Racial Demographics of Callers to CT Safe Connect

Introduction

Clare Dignan who is an investigative journalist claims that “Race [and] wealth dictates which stories of domestic violence get told” (Dignan 2021). The social situation is such that only a small percentage of cases of intimate partner violence receive attention, and many more go virtually unnoticed. If a person of color was facing this type of problem, people would not be concerned. So in this project, we are going to ask: how do the racial demographics of individual callers vary in regions compared to expected demographics from the census? The Connecticut Coalition Against Domestic Violence (CCADV) is Connecticut's leader of victims of domestic violence and those organizations that serve them. Ct safe connect is a program of the CCADV that aims to make it easier for victims of domestic violence to access information, resources, and assistance. The program's goal is to remove barriers so individuals and families can contact Safe Connections from anywhere and receive immediate help for their needs. They communicate in a variety of ways and have multiple options to offer the narrator. The information should be helpful to CCADV identify the specific races’ reports in different regions, then they can help in a more targeted way, for example, if a particular race receives a large number of calls in a region, this can help CCADV to raise awareness and educate the public, thereby improving attitudes toward that race in that area. In addition to that, through the data, we can see that although we all know that Asians are not a large population in the region, they are only reported very rarely, some even in single digits, so why is this the case? Is it because Asians don't like to seek help, or they don't suffer from such problems as violence?

Explore data

Before sharing data findings with readers, I had to mention that the first instance call mainly refers to the first time the client called or was called; the first time that the client was registered in the CT Safe Connect data. No matter the client has called multiple times throughout the year but we only consider the first call registered. Overviewing the total calls in 15 regions, the grand total showed that the Latino population (3906 calls) and white population (4033 calls) had similar calls with only 127 calls different. Asian (215 calls) and multiple (510 calls) populations report much less than the other groups.

Then I reached the Connecticut population in races from Census 2020, however, in the census, there were groups like American Indian and Alaska Native/Native Hawaiian were not included in CT Safe Connect data, so the total population is the number without those numbers, same as Latino and the other-unknown group in CT Safe Connect data.

As shown in Table 1, the black population is the only race with a positive percentage of the difference between actual caller demographic versus expected census demographics, which means the percentage points of CT SafeConnect are higher than the expected census demographics. For example, looking at the black population in the Ansonia region, the CT Safe Connect is 22.73% and the census is 6.44%. Then 22.73 - 6.44 = 16.29 (positive), and this means that CT Safe Connect is 16.29% points higher than expected. On the other way, the percentage with a negative number means percentage points of CT SafeConnect are lower than the percentage points of the census. Using the white population in the Ansonia region, CT Safe Connect is 72.73% and the census is 81.24%. Then 72.73 - 81.24 = -8.56 (negative), and this means that CT Safe Connect is -8.56% points lower than expected.

Region	Black callers versus expected Census population	White callers versus expected Census population	Multiple race callers versus expected Census population	Asian callers versus expected Census population
Ansonia	16	-9	-6	-2
Bridgeport	28	-24	-3	-1
Danbury	2	7	-8	0
Enfield	11	-9	-2	0
Hartford	26	-22	0	-5
Killingly	3	2	-4	-1
Mansfield	2	5	-4	-3
Meriden	20	-15	-2	-3
Middletown	14	-10	-3	-1
New Britain	16	-14	-2	-1
New Haven	29	-23	-3	-3
New London	9	-3	-3	-2
Stamford	27	-18	-7	-2
Torrington	4	-2	-2	0
Waterbury	24	-21	-1	-2

After this, I made four maps for different races to see the races’ first instance call in different regions. The results were quite surprising. As shown in Map 1, Black population is the only group that all the actual caller demographics are higher than the expected census demographics. Black residents accounted for 6.44 percent of all calls while representing only 22.73 percent of the local population in Ansonia, a difference of 16.29 percentage points. In comparison, CT Safe Connect received more calls than the expected census demographics. The regions like Hartford, New Britain, Stamford and Mansfield have the largest different points. They are in the middle of CT, and the most Southwest of CT.

Map 1.

Accounting to the Map 2, In white population, most of the region's actual caller demographics is lower than the expected census demographics. The range between the largest and smallest different points is 30. White residents comprised 84.75 percent of all calls while representing only 78 percent of the local population in Danbury, a difference of 6.75 percentage points. In comparison, CT Safe Connect received more calls than the expected census demographics with 6.75 percentage points. The regions like Danbury (6.75%) and Mansfield (5.07%) have the largest different points. They are in the middlesouth of CT.

Map 2.

Map 3 demonstrated that in the group of people of multiple races, most regions' actual caller percentage of difference is lower than the expected census percentage of difference. There was only one region with a positive number of differences. Multiple residents made up 9.58 percent of all calls while representing only 9.46 percent of the local population in Hartford region, a difference of 0.11 percentage points. In comparison, CT Safe Connect received more calls than the expected census demographics with 0.11 percentage points. The region like Hartford (0.11%) has the largest different points. It is in the middle of CT.

Map 3.

As shown in Map 4, Asian group is the only group with all negative percentage points, which means the actual caller demographics is lower than the expected census demographics. Asian residents added up to 1.61 percent of all calls while representing only 1.07 percent of the local population in Hartford region, a difference of -0.09 percentage points. In comparison, CT Safe Connect received less calls than the expected census demographics with 0.09 percentage points. The largest percentage difference can be seen in Asian groups since there is a negative number. The region like Torrington (-0.09%) has the largest different points. It is in the northwest of CT near New York State.

Map 4.

When we take the races as a whole, the chart shows the highest numbers of different percentages of regions. The Hartford and Danbury region had 2 times the highest report out of 4 in total. Bridgeport, Stamford, New Haven regions appeared one time with rational reason, because they have one of the largest total populations in CT. However, the Enfield and Torrington region had the smallest total population, it is shown on the table one time.

This chart shows when taking the races as a whole the highest numbers of different percentages of regions.

This is a table shows the total population and rank of the population from low to high in 15 regions.

Region	Total population (low to high)	Rank
Mansfield	76001	1
Killingly	81411	2
Enfield	99138	3
Torrington	104002	4
Ansonia	116655	5
Meriden	120549	6
Middletown	159614	7
Waterbury	206295	8
Danbury	209895	9
New Britain	215088	10
New London	252602	11
Bridgeport	279330	12
Stamford	346972	13
New Haven	413101	14
Hartford	608509	15

Uncertainty and Caution

There are some uncertainty or cautions about the data which the readers better know to avoid misinterpreting the results. The most important thing is CT Safe Connect used different categories for races than census 2020. This practice affects the accuracy of the percentages to a certain extent because it rounds up. For example, CT Safe Connect had Latino which the census does not have, and Census had American Indian and Alaska Native/Native Hawaiian and Other Pacific Islander which CT Safe Connect does not have. Although I avoid (divided) those races in the total number of this calculation, there should be some difference in each category like I use African American or Black which census had to compare with the black population which CT Safe Connect had. Also, the first instance calls did not mean the total number of reports about violence, there are always some people who do not like to report this and ask for help. Some people self-identified as either “white”, “multiple”, and this is also an uncertainty point in CT Safe Connect data. In the percentage of the calculation, I used two decimal points instead of lots of decimal points.

Credits

I would like to thank Jack and Myri who helped me to create this data story. Also, thank you for CCADV provided data.

Clare Dignan, “Race, wealth still dictate which stories of domestic violence get told, experts say,” Hartford Courant, Dec 8, 2021