Special Guest Speakers: 

Federated Models 

Johnson: “Everybody understands what centralized means and what decentralized means, federated is a fuzzier notion… You may remember that the way our federated model of government works in the United States is that certain functions are handled by the federal government, like national defense, and the Constitution defines those things that happen at that level. Everything else belongs to the states, so states have the ability to implement their own laws… When we talk about federated in the context of these enterprise data models what it really is fundamentally about is shared authority.”

When managing statewide data models can take a number of different forms. Broadly speaking they fit into three categories: Decentralized, Federated, and Centralized. If you want to learn more about these three models and the differences between them please watch the first webinar in the series. 

Of these three models perhaps the hardest to define and visualize is the federated model. Federated data management models can take a far greater variety of forms than either a purely centralized or decentralized data making them challenging to discuss. However, AppGeo has found that federated models typically possess three main characteristics. First, federated models allow local governments, such as counties or municipalities, to maintain ownership of whatever data they produce. Second, the state, or another overseeing body, provides a framework to aggregate and standardize the local data and then make it available on behalf of the data owners. Third, local governments have the ability to perform their roles themselves or allow contractors or even the state to edit their data with their approval. 

Challenges to Creating a Federated Model 

Nelson: “We didn’t have a process, we didn’t have a standard, and so we agreed first on we need to let those that are gonna submit data know ‘what does right look like?’ We need to have that standard, we need to be able to articulate that standard… We need a process, we need a committee, we need a submission process, we need to do outreach, and so while we didn’t have all those things lined up we pretty quickly identified what we needed. From there we had our starting point.”

In creating the Kansas Next Generation 911 system, the state’s GIS office had a huge amount of work to do to gather data from a variety of partners and agencies. Thankfully they were able to borrow ideas and models from other state data management programs to help create the groundwork for their system. From 2013 to 2020 the program evolved heavily. When the project was started the data they needed was heavily decentralized with data stored inconsistently between stakeholders. However, with funding Kansas could study the problem and identify gaps in the data that needed filling, as well as to bring existing data up to a usable standard. Once basic standards were met, data maintenance could begin. By setting standards and creating specific roles in local government, local data could be standardized and used statewide. The program was able to get lots of local buy-in, with high levels of participation. As an added protection however, in 2019 local participation in the program was mandated by the state.  DASC Clearinghouse used its preexisting local connections, as well as providing training and communicating with local leaders to develop local partnerships. By using a federated model, Kansas used local input to help develop their standards as well as communicate expectations to local partners clearly, while allowing them to maintain their autonomy. 

The Power of Outreach

Battles: “Our main thing was we decided ‘let’s get in front of people and talk to them anytime we can,’ especially our GIS maintainers or our PSAP administrators. We can explain what GIS is and how it fits in the Next Gen 911. One thing that kinda coincidentally happened with this project was I didn’t know what Next Gen 911 was… I feel like we learned along with a lot of our maintainers what Next Gen 911 was, and that helped further build these relationships. We found our early on whether we were talking to a regional meeting or specifically to our data maintainers they’re like ‘what exactly is next gen 911?’ So we went back to the drawing a few years ago and created a video, just what is next gen 911. I think us learning together really did help maintain and build that trust with our partners. Stickers and candy never hurt either.”

Without local data and local buy in, federated models cannot function. Talking to partners and understanding their position within the project, as well as the difficulties that they might face in participating fully is essential to keeping a free flow of quality data. The on-going relationships between local governments and state agencies determine how successful these types of programs can be. Shared authority needs open and constant communication between the parties sharing that authority. Even with a mandate requiring local governments to provide data to the Kansas Next Gen 911 project, they still rely heavily on their local relationships to ensure compliance. 

Are you responsible for working with state and local partners in GIS? We want to hear from you.

Do you want better statewide GIS data? Share your use case using the contact form.