Thursday, November 13, 2014

DeKalb County's new GIS model explores the financial impact of incorporation

The November 12 Task Force on Dekalb County Operations meeting featured a presentation by Dr. Alfie Meek from GA Tech. Dr. Meek is part of a team that has developed a model to help DeKalb County better understand how incorporations impact the county's overall budget. Previous tools focused on revenue, which is easier to calculate as property and commercial tax data and business license data are readily available. This new model creates a more nuanced picture by looking at revenue vs operation costs.

The model focuses on three county funds (271, 272 and 274), as opposed to on all 30+ county funds, because those three funds are the ones that are impacted when incorporation occurs. A fund is a part of the county's budget that focuses on specific functions.

This computer model currently resides at GA Tech but will move to DeKalb County's servers by the end of the month. The public will have access. The model is not perfect but it is a great start, as it allows the users to quickly look at the costs of servicing a certain area.

To help users orient themselves, the model includes some pre-set locations to zoom into areas of interest (e.g. location of contested areas such as Executive Park). To demonstrate how the system works, Dr. Meek drew a polygon around a "test city" and pulled residential and commercial digest data as well as expenses for services to the area.  These expenses are estimated based on a line item review of existing county budgets. As was explained, the model is only as good as the data upon which it is built, and many municipalities (i.e., not just our county) cannot provide the level of detail that would be ideal for this type of simulation. That means very specific data such as where an emergency call originates and the time it takes a county employee to respond to the call and complete required actions (this example related to whether it's costlier to police commercial areas vs. residential areas). Another example discussed road maintenance, where municipalities traditionally estimate maintenance costs based on number of miles and not on a formula that accounts for both miles and number of lanes. Through individual and combined views of the three funds mentioned above, the map changes color to show if an area is a donor (that produces more revenue than it uses in services) or a recipient. Numerical and bar graph breakdowns of revenue vs. operation cost are also displayed. Users can save queries and return to them later (this data is stored at the user's computer).

The model interface allows users to select/deselect blocks to ensure the polygon covers the desired area. The model uses the existing blocks into which the county is divided but annexations rarely respect such tidy boundaries; the model would need higher resolution to provide a more granular view of finances (so, ~2000 blocks vs the current ~200+ block units that describe the whole county). This is an important point as annexations are being drawn to neatly hug the lines of desirable commercial property and the model cannot account for part of a block, just the whole block. The model assumes that a new city will take over all services and there is some flexibility to "add the services back" for example, in the case of police services. It becomes harder to generate "a la carte" scenarios and hopefully the model will continue to be refined.

The model will be applied to maps for proposed cities (the Briarcliff/Lakeside and Tucker maps are supposed to be finalized by November 15). The new model should allow some comparison between county data and data in city feasibility studies. As noted by an audience member, we can expect to see discrepancies as feasibility studies use different methodologies for their calculations. An example was that the DeKalb budget currently allocates about $3-4K per year per acre  for park maintenance but a Vinson Institute study used a $11K/year figure, based on other sources. In other words, we will soon have more information but will still need to wade through some thickets to get to the bottom of things.

Lastly, a note to naysayers: we should refrain from dwelling on why the county did not have this tool already and focus on what this model brings to the table and how it can be built upon to provide increasingly accurate data on county operations. For those who want to know what DeKalb County is doing differently this year, compared to last year, this is an example. Also see links to the Task Force website and Blueprint for DeKalb initiative on the right margin.

Our thanks to Dr. Meek and "Jay" (budget guy) for ably handling the presentation and to the Operations Task Force for getting the ball rolling on this model.