ANNOUNCER: Brick by Brick is made possible thanks to leading support from: ...and many more.

Thank you.

We couldn't do this work without you.

THOMPSON: Hello and welcome to Brick by Brick.

I'm your host, Ann Thompson.

We're so happy you could join us for the television premiere of this multi-platform series from CET and Think TV.

Here, we use solutions journalism to explore credible responses to challenges facing our cities in Southwest Ohio.

Right now, we're focusing on the housing crisis and community development.

These issues are important because housing and our neighborhoods are so foundational in the trajectory of our lives.

Whether you consider yourself poor, well-off, or in the middle somewhere, we all need a safe place where we can also thrive.

You can find additional reporting by our team, explore more solutions via our podcast, and even share your neighborhood vision by engaging with us at CETConnect.org and ThinkTV.org or email us at BrickbyBrick @publicmediaconnect.org.

We can't wait to hear your ideas and what you think about today's topic.

We're looking at a high tech response to the challenge of eviction and ultimately, homelessness in Greater Cincinnati and beyond.

SCOTT: People should care.

A lot of people care about us, you know what I mean?

Just because we're homeless don't mean we're bad people.

JOHNSON: Don't be ashamed because we all have situations.

So don't be ashamed.

Don't hold your head down.

Don't.

Hold your head up high and reach out and get the help that you need.

THOMPSON: Eviction numbers have peaked once again following the pandemic, and there's a record number of people experiencing homelessness across the country.

National experts point to a number of causes, including rising rents, expiring COVID money, and in some states, a strain on the shelter system from the migrant crisis.

Meanwhile, wages haven't kept up.

RIEGEL: And in order to afford modest housing, a two bedroom apartment in the state of Ohio, a person has to earn $20.81 an hour.

THOMPSON: While shelters all over the region try to assist as many families as they can, they are only able to help 14% of those who reach out for help, which means many others are forced to make tough economic decisions each month between paying rent or another urgent bill.

LEONARD: How can we free up other funds, or point you to other services that allow you to free up some money that you might have to make a decision rent or water, or rent or medical bill?

THOMPSON: Now area homeless prevention agencies have a bold idea to keep people housed: identify the vulnerable before they get an eviction notice and proactively offer up assistance.

FINN: This is also very much a mindset shift for a group of nonprofits that are normally used to waiting for somebody to come knock on our door before we considered helping them.

Now we're going to go knock on their door.

THOMPSON: Strategies to End Homelessness and its partners are using predictive data analytics to keep people in their homes.

Coming up, we'll look at how it works, where they got the idea and why there's confidence this innovative solution may move the needle.

Let's get into it.

This is Brick by Brick: Solutions for a Thriving Community.

Hello.

Today's solution is about supporting housing stability amongst our neighbors by preventing the crisis of eviction from taking place.

It's an effort involving predictive data analytics, and while the effort is explicitly working to move the needle to lower evictions, it's hard to ignore the potential secondary ripple effect downstream, a possible reduction in the number of people experiencing homelessness.

Let's start with some context first.

As many have reported, evictions and homelessness are problems that continue to grow.

In Hamilton County, landlords file on average 50 evictions every day.

In 2023 it was close to 13,000.

In Montgomery County, there are about two dozen evictions filed per day.

In 2023 more than 6500 filings took place.

Downstream from that housing issue is homelessness.

When it comes to the unhoused in Ohio, from 2022-23, the number of people experiencing homelessness went up 6%, about half the national trend.

In the Cincinnati area, more than 6000 were experiencing homelessness last year, a slight uptick from years past when we actually saw some declines.

In Montgomery County, it was 5000.

As you can tell from some of these numbers, the latest data lags a bit behind the situation on the ground, which is often worse because not everyone struggling with housing stability is always counted.

Add to that the fact that for every 100 extremely low income people trying to find housing here, there's only 40 units available.

That creates a system where most are paying more than 50% of their income on housing.

That kind of cost burden can be a recipe for crisis.

Amy Riegel, Executive Director of the Coalition on Homelessness and Housing in Ohio, points to wages not staying in line with rising rents.

She's also noticing more people are living on the streets because they can't get into a shelter.

RIEGEL: Individuals who we see either perhaps living in a park or in a doorway, or even in a place that's not fit for human habitation, like a car or abandoned home.

Those numbers are on the rise.

We're seeing more families entering homelessness.

THOMPSON: In her 20 years in the field, Riegel says she's rarely heard of a family ever experiencing unsheltered homelessness, but she is now.

RIEGEL: It was somewhat that marker in our souls of we would never allow a baby to spend a night in unsheltered homelessness.

And now we're hearing reports all across the state of social service workers going out and finding families with young babies living in a car.

THOMPSON: President of Cincinnati Strategies to End Homelessness, Kevin Finn, doesn't want people living on the streets.

Statistics show there are three times as likely to die that way.

The problem is shelters are packed and it's expensive to house people in them.

Finn says we have a good, coordinated system of services in the city, but it's more cost effective to keep people in their own place rather than rehouse them after they're displaced.

With assistance from his agency and others, it's just $1,600 to help somebody find an apartment who was sleeping on a friend's couch, compared to $4,700 to secure housing once somebody is already out on the street.

While eviction and homelessness prevention services are already part of the approach for Strategies to End Homelessness and their partners, Finn and others are now trying something different to keep people in their homes: predictive data analytics.

Developing algorithms and plugging in data to predict things is nothing new.

The health care industry uses it to improve patient care, banks use it to predict whether somebody is too big a risk for a loan, and online stores use it to suggest what you might want to buy next.

Kevin Finn wants to predict who is most likely to become homeless before they do, and offer up assistance to keep them housed.

As you might guess, the more data points, the more accurate the prediction.

He's identified 40 different types of data to plug in, including things like utility shut off notices.

FINN: We already have about a half dozen that we're using, and it's everything from some historical data of people who previously experienced evictions, previously were homeless, had called the Central Access Point Helpline that people call for information about housing resources.

We have data from St. Vincent de Paul about people and services that they have delivered.

THOMPSON: The data is fed into an algorithm managed by partners at the University of Cincinnati.

It will be updated every few months.

For privacy reasons, the information is anonymous to everyone except social workers.

They're given the names of people the computer says need the most help.

Then they work to provide financial assistance and, if necessary, a new place to live.

The City of Cincinnati partially funded this effort with a two year, $2 million grant which supports the social service outreach and assistance.

Additional funds are raised to support the data collection, management, and analysis.

Finn anticipates being able to initially help 160 families a year.

The program started in July.

Finn says many people just need a leg up.

FINN: I've also never met a person, really.

Who's homeless for just one reason.

I mean, typically it was the interplay of 2 or 3.

Or maybe even 4 things that all sort of went wrong.

THOMPSON: Brick by Brick's Hernz Laguerre, Jr. joins me now.

Hey, Hernz.

LAGUERRE: Hey, Ann.

THOMPSON: Hernz, you met a woman who could have potentially benefited from predictive data analytics if the program had been around when she needed it.

LAGUERRE: Yeah, Ann, you know, we generally live in a very reactive society.

Somebody gets an eviction notice, they become homeless.

So what do we do?

We try to get them a homeless shelter.

This program is proactive and can potentially step in before the eviction notice is even written.

I met with the people who are working on this program with data analytics, and I met with a single mother who said a program like this could have prevented her eviction.

On this warm and sunny day, Bethany House Services, a nonprofit organization dedicated to helping the community's most vulnerable families hosted their anniversary cookout.

Families living in the shelter and the community alike enjoyed music and merriment while celebrating 41 years of serving people hit with life's toughest challenges.

People like Devon Cade, who was evicted and faced hardships with her two boys.

CADE: I was at a hard time in my life where, huh, I went homeless because of my rent, so I ended up sleeping in a park for three months.

LAGUERRE: Cade currently stays at the Shelter House with her two boys.

Her limited income couldn't cover her expanding rent, so she got evicted and found herself in an unimaginable position.

CADE: And I feel like I can't get out of this hole because y'all didn't be patient and try to work with people.

I seen a lot of people's homeless all my life, but I would never think that I would go through it.

LAGUERRE: How do people find themselves in these vulnerable situations?

Kevin Finn, CEO of Strategies to End Homelessness, shares how.

FINN: Homelessness and eviction, or people being at risk of eviction tend to be caused by pretty short-term crises.

A lot of times it can be as simple as I can't get to work because the car needs a repair, or, you know, all sorts of things like that that are, you know, fairly short term issues.

LAGUERRE: Unexpected life events can happen to anyone, and when you lack resources, those events put you at a greater risk of homelessness.

The agency Strategies to End Homelessness is working on a solution to step in right when those unexpected events happen.

FINN: Okay, let's catch people before they lose their housing.

LAGUERRE: With the help of funding through the City of Cincinnati's Impact Award, Strategies to End Homelessness is the lead agency in a new system called the Cincinnati Family Housing Stabilization Collaborative.

This program uses data analytics to prevent evictions.

FINN: Let's find people before they have an eviction notice in their hand and head off their sort of new or emerging crisis as early as we can.

And my experience has been that the further upstream you go with these prevention activities, it costs less to help them and you get better outcomes for your investment.

LAGUERRE: The way this works is that data is collected from multiple sources such as Hamilton County, the City of Cincinnati, and organizations alike, masking the names that the data belongs to for privacy.

The data is used to build a profile on a household, using information that shows that they used emergency rental assistance or applied for services at places like the Freestore Foodbank.

Then it gets plugged into an algorithm to see who is likely to get evicted.

Once the system flags a person that is at risk, they are identified.

FINN: We're looking for people that the data tells us are most likely to receive an eviction notice, and then our partner agencies are going to proactively reach out to those families and offer them assistance.

LAGUERRE: The program is new and is still gathering data.

So I asked Lavonia Leonard, who is director of housing at Bethany House Services, one of the partner agencies, about the program's potential with helping those at high risk.

LEONARD: There is an algorithm that says, okay, if an individual has received services from St. Vincent DePaul or Children's Hospital or Job and Family Services or something like that, there is, you know, an algorithm that says they are at risk of potentially getting an eviction or having to come into homelessness, four, six, however many months down the line.

LAGUERRE: A big part of the program is having the most recent contact information for the household, so another phase of the program is set to be released is called Tenant Guard, and it will allow those at risk to register themselves into the system.

LEONARD: Having the right players, getting as many people as involved as we can to have the right conversations and be able to dump the right data into the algorithm to say, yep, Hernz, we need to get Hernz right now.

We need to connect the dots for Hernz to these different, you know, entities.

I think that's the key.

LAGUERRE: Do you feel like your eviction could have been preventable?

CADE: It could have, if I would have got more help or if we had more community resources or stuff like that, certain people wouldn't be so homeless.

LAGUERRE: And the supporting housing can help her shift her focus from her struggles to her aspirations.

CADE: I want a better life.

I want to go back to school since my kids are older now, so let's go around.

Once I get her back on my feet, I want to make sure they're happy.

THOMPSON: Thanks for that story, Hernz.

It seems that the more data you have, the more accurate the prediction is going to be.

LAGUERRE: That's exactly right, Ann.

You know, this program is the most successful when it receives as much information as possible.

So that's why they're trying to connect with entities like Duke Energy and the Public Utilities Commission of Ohio, so that they can see who's the most at high risk.

THOMPSON: And there is another part of the program that you briefly mentioned in your package.

Tell me more about that.

LAGUERRE: Yeah.

You know, the program can't pick up everybody, right?

So there's a way to self-identify yourself if you're having problems with your rent, it's called Tenant Guard.

And if you sign up for the system, it will show you what resources are available for you and your family.

And what's great is that it can tell you about resources you may not have known about in the first place.

You can find that link and other resources on our web page.

THOMPSON: Multimedia journalist Hernz Laguerre, Jr., thanks very much.

LAGUERRE: No problem.

THOMPSON: Still ahead in the program, the use of predictive data analytics isn't just a big idea, it's a proven concept.

LITTLEMORE: Because I think we're all a little scared.

I mean, I was skeptical in the beginning that you could use data analytics to prevent homelessness.

But I think what's different about Mason is we've actually been able to prove that yes, you can.

THOMPSON: What a community in England accomplished to inspire other cities like ours.

Plus, a deep dive into the data with University of Cincinnati economist Gary Painter.

We chat with him in the studio next.

And similar cities think alike.

We hear from a Dayton based professor about how he's working with students to connect the data points around homelessness in Montgomery County.

That's all ahead.

Joining me now in studio is a key player in this predictive data analytics project being led up by Strategies to End Homelessness and others.

He's one of the people behind the scenes working with the data and algorithms, University of Cincinnati economist and Director of the UC Real Estate Center, Gary Painter.

Gary, thanks for joining us.

PAINTER: It's a pleasure to be here with you.

THOMPSON: So you have a difficult job working with nonprofits, helping them figure out how to predict who might be evicted.

How did you do this before?

PAINTER: Well, in the past, a nonprofit would just simply observe that someone showed up again and again and again at their doors and say, "Well, something must be wrong because you've come multiple times."

You know, maybe a decade ago, maybe 15 years ago, people said, "Well, if the first time someone comes in the door, we could give them a survey, then I could assess how vulnerable they really are, what kind of risk factors are in their lives?"

And so that survey was used to guide choices and deploying resources to folks.

But we've also found that both of those methods don't necessarily lead to good outcomes, equitable outcomes.

And they actually take a lot of time.

Either time in terms of seeing someone over and over again or actually time to fill out a survey.

And so that's tends to be how we had approached providing services to people in the past.

THOMPSON: So now you're doing something different.

And where did you start?

PAINTER: Well, what we're trying to do nowadays within the field, across the US and in many different places is to try to actually look at data on service utilization, histories and then to actually use that data to identify people who are at high risk of adverse events.

The work that I'm doing is looking at adverse housing events, whether it's eviction or homelessness or returns to homelessness, and to see if we can intervene before those adverse events happen.

Because we know if someone experiences an adverse event in their life, it's very costly for them.

Things like if you get evicted, then your credit history gets dinged.

It may be increasingly difficult to be stably housed in the future.

THOMPSON: So you mentioned some of those data points.

How many data points do you need for this program to be effective?

PAINTER: Well, we know the more the better.

We don't really know in this new work how much is enough?

You know, can you basically rely on service data from two systems, a health system and, you know, a social service system?

Do you need data on job histories?

Do you need data?

We don't know.

But we do know that the more data that you have, the more effective the model building can be to identify people who are at high risk of these adverse events.

THOMPSON: Is there more than one algorithm and then how often will it be updated depending upon what information you get?

PAINTER: Well, there actually are a plethora of possible algorithms that can be deployed to assess risk and vulnerability in populations.

And so yes, there's absolutely more than one algorithm.

And what our job behind the scenes is to figure out what algorithms are doing a better job of predicting adverse events.

Then we can take those algorithms and deploy them to identify high risk populations.

So whenever we get new data, we recognize that there could be a better algorithm once we have those data.

And so as you have small amounts of data, it's likely that certain algorithms are going to be better at actually assessing risk.

As you get more and more data, I think it opens up the possibilities in terms of what kinds of algorithms will provide more precise risk estimates.

THOMPSON: And will you be updating them, let's say, every three months?

PAINTER: The hope is to be able to, you know, as much in real time as possible to update both the algorithms and the data that feed into the algorithms.

You know, three months is something that I would love to get refreshes of the data every three months, every month.

Recognizing that that is hard to do sometimes when we work with our partners.

Maybe it'll be six months, but our hope is to move into a real time assessment of risk.

THOMPSON: So how do you see predictive data analytics evolving here?

We've talked about it through the lens of eviction and then also you're probably seeing it could help with homelessness.

PAINTER: Yeah, absolutely.

There's any time you're trying to work with a vulnerable population, there's the possibility that using data on previous service histories can provide insights into future service utilization and future vulnerability.

And so while I have expertise in housing and homelessness, there's any numbers of areas where you could see these approaches deployed.

So if you're trying to look at the risk that someone might be taken into foster care, you could imagine, you know, these kinds of algorithms being applied there.

If you're looking at the risk that someone might go back to prison after leaving prison, you could imagine lots of algorithms that could be useful there.

So again, the more data we have, the more questions we can apply those approaches to and hopefully reduce risk and vulnerability in our cities and communities.

THOMPSON: University of Cincinnati economist and Director of the UC Real Estate Center, Gary Painter.

Gary, thanks for being on with Brick by Brick.

PAINTER: Just happy to be here.

Thank you.

THOMPSON: The number of people experiencing homelessness in Maidstone, England, was up 60% in 2018, and the borough southeast of London, population 200,000 and known for manufacturing paper, decided it had to do something.

It teamed with technology partners Ernst and Young and Accenture to try to reduce that number through predictive data analytics.

T here are also U.S. cities who are using predictive data analytics, Los Angeles and even South Bend, Indiana.

But they haven't been doing it as long as Maidstone, England.

We wanted to hear more and talk to their head of housing, John Littlemore.

LITTLEMORE: Well, if you could look at three months ahead, six months ahead and try to help people at that early stage so that they didn't come through at the moment of crisis.

THOMPSON: Maidstone is doing that by aggregating lots of data.

In the beginning, like here in Cincinnati, Littlemore admits it was difficult getting information to predict who needed help.

But once they did, caseworkers could get a lead time of at least eight weeks before somebody goes into a shelter or is out on the streets.

In the first year, their program called One View generated more than 650 alerts.

Of those identified as the highest risk, only 4% became homeless.

It also saved Maidstone about $300,000 in administrative costs and about 3 million in broader societal savings to the community in the first year.

And the borough is starting to use predictive data analytics to help people in other ways, too.

LITTLEMORE: The use of this is, you know, is unlimited really.

I mean, because of what we're doing, our colleagues in social care now are looking to be able to use the product to prevent children having to go into care.

We're using the product to try and predict where people might be subject to domestic abuse.

So, you know, its application is quite limitless.

THOMPSON: Littlemore says Maidstone wishes it had more money to put into the prevention of these problems.

His advice to other cities who want to start using predictive data analytics?

LITTLEMORE: Be ambitious.

Don't be afraid to take on these challenges.

There are very many clever people out there in the private sector that can help you overcome some of the challenges around that.

So work collaboratively with the private sector as well to deliver this change.

But yeah, be bold.

THOMPSON: In the latest assessment, Maidstone had a 98% success rate in achieving its goals, so pretty good results.

Stateside, Los Angeles County and South Bend, Indiana are also reporting positive results.

You can read more about those efforts on our website.

As Cincinnati gets its program off the ground, there's a much smaller effort underway in Montgomery County.

University of Dayton professor of Electrical and Computer Engineering, Raul Ordonez teaches a class called Engineering Systems for the Common Good.

Just like he does with robots, the idea is to produce mathematical models using data.

In this case, it's to help solve social problems.

ORDONEZ: So what if we spend more money on, I don't know, say, more prevention or permanent housing?

So what would happen then?

So that ideally then the county could take these models and then start to study policy making in a more systematic way?

THOMPSON: One reason Ordonez is interested in preventing evictions and ultimately homelessness, is the poverty he saw in his native Ecuador.

Right now, he's trying to gather more Montgomery County data and build partnerships with others, like the University of Cincinnati's Gary Painter.

Obviously, this effort is in its early stages and we'll keep an eye on it at Brick by Brick.

Remember, if you want to learn more about this solution or others that we've started exploring, there are plenty of resources including web articles, video and audio stories, and online extras.

Go to CETConnect.org and ThinkTV.org.

And while you're there, click on one of the big green buttons to give some feedback or answer our current audience questions.

We'd really like to hear from you.

This is the time on the show when we reflect on the solution that we've presented.

For me, there are a couple of key takeaways from our investigation of this solution, including some limitations.

For example, as you've heard, evictions and homelessness are on the rise, and more and more families are out on the street because they can't get into a shelter.

Nonprofits are realizing they have to do something to prevent this and other cases of homelessness.

With predictive data analytics, for the first time, we have a real time snapshot of who is at risk of becoming evicted or homeless and can possibly step in and help them weeks in advance.

There is data that shows it works.

Maidstone, England, says it reduced homelessness by 40%, but there are challenges.

Nonprofits may not be able to reach the people who need help because let's say they don't have the correct contact information.

In addition, it's hard to get data because not all agencies and companies want to give it up, and without it, the system isn't as accurate as it could be.

And the annual cost of this program is unclear and still to be determined if it's sustainable.

We'll get answers to some of these questions as the program develops.

Does it have value?

Kevin Finn of Strategies to End Homelessness says he's hoping to prevent 160 families a year from becoming homeless with current funding.

The communities who are doing this say it is in fact saving them money by keeping people housed.

We'll be following the story here on Brick by Brick.

Coming up on the next episode of Brick by Brick, the Housing Choice voucher program has been around since the 1970s, but is it effective, and could expanding it help respond to the affordable housing crisis as well as reduce poverty?

We'll look at public housing solutions.

Until next month, for Cincinnati based journalist Hernz Laguerre, Jr. And our Dayton based correspondent Amico Moore, I'm Anne Thompson.

We'll be back to explore more solutions soon.

We hope you'll join us then, and maybe even invite a neighbor to watch with you.

Take care.

ANNOUNCER: Brick by Brick is made possible thanks to leading support from: ...and many more.

Thank you.

We couldn't do this work without you.

CAPTIONS: Maverick Captioning CIN OH maverickcaptioning.com