New Orleans Firearm Violence Dashboard

Police data provides a lot of detail about shooting incidents, like where and when they happened. But it often has very little information about the people who are victims of these incidents. Death records from LDH and the Coroner’s Office have more details about these people. However, these datasets usually do not clearly link people to the specific shooting incident and, therefore, do not have the same information as police data. Information recorded in one dataset may not be available in another dataset, or the information may not match across datasets.

For example:

• Death records may not list the exact location of the shooting.
• The date of death may be different from the date of the shooting.
• The place of death may not be where the shooting happened.

By linking all these data sources together, we can:

• get a more complete and accurate count of homicides by shooting;
• fix errors in the data;
• fill in details about harmed people (such as age, race, or address) that are missing in the police data; and
• better understand where people lived compared to where they were killed.

This information can help guide future violence prevention and intervention efforts.

First, we connected victim information from LDH and the Coroner’s Office.

• We used a computer software (R Program) to match records (with the reclin2 package) that appeared similar based on first and last name, sex, date of birth, and date of death.
• We checked the matches by hand to make sure they were correct.
• We added extra matches that were missed because of missing information by reviewing with the NOPD data team.
• We used the Gun Violence Archive to help confirm certain cases where we could not match between sources or had conflicting information between health records and NOPD records.

This gave us a more complete list of the people harmed by firearm violence.

Next, we worked with NOPD to connect victim records to police incident records.

• We learned how NOPD organizes its data.
• We tested our matching process using publicly available 911 data.
• NOPD gave us a spreadsheet of their homicide-by-shooting cases.
• We matched our victim records to their incident records based on age, sex, race, and date of death.

While doing this, we found:

• some cases that should have been included were missing; and
• some cases that were included were not actually firearm related.

We shared these case numbers with NOPD and worked with them to confirm which cases should be included and how the NOPD list could be updated. This collaboration allowed us to build a more inclusive dataset for this project and to support the strengthening of NOPD data systems.

Next, we worked with U.S. Census Data to examine measures related to social and structural conditions in each neighborhood, focusing on factors that research suggests are related to rates of firearm violence. We are continuing to build out additional neighborhood measures related to firearm violence as well, with the goal of adding additional maps in a next phase.

The result is a more accurate and complete list of: 

• the people killed by firearm violence; 
• the incidents connected to those deaths; and
• the characteristics of neighborhoods with higher and lower rates of firearm violence.

For the public dashboard seen here: 

• we only shared summary data (not individual records);
• we shared firearm violence rates and basic demographics by neighborhood and year;
• the maps do not show exact addresses or exact shooting locations; and
• if there was only one homicide in a neighborhood in a year, we did not show that person’s demographic details to protect privacy.

This approach helps protect privacy while still sharing useful information about firearm violence in Orleans Parish.