Racial Profiling in Connecticut?

A first look at the first six months of year 2 data

Story by the Connecticut Data Collaborative
September 21, 2015

B elow are a couple quick visualizations of the statewide results for the new release of data. The two scatterplots explore the post-stop data. In particular, looking at what percent of stops resulted in contraband being found when a consensual search was performed.

The first scatterplot looks at total statewide traffic stops that occurred from October 2013 to October 2014. Each color represents a different race of drivers and the size of the circle is representative of the total number of stops for the given race.

Oct 2013 - Oct 2014

​W​hile less than 1% of the stops for white drivers resulted in a consensual search, contraband was found almost 30% of the time. On the other hand, 2% of black stops resulted in a consensual search and contraband was found approximately 20% of the time. Consensual searches occurred for approximately 2% of Hispanic stops with 19% of the searches resulting in the discovery of contraband.

During the second six month time period, ​again ​less than 1% of white stops resulted in a consensual search with contraband being found ​about ​29% of the time. Hispanic drivers were searched 2.3% of the time with the search finding contraband 24% of the time.

Oct 2014 - Mar 2015

The rate of consensual searches was the same for Asian drivers in both time periods yet the rate in which contraband was found increased 14 percentage points​ in the six months from Oct 2014 to March 2015​.

What these scatterplots do not show you is whether these results are statistically significant when normalized for the number of stops occurring by race. Read the data story for an explanation of the KPT Hit Rate test that was used to examine the Oct 2013-Oct 2014 data for statistically significant differences across race categories.​

Comparing Rates of Warnings and Tickets by Race

From October 2013 to October 2014, Hispanics received a ticket 62% of the time compared to 55% for Black Non-Hispanic drivers and 52% for White Non-Hispanic drivers.

Tickets vs. Warnings by Race, Oct. 2013 - Oct. 2014

During the six months from October 2014 to March 2015, Indian Americans/Native Alaskans had the highest ticket-to-warning ratio, with 60% of stops resulting in a ticket. Hispanics experienced a decrease with 58% of stops resulting in a ticket compared to 51% for Black Non-Hispanic and 47% for White Non-Hispanic drivers.

Tickets vs. Warnings by Race, Oct. 2014 - Mar. 2015

Traffic stops were coded with the following disposition outcomes:

  1. Uniform Arrest Report
  2. Misdemeanor Summons
  3. Infraction Ticket
  4. Written Warning
  5. Verbal Warning
  6. No Disposition

For the purpose of this analysis, we re-coded 2 & 3 (Misdemeanor Summons and Infraction Ticket) to "Ticket" and 4 & 5 (Written Warning and Verbal Warning) to "Warning". Arrests and stops with no disposition were excluded from the totals.