Tech Refactored

Ep. 30 - 2021 Initiative on Closing the Rural Digital Divide

July 28, 2021 Nebraska Governance and Technology Center Season 1 Episode 30
Tech Refactored
Ep. 30 - 2021 Initiative on Closing the Rural Digital Divide
Show Notes Transcript

Over the past several months we collaborated with a group of scholars working on projects to discuss the challenge of the rural digital divide, or internet access gap--from virtual learning in the face of a pandemic to politics to dig permits. We are discussing the projects in a series of one-on-one discussions. This endeavor was undertaken in conjunction with the Space, Cyber, and Telecommunications Law Program at the University of Nebraska College of Law. 

For more info about these projects and the authors visit ngtc.unl.edu/digital-divide

Disclaimer: This transcript is auto-generated and has not been thoroughly reviewed for completeness or accuracy.

[00:00:00] Gus Herwitz: This is Tech Refactored. I'm your host, Gus Herwitz, the Menard Director of the Nebraska Governance and Technology Center at the University of Nebraska. Over the last several months, we've been collaborating with a group of scholars working on projects that does. Discuss the challenges and potential solutions to the rural digital divide.

This endeavor was undertaken in conjunction with a space cyber and telecommunications law program at the College of Law at the University of Nebraska. We're discussing the projects that have been undertaken by these scholars in a series of one-on-one tech refactor discussions, which we have put together for this episode.

Joining us to discuss his contribution to the Digital Divide Initiative is Jacob Manlove. Jacob is an assistant professor of agricultural Economics at Arkansas State Univers. His project is called Impacts of Broadband Availability in Arkansas Public Schools During the Covid 19 Pandemic, A Survival Analysis of Virtual Learning Decisions.

Jacob, welcome to the show. [00:01:00] 

[00:01:00] Jacob Manlove: Absolutely Gus. Thank you. Thank you for having me. 

[00:01:02] Gus Herwitz: So if we could, uh, start with the, I know academics love this question, the 32nd summary of what you're looking at, uh, in this project. 

[00:01:11] Jacob Manlove: Yeah, I'll do my best here to keep it at 30 seconds here. So the, the general idea is that the covid, the pandemic that struck the US changed the structure of many, many things and education is one of those.

And as students and parents and families just in general, we're we're made to force to kind of make the decision, do I do in-person learning? How do I balance the risk of my health and my child's? And community health overall with my decision for education. So really the goal of this is to say, well, how did broadband play into that decision?

Did it have an impact? And if it did, what was the magnitude of it? 

[00:01:46] Gus Herwitz: And the, the motivation, I guess, is in many ways straightforward for, uh, this project. But can, can you tell us a little bit about the motivation for this project in particular, but also perhaps why, uh, uh, you are more generally [00:02:00] interested in the topic?

[00:02:02] Jacob Manlove: Yeah, absolutely. So I think one of the big things when we talk about broadband technology is, is we tend to, to talk about it in terms of economic growth. There's a few baskets that we put it in, and I think education is one that that tends to be overlooked in terms of how it can impact a family, how it can impact your future human capital even based on how it impacts your education.

And so that was really one of the big motivations behind this. And the other thing is just in studying broadband, A lot of the time people either have broadband or they don't. We never really have an opportunity to say, uh, kind of a pure experimental study like we kind of quasi have in this situation here.

And that people are making decisions instantaneously based on these factors. And it really gives a good opportunity to say and measure that to say, Does it have the impact we think it does? And if it does, again, what is the actual implications of that? 

[00:02:53] Gus Herwitz: And one of the important things I think, worth noting about this project, the, the Pandemic obviously, has been one of [00:03:00] the worst things to happen to our country and humanity in a, a good, long while.

And I, I think it's commendable and worth noting that, uh, your project is making some lemonade. That really terrible batch of lemons. I, this is what, uh, an economist would call a natural experiment. You have the opportunity to collect data and look at how people are responding to the pandemic. So I, I think it's notable that you are able to, in the midst of the pandemic, realize, Hey, this gives us a opportunity to look at this important thing.

Let's not let this opportunity go to waste. So, I, I, I, uh, just wanted to call attention to, Before we turn to the, uh, actual substance of, uh, what you do in, uh, the project, can you tell us a bit about how you have approached thinking about the topic and what the lens or methodology you use in the project is, uh, and perhaps what a survival analysis is?

[00:03:55] Jacob Manlove: Yeah, absolutely. So one of the big issues, and of course, uh, anyone in broadband knows that [00:04:00] rural communities are continuously underserved. Um, and there's also a strong correlation between those communities that are underserved in terms of their broadband as well as their educational impacts. And so looking at it that way, um, kind of looking at it through this framework of.

It was essentially a policy change, uh, that students before the pandemic, there was no choice. They had some decision in school choices, uh, whether they went to public, private, what kind of institution they went to where, where state laws allowed that. But now they have an entirely new decision and it's almost as if a policy decision was thrown in and it's being made at the family unit level to.

Do I decide to again, balance the health risk versus education with broadband versus the effects of Covid. And so looking at this kind of, again, through the idea of it being a policy change, the idea was to say in communities that are better served by broadband are essentially, are more students willing to take the idea and do this virtual learning decision. So lots of schools, particularly within the [00:05:00] state of Arkansas, they were given the decision at the family level based on government, uh, mandates from the governor that this is gonna be a family decision. It's not gonna be district by district. Each district within the, in the state, needs to offer a virtual decision and they need to offer an in-person decision.

And then it's down to the family to make that decision. So again, the idea of the survival analysis and say, Well, how long did these students stay away, uh, from the campus? Or did how long did these students stay in virtual learning balanced out against things like the number of active covid cases? Verse again, the real variable of interest is do these students actually have an internet connection, which is really a precursor to doing virtual.

And did it have an impact? 

[00:05:44] Gus Herwitz: And what does this tell us? Uh, what did you, uh, find in the project? Do you have any conclusions at this point? 

[00:05:52] Jacob Manlove: Yeah, so one of the big things is of course, it, it did have an impact. Some, there's a lot of other things that, in founding variables that we saw as well. So politics [00:06:00] played a big role within the decision.

Um, Whether the, whether the particular county, the school district, what their political affiliations were and how they treated covid. We've seen that throughout the entire pandemic and, uh, kind of very polarizing viewpoints. And we see that even kind of follow back in to the education piece as well. But in terms of the, the overall impact of broadband, those districts that had larger adoption, uh, the, the family level, they had broadband within their home.

They were much more willing to take on broadband, or sorry, to take on a virtual learning. Then someone that did not have that. And really that's just showing again that the really, the precursor to do this virtual learning is that you have to have a good, reliable broadband 

[00:06:41] Gus Herwitz: connection. You mentioned that you encountered some confounding factors, and I, I know data is never perfect when you're doing empirical work like this.

Could you tell us a little bit about what some of these unfounding factors are or any, uh, data challenges or other methodological challenges that you, uh, had to, uh, struggle with and. [00:07:00] Yeah, 

[00:07:01] Jacob Manlove: so I, I think one of the big ones like I, I, um, mentioned earlier is there's, there's a large political piece to this in terms of your beliefs and how you respond to this.

There's tons of different ways we can measure that, but again, what's the most appropriate for this? I don't know if we have a good measure of that at the moment. Uh, a couple of other things that we looked at as well, were just the overall school district's performance, so higher performing schools, and I'm sure this is just, uh, highly correlated, just a correlation.

They obviously have better internet already, so they didn't have to make these decisions. It wasn't a decision that was put in front of them to actually make, And the other piece is, is I commend the, the reporting of Covid within the us. I think we did an excellent job, but I think one of the issues with that is it's very cumulative measures and trying to get down.

In a very finite geography, how Covid affected very small pockets of the us. Finding more information about that to kind of subsidize what information I do have, I think could make a big impact [00:08:00] because we think about a district, so I'm in Arkansas, think about Little Rock there. Many districts within Arkansas and measuring it just across the.

Covid might have had an impact on maybe the east side of town just because of the, the culture and the, the response to the pandemic there. But on the other side of town, it could have been completely different. And looking at the way those things were, were handled internally across the geographies, I think also has a large impact.

And that's just not having the data on such a small scale to really measure that, draw that impact out. 

[00:08:35] Gus Herwitz: Any ideas on where this takes you, uh, next steps in, uh, the research or, uh, expansions of the methodology? 

[00:08:42] Jacob Manlove: Absolutely. So, uh, a couple of things I'd like to look at here in the future are, uh, it's no surprise that the districts, the school districts that had the low adoption were rural communities.

I think a really interesting point into this is there's some communities that of course have some anchor institutions that you could. You know, pull up to McDonald's, pull up to your [00:09:00] library, your local park, and connect to wifi. And then that brings into question, is that really enough internet? Or do you have to have the internet at home?

Is it, can we make do with a connection outside of our home? Or does it really need to be in our homes? I think that's the next step in looking at this and saying what type of connection really drove the decision? Was it the fact that I had a home connect? Or was it that it was available to me? And there's some other things that kind of pull into this when we think about it from the rule as well is folks in rural areas, if you want to go to an anchor institution to connect to the internet, it might be a 20 minute drive.

If you're within an urban part of the city, even if you don't have internet, it might be a five minute drive to go get to that particular access points. So again, thinking a little bit more geography, getting a ge, a geographical piece into this is probably the best, uh, the big next step to say how did that impact the decisions as well.

[00:09:53] Gus Herwitz: Okay, uh, any, uh, last thoughts or, uh, comments that you want to add to the discussion? 

[00:09:58] Jacob Manlove: Yeah, I mean, I'm, uh, well [00:10:00] first of all, I wanna thank you Gus for, uh, the opportunity for this and to kind of fund this study. And like you said, uh, just make lemon, uh, lemonade outta lemons here. It's, it's not a great scenario, obviously, it's not something that we want to be in, but in terms of broadband, it really did prevent that- uh, did provide for that natural experiment and

uh, I'm very excited to continue adding to this and seeing where we can look at more variables. Cause right now I know that from what I have, there are plenty more variables that probably impacted this decision. And now the, the question really becomes, What was it? What are the variables I need to be looking for, and are they available, and what impact did they have?

[00:10:40] Gus Herwitz: joining us to discuss their contribution to the digital divided initiative are Christina Beney, the lead author on this paper. Along with her co-authors, Brian Whitacre and Roberto Gallardo. Christina is a PhD student and research assistant at the Oklahoma State University. Brian is a professor and the Jean & Patsy Neustadt [00:11:00] chair in the Department of Agricultural Economics at Oklahoma State University.

And Roberto is the director of the Purdue Center for Regional Development and to Purdue Extension Community and Regionally Economics Special. They'll be presenting their paper. Do dig once and permitting policies, improve fiber availability. Christina, Brian, Roberto, Welcome to the show. Thank you for having us.

Thanks for having us guys. Thank you. So, uh, Christina, I, I'm going to ask you the question I'm asking everyone, which is every academic's favorite question. What's the 32nd summary of what you're looking at in this project? 

[00:11:38] Christina: Sure. Now this research is actually looking at whether two relatively unstudied policy initiatives known as Dig Once and Expedited Permitting have impacted fiber broadband and fixed wireless internet availability throughout the United States.

The two policies have only been adopted by a handful of states thus far with the [00:12:00] state of Iowa having adopted both in 2016. And given the federal internet availability data that we have access to, Iowa has provided a unique way to look at how these policies have influenced internet availability relative to the states that do not have those policies in place.

[00:12:19] Gus Herwitz: Great. Uh, and can you, uh, just briefly tell us what, uh, Dig Once and Expedited permitting, uh, policies are? Sure. So 

[00:12:26] Christina: Dig Once is kind of self named in that state or municipal governments that intend to do any type of road construction or other work where they'll be digging up the roadsides are supposed to let internet service providers know of that pending construction.

That way the internet providers can actually. Utilize that as an opportunity to lay, lay more fiber conduit or the actual cable to, um, take advantage of only having to dig up the roads once expedited permitting, on the other [00:13:00] hand, is just something where the government entities are required to. Abide by a, a quicker timeline in turning over decisions on permitting.

Iowa in particular was at the point where they were taking 120 plus days to make a decision on honoring permits to service providers and under their expedited permitting legislation, I believe they have to keep it to 60 days. 

[00:13:24] Gus Herwitz: And what's the motivation behind this project? Obviously digital divide issues are, are very important, but as a scholar, an academic, you've devoted significant amounts of time to studying this particular question.

Uh, what, what drives you to, uh, this question? 

[00:13:42] Christina: Sure. Even prior to the Covid pandemic, the digital divide has been a concern. Uh, I think Covid further highlighted the need for more and better internet access for all Americans, especially since many were required to work or attend school virtually despite [00:14:00] previous efforts.

At this point in time, millions of Americans still don't have internet access or what's deemed to be acceptable or appropriate Internet access. Especially today under the Biden administration, the federal government is trying to allocate considerable funding to expand broadband internet infrastructure and to make internet more affordable.

So moreover, if, if states have policies at their disposal to, to make these things happen faster or to more efficiently utilize funds, then all the better. So we're particularly interested in, in those types of issues. 

[00:14:34] Gus Herwitz: And you all are economists and are approaching this from an econometric perspective. Uh, can you tell us a little bit about, uh, the methodology that you're using to think about these questions?

[00:14:45] Christina: Sure. So the actual econometric analysis involves what's known as corson, exact matching. It's a process where we take. Observations in the state of Iowa and match them with demographically similar observations in [00:15:00] what we dubbed to be the control states. So those states that either don't have, um, Don't have dig ones or expedited permitting or have received similar levels of funding for broadband initiatives.

We, we run through a, a variety of modeling specifications, but once that matching is completed, we're looking at this through a difference in difference model, which essentially takes observations in Iowa, both before and after the, the policies were put in place to see whether the. The day of, or the date rather, that the policies were implemented has shown a significant divergence in a positive direction, or I guess in any direction, but we're looking for a positive direction from states that have not implemented those policies.

[00:15:51] Gus Herwitz: Brian, Roberto, uh, I'll, I'll come to you with some follow up questions in a moment, but, uh, we should let Christina tell us the exciting part of the paper. [00:16:00] Uh, what does the paper actually say? What's, what are your conclusions at this point from this analysis? 

[00:16:05] Christina: Well, ultimately we, we found that dig once and Expedited Permitting.

Do not have a significant impact on both fiber availability or fixed wireless internet availability. We did see some positive impacts at the national level, so comparing Iowa to all other US states. Um, but ultimately the results don't persist in many of our other model specifications, especially those that, uh, focus on states that are more comparable to Iowa with respect to geographic location and funding amounts and other, other broadband policies. 

[00:16:41] Gus Herwitz: So that, that seems like a surprising, uh, result. Dig once and permitting, they're discussed, uh, quite a bit in the policy sphere. Uh, do you have any sense, uh, what accounts for this?

[00:16:52] Christina:  I think one thing we've noted in this analysis in particular is that we're looking at dig once and expedited permitting together.

There's [00:17:00] not a great way to separate the two of them or to analyze them independently so that could have some bear. But otherwise, I'm, I'm not entirely sure what might cause the, the divergence or the, the lack of results rather. 

[00:17:16] Brian Whitacre: One thing I might add, Gus, if, if it's okay, is, um, so I actually called the Iowa Department of Transportation as we, as we went through this study, and, uh, it seems like although this policy is in place, there's no formal mechanism behind it, so they, there's not a required spreadsheet to have to fill out and send to the providers when they know that construction is coming.

And so there's no real way of, of enforcing this policy to be frank. And so it's more just kind of a, oh, if we know this project's going on, we'll just shoot so and so an email. And again, there's no formal documentation or, or processing has to have, but I think that's, that's something that we might wanna dig into a little bit more for other states that have, that have implemented this policy at a later time.

So I think that that's one thing we, we might kind of point to is why, [00:18:00] why the dig once in particular has not been so, so, uh, success. 

[00:18:03] Gus Herwitz: There's a, a bit of a, uh, if you build it, they will come fallacy in a lot of policies, and I, I think, uh, this project might well be demonstrating an important part of it.

Just having the policy in place might not be enough. Uh, if, as you say, Brian, folks don't know about it, in which case, this sort of project is really demonstrating an important, uh, gap in how we're implementing these. Um, what are the limitations? Are there any data challenges that you had to overcome as you were undertaking this project?

[00:18:37] Brian Whitacre: Well, I'll jump in and maybe Roberto can can speak to this cause he's our, our data guru, but, So essentially, I mean, you, you all know though the issues with the FCC form 4 77 data, it's available at the census block level, but it assumes every household with any block has service if the provider says that that block has service.

Um, and so what we did was ultimately aggregate those to the [00:19:00] census tracked level because we felt like that was a broader, a better, uh, geographic unit for who's actually making the decision. To potentially provide fiber or fixed wireless. They're making it a kind of a broader decision than an individual block.

And so we actually came across some interesting patterns as we did that. We actually showed a drop in fiber between 2017 and 18 and a drop in fixed wireless, uh, availability over those same years. And this has actually been replicated by some other researchers, uh, out aside from us. And so it does look.

There's these kind of strange trends you wouldn't expect to happen. And so we're still trying to, to dig into that a little bit more. You know, we don't think people went in and tore up fiber, um, but we're just not sure if it's actually, you know, people, providers leaving the market or being renamed or, or what the case is.

It does look like there actually was some consolidation in the fixed wireless market. Where, uh, you know, maybe a mom and pop tower, uh, went out of business and, and, and, and gave their their lines to someone else, or, so we're still looking at that. That was [00:20:00] certainly a data issue and we're, you know, the matching that we're using really kind of wants to look at the trends pre, uh, policy implementation, and we didn't get a great match there.

We didn't get a, you know, a- a- we saw the trends in Iowa and we wanted to construct other states that had exactly the same trends, but we just couldn't get it because there was a lot of zeros, uh, with fiber availability and a lot of, you know, places with good fiber availability. But we just didn't, weren't able to match that trend identically as we kind of wanted to.

[00:20:32] Roberto Gallardo: Yeah. And in addition to that, the census track, you know, has a lot of, uh, demographic and economic information. And so for the, to, to really control for other factors is that we aggregate it also into the census track. But like Brian mentioned, I am intrigued. I recheck the numbers multiple times, and you just see these drops out of nowhere.

We did some preliminary descriptive analyses and there's no pattern to it, so we really don't know what's going on, and that's something that [00:21:00] I'm intrigued to learn more about. I'm guessing it's a reporting issue. I'm not very familiar with how much the FCC actually crosschecks or validates the data even.

within the, the, the providers, not with the customers. We know they don't do that, but within the providers, do they just receive that and if there's just only a massive mistake, then they flag it and and come back? We don't know. Or at least, I don't know. I have read in the past of very big errors in the data that the FCC took for granted, published the data, and then third party folks came back and.

You know what, This is not right. And so the FCC then went back and the providers had to backtrack that information. I'm guessing there is something around that here where the provider may have, you know, reported something some year and then another thing next year, and then you see that huge drop, which doesn't make sense really.

[00:21:54] Gus Herwitz: Well, perhaps, uh, that response is forecasting the response to my next question. What's [00:22:00] next? Uh, where, where's this research taking you? Or do you have other projects in the pipeline? 

[00:22:05] Brian Whitacre: We have published some, some research before Roberto and I, and again, you all were great to sponsor that at, at UNL. That show that some state policies do appear to have an impact on rural broadband availability.

And so this one, it doesn't look like there's a, a great positive impact there. And so we certainly wanna publish this in with the legislators and people who are looking to, as Christina said, pass policies that can have an, a positive impact on, on availability. So I think we're gonna continue to, to dig into this policy angle that that is of interest to a lot of people.

You know what other policies are out. That that might work. Certainly, you know, of interest to, to people as, as the funding does come down from the federal government and wanna. We'll make sure we have legislation in place that, that makes it as easy as possible. 

[00:22:49] Brian Whitacre: And I'll just throw one more thing out there that, you know, So here in Oklahoma we don't have a, a Dig Wants policy.

You know, it's, it was brought up, I, I'm part of our, our little legislative team that's looking at how we can make broadband. [00:23:00] Available in, in rural parts of our state. And the issue is, well we already, you know, the Department of Transportation already has good relationships with the ISPs and so they just already let 'em know about it.

We don't need a formal legislative policy to to, to do that. And so I think that's maybe something that's also kind of hiding the results. So if we can get at some way of uncovering how good that relationship actually. I think that would be, would be interesting, but it doesn't look like just simply passing this Dig one policy is gonna have a, a huge impact on the, the, these two particular technologies that we're interested in.

[00:23:31] Gus Herwitz: Well, Christine, Brian, Roberto, thank you, uh, very much for joining us and really for this, uh, uh, really interesting, uh, and important project. Uh, I look forward to seeing where it continues to go.

Joining us to discuss their contributions to the Digital Divide Initiative are Angela Hollman and Tim Obermier. Angela is an associate professor of cyber systems at the University of Nebraska at [00:24:00] Carney, and Tim is a professor of information networking and telecommunications at the University of Nebraska At Carney, they'll be presenting their paper, Reverse engineering rural measures.

Is this the solution for real time speed monitoring? Welcome to the show, Angela and Tim. Hi Gu. Thank you. Thank you. So Angela, let's start with you. Uh, the impossible question I'm asking, uh, everyone, academics love this question. What's the 32nd summary of what you're looking at in this project? 

[00:24:30] Angela Hollman: Well, in this project, we are looking to try and find, I will throw out the word quantitatively, the needs of a resident or business owner for their internet connection.

So this is a question that nobody has really had the data to answer before, and we think that our annualized data in our project of rule measures does have the keys to start to answer. Like I said before, quantitatively this 

[00:24:57] Gus Herwitz: question. And Tim, [00:25:00] what's the, the motivation behind this project? I know you and Angela have been working on this particular project, but also other iterations of this project for, for quite some time.

What drives the two of you to spend so much of your professional lives working on this question? . That's 

[00:25:16] Tim Obermier: excellent question Gus. Uh, and a fun question to respond to. Actually, we can go back probably close to 10 years and being academics, I could stretch this response out for probably 30 minutes, but we don't have that much time.

The bottom line is, you know, 10 years ago I got really interested in what, what did internet actually cost people and the comparisons between the rural environment, the rural residents, and the urban residents. And we actually found out in some, some initial. Research that the rural residents paid upwards of 170% more for their internet access than did residents in in urban environments.

And then we got kind of curious, uh, and Angela jumped on board and we, we started looking, uh, a little bit more in depth of the impact of [00:26:00] tariffs on the cost of internet as. Were involving from a telephone age to a, uh, internet age. We still have tariffs, as you well know. Uh, and they were impacting as we found out the cost of rural internet access.

Those things are, are highly motivational for us. We, we really wanna see what's happening in the world and compare that, uh, to those, those urban residents. Ultimately we started to ask the question, could we de develop a device, Build a device that could quantitatively measure what we're looking at, rather than just looking at advertised rates and costs, uh, that the providers have out on their websites.

And Angela successfully engineered and designed the, the device we call a quantitative throughput device. And at that point, she took over the leadership of the project and we're doing some really fun things with that, and that's why we're here today to talk about that device. The capabilities that, that it's providing for us.

But we also have a passion for [00:27:00] this, and the, the underlying passion is we both come from a rural environment. We grew up in a rural environment, and I have to add this, I, I sat around the table and listened to my grandparents talk about the days of when. Their, their farms received electricity and, and the impacts that that had upon their ability to interact with the outside world.

And that, for me is a very strong passion in this world of the internet because we're, we're living it right now. We're watching communities that. That are going to suffer if they don't have good quality internet access. And so therein lies that passion. We're both residents in the rural environment. Uh, we both have struggled to get good quality internet connections.

And so it's fun to research this and help out those people in the rural areas. 

[00:27:49] Gus Herwitz: So a quick follow up question. You used the word tariff, and I expect many listeners to hear the word tariff and they think international trade and fees placed on products going into or out of [00:28:00] a country. I mean, in the telecommunications context, can you just ex.

Explain what a tariff is to the listeners. 

[00:28:05] Tim Obermier: Certainly, yes. Well, a tariff is a fee, um, basically that, uh, a company has to identify publicly, uh, and it has to be essentially approved publicly in, in many cases so that it, the company is actually charging the customer a fair price. They got started many, many years ago when companies were, were actually goes clear back to grain handling actually, and, and.

People need to be charged a fair price. And when you have a, a provider that is a monopoly type providers, we had in the telecom days, we had to have government oversight to make sure those prices were fair. And so those tariffs were advertised, uh, to let people know what they were going to be charged. We, as we involved a course into the internet, uh, age, we had some of those tariffs that were impacting telecommunications services that, uh, also provided internet services.

The folks were being [00:29:00] charged based upon these published tariffs. 

[00:29:03] Gus Herwitz: Yeah, so the, the key with tariffs, perhaps rubber meets road, is they're published, These are government regulated published prices. So the government says it costs $30 a month for this. And if you want to change that price, you need to go through the government.

If you want to offer a lower price, sometimes you're not allowed to. So that's the understanding of a, a tariff at the highest level. 

[00:29:25] Tim Obermier: And I like your explanation better than mine. It's much more simple. , 

[00:29:30] Gus Herwitz: I'm the one always pushing for these 32nd answers. I'd better be able to deliver on my own. Uh, so Angela, Tim mentioned the device that you engineered.

To the next question, can you tell us a bit about the methodology and the approach that you've taken to this research question? 

[00:29:46] Angela Hollman: Sure. When we were looking at a way to get these measurements, we were, we were trying to determine like, what are we seeking, What are we seeking to have access to that we currently don't within our current mapping system?

And I [00:30:00] think the real answer to that was we really don't know what's going on. During all times of the day in a person's household. And then we don't know how satisfied they are with the cost of what they're paying and some, some little deeper quality of life questions that we thought the internet may be affecting, may or may not be affecting.

Right. So we really sought to. Think about how we could collect that data and we came up with a twofold answer of we're going to have, have a device that's independent of anything else in their network that connects to their network for as close as we can be to the service provider. And it's gonna have to measure at least over a weekend and some of the weekday so we can really get at those peak times and we can really get at the non busy times.

And we. Really good comparison and we can see what's going on. And then we also wanted to get those social science metrics. We wanted to see quality of life and like I said, and the cost was still something that we were interested in at that point. So we developed a survey that they would take [00:31:00] alongside and then we can link them together, um, with a unique identifier that's both on the device that they then enter into the survey.

And then the third piece was tracking their physical location. So we wanted to get granular enough so we knew that device was placed at this street address number without the, Oh, it may be within one to one and a half miles. It may be only on the county. We really wanted to dive down, so we knew it's at this.

And they were testing it over the week, and that is directly linked to the social science metrics, 

[00:31:34] Gus Herwitz: so that this is such an important project. I, I just have to jump in to say, we, we talk about how much bandwidth we need and no one really has. Much understanding or no, There are those who do, but most consumers don't really understand what's a bit, what's a megabit, what's a gigabit, How much do I need?

And if you don't understand that, do you want more bandwidth? Yes. How much do you want to pay for it? Less? How many applications do you [00:32:00] want to be able to use? All of them. I, what, how fast a car do you want? As fast as possible. How safe should it be? As safe as possible. How much it cost, it should be free.

We, we need to have some more nuanced understanding when we're trying to assess what we want and need. And your, your project is taking such a great on the ground approach. Tying in all these questions, which I guess leads to the next question. What are your results? What's the, the project showing at this point?

[00:32:27] Angela Hollman: Well, the project has shown at this point that our methodology. Is is working. Um, we've worked through some, some user issues, but I also think too that one of the interesting side notes that this project is doing is what you just mentioned, Gus, is that some of the people that are getting handed this device are having a little bit more awareness.

Of how much is flowing through their network and how satisfied they are with that. And they can see that exact metric that's popping up for them that we allow the users to, to view. So I think that's kind of a side, kind of a [00:33:00] side note that's not as sciencey. That is a really cool piece of this project.

And that happens not only to the people that actually purchase. Paint, but to the people that are handing out the devices too, are getting some of that awareness too. So like I said, that's one of the non sciency pieces, but one of the sciencey pieces that we found are, we're finding these really interesting usages when we reverse that process of this is how much we're measuring, but, oh wait, we also have this social.

Piece where we know how much the connection should be, and then we can see a max going across on the off peak time. So to us that meant we can start to see these needs. We can take this section of people that have installed this device and are satisfied with their internet. And we can look at exactly how much right there that satisfaction index is.

You know, okay, it's 10 megs per house, or it's 20 megs per house. And then furthermore, we have them report on how many household members are in the house, [00:34:00] which I do think plays some importance. I think the age plays some importance in there. And I think that our data is, I like we, we have pilot data.

Right now I wanna call it pilot data because I don't think we have enough to say, Oh, we can generalize this. But our data now is showing some really interesting, some really interesting, I don't know, some really interesting stuff. . 

[00:34:22] Gus Herwitz: Can, can you my, my next question is going to be, uh, why are the next steps in the project?

But, uh, I, I have to ask, can you give us any preview with a caveat? This is all rough, preliminary data. 

[00:34:34] Angela Hollman: Okay, so. We found, and I, I will say, I wanna call this an optimistic view because technically we know that internet speed is really bursty. And so when we analyzed the data in this project, we took a very optimistic view, which means we looked at the max amount that they were getting.

During a busy time on the network and then we kind of crunched that against, um, what they should be receiving to get how much they were using. Somebody could say, No, you [00:35:00] should look at the minimum. And I would say that's a pessimistic approach and we can definitely do that in the future. So this is an optimistic look at how much somebody needs and how many megs somebody needs for, for their internet connection.

But overall, we found that the throughput per household member was around 10.7 megs. Their max average download during these times when we measured. That's rough pilot data, but I think it gives us a place to start and it's a place that we, a number that we haven't had before. 

[00:35:32] Gus Herwitz: Tim, did you want to jump in?

[00:35:33] Tim Obermier: Well, Angela, could you also share the contrast, uh, in what the individuals think they need for a download? I, I think that is a very interesting contrast. Yeah, 

[00:35:46] Angela Hollman: it, it definitely is, and it, I think that goes to what Gus was talking about earlier on awareness. I did wanna clarify too that when I throw out that 10 meg, that was per household, which was a little surprising to me.

But again, this is an optimistic look. So [00:36:00] it means we were taking those max times instead of them in, and that would be higher if we took that more pessimistic approach. So what people think they actually need lies somewhere between 33 to 35 megs is what our show study is showing. So there's definitely a, a pretty big discrepancy in what our data is showing optimistically that people actually need versus what 

[00:36:27] Gus Herwitz: they think they.

So that's, uh, a fascinating result and I, I expect there's a lot of detail in the data bimodal or tri modal distributions and different things like that. So I, I really look forward to seeing, uh, how you continue to work with this, uh, project and ultimately publish the results and get the results out there.

Which leads us to the next question. What's next with this project to both, this one in particular and the research line general? Well, I guess I 

[00:36:55] Tim Obermier: could probably jump in here real quick and just say that we are actively working [00:37:00] several projects across the state and distributing units in various partnerships that we have.

So we're, we're, That process is enabling us to build the database so that we can make more accurate analysis for this particular construct. 

[00:37:17] Angela Hollman: I also wanted to put in that we have recently also hired a postdoc in the GIS space, and we're really excited about some of the analyses that he can bring to our data to break that down across our geographic space.

Not only you know, in Nebraska, but as we look to this for a model of potentially expanding outside of Nebraska and nobody else. Is really doing this granular of speed testing, if you wanna just minimalize it to that across the nation and the demographics that we're also collecting and matching it with the quality of life.

And so having that in the geographic space too, will- will really be interesting. 

[00:37:59] Gus Herwitz: Yeah. [00:38:00] I'll, I'll just end by saying, well, I'll end in a moment by saying thank you. But this again, is such an important uh, uh, paper. And anecdotally as I've been talking to network engineers who work at ISPs during, uh, the Pandemic, uh, they've had a lot of calls about folks, I'm not getting enough speed.

I upgraded my internet connection to the house. Why don't I get good performance? And the, the number one thing that they need to do when they visit folks houses is replace their wifi route. Upgrading it from an 11 megabit per second, 802.11b technology to newer technologies. People are using technologies that are simply not capable of higher speed internet, and that goes to the importance of your approach of having a device that connects directly to or as close as possible to the internet access point in the house.

So that you're actually measuring what you're trying to measure. Uh, it's such a hard problem, uh, and you're doing such a, a great job trying to bring in all these different factors, uh, [00:39:00] sociological, technological, economic, to really get at this. Thank you very much for, uh, the discussion and the really great work.

I, I looked forward to, uh, continuing to see where it goes. Thank you, Gus. 

[00:39:10] Tim Obermier: Appreciate it. 

[00:39:12] Angela Hollman: Thanks for having us Gus.

[00:39:14] Gus Herwitz: And thank you for listening. I've been your host, Gus Herwitz. Thanks everyone for joining us for this episode of Tech Refactored. If you want to learn more about what we're doing here at the Nebraska Governance and Technology Center, you can go to our website at ngtc.unl.edu, or you can follow us on Twitter at unl underscore ngtc.

This podcast is part of the Menard Governance and Technology Programming Series hosted by the Nebraska Governance and Technology Center. The Nebraska Governance and Technology Center is a partnership led by the Nebraska College of Law in collaboration with the Colleges of Engineering Business and journalism and Mass Communications at the University of Nebraska.

Colin McCarthy produced and recorded our theme music. Casey Richter provided technical assistance and advice. Elsbeth Magilton is our executive producer, and Lysandra [00:40:00] Marquez is our associate producer. Until next time, the metrics are coming from inside the house.