---
id: "workshop:smart-data-fair-cities-evidence-based-approaches-to-city-management"
type: "InnovateUS workshop"
title: "Smart Data, Fair Cities: Evidence-Based Approaches to City Management"
url: "https://innovate-us.org/smart-data-fair-cities-evidence-based-approaches-to-city-management"
date: "2025-10-16"
duration: "60 minutes"
language: "English"
instructors: ["Sonita Singh"]
recording: "https://www.youtube.com/watch?v=-g-vVK9YLNY"
transcript: "automatic, from the recording"
source: "InnovateUS (innovate-us.org)"
licence: "CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/), as stated on innovate-us.org"
---

# Smart Data, Fair Cities: Evidence-Based Approaches to City Management

InnovateUS workshop · 16 October 2025 · 60 minutes

Cities today generate more data than ever—on housing, transportation, public health, labor, safety, and beyond. But raw numbers alone don’t create better governance. Public professionals need strategies to manage, interpret, and act on this data in ways that are both effective and equitable.

In this workshop, Dr. Sonita Singh will introduce participants to principles of data-informed city management, with a focus on how evidence-based approaches can strengthen public services and address inequities. Drawing from her work with I-STARRT and the City of New Orleans, she will share how combining quantitative and qualitative insights—numbers and lived experiences—can help governments not just treat symptoms, but tackle systemic challenges at their roots.

Participants will leave with practical tools and frameworks for using city data to plan, evaluate, and course-correct policies, while building trust and transparency with residents.

## Learning goals

- Understand the core challenges and opportunities in managing city data

- Identify strategies for integrating evidence-based research into city planning and service delivery

- Apply frameworks for combining quantitative indicators with qualitative community input

- Recognize how data-driven equity models can support fairer and more effective governance

## Speakers

- **Sonita Singh**, President and CEO, I-STARRT

Recording: https://www.youtube.com/watch?v=-g-vVK9YLNY

## Transcript

Automatic transcript of the recording: names and terms can be misheard, and the times can be a few seconds off. Each time links to that moment of the video. The transcript alone: https://innovate-us.org/smart-data-fair-cities-evidence-based-approaches-to-city-management.txt (plain text) and https://innovate-us.org/smart-data-fair-cities-evidence-based-approaches-to-city-management.json (every segment with its start and end second).

**Sonita Singh**

[0:03](https://youtu.be/-g-vVK9YLNY?t=3) Okay, hi everybody, good to see you. I have had such a great time with InnovateUS and I think it's going to keep going. I will say that this presentation, Smart Data, Fair Cities, I'm going to say that this is really just the beginning. So this is an orientation around thinking about how you can use data in your city management and how you can highlight that that data will begin a process of creating equity in those cities. So I'm going to introduce some ideas to you and I think as we're moving along and we can put it in the chat, we could also do a bit of a follow-up around this and get deep into some case studies and deep into some data. I think it's always good to get a strong idea and orientation around what thoughts and what's available and how we might be able to mold that. But as we get deeper, I'd like you to have this real functional ability to pull data, pull multiple types of data and play with them and understand how one set of data may impact another and how data can be dynamic and complex.

[1:27](https://youtu.be/-g-vVK9YLNY?t=87) So let's move forward. Just a little bit about me. I love the introduction, very humbled and very thankful. One way to think about me, one is that I look at AI and machine learning as a tremendous tool for good. I'm a Star Trek baby and we couldn't have these advanced societies without getting some of our more complex problems under wraps because we have the technology to do it now. So I want you guys to think about this moment, think about this time. Yes, there are a lot of negative elements about AI and machine learning that feel like it is overwhelming and a little disempowering, but I'd like to flip the script and say, wow, there's a tremendous amount of data that we can use to our benefit. And this is a little bit of a introduction into that. Another thing to think about with me is that I've spent a lot of time in vulnerable populations and in vulnerable settings, meaning there may have been a real shortfall of funding, of resources, of thoughtfulness around how systems are developed. So I like to fill in those gaps. I like to think about how do we address problems systemically. And so you'll see a lot of that in this presentation and as we move further along in these sessions. Finally, I look at urban environments and I look at them from a complex dynamic systems framework. So we think about that in the space of like tipping points and bifurcation state changes, understanding that there are patterns that happen in the urban environment and things are really joined and combined. Okay. So first thing we have to think about is this is deep work. Everything that we do,

[3:34](https://youtu.be/-g-vVK9YLNY?t=214) everything you've done today, last week, for the last year, it's not easy. It's complex. It's not always fun. So we're going to have a little bit of fun today because I've been having a stressed out time. I don't know about y'all, but we're going to keep it light. I'm here in New Orleans. New Orleans is known for doing a good job at having fun. So we're going to reflect that in our presentation. So things to think about when you're looking at city data. We're going to talk about, you know, I have a whole itinerary mashed out, but it really comes down to this. Working with data can get very convoluted and very bizarre really fast. So it's good to have someone to handhold, someone who understands. So always whenever you're thinking about doing a data project, your city managers or those people who do data and they, you know, crunch that data pretty regularly, they will always appreciate you addressing them from the perspective of how, from the perspective of curiosity. How can we do this? How can we bring data together to make a better environment, to do a better job? And they will always tell you that there are things that you will need to avoid. And there are things that are actually really fun and beautiful about this process. I'm sorry. I really sound very nerdy right now. Okay.

[5:09](https://youtu.be/-g-vVK9YLNY?t=309) All right. So what are we going to do today? We're going to talk a little bit about the what, whys, hows, wheres of city data. We're going to think about the core benefits and challenges in managing city data. Where is the real benefit in this and how does it play out? And then some of these challenges you're going to meet. I like to go into this conversation around evidence. Oftentimes we make our city strategies based off of what sounds good, what looks like it might work. But we don't always have evidence for that. So when we think about city planning and service delivery, we want to be able to say, hey, we've used these techniques and they've delivered for us. Or other cities have used these techniques and it delivered for them. So we want to bring those best strategies and we want to be able to show that over time, this investment, this work around data is actually giving us some outputs. So I think ultimately when you're introducing using city level data and manipulating that data and trying to say, okay, I believe we should take strategy A, B, or C, you want to be able to evidence it first. So we'll talk a bit about that.

[6:34](https://youtu.be/-g-vVK9YLNY?t=394) There are elements of using data at the city level, especially now with large language models, that it's just almost no end to what you can do. So you have a fair amount of quantitative data that you can parse through, play with, you can get some numbers, you can do some regressions. At this point you can do anything that chat tells you to do. But also there's the qualitative data and that type of data is what people say. It's what comes out in the news. Right now I'm doing a project that's scraping news from multiple countries in multiple languages. What are people talking about? What's really important to them? And there are also ways to combine those two. You can bring the quantitative data and the qualitative data in the same place and be able to say something more, something new that you had not been able to say before. Of course, as I'm saying, when you look at what your population is saying and you look at what their problems are, it is an easy segue to build equity for them. Structural equity we can talk about. There are different types of equity, but also just being heard and being able to show your constituency you've been complaining about the same thing for the last four or five years. This is us showing you that we're addressing this. We're managing this. We're thinking about it in our development.

[8:16](https://youtu.be/-g-vVK9YLNY?t=496) Okay. So everybody talks about silos, but I was just in the country and silos are really fun to look at. They're storage units. Obviously they keep a lot of America alive. We can't really do our grain work or agricultural work without silos, without a place to kind of take our data and stack it, but it can be overwhelming. Most of the time, the part that's most overwhelming is that the data sits by itself. It's just one type of data or a couple types of data that never really fully integrate in a functional way with other data. I'll give you some examples. We've got 311, which I love to work with. I'm using New York City as an exemplary city right now because they really put most of their data out. It's mostly open source. I feel like many of you cities, if you're representing a city, your data is open source and nobody uses it, or only a couple people use it that have the finesse, but what I'd like you to do is spend a little time going through either your city or a large city's transparency commitment via their city-level data. You can just bring that data down in an Excel sheet.

[9:50](https://youtu.be/-g-vVK9YLNY?t=590) It doesn't have to go into some large software to get consumed and digested and spit out into something you don't understand. I think it's very easy at this point. I will demonstrate in another session how you can just drop that data right into ChatGPT and ask it questions and say, okay, I've got 311 data for all of 2025 thus far. What are the biggest complaints? What zip codes are they coming from? What are people most worried about? Well, check it out. That 311 data might say, oh man, we have a ton of people who speed through my neighborhood. They're speeding through the neighborhood because there's construction. That construction causes detours and now this fast traffic is moving through a residential area. Well, that data would sit in the transportation data. So now you're seeing, oh wow, we've got transportation data that coalesces or aligns with our 311 data. So this is a good example where data can connect together. And if you just looked at it by itself, you might not see, oh, this construction sitting here, but it's creating a secondary problem. Another example could be our public health issues. We've got a bunch of new cases of COVID.

[11:24](https://youtu.be/-g-vVK9YLNY?t=684) We've got other things that are sitting on the horizon and we need to say, all right, where do we think about emergency dispatch? Where do we think about where our most critical systems should be deployed and where should they be sentinel? The darker side of this would be profiling and the police, where the police are always running up and down the higher crime areas. Well, you can do that in a proactive way where you know that there are problems at the city level that might be burgeoning. They're growing. How can we send those care teams out ahead of time to start gathering qualitative data to fix these problems before they get out of hand? So I just want you to think about your unit of operation, however you're coming to this problem. Think about the data layers that might seem like they're in different places, but they could be very relevant to you. If you're thinking about it and something really pops up in your mind, please throw it in the chat because I think everybody on the call and in the future will be like, you know what? I've never thought about aligning my heat island data with my emergency room visits for dehydration. These are things that could really sit on top of each other and you can look at those patterns.

[13:05](https://youtu.be/-g-vVK9YLNY?t=785) Another problem is, as we just mentioned, when you have all this data, sure you've got it, but it requires some sophistication to unite it and to unite it in a meaningful way. Most of the time you're going to get this information, you'll look at it and you're going to glaze over and you're like, what am I looking at? And what does this mean? And this is too much. And I will ask you to take a deep breath and start contemplating what are the threads that tie this together? What are certain elements of your data, your city level data, what can we see between them and how can we look at them in one unit? So let me take a second to hear from people either in the chat or come off mute. What are some problems that you're dealing with and you have maybe two elements of data and you want to tie them together? You can name something that you have a question about, or you can name something that you've solved the issue for.

[14:26](https://youtu.be/-g-vVK9YLNY?t=866) Okay, I see safety, security, downtown, which might have been for an earlier point, but I think it's a great point. Video sensors. I'm curious, do you want to give a little more detail about video sensor and security downtown? You can text or talk. All right, I'm seeing auto safety data and crash data. Excellent. And if you're in a city that As a fair amount of video sensors, you will be able to see some things about that. Yeah, these are excellent. So I'll say that there are a couple of elements that pop up to me within these examples. One, safety and security in an urban environment, particularly in your downtown areas. That is a lot of information. That's a lot of data. Those video sensors, we think about privacy. We think about your ability to see multiple sort of area targets at one time. And maybe if you're having a public safety issue, and we're going to get into public safety in a second, if you have an issue around that, you want to be sure when we're talking about ethics, right, that you're thinking about how you can monitor that information and sustain ethics. And I'll just make a really quick note about AI and machine learning.

[16:04](https://youtu.be/-g-vVK9YLNY?t=964) Our AI has now the capability of facial recognition. And as you're using that downtown security, and you're seeing things happen, you also want to think about what are the limits of facial recognition. Your use of AI and machine learning, it does a really fantastic job of understanding objects, objects in space. Maybe it's recognizing the traffic is too fast for one particular area, and they'd like to slow the traffic down. And you want to make the area more walkable. By making the area more walkable, it's safer, the property value increases, so on and so forth. So a really interesting way to think about that would be to look at cities that have created structures that slow the throughput of the traffic down, and in that, in doing so, have improved the usability of that space. So there's a lot of ways to slice that pie.

[17:17](https://youtu.be/-g-vVK9YLNY?t=1037) And I think one of the ways that we like to think about information and data, we'll get into it in a second, in this case study, is geographically. So let's look at a case study around housing instability and displacement risk. So we have an issue in New Orleans, we have a lot of people who come from maybe even colder environments because it's still very warm in the winter, but these people may have come unhoused, or because we have such a tremendous gentrification issue here, we have a lot of people who are becoming unhoused or housing unstable in transition. That number is growing pretty regularly. So when we think about some of the drivers of a complex issue such as this, they're not just one type of data. There are many types of data that roll into this issue and give some thoughtfulness to it. And as you're contemplating the specific problems that you are facing, I want you to kind of do this exercise with me. Take a minute and layer the data that are relevant to the problem you're thinking of. So if we're looking at housing instability, things that we know that we can get data on off bat. We know that we can look at a parcel level, which is at a house level. We can always aggregate up, look at the block group level, we can look at the census tract. I actually always avoid looking at the census tract or even the neighborhood because it aggregates too many things into one place. So sometimes you got to get a little granular, sometimes not as much. But we know that we can get the property tax, we can look at the assessor data, we know we can get that 311 information, we can pull that

[19:19](https://youtu.be/-g-vVK9YLNY?t=1159) right on down into an Excel CSV file. We can look at inspections and violations and we can map that. We can look at eviction rates, not just rates, but the actual filings and put those filings at their geographic level. So evictions, foreclosures, we can see how much the rent is. And of course, we can see what the average mortgage rates look like. And we can also parse that by who are the wealthier income tracts. Does this align with race? Does it align with how long you've owned your home or the average neighborhood has owned their home? So now you can see we can really dig down deeper. Now I will say that we've all seen a heat map. We've seen a dozen million billion heat maps and we're really happy that heat maps exist. We're really happy that we're able to go a little deeper and we've all either read or created these reports that show, OK, it's these people.

[20:34](https://youtu.be/-g-vVK9YLNY?t=1234) These are the trends. This is what it looks like right now. This is what we think we should do. And here are some solutions that we think may work. Now, what I find is that this approach, while it's maybe the best we're able to do, our ability to analyze and be precise with data has really improved over time. And where you sit in your seat right now with your chat GBT, reading your emails and taking some of the stronger language out, that chat also can do this work with you. It can also show you how if you have higher risk areas. What are some of the other factors that feed that risk? And then we can get a bit more complex. So that's the thing that I want to walk you through. And I'm going to ask you to take a deep breath again and get ready because we're going to get dynamic with it. OK, so as I was saying, I really do love that we've been able to think deeper around data, visualize it with big, pretty colors that give some trends.

[21:56](https://youtu.be/-g-vVK9YLNY?t=1316) But I think that we can do better and get a little more into what it's going to take to turn some of these conditions around. So this idea of finding the thread of data that ties our problem together, the reason we want to be precise, the reason we want to be thoughtful and complex is so we can get to solutions that doesn't solve the problem just for one group of people or two. It solves the problem or reaches towards a solution that can be homogeneous for the group. And that's something that even in public health, we often try solutions that might fit one target group, but doesn't really shift the overall situation. And that's how I want you to think about this evidence-based approach. That's how I want you to think about complex solutions for complex problems, really. Yes, they are complex. Your problems are complex. My problems, I promise you, they have layers. And, you know, one of the things we think about and just as we're prepping to get into this part of the conversation, I want you also to think through this lens, not this is complex in that it's not solvable. I want you to see your vision and your thoughtfulness around the problem in deep empathy. Think about the client that you care about. Think about the population that you're focused on and think about all of the issues that they are trying to solve.

[23:46](https://youtu.be/-g-vVK9YLNY?t=1426) One of the projects that I've done some time ago, working with women who are HIV positive, who are pregnant, who are low income, and have had a history of exchanging sex for survival. We use that term survival sex to contemplate when you're really just trying to eat. And find a place to sleep. Okay? So the way you would solve the problem of someone like that, you're not A, going to be judgmental, right? And B, you're not going to fix it with just throwing money at their problem. You're going to need a full team. You're going to need to be comprehensive and thoughtful in the way you work with that population. And if you guys ever want to get into, you know, sort of the weeds of how we do that, it's really beautiful. But we also think of things comprehensively, meaning there is a psychological and psychiatric care team available. There's peer support available. There's a social worker there to deal with the direct needs. There's special funds for transitional housing, moving to permanent housing. There's obviously a physiological element that does not exist in a vacuum, meaning it's not just one doctor. And it's not just one nurse or person who follows you to try to make sure you're staying on your medication. It's a whole group of people who are sticking with you, sticking with you, trying to make sure you're not failing. So I really just want you to see in this idea, it's not one solution for one problem. Most of the things that we're looking at and dealing with with our population every day, there are large entrenched drivers that are

[25:50](https://youtu.be/-g-vVK9YLNY?t=1550) keeping them in place. And I want you to put your mind around that. You would be surprised if you just add two or three layers of data to the problem you're looking at and bring them all in one forum. You will be absolutely blown away that there may be solutions that are sitting right in front of you. Okay. A perfect example of this is some research that I've done actually several years ago where I was looking at high-risk populations, again, multi-vulnerable. These were formerly incarcerated people who were being released. They were considered high-risk, high-risk for committing crimes and high-risk for recidivism. But we were releasing them in the highest risk areas, so areas of very high urban insecurity. So this is New Orleans. And when we think about one issue, so structural inequity, in this case, we were looking at access to education, education to workforce, workforce pipeline. We're looking at employability, employment, infrastructure, utility, transportation. We're looking at just the basics, right? And we're seeing that there were real issues in very specific populations in very specific areas.

[27:21](https://youtu.be/-g-vVK9YLNY?t=1641) Not only that, but those people were having a hand-to-mouth existence in those very same geographic locations, meaning their rent-to-income ratio was at or higher than 50 percent. Their access to work through private transportation, meaning a car, was extremely low. Their ability to own a home was extremely low. Female hardship really referred to women that were female-headed households with children under 18, which also coincided with these other variables of vulnerability. Crime and alcohol is ubiquitous throughout the city of New Orleans, but when we break that down, crime and alcohol, I'm going to say alcohol outlets. So the concentration of alcohol outlets, whether you know it or not, is firmly connected to STD rates, to low public health outcomes, and to higher crime. And in New Orleans, we're still suffering from Katrina, so there's still a fair amount of blight. And one thing that we were recognizing is that most of our population that were being released were released to the most vulnerable areas in there. So when we think about data, we can think about it geographically.

[28:59](https://youtu.be/-g-vVK9YLNY?t=1739) We can get to layers of sensitivity or layers of risk that underline our issues that we might not be understanding, oh my gosh, this is directly related to my problem. And that is another place where qualitative data really fits in. When your town halls happen and everybody feels really good after it's over, but you still have the same problem the next mayor and the mayor after that and the mayor after that, it's because you may not have addressed your primary issues. You may not have layered and understood that those issues are happening together. They're dynamic. They feed off of each other. That sounds like a lot, but I want to say that it happens in every city. These are issues that are occurring in every city. So if you're at all interested in seeing how your city sort of concentrates disadvantaged...

[30:10](https://youtu.be/-g-vVK9YLNY?t=1810) Now, I gave you guys a little sneak peek, but I'm going to say that while we can look at things geographically, you can look over time as well. You can look over time, you can look over space, or you can look over time-space, which is my new favorite. ArcGIS has come out with something called time-space cubes, or space-time cubes, whichever I am obsessed. you like. I love this capability. One thing that's truly important in the world of fixing problems are recognizing patterns. Some of those patterns may occur within a time-space, like heat islands. If you're worried about construction, and you have a whole neighborhood, or you have roads to rebuild, and you have, say, a heat island, your workers cannot work when those temperatures are above a certain element.

[31:21](https://youtu.be/-g-vVK9YLNY?t=1881) There are all of these temporal pieces, but there are also patterns that happen over a long time. We can see that in gentrification. That's really an easy way to think about, oh, this neighborhood didn't become gentrified overnight, but we can see over a period of months and years, the people who originally were in this neighborhood cannot afford to live here anymore. That fed our housing instability issue. That is why time, being temporal and spatial, is really important. One of the elements that demonstrates this are space-time cubes. I'm just going to take a little minute to dig into what is a space-time cube, and why the way it shares a pattern is really important. Let's imagine that this cube sits over a part of your city.

[32:25](https://youtu.be/-g-vVK9YLNY?t=1945) You're able to break out that part of your city by little geographic units. You want to say, hey, I really want to understand the patterns happening across this area, because things are really changing. The dynamics, the groups that are living here, what they need, how they exist, all of that is really changing very quickly. I need to understand that in real time. What time-space cubes are able to do, space-time cubes, I flip them all the time, based off of what I'm focused on. You can see on the color coding, if you have a new hotspot, or you have a persistent hotspot. Let's just think about this. This little column right here, it's been hot out of 1, 2, 3, 4, 5, 6, 7 time units.

[33:31](https://youtu.be/-g-vVK9YLNY?t=2011) It's been hot four times. That gives you a sense, like if you were to say those were a year, that gives you a sense that that place is persistent in its issues. Let's look at a new hotspot. Here, this place has been the same over time, and then, uh-oh, something's happened. There's a new change. Now we can recognize we're at the beginning of that change. It is dynamic, all of these ways that things are changing. Now you've got a part of town that maybe is less at risk. It's been less at risk, but it is persistently less at risk, or it's even doing better. Let's say it's been at a less risk level for a long time, but now it's really moving to the positive. Now you're getting the best outcomes you can think of.

[34:33](https://youtu.be/-g-vVK9YLNY?t=2073) You have an intensifying cold spot or low-risk spot. Now we can just go down the line and see, oh, wow, there's places that are hotspots and they're diminishing. They're getting better, but maybe they're not really in the low risk. Maybe they're not doing that well. See how, over time, we can think about where we're going to do our interventions, how we're going to do them, and I'm going to take it all the way to the top. You can do your analyses across multiple indicators in these space-time cubes. Say you want to look at this housing issue, housing instability. You want to look at the costs of houses in specific neighborhoods. You want to see that over time, but you want to look at that in reference to new taxes that have just been passed, maybe two or three years ago.

[35:46](https://youtu.be/-g-vVK9YLNY?t=2146) The beginning of that new tax code might be when you start seeing some real dynamics in each of these geographic areas. Now you can see that that may have had an impact in very specific places. When you are working in these arenas, again, it might look like it's really difficult math, These are just patterns over space and time. but it's not. You can take all of your problems and set them across and say, oh, I want to look at how this happens in relation to this, saying, I want to look at this and how it happens in relation to one, two, three, four indicators. Now you're able to do that really with ease, almost in real time. Okay, and I think when we talked about that earlier conversation about layering complexity and looking at the outcomes of that complexity, I want to just finish with two potential ways to drop into how are we really going to solve this problem and to show that you can be very specific around the outcomes that succeed or fail, I'm sorry, the target populations that succeed or fail in these dynamic settings.

[37:22](https://youtu.be/-g-vVK9YLNY?t=2242) So this visual are my population. In this particular research I was doing with the Department of Justice, Bureau of Justice Assistance and our statewide jails, prisons, day reporting centers, we were looking to see how many previously incarcerated people recidivated, went back into prison. And something that really came out in the findings was that this blue and the deepening blue were the most insecure areas. They had all of those levels of vulnerability stacked and deeply concentrated. And you almost, and the green areas are where you really just don't see those problems. You don't see those issues. Everything seems really to be functioning very well and the city is doing really well Well, here we can see, number one, my releasing population only went to the most vulnerable by these populations.

[38:35](https://youtu.be/-g-vVK9YLNY?t=2315) areas. And my people who went back to prison, so the yellow is the folks who returned and within, I think this was about a year's time, had not recidivated, but the red absolutely recidivated. And we see this separation. We see that they're only going back to the highest risk neighborhoods. And those people who went back to prison were only in the highest risk neighborhoods. There's another element to this that I want to just lay on you just to take it a step further. We can look at those dynamics over time. We can look at what drives these outcomes that we care about up and what drives them down. And this was just a little more advanced analysis on that population saying, look, if the client had a higher amount of risk themselves, just by themselves, and they had a very high level of urban insecurity, what is their likelihood of going back to prison?

[39:48](https://youtu.be/-g-vVK9YLNY?t=2388) And in this way, we can really get into these layers of complexity. It might not be just an issue of we just need to put better vouchers out. We just need to get more dollars to fix the roads and make the recreation better. There's other things at risk here. Maybe we need to do more employability for younger people. Maybe we find that the crux or the real issue is that this entire area doesn't have a good people-to-work pipeline or people are underemployed. So I'm just really imploring you as a group to think about the issues that are out there, how multilayered they can be, and how you can parse those out and start looking, oh, wow, in this group, we can start thinking about these problems and getting a lot more specific to them.

[41:05](https://youtu.be/-g-vVK9YLNY?t=2465) I am at time. I can't wait to hear your questions. My organization is a nonprofit. It's the longest acronym ever. It's called ISTART. But what we are doing right now is just this. We're looking at complex issues, and we're parsing out some of this vulnerability. We're parsing out some of these risk drivers. We're parsing out what are the structural drivers that are keeping our populations in the same situations or worsening situations over time. And we're thinking about this all the time. So as I mentioned, if you're looking at a problem that feels very complex to you, you've thrown this intervention, that intervention, the other intervention at it, all three interventions at the same time, but you're still seeing the same outcomes, I would love to talk to you about it and think around those issues together.

[42:06](https://youtu.be/-g-vVK9YLNY?t=2526) Thanks, guys. All right.

* * *

Original page: https://innovate-us.org/smart-data-fair-cities-evidence-based-approaches-to-city-management · Licence: CC BY-SA 4.0 (https://creativecommons.org/licenses/by-sa/4.0/), as stated on innovate-us.org
