Clark County Nevada, home to the city of Las Vegas, has experienced rapid population growth. In 1920, the county had a population of less than 5 thousand. A century later, it’s grown to nearly 2.5 million. The 2023 population is estimated to be 2.4 million, a 25% increase from the 1.9 million of a decade prior. This trend is only projected to continue.

Such an increase in population can strain a city's infrastructure if not properly planned for. Not just in the basics of housing and jobs and the underlying utility grids that supports them, but social services too. Schools, health care facilities, fire stations, and police centers need to be constructed or expanded.

To adequately prepare for the future, the city will need to plan where best place these efforts. The question is, did the city properly plan for today, and if not where would be the best place for improvement?

Data Sources

Multiple data sources are needed to be combined, compared, and otherwise analyzed to produce a useful result. Primarily there's a need for census data, to identify socio-economic factors across the study area. Additionally, a there's a requirement for current crime statistics. For this, the Las Vegas Metro Police Department maintains a dataset for all calls for service within the past 30 days. Finally, a list of current police facility locations.

These sources can be found in the following locations:

• Census Data:

  • US Census TIGER Tract Data - https://www.census.gov/geographies/mapping-files/time-series/geo/tiger-data.html

• Crime Data:

  • ArcGIS Portal - LVMPD Reported NIBRS Calls for Service - Last 30 Days

• Police Station Locations:

  • Clark County Geographic Information Systems Management Office Hub Site - https://clarkcountygis-ccgismo.hub.arcgis.com/datasets/ccgismo::law-enforcement/explore

Process

To produce an answer, this analysis is broken in to three major parts.

1. Analyze Population Metrics

2. Analyze Police Data

3. Compile Results

Step 1: Analyzing Population Metrics

While there are many factors that attribute to crime, some can be more significant than others. Of them, this analysis will look expressly at Population Density, Poverty, and Educational Attainment.

For each of these attributes, specific data fields from the relevant tables are joined with the ACS_TRACT_NEVADA feature class, using the GEOID_data and GEOID fields, filtering the data to just those census tracts that make up Clark County Nevada, those tracts with a 003 in the COUNTYFP field. The specific data fields are then used in various calculations to produce information for the creation of a choropleth.

NOTE: While Clark County is significantly larger than the city of Las Vegas, due to the size of many of the census tracts and that the majority of the county population exists within the area making up the four distinct municipalities that are collectively known as Las Vegas, the maps produced will be focused on the greater Las Vegas area.

1. Population Density. If criminals make up a static percentage of a population, then crime is directly proportional to the population. Even should it not be static, a higher population results in more "targets of opportunity" for criminals to engage. This analysis will assume that a higher population count will have a higher level of crime.

• Census Table: X01_AGE_AND_SEX

• Data Field: B01001e1 - Total Population

• Calculated Field: Total Population / Polygonal Square Area (in miles)

2. Poverty Level. Excluding crimes of passion, it could be assumed that most crimes are crimes of need. Objects stolen can be sold to pay for food, housing, or clothing. Property destruction and assaults can occur during such thefts, whether they're successful or not. Assuming that people below the poverty level would have a greater need for basic human necessities, this analysis will assume that an area with a higher percentage of its population below the poverty level will have a higher percentage of crime.

• Census Table: X17_Poverty

• Data Field: B17001e1 - Total Population
  B17001e2 - Total Population Below Poverty Level

• Calculated Field: Total Below Poverty / Total Population

3. Educational Attainment. Although individuals of all intelligence levels can and do commit crimes, the types of crimes committed are different. Senior business executives, having graduated from multi-year colleges, are more likely to commit fraud or tax evasion than to steal a car, for instance, and such a crime would not be under the jurisdiction a local police department, but rather a federal agent and not something in scope for this analyis. Under that idea, this analysis will assume that areas with lower educational attainment will have higher crime. Specifically it'll focus on those without a high school diploma or equivalent.

• Census Table: X15_Educational_Attainment

• Data Field: B15003e1 - Total Population Over 25
  B15003e2 > B15003e16 - Sum of those 25 or over whose highest educational attainment is 12th grade or lower, without a diploma.

• Calculated Field: Sum of Those Without a Diploma / Total Population Over 25

Step 2: Analyzing Police Data

While the entire Las Vegas Metro Police Department is spread across the greater Las Vegas area, not all police stations are equipped equally. While the data for each station's manning and resources is not available, the distance a single car can cover within a time period is. Loading the list of current police stations into the ‘facilities’ layer of the ArcGIS Network Service Area Analysis tool, a coverage area for the LVMPD can be established, showing areas within 5, 10, and 15 minutes of drive time from each station.

Of course, the ability for police to respond is only half of the equation. The demand for their service is the other half.

Clark County’s GIS Management Office provides products containing the locations and types of each service call responded to, covering the past several years. These products are point-based feature classes based on the addresses of each service call. Limiting the data to a specific time period, January through December 2022, and then performing a count-based Spatial Join with the census tract data, the number of calls from each census tract can be determined. Averaging each across the 12 month period provides a useful metric to highlight areas that have a high current demand for police action, which can then be shown as a choropleth map, similar to the population based maps from earlier.

Step 3: Compiling Results

Each census tract can be given a ranking for each of the data attributes. Adding a field to each class to hold the ranking, each rank field can be joined to the original census tract feature layer.

A new field can then be created, calculated from each of the rankings and given a weight-based multiplier:

• Population Density: 10% - While population density can play a factor in the potential for crime, plenty of population dense areas have low crime. Thus, it is given a low ranking.

• Poverty: 15% - Poverty can lead to crime, but it is not the only factor and, for the purpose of this analysis, will be assumed to not be the most significant factor.

• Education: 20% - This analysis will assume that those who commit crime do so, because they're not smart enough to figure out another way out of their situation, not smart enough to understand the consequences, or not smart enough to realistically accept their own capabilities, and that this plays a larger factor than that of poverty.

• High Crime Areas: 25% - This analysis is focused on determining where additional police forces should be based at and, as such, existing crime plays a bigger factor than potential crime.

• Distance from Existing Stations: 30% - Considering that a key purpose of this analysis is to determine if a new police station would benefit the city, this ranking is given the value of most import.

Using these metrics, an equation can be written to provide a final ranking for each of the census tracts:

(!RankedDensity! * 0.10) + (!RankedPoverty! * 0.15) + (!RankedEducation! * 0.20) + (!RankedCrime! * 0.25) + (!RankedDist! * 0.30)

With each tract ranked from 1 to 5, from most probable crime locations to least, there distinct areas that could benefit from additional police forces. Although distance from existing stations was given the most weight, many of the areas are still within the 10 minute response time. Ideally, a new police station should be built outside of that coverage area, and combining the weighted results with the police station coverage area highlights a couple of specific areas that meet that criteria.

Results

Comparing the data highlights potential areas where the city could construct additional police stations. To determine if these locations are enough to significantly increase the coverage area, a test needs to be performed.

Creating a new feature layer to house the potential sites, a second round of network analysis can be performed. Using the same criteria as before, a range covering 10 minutes driving time from each station is created.

Overlaying the coverage of potential sites, with the coverage of the existing stations, the new locations appear to fill in the gaps well. Whether they'll be enough to support the future growth of the city is hard to tell, but when the current need isn't yet met, these locations do provide for significantly increased security coverage.