Exploring Customer Service Requests (311)

The above map summarizes 2024 Dirty Alley and Dirty Street complaints by Community Statistical Area (CSA). The color of the polygons represents the time taken by the City of Baltimore to resolve and close the tickets, while the extrusion height of a CSA polygon represents the 2023 residential vacancy rate. Clicking on a polygon expands a pop-up, which shows the name and numbers for that CSA.

For each 3-1-1 report, the number of days to close (defined as CloseDate - CreatedDate) was computed. Then, for each CSA, the mean number of days to close was computed. The color scheme is continuous based on the number of standard deviations above or below the mean number of days across the City: see Figure 1.

Variable	Mean	SD	Median	Range
Days to Close	3.39	0.41	3.37	[2.56, 4.48]
Vacancy Rate	5.94%	7.69%	2.62%	[0.03%, 28.47%]

Figure 1: Descriptive Statistics of days to close and vacancy rate.

In the map, there does not seem to be much of a relationship between time to close and vacancy rate. In fact, it seems that the relationship is closer to strictly geographic in nature: visual inspection demonstrates an apparent line from northwest to southeast across the City. CSAs west and south of the line receive slower service than those north and east of the line, even though there are areas of higher vacancy on both sides of this divide. The slowest response is in South Baltimore, which has the second-lowest vacancy rate (0.12%); the fastest response is in Southeastern, which has a vacancy rate (1.03%) well below both the median and mean for the City as a whole.

This relationship (or lack thereof) is further demonstrated in Figure 2. The low \(R^2\) value and high \(p\)-value both demonstrate that there is no significant relationship between the mean time to close and the vacancy rate within a CSA.

Figure 2: Linear relationship between days to close and vacancy rate.

In 2019, Christine Zhang and Talia Richman described Baltimore’s approach to responding to 3-1-1 calls for the Baltimore Sun. In the article, the authors detail the City’s approach to dealing with service requests: “the Department of Public Works’ Bureau of Solid Waste divided the city into five geographic sections [each with] a division chief and resources to respond to service requests.” Division of jurisdictions into smaller units to better handle local issues is a common approach in local governance; it also explains some of the patterns described in the map.

The Southwest section of the City, including neighborhoods such as Curtis Bay and Cherry Hill, is more commercial than many of the other divisions. Commercial activity generates much larger quantities of refuse and often leads to higher instances of illegal dumping. Zhang and Richman mention that though the City planned to use more resources in the Southwest division, staffing shortages prevented them from being able to do so.

The areas east and (especially) north of the Inner Harbor are much more residential: it’s just not possible for them to generate as much trash as the commercial and industrial areas. In addition, because they have more residential properties, the vacancy rates in these locations are more stable, and averaged out over more homes.

These three factors are particularly important in disentangling the relationship between 3-1-1 call resolution time and residential vacancy—Baltimore’s Open Data portal allows the public to examine how City processes could be improved, and why processes change across space.