Demographic Destiny: Voting Trends in America’s Fastest Growing Cities are Determining our Political Future
An Overview and First Section in a Study Examining the Political Shift in America's Fastest Changing Metropolitan Areas
The United States is driven by transformation, migration, and growth. A people predestined to move. Riding a tide of circumstance, Americans are pushed from one place to the next, into new neighborhoods, beyond state lines, or across the country. Some places grow while others shrink, altering and reinventing these area’s very composition along the way. Occurring during a time of political tumult these demographic shifts have played a significant role in determining the nation's present position and will continue to do so into the future. To better understand how these changes are impacting the American political landscape this study looks to closely detail the voting behavior and trends taking place in the areas experiencing the greatest amount of change currently. To do so, I sample the 20 fastest growing metropolitan statistical areas in the country and chart the voting behavior of each through the last four contemporary presidential election cycles.
What is a Metropolitan Statistical Area?
Cities are the primary economic and social engines within any region. As such, their growth and change impacts well beyond the city limits and deep into the suburban and rural communities beyond. To account for this, this study utilizes the Metropolitan Statistical Area (MSA) as a base unit. As defined by the Census Bureau, “The general concept of a metropolitan or micropolitan statistical area is that of a core area containing a substantial population nucleus, together with adjacent communities having a high degree of economic and social integration with that core.” All voting information collected accounts for the full results of every county included within each MSA. This wide sampling provides a close look into voting trends that extend well beyond urban communities and into the suburban and rural periphery. While there is a somewhat arbitrariness to what is or is not a part of a MSA, the hope is that the more information included the clearer the picture becomes. Likewise, there are clear differences in size and scale of counties between different states. Nevertheless the MSA’s samples here provide a wide sampling of differently sized counties and the trends they exhibit. County to county comparisons in later sections will provide further insight and delineate these differences.
The full name of each MSA will be listed in the first legend (Fig 1) as well as linked to full census listing when first referenced in text. For ease of reading each MSA will be referenced by the principle city throughout. (Ex. Austin-Round Rock-Georgetown, TX -> Austin, TX). Each metro will be color coded by home state which will be used throughout the graphs. Substack does not offer Datawrapper embedding currently. To access fully interactive charts with detail data please click the links below each image. Further citations are listed at the bottom of post.
Where in the world are we talking about?
Taking a glance at the map (Fig. 2 there is a definite trend. All but Denver, Seattle, and Salt Lake City are located in the Sunbelt. This trend is the continuation of multi-generational migrations away from the Northeast and Midwest starting after World-War II which is now on-going in southern cities. Starting in the southeast, this group includes Atlanta, Charlotte, and Raleigh which makes up the projected Southeastern Megalopolis. Nearby are Nashville and coastal Charleston. Florida, which has spent the better half of the last decade as the 3rd most populated state, is host to four of these metros (Jacksonville, Northport, Orlando, and Tampa). Not to be outdone Texas hosts five of it’s own metros (Austin, Dallas, Houston, McAllen, and San Antonio) with Oklahoma City nearby. Moving further into the desert Phoenix and Las Vegas round out the list both bordering the most populated state in California.
Cumulatively these 20 Metros have grown by nearly 9 million over the last 10 years, and now are home to over 58 million Americans in total. Likewise these metros make up significant portions of the states themselves. Las Vegas constitutes 73.6% of Nevada. The five Texan metros, which are home to over 20 million, make up 70% of the 2nd most populated state. Similarly, Phoenix is home to the 68% of all Arizonans. Atlanta, the largest of these metros outside of Texas, makes up 56% of all Georgians. Denver and Seattle both account for over half of their state populations. The Floridian metros, Salt lake City and Oklahoma City all represent at-least a third of their states, with only Nashville (28.3%) and Charleston (15.6%) short of that mark in their home states.
The size of these metropolitan areas is made ever more important by the electoral significance of the states they inhabit. The growth they have experienced has only increased their importance, with Arizona, Colorado, North Carolina, Florida, and Texas all set to pick up congressional seats and electoral votes after the 2020 census redistricting. The change they experienced has already had an impact. Arizona and Georgia flipping from red to blue in the 2020 presidential election had much to do with the change in Phoenix and Atlanta. Similarly, North Carolina, a perennial red state which only had gone blue by narrow margins in 2008, but in 2020 the candidates were separated by less than 75,000 votes. Nevada, which is solidly blue in recent history, has no less received swing state treatment on the campaign trail. Florida, the battleground state of battleground states, will receive two additional congressional seats after the census becoming the 3rd state in electoral vote total. Texas, the 2nd most electorally significant state and projected to receive three congressional. Heralded as the future ace in the hole for the DNC, the Lonestar State is set to become the most decisive state for the future of American elections.
The Voting Metrics
There are three central metrics measured here to track the voting trends of these different regions. First and foremost, voting margin between the candidates and the change in this figure between elections indicates the shifting party preference of the voting population in each metro. Secondly, voter registration growth indicates how many additional voters sign up to vote between elections. On paper this should best match population growth, but due to varying degrees of voter participation from state to state and county to county voter registration varies significantly. Third, and the other half of voter registration, voter participation. There are multiple ways to measure voter turnout. What is used in this study is participation of registered voters. This shows what portion of those who register to vote actually end up casting a ballot on election day.
Each section to follow will include a summary of what is plotted in the graphs below as well as my own interpretation of the trends shown afterwards. I would encourage readers to explore each graph, and develop their own interpretation of what is shown. Voter statistics are more complicated than what many may think, and there are certainly many factors that influence direction. That being said there are certainly some undeniable directionality in the general trend.
Voting Margin as Percent of Voting Population
Voting margin by percentage indicates the difference in results between candidates of the two primary political parties. The positive y-axis indicates results in favor of the Democratic candidate with the negative y-axis indicating results in favor of republican candidates. (Fig. 3) plots each metro results accordingly. To get a better sense of general trends (Fig. 4) plots the election result margin relative to 2008 results. The positive y-axis on this chart indicates a shift in margin toward Democratic candidates and negative y-axis toward republican candidates, but does not indicate the definite result for the candidates in each election.
Metros that Followed the Trend
If we now consider specifically the metros that had voted more in favor of the democratic candidate in ‘16 relative to ‘08 and continued along that trend into ‘20 this trend becomes more apparent. 8 of 20 metros are represented in this group which included Atlanta, Austin, Dallas, Houston, Oklahoma City, Phoenix, Raleigh, and San Antonio. This group includes the four most populated metros (Dallas, Houston, Atlanta, Phoenix) and four of the 5 metros within Texas. On average these metros had shifted in favor of the republican candidate (-5.70%) between ‘08 and ‘12. However, by ‘12 the boomeranged back over 11% in favor of the Democratic candidate (+5.35). In ‘20 this trend continued moving further in favor of the Democratic nominee (+10.7). In ‘08 only 3 of these metros had voted in favor of the Democratic candidate, but by 2020 all but Oklahoma City had. Orlando is not within the group but very well could be as it had shifted to in both ‘16 to ‘20 relative to ‘08 despite moving slightly in favor of the republican candidate from ‘16 to ‘20.
Nearly following the Trend
Another 6 sampled metro areas had shifted further in favor of the Democratic candidates by ‘20 relative in ‘08 despite shifting further in favor of republican candidates in both ‘12 and ‘16. This includes Seattle and Denver of which both represent two of the most liberal metros both within liberal states. On the other hand Nashville and Jacksonville represent two of the most conservative metros in the sample. The former a part of a solidly red state, and a later within Florida which had shifted further right from ‘16 to ‘20. Florida exhibits many state wide trends distinct from the rest of the sun belt which will be examined later in a statewide analysis. The remaining two metro’s Charlotte and Charleston have made two of the least dramatic swings of all metro but still follow the general trend.
The Right Swinging Metros
Las Vegas, Tampa, and Northport have bucked the general trend completely. The two neighboring Floridian metros have remained conservative through the last two elections and played a major part in keeping the state red. Alternatively, Las Vegas started in ‘08 as one of the most liberal metros only and has only become increasingly competitive since then.
McAllen is Exceptional
The metro that rounds off the list as the 20th fastest growing metro stands distinct from both it’s Texan brothers as well as the group as general. The city sitting on the Mexican border began as the most liberal city among the whole group, but as the largest city in the Rio Grande Valley represents the furthest right shifting region in the country in 2020. McAllen will be a major outlier throughout the sample.
Voting Margin by Votes
While comparing the voting margin in percentage gives insight into the relative change each metro is undergoing, the weight each metro population is lost. While many of these areas represent significant sections of their home states some metros represent much larger communities comparatively. To get a better sense of scale (Fig. 5) and (Fig. 6) plots voting margin by actual vote count. (Fig. 5) correlates to (Fig. 3), now showing voting margin by vote count with the positive y-axis in favor of the democratic candidate and negative y-axis in favor of the republican candidate. (Fig. 6) correlates with (Fig. 4), now showing change in voting margin by vote count with the same positive and negative axis assignments.
Detailing Growth
As areas specifically undergoing major growth, plotting change in vote count margin shows that growth that otherwise doesn’t appear in percent margin. Put another way Percent margin is always shown as part of the total voting population and therefore doesn’t show the increase in registered voters or change in voter turnout that makes up that total voting population. As an example the margin between candidates in Atlanta between ‘16 and ‘20 had grown from 5.34% to 14.11% or by a factor of 2.64. In the same period the vote margin had increased from nearly 108,000 to nearly 360,000 or by a factor of 3.33. This is important as these metropolitan area’s growth outpaces the states around them. As they grow they become an ever more substantial portion of the whole state population. Of course the general trend still matches the previous section, but simply gives a separate perspective. This is given additional context when considering the next section, voter registration.
Voter Registration Growth
On paper, voter registration should most closely follow population growth. However, as demonstrated in the graphs and charts below this is certainly not the case. (Fig. 7) compares voter population growth as shown in (Fig. 1) to voter registration growth. It is important to note that the census period used to measure growth (2010-2019) is slightly different from registered voter growth (2012-2020). With one fewer year included in registered voter growth this chart displays a conservative view, and clearly indicates which metros had registered voter growth that outpaced population growth. (Fig. 8) charts voter registration percentage growth between elections. (Fig. 9) depicts voter registration growth between elections in total voters.
Some Metros Register More Voters and it Matters
Three metros, Phoenix, Atlanta, and Salt Lake City, had voter registration growth that outpaced population growth by over 25%. All three were among the top 5 greatest left swinging metropolitan areas from ‘08 to ‘20. Phoenix and Atlanta were both in the top 3 in registered voter count, and both represent over 50% of their state populations. Subsequently, Arizona and Georgia both had swung blue in 2020.
The five Texan metro’s voter registration growth had all outpaced population growth by a noticeable margins relative to other metros: McAllen by 16.4%, San Antonio by 11.74%, Dallas by 9.60%, Austin by 9.37% and, Houston by 7.20%. The latter four were among the top nine most left swinging metros filling out that group with the previous section. The exception, McAllen, was the most rightward swing of all metros. These metros represent well over 60% of the state’s population. The Texan presidential election in 2020 was the closest in contemporary history.
The remaining metros located within swing states tended to have voter registration growth that outpaced population growth. That would include the four Floridian metros as well as Las Vegas. North Carolina stands as the one exception among swing states with both Raleigh and Charleston showing similar population and voter registration growth.
Denver, Oklahoma City, and Nashville all have relatively close population and voter registration growth. Colorado, Oklahoma, and Tennessee are all relatively non-competitive compared to much of the rest of the field.
Seattle and Charleston seem to buck the trends shown among the group. Seattle, in relatively non-competitive Washington, voter registration noticeably outpace population growth. Seattle did have the second most dramatic swing to the right next to McAllen. Charleston, among the smallest metro, is in a relatively non-competitive South Carolina but has voter registration that outpaced its population growth. It is also worth noting that Charleston’s voter registration numbers have risen at a near linear rate over the past four elections as shown in (Fig. 8).
Swing State Amplification
It would seem that the relative competitiveness of a state in an election, and the additional campaigning, advertisements, and door to door action that goes along with it, brings about additional voter registration that outpaces the growth expected. This can be clearly seen in states that are of more recent interest particularly Arizona, Georgia, and Texas. When this is the case the ongoing trends in party preference shown within those metros are amplified as the newly registered voter becomes a more prominent part of the voting population.
Voter Participation
The final trend examined, which goes hand and hand with voter registration, is voter participation. This is measured by registered voters who voted in the charts below. Voter participation or voter turnout is often measured as the portion of votes cast by the voting age population (VAP) or voting eligible population (VEP). This metric shows what percentage of those who could vote do so. The difference from metro to metro in voting registration growth indicate separate populations vary in their tendency to register to vote. The census currently produces national and state level voter eligible population estimates during election, but do not do so for metropolitan, county, or municipality scale. Determining those figures would be helpful in detailing what portion of the metro population are available to register, and would be a key area for continued research.
Nonetheless, the voter participation rates measured here and shown below give good detail in the political participation of the metro as well as state populations. (Fig. 10) plots registered voter participation rate of each population over the past four elections as well as the average of all metros. The tables in (Fig. 11) compares the registered voter participation rate of each metro next to that of the state. (Fig. 12) Shows the difference between the metro and state registered voter participation along with the metro average and average difference across elections.
The General Trend
Election-to-election participation rate seems to largely stay consistent across each metro. The then historic ‘08 election had relatively high participation across the board. Participation in the subsequent ‘12 and ‘16 elections waned slightly comparatively. The now historic ‘20 election, which included the greatest measures to ease voting methods, had predictably high voter participation despite the relatively high rate of growth in voter registration. This can be shown in the metro averages with ‘08 at 69.7%, ‘12 at 68%, ‘16 at 68.8%, and ‘20 at 74%.
Turning attention to metro-to-state comparisons, it would seem in general, that the registered voter participation rate of each metro largely follows the voter registered voter participation rate of their states. The average difference between the metro registered participation rate and state registered participation rate across all elections was 0.7%. This result is expected as many of these metros represent significant portions of the state’s total voting population. Only 6 metros had registered voting participation rates that deviated at least 3% above states average. Austin, Raleigh, Northport, Nashville, Atlanta, and Salt Lake City. Only Nashville (+5.3%) and Salt Lake City (+7.5%) were significantly above state average. Similarly, only 2 metros were significantly below state registered voter participation rates: Charleston (-3.7%) and, the largest outlier, McAllen (-12.9%).
Who Votes and Who Doesn’t
As metros tend to follow their state registered voter participation rates the graph plotted in Fig. 9 outlines the relative level of participation between one another. To give some order to this comparison I will follow the order of the voter registration section.
The three areas with the highest voter registration growth Phoenix, Atlanta, and Salt Lake City differ in voter participation. Salt Lake City sits well above the field with the highest registered voter participation rate, however Utah as a state does have some of the lowest voter turnout in the nation. Utahans are unlikely to register to vote, but those who do are very likely to vote. Phoenix sits at the top of the next block of states with a stable and high voter participation rate. Atlanta has one of the more interesting lines of the bunch. Between ‘12 and ‘16 the metro registered a staggering number of voters as shown in (Fig.) 8, but during that period of time voter participation plummeted. While it seems a large number registered to vote fewer actually went to the polls. They continued to register a large number of new voters during ‘20, and while voter participation did not return to where it once was the rate did rise during significantly. This is once again in the election where the state flipped from red to blue.
Texan metros had relatively low registered voter participation which remains true state wide. Filling out the bottom of all metros throughout the elections, it was only in the the ‘20 election that Texans metro came close to the rest of the field. McAllen once again sits as a clear outlier to the rest of the field with incredibly low voter participation. With that said, McAllen had its highest voter participation in the election where it experienced the largest voter swing of any place in the country.
The remaining metros in swing states had fairly moderate changes in voter participation that follow fairly closely to state and national trends. Florida, a state with extensive early vote and vote by mail for close to two decades, showed very low change in participation across the four metros. North Carolina, which had the steadiest voter registration to population growth, also had fairly flat registered voter participation. Las Vegas is a different story. It saw noticeably higher voter participation in elections in which Trump was a candidate. Both of which were elections in where the metro moved relatively right in comparison to the general trend.
To finish out the field with metros in relatively non-competitive states that experience relatively low voter registration growth, Nashville, Oklahoma City, and Denver, all have relatively consistent voter participation. Charleston has moderate voter participation that does not seem to follow any trend with voter registration growth. Seattle has notable high and and consistent voter participation.
What’s To Come
This information so far has delivered a strong overview of the general trends shown in each metro. This being said, there are additional trends within the communities that make up these metros are not necessarily shown from this broad view. Subsequent parts will give a closer look into these trends to give a more thorough review of these quickly changing and politically important part of the world. This will include:
Where We’re Divided: The Growing Polarity between Urban, Suburban and Rural America
In depth review of metro change in key states like Georgia, Florida, and Texas
A concluding forecast that gives a prediction of the political implications these metro changes will have on the country
For this and more please consider subscribing and sharing. Thank You!
Citations
Luckily all the information used here is public, but unfortunately collecting that information takes quite some time. Metro population figures all come from the U.S. Census Bureau. The site citypopulation.de was a useful tool in determining the counties in each metro area. State and county level election results can be found at Ballotpedia. Election to election registered voter figures and vote total can be found on a state by state basis through secretary of state or state election board databases. I have collected all of these figures and have put them all together on one massive spreadsheet. I have the full intention to release this spreadsheet after the full project is completed. However, as it currently has important figures for subsequent parts not shown in this post I will not be sharing it in full yet. For any interested in seeing parts of this spreadsheet earlier feel free to email me @ryan.givens2@gmail.com