Using data visualizations for small college peer comparisons while managing through tough times

Hi, it's Gary Stocker back again with another episode of This Week in College Viability. We're doing kind of a video version this week, and we have a guest who's going to show us some pretty neat software used to analyze college financial performance, enrollment trends, and much, much more. Our guest today is Matt Hendricks. He's the founder of Perspective Data Science. And Matt, welcome to the podcast.

Hey Gary, thanks for having me. Happy to be on the show.

Excellent. So let's jump right into the questions because we're about getting right to the data here at College Viability all the time. I say, let's go to the data. So Matt, I believe we're clearly in the era of Moneyball in higher education. Of course, for those not familiar, Moneyball started with the Oakland Athletics back in the 1990s, and it was an innovative way to use data to run their business. And Matt, what are some of the innovative ways you're seeing colleges use data?

Yeah. So I, I guess I would back up a little bit. Like, so, so in Major League Baseball, they use data forever, right? Like what, what changed in my view with Moneyball was that they started at least one front office, Oakland, the famous case, started using advanced techniques fairly advanced techniques like regression analysis and machine learning to sort of step back and ask big questions first and then sort of test conventional wisdom and then, and then start going from there. So basically from what my read on it is, they started running regressions, right? Like what, what player statistics are most closely predictive of winning games, right? And they found things that were sort of contrary to the conventional wisdom on-base percentage, one variable that stood out. So using regression analysis for that and maybe some machine learning also. And then they took it a step further.

Okay.

Like given the characteristics that are the player stats that are most strongly related to winning games, which players are sort of undervalued out in the marketplace, right? And we'll just put a whole team together of these underpaid players based on, you know, their output. So that was the really novel thing with Moneyball, right? Starting to apply like advanced methods relatively advanced methods, regression, machine learning. I doubt they were doing AI in the '90s, but definitely started doing regression. So that's, that's really where I want higher ed to get. And I think you agree with me, Gary, on that. It's not so much just using data, it's using the advanced techniques that we've been, you know, teaching for years. I taught econometrics for 12 years, right? This is one of the things we teach, regression analysis. And it's a really powerful technique and fairly intuitive as well. So I think really anybody can, can use it and benefit from it. But that's what I'm trying to do is, is use regression analysis, use machine learning, use AI to ask big questions, test the conventional wisdom, find anomalies with operations at a particular university, which you know, begs questions about why do we look so different from what, you know, an AI would expect or from what the regression is telling us we should look like given our characteristics. Uh, those are— that's the really the breakthrough. So what innovative— was your question was what innovative ways are universities using? I mean, I hope— I, I don't know exactly what, um, schools are doing, but I suspect most of them are not, or a lot of them are not, uh, using advanced, you know, machine learning techniques and regression techniques. I say that because a lot of schools, sadly, don't even have institutional research departments. And when they do, they seem to be sometimes underfunded. Like you got a part-time faculty member doing it, that kind of thing. But clearly, like the, you know, top— the right tail is probably doing similar stuff to what we're doing with our apps. But honestly, what I'm trying to do, and I think what you're trying to do, is, is bring this advanced analytics to the, you know, left tail, the schools that don't have institutional research, that, that don't have the— that think they don't have the resources to really employ advanced techniques. They can, right? We can do this for them for cheap. And that's really, I think, where I'd like to get. I'd like every school to be using Moneyball, using advanced techniques and learning from their data.

So, so, Matt, it's fair to say that colleges really aren't even in general, colleges aren't even general close to maximizing the way they can use data and even more importantly predictive analytics.

Yeah. Yeah. I don't, yeah. Like I said, I, I think some schools are happy to just kind of, you know, play a little bit with iPads, do some means. You know, they'll pick a lot of times I see, you know, pretty shoddy analysis where, you know, it's really, it seems like the motivation for some of the work that I see put out in the media and stuff like that is just really to try to make the school look good rather than really taking an honest look at their operations and learning from the data. It's not everybody. I don't want to be really negative, but, you know, I've seen cases where, you know, they're, they're hailing this massive freshman class that came in. Yeah, it's bigger than the previous year's freshman class, but it's far smaller than 3 years ago. Right. Like, and then just leave that fact out. Right. So it's kind of like rather than really being honest and, and digging into the data to try to learn from it, it's more of kind of like, how can we spin this to help to, to make us look good? Right?

Like, I don't think that's how I'm, I'll take the role, I'll take the role, Matt, of being a critic. You know, I call that management by PR. Alright?

Yeah.

And then that's exactly what it is. You did a good, you did a good job of describing that. And of course, these folks are entitled to spin data. We all do it. But that's kind of the role that I think I play and you play as well as well. All right, spin it any way you want, but let's look at the real data. Let's look at the real trends because that's what's more important than what you're trying to spin. Um, tell us then about Perspective Data Science. I think I botched that name at the, at the start. My apologies.

Yeah, Perspective Data Science. Um, so, so like I said, so my background, I'm, I'm an economist. I've worked as a professor for— I was at the University of Tulsa for 12 years, and even before that I was teaching econometrics. My, my training really is in using data, right? I'm an applied economist, PhD in applied economics, which really means a focus on data analytics, machine learning. This new stuff with AI is blowing my mind. I'm trying to stay on the cutting edge of that. So basically what happened with Perspective Data Science is I like to study education policy, and that's what I've done my whole career. My dissertation's on education policy, but I had been focused in K through 12 and actually pre-K and Head Start programs. Until maybe about 2017 or '18. I can't remember the exact switch I made, but basically what happened is I was a professor at the University of Tulsa and all of a sudden the administration pulled out our retirement package and we went from being sort of blissfully unaware of financial issues to all of a sudden like the school is in dire straits, right? They decided to cut most of our academic programs and took our retirement benefits and decided, wow, I should probably look at the higher ed data and see what's really happening. Just because a lot of my colleagues were really in the dark. Most of us were in the dark that kind of wanted to know. I expected to basically confirm everything that the administration was saying, but that turned out to not be the case, right? There were a lot of untested assumptions. And I think this happens a lot. And it's an honest mistake, honestly, you know, your school's struggling, what's an administration automatically going to assume? Well, we got to cut programs, right? We got to cut costs. And the only way to cut costs, um, they assume, is by cutting programs and faculty. Well, turned out to be a false assumption in the case of University of Tulsa. It's not always a false assumption, but basically our instructional spending looked normal, right? And, and so did our faculty size for the most part. What wasn't normal was a lot of other spending, right, on the administrative side. What I call administrative is not necessarily upper admin. It's just non-instructional stuff, a lot of extra staff. So basically, that was when Perspective Data Science was born, right? I started doing benchmarking work, some similar things I'd done for K through 12, and I've been doing it ever since, right? It's a— there's a lot of schools that need a lot of help, including my own school at one point. They're doing great now. But, um, yeah, I just saw a need. So I'm sure the story that you have is similar, right? You saw a need.

Yeah. And that's, that's what I was going to get ready to follow up with. Faculty are still in the dark when colleges make any kind of change to their business model. Faculty in most but not all situations go a little nuts because you didn't tell us, we didn't know. And that's, that's sadly still the case. And for those listening, I want you to know that we're going to use Matt's model here in a couple minutes to focus on Eureka College. In central Illinois to talk about how he uses it and look at the challenges that that college faces as well. So stay tuned for that here in a couple of minutes. And let's jump back into it then, Matt. Simple question, he says teasingly. Is it possible to predict which colleges will close?

Yeah, I think so. I actually have an AI model that can do just that. Actually, I estimated a model that was predicting mergers and acquisitions, but it's really the same model to predict closures. Yeah, it can be done. The question is whether it's accurate, right? And we can make the prediction. I think, I think so. I mean, you have, everybody has a credit rating, right? And it's pretty good at predicting whether you're going to go bankrupt, right? That's why banks use it. It's been tested over time. We have a lot more financial data on universities, right? And we can do probably a much better job of predicting university bankruptcies than we can an individual, like, you know, person, right, applying for a loan. So yes, definitely can be done. Issue is the IPEDS data is not great for predicting closures because it actually doesn't have at least two of the most crucial variables, which is one is change in net assets, unrestricted change in net assets from operations. You need that. That's really the bottom line number for a university. The other one is change, a net change in cash from operating activities. Neither of those are in IPEDS, although I should say the unrestricted change in net assets is now showing up in IPEDS, but you're only going to have like 1 or 2 years now. Yeah, thankfully. But the cash flow is not there. And that's actually the cash flow is the best predictor, right? Once the school goes negative on net cash from operating activities, then lots of problems can't be ignored, right? You have to either borrow or sell assets to pay your bills, right? To pay payroll. So that's when everything goes south quickly. So yeah, you have those variables. You can make great predictions in my opinion.

So Forbes has their annual financial grades, A through D minus. You're familiar with that?

Not really. I may have seen it a while back, but—

So the reason I bring that up, Matt, is they come up with the like two values. They give it a grade, A-, E+, whatever, and then they assign a 0 to 5 point scale value for that. So this college is 4.2, this college is 2.3. Is there value in assigning a relative risk number like Forbes does? They have their own algorithm, their own approach. Is there value in doing that? So your college is 6.1, my college is 1.2.

Yeah, I, I think so, if it's accurate, right? Like, if it's credible, uh, I think it's, it's a good thing and they should do it, right? I, I, that's part of the reason why We do what we do, right, Gary? We want everybody to be informed, right?

Absolutely.

One of the problems in higher ed is that like there, there is a lot of uncertainty. There's a lot of lack of information and it causes problems, right? I mean, I see it on the faculty side, right? The faculty don't know why the school is doing what it's doing. You'll just sort of like fill in the information gap with all sorts of different theories, right? Conspiracy theories, all sorts of stuff. I mean, usually the truth is It's just a lack of information, right? I mean, even the administration, they certainly won't admit it, but maybe they have a lack of information as well. So it creates lots of problems, right? And lack of trust. And it really can crush a university and actually can crush enrollment too, right? If you have low morale, your faculty is mad, they're leaving, the students are going to sense that and leave as well. So I think more information is better. And so I like what Forbes is doing as long as it's an accurate and sort of transparent metric, right?

Okay. So before we get to Eureka College and, and your model, one more question. And that Moneyball, we talked about that at the top of the show. Moneyball has changed all sports, particularly at the college and professional levels. There's no way around it. They've got data that just shows both history and predictive value of a player, of a team, of a record, of who's going to win the Super Bowl, all those kind of things. Just your own opinion, Matt, what's, what's the Moneyball model's possibility in changing higher education?

Yeah, so I started, I, I've started creating some that I thought would be most valuable. One is, you know, just running regressions in a peer set and say, okay, you know, this is run a regression line of say like total expenses, right? And regress that on a bunch of school characteristics in the peer set and then get a predicted value for every school, right? But Usually just focus on the target school and then look at that and say, okay, this is our actual spending. This is what the model predicts that we should be spending. You know, why are we different? Do we have a problem? Those are the kinds of things that I think the sort of Moneyball mentality where they would fit in, into higher ed. And that's really what they do with players, right? They're looking for players that have big residuals, right? That have low salaries, actual low salaries, but are really predicted to have high salaries because of their productivity. Do similar, similar sort of residual analysis in higher ed. It's really useful to identify areas that need a closer look. You know, is our, is our staffing not normal? You know, do we have too many people? Are the salaries not normal? Do our, does our enrollment funnel look weird? I mean, do we have just really low applications, right? These are a lot, lot of things to test with just basic regressions, and it can be quite useful to sort of pinpoint where to, where to sort of focus policy decisions and how to sort of right a ship, uh, for a college.

Okay. And go ahead and show your screen because we're going to now bring up the, uh, the app. And, and while Matt brings that up onto the screen, we're going to focus on Eureka College. It's a small private college in central Illinois, and both from the college viability app that I have created, um, they have some very negative trends. Associated with tuition and fee revenue, enrollment, the graduation rates are low, and lots of things. But my approach is a little bit different, a lot different than Matt's. Matt's is more predictive, analytical. Mine is more trend analysis kind of stuff. And Matt, just, I know before we get started on this, the peer group, the peer institutions is one I see it's on the screen here. Just before, before my first question, tell us how quickly, tell us how you identify peer institutions for Eureka College?

Yeah, this is, uh, just a pretty basic approach. Basically, I have an algorithm. It's a really common algorithm used in machine learning called k-nearest neighbor. Part of that is to find your nearest neighbors, right? It's basically the idea is find similar universities and then compare yourself to the, to those universities. So, um, this peer set you're seeing right here was produced by k-nearest neighbor algorithm, uh, specifically the Euclidean distance part of that algorithm. And basically basically what it does is takes probably a couple dozen variables, you know, revenues, size, Carnegie classification, lots of different things go into the algorithm and the algorithm calculates a distance measure, right? And actually have a graph of that. So the target school is your Eureka College here. And then every school in the world in the sample gets a calculated distance measure, right? Like how similar is that school to Eureka Eureka College.

So it's not necessarily physical distance, it is a distance measure.

Yeah, it's distance in terms of similarity. It's like a similarity score, right? And basically what I do is cut it. That's what the K and the K part is, right? Cut it at a certain number of peers here.

Man, if you would go back, Matt, if you would go back to that list, because I had an observation as I was looking at this recently. As I looked at this, there are at least a handful of colleges in this peer group for Eureka College. That have already closed.

Yeah.

And there's at least that many that in my analysis and maybe in yours are, are almost for sure to close in the coming months and years as well. What does that tell us?

It tells us they're in a group of, uh, schools that are at risk, right? Um, their peers, according to them, and they only have 500 students, they, you know, have very, very small endowment. They're in the really high-risk group of schools, right? Some of them have already closed, right? This algorithm is matching based on— so obviously these schools are still alive for the data that they were getting matched on. But yeah, they're in a high-risk category. And so they need to be— a lot of my analysis, right, is just showing, okay, what would they need to do to be average in the peer set, right? So they would need to— if they really actually wanted to work with me on this, I would suggest they not be average in this peer set, right? Try to be better. Than average in this peer set. You can see it on the key financials page, right? You definitely don't want your net income margin to go negative, right? And the peer set's actually going negative. So there's a bunch of schools in here that are really struggling. And actually, the scary thing is Eureka is doing worse, right, on that key financial metric.

Yeah. And even looking at some— Ohio Valley has closed, they're negative. Lion College has been in trouble for a long time. The only thing that saved them was a sudden donation from somebody. Iowa West St. is closed, McMurray is closed, and I think Northland College is effectively closing, although they're trying to fight it. And so that's an interesting observation. All right, let's use the data, use your visualization as appropriate, but share with us the top. All right, you get a godlike presence here, Matt.

Yeah.

Share with us.

So this is what I do every day is I make these dashboards and then I write reports, right? So this is great. This, what I'm looking at here is net operating income. Right. So everything was going fine for Eureka College. The red trend here, you want this to be at like +1% to +5% range. And that's where they're sort of bouncing around until 2019. And then things get bad here, right? So they have a couple of years. Well, all the years are negative after 2019 and a big drop here. So this is really concerning, especially for fiscal year 2021. And most schools were doing quite well here. Because of a lot of government aid from COVID Right.

And now let me just, just jump in with a clarifying question because I'm trying to think from our viewers we'll talk about. So your data is from iPads and Eureka College itself submitted this data, right?

Actually, this— what you're looking at right here is actually from the audited financial statements.

Okay. Same thing, though. Same thing. Yeah.

Yeah. I should mention that. So this is a dataset that I've created actually my AI. I trained an AI to basically read audits and pull data out. And so what it's pulling out here is change in net assets, unrestricted change in net assets from operations, which as I mentioned, is not— hasn't been traditionally available in IPEDS. And then it's dividing that by operating revenue to get this what we call net operating income margin. And normal value is obviously plus 1%. All right. And they're not normal. Since about 2019. So this is where I usually start, right? Like, where did things go south and what happened? Why did it go south? Right. That makes sense.

Yep. So you also mentioned in some of the notes you sent me in advance of our recording this morning, a rapid increase in expenses.

Mm-hmm. Yeah. So, so basically I was trying to figure out why they went negative after 2019. And it looks like both costs— you look at revenues, you look at expenses. And it seems they have issues on both sides, right? The revenues are lower, their expenses are higher. And so I can show you expenses, the trend here. So 2019 is they're at about $15 million in expenses and fiscal year 2023, they're at $16.5 million. So what is that, $1.5 million increase in expenses? right? And then do the same thing on the revenue side. They have a reduction in $2 million. So really that's what's creating the $3.5 million deficit, right? You've got higher expenses, lower revenue. So then the question is what, what's causing that, right?

Right, right.

Yeah. So what's causing the expense side? You know, I don't know for sure, but I can certainly point to some areas that need to check. One thing that's happened here, I'll just show you institutional support. Institutional support spending has gone up quite a bit since 2019. It was $4.4 million and now it's $5.3 million. So that's almost the full $1.5 million right there. So that actually turns out to be most of where the extra spending is jumping in.

And define institutional support for the viewers.

Yeah, institutional support includes your president's office, provost's office, all of like anything that is like institutional level, like management, your security, your IT infrastructure usually will be in there. But there's like, there's a whole list I have down here actually. They're supposed to follow these guidelines, president, provost, all the VPs, advancement office, business people.

Okay.

And so on. So, right.

So a summary could be for some reason that Eureka College has spent a lot of money the last couple of years on non-academic expenses.

Yeah. Yeah. So it's a red flag, right? So you just, so the, this is how we use the dashboard, right? Like, okay, why did it go up? So we can, we can dig deeper really and figure out like what, what's happening. One, one thing that I suspect is happening is that they've hired a lot of contractors and this is from 990s, by the way. We can see on the 990s how many contractors a university has hired. And you see that there's a sort of explosion after 2019. It went from 4, which was sort of their norm for contractors. And actually that's actually normal. The peer group average here is 5 in 2022 and then kind of hovers around 5, but also explodes after 2019. So I suspect that's the problem. And, you know, may not be an issue if you look at like staffing, which I also can have in the dashboard.

Right.

So one of the things I was wondering is, well, maybe they actually improved efficiency by hiring a bunch of contractors and sort of substituting the contractors for staff. But that is not the case here either. Their staff size actually increases after 2019 as well. So lots of questions to ask there, like were these contractors necessary? What can we do without them? Those kinds of questions are really important, right? Because you can save $1.5 million if you sort of get back at least, right? Each of these contractors getting paid $100K. So if you get back to like 5 instead of 15 on last read, you get close to saving that $1.5 million.

So in your experience, and I know you do these, these visualizations for other colleges, is Eureka typical or atypical and what's going on?

So actually, sadly, most of the, most of the work I do for making these dashboards, I make them for big schools. And it's fine. Like, you know, but honestly, the big schools are not in trouble financially generally. And even if they do have financial issues, it's quite easy for them to, to weather that storm. So honestly, I want to do a lot more small college work. But yeah, Eureka looks similar to other schools that I've seen that have had financial issues. I've made some for small schools that, you know, they, they start running deficits and they do it for several years and they don't. I mean, to me, for basic, based on like press releases and conversations I had with like alumni and faculty, it just seems like they don't really know what the problem is, or they, maybe they're just like hoping it will go away without really changing anything, right? Or not changing the right thing. So in that sense, they are typical, right? It's not always what you expect to be the issue, right? Maybe they're thinking, oh, we got to cut faculty and programs, right? To balance the budget. Well, here it looks like that's not actually the solution. The solution is probably more like cutting cutting financial staff and contractors, right? And once they do that, they would become cash positive, right? And then you got to get revenue up, right? And once they get that revenue up, then everything's great again. Hopefully that would be the sort of plan I would have for them.

But so you talk about revenue up and I know you listen to some of my stuff and you're generally familiar with the market and that Driving tuition and fee revenues in this tough market is going to be problematic for most private colleges. In large part, it may not be an enrollment number, but it might— it's probably the net revenue number from tuition, because colleges, as you and I both know, are discounting tuition tremendously. 55%, something like that, is the average tuition discount. So for Eureka or any other small college to, to grow its enrollment to the point that it contributes positively to its financial health and even its viability. Is, is, is that a reasonable scenario in your opinion?

Yeah, I think so. I mean, you can already see evidence that they have the capability here. Well, so first of all, their price looks good. That's always something I check, right? This is net price. This is actually the price that students pay. And so one, one thing you can see here with the red numbers here, and it's actually, this is a regression result that I can show you further down here. Actually, you can see right here. So the net price for a student, a Title IV student that attends Eureka, including full room and board after getting all their aid, the discount we call it, right? Right. They're paying about $23,000 out of pocket per year to attend Eureka, which is actually quite a strong— it's pretty high. It's like 12, 13% higher than what they pay at these peer institutions on average. So the, the, what that tells me is that their price looks fine in terms of a revenue standpoint. Uh, although it may be too high, right? Like this is another thing that, you know, folks don't often understand is that, um, you know, they may not know that their price is too high or too low. It's not super obvious unless you do this kind of benchmarking. Uh, but interesting thing is their price for, uh, lower income Title IV students is price is about on mark, actually a little low, but their price for high-income students is a bit high. So to me, that says there could be something wrong with the sort of like either need-based aid or merit-based aid. It can kind of muddy the water both ways. So one thing I worry about with schools that have high prices is that they could have a low yield rate. They could have high attrition just because of that, right? Like as an economist, we imagine our consumers are like constantly looking for a better deal to maximize utility, right? And so it could be that a lot of these students are realizing, hey, I can go to this other school over here in Illinois or even a neighboring state and pay a lot less than I am at Eureka. So that could perhaps explain some of the attrition issues. They have low yield, that kind of thing. And they have a nice natural experiment to check it, right? If you want to do Moneyball, here's your chance, right? You can check the attrition rate for this group of students that have normal prices compare that to the attrition of this group of students, right? And then also exploit the fact that prices have bounced around, right? Their net price went up substantially after about 2016, '15. So there's a nice natural experiment to test how sensitive are your students to changes in net price, right? Maybe it's, maybe price is driving the low retention, the low yield rate. These are the kinds of things that you do for Moneyball, right? You ask questions. You test it with regression. And so this dashboard doesn't do that. But, you know, you follow the dashboard with these sort of targeted analyses that are motivated by the dashboard.

So, Matt, I'm going to make you a godlike figure one more time with— all right, with one final question here. And let's create this scenario where you're standing in front of every small private college president in the country. What's the one top message you would give to them in terms of managing their business based on the work that you do?

Yeah. So I think number one is just use the moneyball technique, right? Like, don't— they may not even know it, but don't make assumptions or, or guess about important policies. Really any policy, right? Most of the time, in fact, almost every time, you can first run a rigorous analysis and check all of your assumptions, right? So it gets a lot of folks in trouble when they make a big decision, say, to cut programs or to cut this or that or invest in a certain program, but they didn't really test a lot of the assumptions, right? Maybe your assumption that, you know, we need to cut academic programs is wrong. Maybe the issue is spending other in a different spot in the university. You're going to look foolish in the end by making a decision based on something that you could have tested easily, right? So just don't assume, don't test. Another thing is, even if you do guess right, you're still going to have a problem, right? Like Gary already mentioned, right? If you decide, even if correctly, you say, hey, we're going to cut half of our academic programs, half of our faculty, and that's going to get us a balanced budget. Even if that is the right move to make, you're going to have a really hard time selling that to your faculty, to your board, to your alumni if you don't have something like this dashboard or something like Gary's analysis in the background that you can say, hey, look, this is why I'm making the decision I am, right? This is why we need to cut, you know, whatever major you're talking about, right? We just don't have the money. This you can see for yourself. Here's the analysis. Here's all the data. And then the major benefit of that is that you can get buy-in to your policies, even if you know. And another thing is if you share it and you open it like that, possibility that you're making the wrong move, right? Maybe you have a problem with your analytics. And so I guess another recommendation I would have for a university president is just to be open to potentially being wrong, right? And if you're wrong, change course, right? Just make decisions based on the data, not based on your gut. And don't be afraid to say, hey, my initial assessment was wrong, right? You guys corrected the analysis. Let's do the right thing, right? I think in the end, it probably wouldn't cost you your presidency, right? You may actually build more, you may build more trust with the board and the faculty if you're humble and following the data and being rigorous in the way that you make decisions. So that I recommend that approach.

And I might add to the president, they might even survive the no confidence vote.

Yeah, exactly right. Yeah, you might not get it. So the great thing about being open and sharing exactly what you're doing is that you'll build a lot of support just with reasonable people. Most faculty are data people, right? They're used to making decisions. They teach this every day, right? Used to making decisions based on evidence. So I think if you're open and show exactly why you're doing what you're doing, most faculty members are going to say, yeah, that's, I mean, it's too bad that we have to cut all these things, but it looks like the right move. Right? I think at the end of the day, most faculty would sort of be on board.

Yeah.

You know, you're always going to have folks that are, you know, going to be against it, whatever. But you're not going to, you know, you don't have to worry about a small minority that sort of dismisses evidence.

Okay.

That makes sense.

Well, Matt Hendricks, the founder of Perspective Data Science, has been my guest today. Matt, could we go 7 hours more? Because I'd love to listen to you.

Yeah. Yeah. This is, I love this.

7 hours.

Can I say one more thing to small school presidents? No, I think there's a lot of fear. Like, so small school presidents, I would say one thing to them, just hang on, right? So here's the thing. Here's why they should be a bit optimistic. All right. And, and feel like they're in control because basically you have like a general equilibrium thing happening right now where some schools are going to close. Right. Um, and there's a lot of, a lot of fear about this, like enrollment cliff, and that's going to kill all the small schools. I don't think that's going to happen. And here's the reason. There's always going to be demand for small schools, right? I was— I went to a small school by choice, right? I wanted to be in a small school where I have like a relationship with the professors. There's always going to be demand for that. So what's probably going to happen in the long run here is my guess that a lot of small schools will close. And those kids that leave those small schools that close will go to another small school. On that, right? So that what the implication of that is, is that you don't necessarily have to like outrun the bear, so to speak, right? You don't have to outrun the enrollment cliff. All you have to do is survive longer than your peer institutions, right? Your other small schools, if you can, and you can do that, that's totally in your control, right? If you survive longer, you're more efficient, you're more effective than your, than other small schools, you will likely survive this. And thrive on the other side. That's, that's sort of my take. And so I want them to have hope and, and really bear down, right? Just get after it and survive. Right. You should be in survival mode at this point.

Well, again, Matt Hendricks, the founder of Perspective Data Science, has been my guest today. Matt, so much thanks. Let's talk about doing that again sometime. Yeah. And, and to those listening, we appreciate your time and investment. My name is Gary Stocker with College Viability. We'll talk again soon.