Saturday, February 18, 2017

2016 Hate Group Concentration Predicts Trump % of the Vote

This week the Southern Poverty Law Center came out with it's hate map for 2016.  They found an increase in the total number of hate groups in the US from 892 in 2015 to 917 in 2016.  For the last three years I have been adjusting the hate group numbers for the population in the US and in each state as the number of groups per million residents.  The hate group rate did increase from 2.78 to 2.84 groups per million residents in 2016.  This increase is being driven by primarily by increases in anti muslim hate groups.  You can see the numbers from 2016 and 2015 in the table at the bottom of the post.  The number of hate groups is still below the all time high of 1,018 groups in 2011 which gives a hate group rate of 3.27 groups per million. 

The District of Columbia has the highest rate of hate groups with 30.83 groups per million residents and hate crimes at 96.69 incidents per million residents.  DC also had the largest increase in the hate group rate with an increase of 4.05 groups per million residents.  The Midwestern and Mountain states of Montana, Idaho, and South Dakota have the next highest rates and rate increases.  

The states with the largest decreases in the hate group rate were Vermont (-3.19 g/million), Oklahoma (-2.82 g/million), West Virginia (-2.15 g/million), South Carolina (-2.07 g/million), and Arkansas (-2.03 g/million).  These states were ones with some of the highest rates of hate groups in 2015.  The next step is to look at how are these rates associated with Donald Trump's % of the vote in 2016.



Regression models were run with Trump's % of the vote as the dependent variable and the 2016 hate group rate and the hate crime incidents in 2015 (the most recent year available. as the dependent variables.  The hate group concentration was a significant predictor but hate crime incidents were not.  The regression equation is given by:

Trump % = 42.84 + 2.33*(Hate Group Concentration)

This means that for those states with zero hate groups (Alaska and Hawaii), Trump would expect to have 42.8% of the vote.  For every increase in the hate group concentration by one group per million residents for a state, Trump would expect an additional 2.33% of the vote.  The graph above shows the fit of the regression line created by this equation.  The regression model accounts for 20.1% of the variability in Trump's vote.  

Washington, DC was not included in the model because it is an extremely high outlier in hate groups (30.83 groups/million) and hate crime incidents (96.69 incidents/million) and an extremely low outlier in Trump's % of the vote (4.1%).  DC also has the highest population density of any state/territory in the electoral college.  For these reasons plus the fact that DC is technically not a state, it was excluded from the model.  The District would make an interesting area for social science research.

**Update**

I have received comments on Facebook that I implied that the concentration of hate groups in each state caused voters in that state to vote for him.  I will state that prediction does not mean causationA significant correlation or regression means that an increase in the one or more predictor variables considered coincides with a change in the outcome variable.  There is always a potential third variable that was not considered that can explain the change in both variables.  These include poverty and education.  I will consider these variables in my next post.


State Name
Hate groups 2016
Pop
2016
Hate groups per million '16
Hate Crime
Incidents
Per million
'15
Hate
Groups per
Million '15
Trump
%
Change
in rate
US
917
323,127,513
2.84
18.20
2.78
46.7
0.06
District of Columbia
21
681,170
30.83
96.69
26.78
4.1
4.05
Montana
10
1,042,520
9.59
43.56
5.81
56.5
3.78
South Dakota
7
865,454
8.09
18.64
5.82
61.5
2.26
Idaho
12
1,683,140
7.13
20.54
4.83
59.2
2.30
Mississippi
18
2,988,726
6.02
0.00
6.35
58.3
-0.33
Tennessee
38
6,651,194
5.71
33.48
6.21
61.1
-0.50
Alabama
27
4,863,300
5.55
2.06
4.53
62.9
1.02
Arkansas
16
2,988,248
5.35
1.68
7.39
60.4
-2.03
Kentucky
23
4,436,974
5.18
42.48
2.94
62.5
2.25
Virginia
39
8,411,808
4.64
18.85
3.82
45
0.82
New Hampshire
6
1,334,795
4.50
9.77
4.51
47.2
-0.01
Delaware
4
952,065
4.20
11.63
5.29
41.9
-1.08
Missouri
24
6,093,000
3.94
16.44
3.62
57.1
0.32
Indiana
26
6,633,053
3.92
9.52
2.42
57.2
1.50
Wyoming
2
585,501
3.42
3.41
3.41
70.1
0.00
Pennsylvania
40
12,784,227
3.13
5.00
3.12
48.8
0.00
Georgia
32
10,310,371
3.10
4.31
3.82
51.3
-0.71
Florida
63
20,612,439
3.06
3.55
2.86
49.1
0.20
North Carolina
31
10,146,788
3.06
16.03
2.89
50.5
0.17
Ohio
35
11,614,373
3.01
35.82
2.93
52.1
0.09
Maryland
18
6,016,447
2.99
6.83
2.50
35.3
0.49
Louisiana
14
4,681,666
2.99
8.14
4.50
58.1
-1.51
Colorado
16
5,540,545
2.89
19.61
2.93
44.4
-0.04
Washington
21
7,288,000
2.88
38.35
1.12
38.2
1.77
Michigan
28
9,928,300
2.82
31.14
1.91
47.6
0.91
Oregon
11
4,093,465
2.69
16.13
2.23
41.1
0.45
Nebraska
5
1,907,116
2.62
10.02
2.64
60.3
-0.02
Arizona
18
6,931,071
2.60
40.42
2.64
49.5
-0.04
Illinois
32
12,801,539
2.50
7.00
1.79
39.4
0.71
South Carolina
12
4,961,119
2.42
11.23
4.49
54.9
-2.07
Kansas
7
2,907,289
2.41
21.29
1.72
57.2
0.69
New York
47
19,745,289
2.38
25.26
2.22
37.5
0.16
Maine
3
1,331,479
2.25
28.59
1.50
45.2
0.75
West Virginia
4
1,831,102
2.18
22.23
4.34
68.7
-2.15
California
79
39,250,017
2.01
21.38
1.74
32.7
0.28
Texas
55
27,862,596
1.97
6.95
3.06
52.6
-1.08
Minnesota
10
5,519,952
1.81
19.86
1.09
45.4
0.72
Massachusetts
12
6,811,779
1.76
60.49
1.47
33.5
0.29
New Jersey
15
8,944,469
1.68
36.84
2.34
41.8
-0.67
Vermont
1
624,594
1.60
12.78
4.79
32.6
-3.19
Wisconsin
9
5,778,708
1.56
7.45
1.56
47.9
0.00
Oklahoma
6
3,923,561
1.53
9.46
4.35
65.3
-2.82
Connecticut
5
3,576,452
1.40
25.90
0.56
41.2
0.84
Nevada
4
2,940,058
1.36
20.06
1.38
45.5
-0.02
North Dakota
1
757,952
1.32
47.56
2.64
64.1
-1.32
Iowa
4
3,134,693
1.28
1.92
1.60
51.8
-0.32
Utah
3
3,051,217
0.98
15.69
1.00
45.9
-0.02
New Mexico
2
2,081,015
0.96
6.23
1.44
40
-0.48
Rhode Island
1
1,056,426
0.95
17.04
2.84
39.8
-1.89
Alaska
0
741,894
0
10.83
0.00
52.9
0.00
Hawaii
0
1,428,557
0
0.00
0.00
30
0.00
**Related Posts**

Concentration of Hate Groups Predict Hate Crimes (if you consider DC) and Trump Vote (if you don't)

 

More Hate Groups in States Where Trump and Clinton Win (and in DC Where He Lost)


SPLC Hate Group Update: Washington, DC has an Increase in Activity


 

A Wave of Hate Groups in California? No in Washington, DC