One of the more memorable elections of recent memory is finally over, and once again many of the prognosticators made what turned out to be bad predictions. Take voter registration, for example. Some commenters thought that a surge in urban area registrations would result in a Democratic majority in the state’s House of Representatives. Not only were they wrong about the shift in party support, but there is also reason to believe they were wrong, in part, about the surge in voter registrations.
The Texas secretary of state (SOS) has data on statewide voter registration going back decades. The column chart shows the number of voters registered for the November general election for every presidential election year from 1976 to 2020. The orange columns give the U.S. Census Bureau’s estimated population of the state for that year (except for census years). Population data for 2020 is not available currently.
Although there was an increase of 12.3% in the number of registered voters from 2016 to 2020, a double-digit increase is not exactly unusual. From 1980 to 1984, the number of registered voters increased by 19% as seen in Table 1. Registrations soared from 1992 to 1996 by 24.9% and then increased an additional 17.3% by 2000. Growth in the number of registered voters tapered off after 2000, reaching a low of 0.5% from 2008 to 2012 before hitting double digits again in 2016 with a 10.7% increase.
However, the increase in the number of registered voters was concentrated in certain parts of the state. Map 1 shows the percentage change by county in the number of registered voters from November 2016 to November 2020. Though many counties experienced a decrease, those near the Texas Triangle (encompassing Houston, Dallas-Fort Worth, San Antonio and Austin) had significant growth — the same counties that usually see the greatest numerical population growth. Similar growth occurred in counties in the Valley, in El Paso County and in a group of counties overlying the Permian Basin near the southeast corner of New Mexico.
Table 2 compares the percentage growth in population estimates from 2016 to 2019 (2020 population estimates are not available) to the percentage growth in the number of registered voters from 2016 to 2020. The table is limited to the 30 counties over 50,000 population with the largest increase in the number of registered voters from 2016 to 2020. The table includes two columns for change in population. The first shows the change in total population while the second shows the increase in the voting population — those 18 years and older.
For many of these counties, at least half of the growth in voter registration can be attributed to the increase in the number of people of voting age; exceptions include Nacogdoches County, which decreased in both total population and adult population.
Even though the time periods do not exactly match, the data in Table 2 combined with the fact that many rural counties are losing population as many of their residents relocate to urban areas can explain why many urban and suburban counties had such large increases in the number of registered voters. Rather than an influx of voters from more liberal states, the additional voters often came from conservative rural Texas.
Early voting and voting by mail also made headlines across the state. Map 2
shows the percentage of registered voters who voted early for each county. Early voting in person was so prevalent that it accounted for more than 50% of the total votes cast in 80 counties! Though these were largely the urban and suburban counties of the Texas Triangle, the list includes many counties across the state. The Panhandle’s Wheeler County topped that list with 86.17% of its voters having voted early.
South Texas and far West Texas (excepting Jeff Davis County) had relatively few people voting early in person. Perhaps people were avoiding the early voting lines due to the large number of COVID-19 cases in these counties.
Mail-in ballots became a significant source of controversy during this year’s general election. Out of almost 17 million registered voters, just 958,256 mail-in ballots were cast statewide, mostly in the larger counties.
While the largest number of mail-in ballots were cast in the largest counties, they were more likely to be used by a higher percentage of the registered voters in rural areas. For example, mail-in ballots were used by more than 10% of the registered voters in six counties: Polk, Duval, Jim Hogg, Terrell, Coke and Gillespie.
Map 3 shows the percentage of registered voters who cast mail-in ballots in each county. Unlike the previous map, this one does not show the same focus around high-growth and urban counties. While it is tempting to discuss the line of high percentage counties running from the Panhandle southeast to both the Eagle Ford Shale area and Houston, other than geography there is no obvious connection between those counties.
Then there is the stretch of counties to the west of this band where few people voted by mail. The outliers in the Big Bend area (Jeff Davis, Brewster and Terrell) might be explained by the size of the counties, which might have made voting by mail an attractive alternative to a long drive — except that some of their equally large neighbors (e.g., Hudspeth, Culberson, Presidio) often received relatively few mail-in ballots.
One thing the pundits got right is that population growth is affecting the state. The large population increases in urban areas can explain a significant part of the increased voter registration in those counties. Similarly, the population losses in many rural counties likely resulted in many of them experiencing a drop in voter registrations as residents largely migrated to urban areas. Add in the anemic growth in voter registrations from 2000 to 2012, and the recent surge no longer appears all that surprising. In addition, population change combined with the current pandemic at least partially explains the increase many counties saw in early voting — both in person and by mail.
Given the controversies surrounding this year’s general election, perhaps it is not surprising that 337 bills affecting the Election Code had been filed in the Legislature as of 2 p.m. Dec. 3 — more than a month before the next legislative session starts. Only some of those bills will affect voter registration, early voting or mail-in ballots. Regardless of what bills are passed, county officials will continue to make voting as simple as possible for all eligible voters while following state and federal law.