frequently, that a lot of states, their agencies are siloed and their data is siloed and it's not integrated well or doesn't have the type of data sharing infrastructure that will allow for these predictive risk models to kind of analyze and accumulate all of the data across the board, across, you know, education, healthcare, workforce, Medicaid, anything that would involve child welfare. Do you find that states need to first improve their data infrastructure in order to be able to adopt this tool? Or do you think most states are already on the path where this tool can be helpful now?

[B]

Well, I think we're on a path that can be helpful now. And I think we've heard enough from states that are ready to test this and ready to go live. They just needed a little push. As I said, I came from Idaho. We were the first state to go live with predictive risk modeling statewide. Idaho, as you know, is a small state, 2 million people, rural, with some disparate data collection systems and, and say empirically, if we got it done at Idaho, I'm confident anyway, what surprised me the most, there's vendors in this space. We worked with a particular vendor who was very hands on and helped us do the data mapping and set this up the right way. I was surprised at the cost. Wasn't anywhere in the ballpark of what you expect. It was a lot lower than I expected, to be perfectly honest. So when we said we made $6 million available for state grants, we think that's going to help at least 10 states go live, allowed those 10 states to try different use cases with a goal of then informing and encouraging other states. In Idaho, you know, I went live with a grant. I think the Casey Family foundation gave us a grant to do the initial upfront costs and we wanted to see if it worked. We wanted to see if we thought it improved our response. We wanted to see if we thought it improved child welfare outcomes. Obviously, I've since left the agency, but I did get a note from the team recently telling me that they found it to work and that they're going to continue to fund it ongoing with their own resources. Which is the strongest sign that it worked. That's great. Oftentimes competitive grants, you kind of feel like Johnny Appleseed, you're trying to plant some seeds and hope trees and people will continue to harvest them. And just based on my own experience having gone through this at the state level, I'm confident this is going to plant some trees that will grow and will Undoubtedly, with successful implementation from the states improve child welfare outcomes, there certainly

[C]

seems to be an appetite for that. And just to clarify, so states have until the end of this month, June 30, to apply and register for the Home for Every Child Innovation Challenge. And with that, they have to agree or not agree to sign on to the Home for Every Child Innovation initiative. And then the competition, the Innovation Challenge, starts in October. Do I have that correct?

[B]

That's correct. I mean, theoretically, a state could join a Home for Every Child Innovation Challenge and not formally, you know, enter into a PIP with us, but they would have to provide the monthly data for. For tracking. So I don't know what the advantage would be, but mostly correct. Yeah.

[C]

Okay, great. And then tell me a little bit about the competition. How long does it last and what are the Prizes?

[B]

Goes through September 30th next year. So I believe that the top prize for the state that achieves the highest gets 3 million. And then I believe second prize is 2 million. I think the states with the highest increase is a million each.

[C]

I mean, America loves competition and I think it breeds a healthy marketplace to be able to say that you are the best at something. But I mean, it lifts everyone up to be able to just even make a measurable stride for improvement. So I think this is a great way to kind of gamify this effort and really make it and incentivize states to participate, which it seems to be, there's a lot of enthusiasm for.

[B]

I joked, I mean, we've done about 20 events with governors on this, and the thing that comes up most frequently is that spirit of competition, healthy competition. You know, we had one governor put it saying, you know, well, we hope to win, but more importantly, we want all kids to win. We know this is going to be a win for all kids. We've been pleased with the response from states to date. Diverse states, red states, blue states, mountainous states, rural states, you know, everything in between. But olive branch remains extended and invitation remains open to all remaining states and would love to get as many participating as possible.

[A]

Thank you, Assistant Secretary Adams, for participating in that conversation. You can subscribe to the priorities [e-mail] and wherever you get your podcast while you're there, be sure to leave a review or rating on the podcast page. That small extra step helps more people like you find the show. This podcast is a production of Satisfaction Scoop News group in Washington, D.C. our producer, Carlin Fisher puts it together. Until next week, I'm Sophia Foxoell. Thanks for listening.