Stefanie Costa Leabo, Chief Data Officer for the City of Boston, talks about her non-linear career path (she considered becoming a political science professor at one point), the tangible connection between her work and what happens outside of city hall, and how sharing data helps Boston's residents better understand what's going on in their city.
Anthony Deighton: Welcome to DataMasters. I'm Anthony Deighton, Chief Product Officer at Tamr. My guest today is Stephanie Costa Leabo, Chief Data Officer for the City of Boston. Stephanie leads the city's efforts to use data to make government more effective and deliver better outcomes for constituents. Stephanie took a nonlinear path to the Chief Data Officer role, and you'll hear about that journey. She considered becoming a political science professor at one point. She was an analyst for the City of Somerville before joining the City of Boston as an analyst and eventually becoming Chief Data Officer.
Anthony Deighton: Stephanie will also talk about the various projects she's worked on, including Boston's City Score Initiative, what opportunities arise when government data silos are broken down, and the tangible connection between her work and data. Stephanie, thank you for joining us today and welcome to DataMasters.
Stefanie Costa Leabo: Thanks for having me.
Anthony Deighton: So maybe we could begin a little bit with actually with your career path, because you've been amazingly successful as you've moved from Somerville to Boston, and now as Chief Data Officer. Perhaps you could share with the audience your journey and maybe imagine some of our listeners are aspiring Chief Data Officers themselves. What could they learn from your career history?
Stefanie Costa Leabo: Yeah, absolutely. I think the first thing I would say is that I never expected to be in this line of work. If you had asked me in high school or even college. Chief Data Officer positions, especially in government didn't really exist while I was in school, so I think I'm very fortunate to have found my way here, but it was perhaps a nontraditional or nonlinear path.
Stefanie Costa Leabo: I studied political science as an undergrad and towards the end of my undergraduate career had some faculty reach out to me and say that they thought that I might be a good candidate for postgraduate study in political science to ultimately become a political science professor. So after I graduated from UMass, I went to George Washington University to start a PhD program. And I was really fortunate to in addition to getting a great political science education, to be in a program that had a really heavy quantitative methods and econometrics component as well.
Stefanie Costa Leabo: So I got a really solid educational foundation in how to use data to answer questions, particularly in a social science type context. And just over the course of my PhD studies, I reached a point where I realized I really wanted to have more of a connection between my work and its impact on the world around me. I wanted something that felt more tangible and I was fortunate enough to find a position working for the City of Somerville as an analyst. And from that point on, I was just absolutely hooked on local government. In my mind, it's one of the best places to be able to experience that tangible connection between your work and its impact.
Stefanie Costa Leabo: Everything is about working to improve your community and the services that are provided to yourself and your neighbors. You are kind of benefiting and experiencing what all of your constituents are. And so you can learn through your experience of moving throughout your community and draw on that and see where there are opportunities to perhaps make changes and improvements that will then affect others. And it's really easy to draw a line between what you do in your job, to what you see happening outside of City Hall on the streets, and that's just a really fantastic feeling.
Anthony Deighton: I love this idea of the tangible nature of the work, because at some level data isn't tangible at all, it's just totally ephemeral. And yet you've been able to create this connection between this ephemeral idea of data and this very tangible experience of your day to day life living somewhere, being a part of a city and the services that it provides. And it strikes me that the City of Boston in particular, and especially in your role leading data initiatives there is, in a way, it's the city has also grown up in terms of how it's used data. And you've been a big part of that. And your career has been a big beneficiary of that. Maybe you could share a little bit about, in your time at the City of Boston, how their relationship with data and how they see it affecting citizen lives in that really tangible way, how that's changed.
Stefanie Costa Leabo: Yeah, yeah. So I think another really key point of my professional story is just lucky timing. So when I started working in local government in these work data-driven roles, data driven decision making in government, particularly in local government, was still relatively new. And so I had this really wide open blank piece of paper in front of me. There really wasn't a playbook to follow or sort of a particular path that things needed to progress on. And so it provided a lot of opportunity to advance fairly quickly within this field. And I think what really made sense for me in this work is that what so many cities and towns need right now, or what Boston has needed over the last few years, is someone who understands the the business case and the value proposition of data and how it can be used to make the city more efficient and improve services, rather than someone with perhaps a more traditional computer science background.
Stefanie Costa Leabo: So that is not me by a long shot. I don't have that more traditional technical background. My background is much more functional and focused on sort of the outward facing components of using data. And so my hope in this role is that I can help shift the culture and practice of data driven decision making in Boston to the point where five to 10 years from now I kind of work myself out of a job, and I'm no longer the right person for this role and can hand over the reigns to someone who can take the team even further from a technical standpoint. But I really feel like you need to have a solid foundation and strong buy-in across the organization before you're able to be truly successful on that deeper technical work.
Stefanie Costa Leabo: And that's not to say we haven't made great strides from a technical standpoint. We've implemented the city's first centralized data warehouse. And have really focused on building out our in-house data engineering capacity as a team, which is quite unique for municipal data analytics teams in particular. So we've been doing that work this whole time, but in terms of what the leadership strategy for the team looks like, I think what Boston needs right now and what most government CDO positions need right now is someone who can make the business case for why data is important and how it can be helpful across the organization.
Anthony Deighton: I think that's an important point, and I agree with you. Many city governments would benefit from taking that sort of value driven approach to thinking about data, but I'm sure our listeners would agree that that's probably true across many different domains. Almost every business would benefit for thinking about their data driven initiatives from a value perspective. And I think everyone probably can also empathize with this feeling of not having a playbook and not having... you're starting with a blank sheet of paper. Many organizations are quite immature as they handle sort of this deluge of data that they create and thinking through how to get value out of that data and how to bring that data together, how to figure out how to break down those data silos.
Anthony Deighton: And maybe you could share from your perspective, especially from this perspective of data sitting in silos, less from a technical perspective, as you point out, more from the value perspective, some of the value that you've seen in Boston, where you can break those silos down, bringing data together that wasn't previously brought together, and then show that value within the government and also back to the constituents.
Stefanie Costa Leabo: Yeah, absolutely. I think a really great example of all of that is the work that the analytics team has done to support the city in the implementation of it's relatively recent short term rental ordinance. So for a little bit of background context, the city passed an ordinance that impose regulations on short term rentals. So what people commonly refer to as Airbnb's, but there are a number of companies in this space because of the impact that it was having on our housing and our rental prices in the city. And so our work started before the ordinance was developed in working with our department of neighborhood development to see what we could do to quantify the impact that short term rentals were having on our housing market. And we were able to demonstrate that the growth of short term rentals was increasing rental prices, decreasing available housing stock, and really contributing to an overall housing stock shortage and housing instability that the entire greater Boston region is really struggling with right now.
Stefanie Costa Leabo: But then when it came to implementing this ordinance, it's been really, really interesting because data is so critical to every aspect of the implementation and enforcement of this ordinance in a way that I have never experienced before. So to just give you a few examples of the work that we had to do to make this happen, and specifically to speak to your question about silos, we had to create from scratch an eligibility dataset. So what this was, was an interpretation. It was a data interpretation of all of the criteria that were laid out in this ordinance for which residential properties could or could not be eligible to register as short term rentals. And so we had to combine many different data sources that had never been linked together before. So we needed information from our assessing data that gave us information on the characteristics of a given property.
Stefanie Costa Leabo: Was it a single family home? Is it an apartment building? Is it a condo? How many units does it have? All of these things were relevant to determining eligibility. In addition, we needed to link up other data sources from our code enforcement dataset, because if you have an open code enforcement violation or have had more than three violations within the past six months, you are no longer eligible to be an active short term rental. And so there were five or six of these core datasets that we needed to link together. And what we ultimately created was an eligibility dataset for every single residential housing unit in the city that would tell you whether or not that unit was eligible to be a short term rental and for which type of registration it was eligible for.
Stefanie Costa Leabo: And that data is all up on Analyze Boston, which is our open data portal. And so it's shared with the public, but also through Analyze Boston, it's shared with Inspectional Services, and Inspectional Services is using this data every single day as they're reviewing applications for short term rental licenses. It's one of the first checks that they do to see if a property is eligible or not. Now ISD has the ability to override this data if a homeowner is able to provide proof that our initial assessment was incorrect, but it's been really foundational to being able to evaluate and register applicants for these new licensed types. So it's [crosstalk 00:14:37]
Anthony Deighton: That's really fascinating, because I think what you're sharing there, this is something that everybody can really empathize with. This idea of creating change, not only within the City of Boston, but actually within the whole community about how we think about both the positives of short term rentals, but also potentially some of those hidden costs, and those costs are hidden in those disparate data that, to your point, have never been brought together. And I think that's also another really fascinating side to the story, which is that you're sharing the results of that analysis publicly, in this very open and public way. And maybe that's something you could share a little bit with, I think... Especially for commercial organizations, they don't often think this way about taking data and making it publicly available, but as a government, especially when accountable to its constituents, you think about that in a very different way and maybe that's something worth sharing, that difference.
Stefanie Costa Leabo: Absolutely. Yeah. I think one of the things that I've enjoyed most about doing data work in government is that we don't have this concern or pressure over protecting IP or a bottom line. And so we're able to be much more open about the work that we do. And I would say this takes two main forms. So the first, as you mentioned, is open data. So we run Analyze Boston at data.boston.gov, which has over 150 datasets that anyone is welcome to explore, analyze, visualize, et cetera. And we're constantly looking to add more data there. And so, I think generally that's just a really great tool. That's part of the city's commitment to being open and transparent about city services and what the city is doing.
Stefanie Costa Leabo: So you're able to look up assessing information. You can take a look at all of the open building permits. You can analyze all of the non-emergency requests that come in through our 311 system. So it really gives you the opportunity to dig into the nuts and bolts of what's happening behind the scenes in the city.
Stefanie Costa Leabo: And then the other thing that we do, is we collaborate with other cities and other government agencies, and again, it's great that we don't have to be concerned about protecting a bottom line, and we have the opportunity to be really open and share with other communities, either something that we did that we found really effective and just want others to know about or vice versa, to be able to replicate great work that other have done, and that sharing is really a core part of what we do and allows us to avoid having to reinvent the wheel, because local governments across the country, while they might look somewhat different in form, we face a lot of the same common challenges. And so there's a lot that we can learn from one another and a lot of additional time, effort and money that we can save by sharing what what works and just not starting from scratch every time we face a new challenge.
Anthony Deighton: Yeah. So I think something, certainly, the private sector could learn there as well is there's more that makes us the same, than that makes us different. And that by sharing the data and the insights that you gain from it, you could potentially drive a significant value inside the organization. And it strikes me that CityScore is another good example of that. A way of creating that accountability between government and the constituents, and also sharing between cities, and also between government agencies, how effective the city is being at providing services. And again, strikes me as something that's very different about the way you approach the task and the work as Chief Data Officer in the City of Boston, versus how we might think about a chief data officer in a commercial enterprise. You're literally taking your performance management dashboards and publishing them for the world. Maybe you could share a little bit about, maybe first, what is CityScore as not everyone may know. And then also just how you think about that differently because it's so public and so valuable.
Stefanie Costa Leabo: Yeah, absolutely. So CityScore was actually the first project that I worked on as an analyst. I was assigned it my first day back in 2015. So CityScore has story [crosstalk 00:20:07]
Anthony Deighton: The first day job.
Stefanie Costa Leabo: So CityScore has been one of the most consistent and constant things throughout my tenure here in Boston, which is kind of cool. But at the end of the day, what CityScore is, is a performance metric, performance indicator platform that is really designed for a mayoral and executive level. And what it does, is it normalizes the series of metrics across lots of different government functions and scores them, and then aggregates that score for different time periods. So you'll have a score for the day, the previous week, the previous month and the previous quarter. And so it functions is a high level indicator tool for the mayor and cabinet chiefs to be able to have an at a glance way of assessing how well the city is doing across a variety of indicators.
Stefanie Costa Leabo: And that was sort of the initial ask from the mayor's office. It was very open ended, but they wanted a way to be able to quickly assess how well the city was doing on any given day. We have hundreds of performance metrics across all of the departments in the city, and we have developed a lot of dashboards that present them, and the mayor has those on his dashboard, but when you're Mayor Walsh and you have maybe five minutes to yourself in your office on a good day, it's just really difficult to take in 80 to 100 metrics that are flashing across your screen on a series of dashboards. Those are really great for the cabinet chiefs and the department heads who lead that work, but even for those of us that work in performance management, it's a lot to keep track of.
Stefanie Costa Leabo: How is this metric defined? What is the target? Is 80% good? Is 80% bad? And so what we wanted to do was create something that would really quickly and easily allow anyone to know, "Okay, this metric is doing well. This one maybe need some attention." So that was the intention behind CityScore. And so in addition to having that on the mayor's dashboard... So he has a series of screens in his office with rotating dashboards, with various metrics and data visualizations that allow him to stay up to date with what's happening across the city. We also publish city score on boston.gov at boston.gov/cityscore. So anyone that goes to that site can see the exact same data that the mayor is seeing at that time. And so it's both an internal performance management tool, but the hope was also that through external engagement from the public, that it could be an accountability tool as well.
Anthony Deighton: Again, I think that's amazing, and I think, again, for the people thinking about the challenge of creating accountability inside an organization, again, I think you have that problem times two, you need to create accountability within city government, but also between city government and the public. And I think it's a really exciting and bold initiative to take those metrics. I think it's actually really fascinating that the mayor looks at the same metrics... That I can go to the city website and go find that those two things are the same. Creates a tremendous sense of alignment and a common purpose between governments and the citizens. So using that idea, and maybe to link your work today back to your academic background, I know you studied really this linkage between data and democracy, political science and democracy. Maybe share some thoughts about, in your perspective, about how your approach to data, to open data, to creating accountability maybe operates, in a broader sense, to create democratic institutions, and break down some of the strife we see on the world with non-accountable governments?
Stefanie Costa Leabo: Yeah, it's a really great question. I think the most obvious place that this is happening is, as we discussed earlier, through open data. And so it's a great tool for transparency. We get a lot of engagement with our open data from the public, particularly from the academic community. Students are frequently using Analyzed Boston to do projects through their coursework. Academic researchers are using it in their research, in their publications. So we get a lot of great engagement, particularly from the academic community, I would say. Although we do get questions. You can submit questions to us through our open data portal. And I would say we get at least a handful of questions about open data every single week from just average constituents who are curious about what they're finding.
Stefanie Costa Leabo: But I think what's interesting at that point is... There's two components, I think, that you need for open data to be truly effective and have the potential to be transformative. So access and availability is really important and that's the first step. The data has to be there. It has to be publicly available, but then the second part is data literacy. So I think there's a reason that we see most of our engagement with open data coming from the academic community. They are the ones that are often the most prepared and have the right skills to be able to leverage the data that we make accessible. So I think for open data to really become more of this force for transparency and accountability, we need to see an increase in data literacy across the general population.
Stefanie Costa Leabo: And there are two things that our team has done in this area to try and move the needle there. So our open data portal has existed since 2012. It just went through a major revamp in 2017, and we are still working to improve the documentation and the context that's provided with our data to make it more readily usable. You need to have all of the surrounding context of a dataset to know what the fields mean, and how you should be using it, and how it relates back to either the services or the business processes that that generated. So there's a lot of work still to come from us on improving that documentation and just packaging the datasets in a more usable way.
Stefanie Costa Leabo: Another thing that we've done... So I'm going to blank on my dates here, because the years kind of blurred together, but I want to say it was around 2018. We did a partnership with the Emerson Engagement Lab to create an online course for librarians, in particular, to teach them how to become data ambassadors. So this course was all built around Analyze Boston and was designed to help librarians, particularly in the Boston Public Library system, but I think there were university librarians and Boston public school librarians involved as well, but to teach them more about open data and how it can be seen as just another research resource that librarians can point their constituents to when they come to the libraries, and how they can help their patrons leverage open data as they're trying to answer questions about the city. And so that was sort of a pilot that we did. It was really successful, and I think we're hoping in the future to build out that type of programming in a more robust way, and to also start training internal city staff on those practices as well.
Anthony Deighton: I think that's totally fascinating, because I think what you point out is, it's not just enough to throw the data out there, but you need to give people the tools to make meaning of that data. Documentation, education are obviously really important dimensions of that, and clearly investments that you and the city have made, and that's great. I wonder, also, if the siloed nature of government contributes to that challenge. If you think about it, governments, by their nature, are very siloed. Boston is a sort of standalone unit compared to Massachusetts, which is a standalone unit compared to the United States, which is a standalone unit compared to the world. These are natural silos. We naturally create silos within our governments. But I wonder if there's also a burgeoning opportunity to make meaning out of the data that, in your case, the City of Boston's collecting, by breaking those silos down and thinking about how we resolve and reconcile data between departments, if you think about within the city, but also between cities, between the city and the federal government, and maybe you could comment a little bit on your view of the cost or impact of not having data reconciled between these silos might have, what opportunities might exist if we could break those silos.
Stefanie Costa Leabo: Yeah, you're absolutely right. That government just tends to be siloed. That is not unique, I think, in any particular [crosstalk 00:31:30]
Anthony Deighton: Commercial organizations suffer from this as well.
Stefanie Costa Leabo: Yeah. And so, I think what was really great when Mayor Walsh created the analytics team in 2015, we were the citywide analytics team. And I think that citywide was chosen for a reason. The intent was to signal that we would work across all departments to try and break down these silos and be proactive in bringing data together to help create as rich a picture of what's going on in Boston as possible. And I would say there's a functional component of this and a technology component. So sometimes there's a disconnect between people or programs, and when someone just isn't aware that there's information in another team or another department that could be helpful to the work that they're trying to do. So sometimes it's just a lack of awareness of what's out there.
Stefanie Costa Leabo: And then sometimes people are actively trying to share their data, but the technology that they're working with makes it really challenging. So there's definitely a technology modernization component to all of this as well. And I would say one of the most significant projects that our team has taken, it's more of a backend project. So a couple of years ago, we launched the city's first centralized data warehouse, which is managed by our team. And that was really our first step in working to ultimately resolve these types of issues of data siloization in particular. And so we continue to get access to more and more data from across the city, and we keep it in our warehouse, and it gives us the ability to join data together that would otherwise be separate. I think that short term rental project is a great example of how you can leverage data from multiple departments to create this more complete picture of what's happening on the ground in the city.
Anthony Deighton: Yeah. Silos naturally create barriers between data and the work you're doing is really breaking that down.
Stefanie Costa Leabo: Yes, exactly. I think another key area that we're focused on in this context is on addressing, and just in general data quality. So depending on the application or system that data is being entered into, and at sometimes depending on the individual that's doing that data entry, you could see a handful or 10 different ways of spelling out the same address. And sometimes the systems that we have are not sophisticated enough to reconcile and to match those. And so we've been doing a lot of work to clean up our advertising data and really standardize it, as well as to put some data processes, some functions in place that allow us to do better matching of inconsistent addresses, as well as, in some cases, names. So if we're talking about being able to match property owners across multiple systems, that has really improved our ability to get a complete picture of what's going on in the city in terms of costs or opportunity for value. What that allows us to do is to make sure that we are issuing permits or code enforcement violations to the correct address and the correct owner, which increases the likelihood that those things getting paid. And so there can be a very real financial too…
Anthony Deighton: A very real and direct benefit to matching the quality of the data.
Stefanie Costa Leabo: Absolutely.
Anthony Deighton: But I think to your point, there's a tremendous amount of value, not only when you bring data together, but then as you bring that data together, you realize potentially the quality of that data isn't what you thought, and there's a need and value in cleaning it up.
Anthony Deighton: Absolutely.
Anthony Deighton: So I know we're out of time. And I want to thank our guest, Stephanie Costa Leabo, Chief Data Officer for the City of Boston. Also, want to thank you for your work. You're adding a huge amount of value to the City of Boston and making it... As a resident of said city, I personally value and appreciate the work that you're doing.
Stefanie Costa Leabo: Great. Thank you, Anthony. It was really great. Yeah.