Election Recap 2018: Demographics and Education Levels in Flipped Districts
Date: November 20, 2018
*This post has been updated to include updated vote total numbers as well as to include more districts that have flipped since November 20, 2018. The vote totals and districts include reflect the numbers as reported by the New York Times as of December 14, 2018.
A month after the midterm election, the immigration narrative in the House has become even clearer. Anti-immigrant candidates lost in 2018’s swing districts and that has helped Democrats gain control of the House of Representatives. So far, 43 districts have flipped from Republican to Democratic control, while only three districts flipped from Democratic to Republican. Teasing out precisely why they won these races is a difficult task as there are many factors that went into each result, but there are certain themes that were common to nearly every one of these districts. Chief among them was demographic change in many of these suburban and swing districts. To understand this change, New American Economy analyzed U.S. Census Bureau data from 2013 to 2018 and found that over the past five years, nearly all districts that flipped* are much more educated and much more diverse than they were even just five years ago.
Voters in the districts that flipped are more educated:
- In all districts that flipped, the number of college-educated adults increased sharply. In all but two there were at least 10,000 more college-educated adults in 2018 than there were in 2013.
- In all districts that flipped, the share of the population with at least a bachelor’s degree increased. In the district with the greatest change, California’s 45th district, the college-educated share of the population grew by 8.5 percentage points in just 5 years, or 75,868 people. Even in the district with the smallest change, California’s 21st district, the college-educated share in population increased by 0.1 percentage points, or 2,771 people, more than three times the margin of victory for Democrat T.J. Cox.
- In 23 of the districts that flipped, the number of new, college-educated residents outnumbers the margin of victory in this year’s midterm elections.
Table 1: Top 5 Districts by Change in Share of Population and Overall Number with at Least a Bachelor’s Degree, 2013-2018
Congressional District | Increase in Share of adult population with at least a BA | Increase in Adults with at least a Bachelors’ Degree | Margin of Victory, 2018 |
California 45 | 8.5% | 75,858 | 12,523 |
Illinois 14 | 7.0% | 37,423 | 14,871 |
California 49 | 6.2% | 36,501 | 37,215 |
Georgia 6 | 5.9% | 45,686 | 3,264 |
Iowa 1 | 5.0% | 29,201 | 16,900 |
Source: American Community Survey, 2013-2016, and author’s estimates. In keeping with U.S. Census norms, education attainment looks at the adult population over the age of 24. Margin of victory are those reported by the New York Times as of December 14, 2018.
Voters in the districts that flipped are more diverse:
- In all but four of the districts that flipped, the share of Asian American and Hispanic American eligible voters increased in the last 5 years.
- In 24 of the districts, the number of Asian American and Hispanic American eligible voters increased by more than 10,000 in just 5 years.
- In 23 districts, the number of new Asian American or Hispanic American eligible voters was higher than the margin of victory in 2018.
Many of these new voters are immigrants:
- In all but three of the districts that flipped the number of foreign-born eligible voters grew over the last 5 years.
- In 26 of those districts, the share of the electorate that is foreign-born increased.
- In several key districts that flipped (CA-49, VA-10, TX-32, and NY-11), the number of new foreign-born eligible voters increased by more than 20,000 in the last 5 years.
Table 2: Top 5 Districts by Change in Share of Electorate and Overall Number of Voters, Asian or Hispanic, 2013-2018
Congressional District | Increase in Share of Electorate, Asian or Hispanic | Increase in Asian or Hispanic Eligible Voters | Margin of Victory, 2018 |
California 21 | 7.8% | 59,356 | 862 |
California 25 | 7.5% | 77,699 | 21,396 |
California 39 | 6.2% | 45,521 | 7,610 |
California 49 | 5.6% | 50,437 | 37,215 |
Virginia 10 | 5.5% | 53,379 | 45,515 |
Source: American Community Survey, 2013-2016, and author’s estimates. Margin of victory totals from New York Times, accessed on December 14, 2018.
Table 3: Top 5 Districts by Change in Share of Electorate and Overall Number of Voters, Foreign-Born, 2013-2018
Congressional District | Increase in Share of Electorate, Foreign-Born | Increase in Foreign-Born Eligible Voters | Margin of Victory, 2018 |
California 49 | 5.6% | 32,982 | 37,215 |
Texas 32 | 4.8% | 25,933 | 16,966 |
New York 11 | 4.4.% | 23,005 | 10,793 |
New Jersey 7 | 3.7% | 16,753 | 16,200 |
Virginia 10 | 3.5% | 27,776 | 45,515 |
Source: American Community Survey, 2013-2016, and author’s estimates. Margin of victory totals from New York Times, accessed on December 16, 2018.
At the same time, the share of white voters in these districts declined:
- In all but one of the districts that flipped, the share of the white vote shrank.
- Several districts that flipped—CA-39, NY-19, NY-11, NJ-7, NM-2, and NY-22—had more than 15,000 fewer white eligible voters than they did just 5 years ago.
- In eight districts the decline in the number of white eligible voters alone was larger than the margin of victory in 2018.
Table 4: Top 5 Districts by Decline in Share of Electorate and Overall Number of Voters, Non-Hispanic White, 2013-2018
Congressional District | Increase in Share of Electorate, White | Decrease in White Eligible Voters | Margin of Victory, 2018 |
California 39 | -7.0% | -32,884 | 7,610 |
New York 19 | -0.7% | -24,588 | 7,573 |
New York 11 | -2.1% | -23,177 | 10,793 |
New Jersey 7 | -0.1% | -19,662 | 16,200 |
New Mexico 2 | -1.7% | -17,914 | 3,722 |
Source: American Community Survey, 2013-2016, and author’s estimates. White is defined as all eligible voters that identify as non-Hispanic white. Margin of victory totals from New York Times, accessed on November 16, 2018.
In many of these districts, the Republican candidate ran on an explicitly anti-immigrant platform. And while it’s impossible to know exactly why any voter votes the way they do, it is clear that among these rapidly growing groups – college-educated, Hispanic, and Asian American voters – that message did not resonate. In Virginia’s 10th district, incumbent Rep. Barbara Comstock ran anti-immigrant ads that focused on MS-13 in an area that has added more immigrant residents than any other major county in the United States in recent years, including 13,000 immigrant voters in the last two years alone. Add that to the nearly 38,000 new college-educated voters this district has added since 2013 and it is not surprising Rep. Comstock lost to pro-immigrant Democrat Jennifer Wexton by more than 45,000 votes. Similarly, in Iowa’s 1st district, Republican incumbent Rod Blum ran anti-immigrant ads marked by photos of Hispanic men with tattoos and claims of “criminal illegal aliens on our streets.” That message did not resonate with voters in this area, which has added more than 29,000 college-educated voters in the last five years and 1,400 Hispanic American and Asian American voters in the last two years. Blum lost to Democrat Abby Finkenauer by 16,900 votes.
To learn more about how anti-immigrant ads turned off key voters, check out our post-election analysis here. To see how demographics have changed in 2018’s most competitive districts just in the last two years, see our latest research here.
*This analysis includes the following 35 districts that flipped from Republican to Democratic as of December 14, 2018, and that were not redistricted since 2016:
AZ-2, CA-10, CA-21, CA-25, CA-39, CA-45, CA-48, CA-49, CO-6, GA-6, IL-6, IL-14, IA-1, IA-3, KS-3, ME-2, MI-8, MI-11, MN-2, MN-3, NJ-2, NJ-3, NJ-7, NJ-11, NM-2, NY-11, NY-19, NY-22, OK-5, SC-1, TX-7, TX-32, UT-4, VA-10, WA-8.