Richard Florida is a co-founder and editor at large of CityLab and a senior editor at The Atlantic. He is a university professor in the University of Toronto’s School of Cities and Rotman School of Management, and a distinguished fellow at New York University’s Schack Institute of Real Estate and visiting fellow at Florida International University.
States with more working-class voters are solidly red; those with a dominant creative class are solidly blue; service-class heavy states aren’t easily defined.
We typically divide the electoral map into red and blue states, and class is a feature, if not the key feature, in that divide. What we are witnessing is nothing less than a great inversion of America’s political geography. Dating back to FDR and the New Deal, the blue-collar working class once provided the backbone of the Democratic electorate, but today, states with larger working-class populations have swung solidly into the Republican camp. And, blue states have become those where the knowledge, professionals, and cultural workers that make up the creative class predominate.
That’s the key takeaway from an analysis of the connection between class and American politics I conducted with Patrick Adler and Charlotta Mellander. Our analysis looked at the role of class (defined as the kinds of work people do) and voting in the last three presidential elections.
We looked at the correlations between the share of workers that make up the three major classes—the blue-collar working class, the knowledge-based creative class, and the even larger service class—and state voting patterns. The table below details the top states (including Washington, D.C.) with the largest share of the three major classes.
Top Five States With the Highest Share of Workers From Each Class
Then, we drilled down further examining the correlations between state-by-state voting patterns and the 22 major occupational groups, and the more than 800 individual occupations that make up these classes. As usual, I note that correlation does not mean causation, but only points to associations between variables. Still, the patterns we document suggest the powerful role of class in defining America’s political geography.
States with larger working-class populations positively correlated with voting for Republican presidents. This correlation saw a sizable jump in 2012 and increased again in 2016, but more modestly. This is the culmination of a long-running shift, first identified in the 1970s, by Republican strategist Kevin Phillip’s identification of the so-called “silent majority” of socially conservative blue-collar voters.
How Occupational Class Has Correlated With Vote in Presidential Elections
|State-Level Occupational Class||Obama 2008||Obama 2012||Clinton 2016||McCain 2008||Romney 2012||Trump 2016|
Creative classes correlate with voting for Democratic presidents across the three most recent presidential elections. Again, we see a big jump in 2012 and a smaller one in 2016. These correlations are similar in strength to other markers of class, like education and income. More affluent states with greater shares of college graduates skew blue, while less economically advantaged states with less-educated populations trend red. This blue, creative-class pattern is in line with John Judis and Ruy Texiera’s idea about the increasingly liberal orientation of “ideopolis” cities in which knowledge workers cluster.
But the pattern for service-class locations is more mixed. The service class is the largest class by far, composed of more than 70 million members—more than 45 percent of the workforce—whose members toil in low-wage, precarious work in retail shops, office work, and food service. States with greater shares of service-class workers lean slightly Democratic, but not nearly to the degree creative-class heavy states fall into the Democratic camp or working-class heavy states line up for the Republicans.
In the 2016 election, for example, the service class of the workforce was much more modestly correlated with Clinton support and more modestly negatively correlated with Trump support. Part of the reason is that service-class jobs are more spread out across the nation, and part of it is that service-class jobs tend to cluster alongside professional and knowledge-based jobs in larger cities and metro areas.
How Occupational Composition Correlated With Political Vote in the 2016 Presidential Election
|Business & finance||.74||-.75|
|Arts, design, entertainment, & media||.68||-.73|
|Computers & math||.65||-.67|
|Life, physical, & social science||.45||-.54|
|Community & social services||.27||-.38|
|Architecture & engineering||.16||-.26|
|Education, training, & library||.14||-.20|
|Building & grounds cleaning||.05||-.09|
|Personal care & service||.05||-.11|
|Food preparation & service||-.03||-.07|
|Farming, fishing, & forestry||-.11||.06|
|Office & administrative support||-.22||.23|
|Construction & extraction||-.57||.45|
|Transportation & material moving||-.64||.71|
|Installation, maintenance, & repair||-.84||.83|
See how business and finance jobs correlated with the Clinton vote, compared to how installation, maintenance, and repair jobs correlated with the Trump vote.
America’s class-based political geography comes into sharper view when we look at the 22 major occupational groups that make up these classes. States with large shares of working-class occupations like installation; maintenance and repair; and construction and extraction are solidly red. Interestingly, the occupations which are most closely connected to blue states are among the very highest paying professions, such as business and finance, followed by arts, design, entertainment and media; and computers and math occupations.
Locations with larger shares of working-class occupations again show up solidly in the Republican party. But now a couple of interesting cross-class patterns become apparent. Two service-class occupational geographies line up more modestly in the Democratic column. States with larger shares of higher wage, more unionized occupations like protective services and community and social service occupations, and states with lower-wage, less-unionized jobs like healthcare support and personal care and service occupations, both trend blue. And there is also one creative-class geography that lines up red: doctors and healthcare practitioners. This may reflect their opposition to Obama’s healthcare reforms.
Our class-based political geography becomes even more interesting when we zero in on the more than 800 specific occupations that comprise the U.S. economy. The red state pattern is relatively straightforward. Support for Trump is highly correlated with the state-wide share of blue-collar working class occupations like welders, tractor trailer drivers, bus and truck mechanics, and so on. But Trump support is also correlated with larger shares of service class occupations like cafeteria cooks, parts salespeople, and tellers. Only a few creative-class occupations, such as radiologic technologists and occupational health and safety technicians, correlate with Trump support.
Occupations Most Correlated With Trump Votes
|Welders, Cutters, Solderers, and Brazers||Working||.74|
|Heavy and Tractor-Trailer Truck Drivers||Working||.71|
|Cooks, Institution and Cafeteria||Service||.68|
|Industrial Machinery Mechanics||Working||.66|
|Welding, Soldering, and Brazing Machine Setters, Operators, and Tenders||Working||.64|
|Tire Repairers and Changers||Working||.64|
|Bus and Truck Mechanics and Diesel Engine Specialists||Working||.63|
|Water and Wastewater Treatment Plant and System Operators||Working||.60|
|Electrical Power-Line Installers and Repairers||Working||.59|
|Morticians, Undertakers, and Funeral Directors||Service||.59|
|First-Line Supervisors of Mechanics, Installers, and Repairers||Working||.57|
|Chemical Equipment Operators and Tenders||Working||.56|
|Maintenance Workers, Machinery||Working||.55|
|Electric Motor, Power Tool, and Related Repairers||Working||.55|
|Occupational Health and Safety Technicians||Creative||.54|
|Meter Readers, Utilities||Service||.53|
The pattern for blue states is a bit more mixed. Clinton support was highly correlated with creative class occupations like medical scientists, market researchers, lawyers, and computer and information system managers. But Clinton support also correlated with some service-class occupations like manicurists/pedicurists and preschool teachers. The only working class occupations to be correlated with the Clinton vote are bus drivers and other transit workers.
Occupations Most Correlated with Clinton Votes
|Manicurists and Pedicurists||Service||.71|
|Medical Scientists, Except Epidemiologists||Creative||.66|
|Preschool Teachers, Except Special Education||Creative||.63|
|Parking Lot Attendants||Service||.63|
|Bus Drivers, Transit and Intercity||Working||.61|
|Market Research Analysts and Marketing Specialists||Creative||.60|
|Self-Enrichment Education Teachers||Creative||.58|
|Computer and Information Systems Managers||Creative||.55|
|Public Relations and Fundraising Managers||Creative||.54|
|Producers and Directors||Creative||.53|
|Computer Systems Analysts||Creative||.53|
|Financial Specialists, All Other||Creative||.53|
|Software Developers, Systems Software||Creative||S.52|
|Personal Financial Advisors||Creative||S.52|
|Architects, Except Landscape and Naval||Creative||.50|
It appears that service-class geographies are most up for grabs politically. Indeed, as states with a large working-class share have largely abandoned it, the Democratic party’s future would seem to lie in a cross-class coalition of the service and creative class areas and voters.
Job categories like retail sales, customer service, personal-care aides, maids and housekeepers, food service workers and more employ millions upon millions of Americans. These jobs are disproportionately held by women, immigrants, and people of color. These are precisely the kinds of occupations and workers that could be galvanized into a Democratic coalition by policies aimed at higher minimum wages, job upgrading, affordable housing, accessible and affordable healthcare, protecting immigrant and minority rights, and a more robust social safety net for less advantaged groups.
The Largest Occupations Not Correlated With Vote Share ( Most Ubiquitous Occupations)
|Customer Service Representatives||Service||-.04||2,723,850|
|Secretaries and Administrative Assistants, Except Legal, Medical, and Executive||Service||-.10||2,320,250|
|General and Operations Managers||Creative||-.07||2,198,270|
|Stock Clerks and Order Fillers||Service||.09||2,035,360|
|Personal Care Aides||Service||.05||1,497,740|
|Secondary School Teachers, Except Special and Career/Technical Education||Creative||-.07||1,013,660|
|Middle School Teachers, Except Special and Career/Technical Education||Creative||-.03||627,930|
|Packaging and Filling Machine Operators and Tenders||Working||-.09||391,400|
|Medical and Health Services Managers||Creative||.07||333,120|
|Sales Representatives, Wholesale and Manufacturing, Technical and Scientific Products||Service||.08||330,460|
|Production, Planning, and Expediting Clerks||Service||.00||323,240|
|Child, Family, and School Social Workers||Creative||.07||302,560|
|Heating, Air Conditioning, and Refrigeration Mechanics and Installers||Working||-.10||296,190|
|Educational, Guidance, School, and Vocational Counselors||Creative||.07||262,380|
Forging such cross-class coalitions is an idea that is making headway among some Democratic strategists. Political consultant Stanley Greenberg has pointed to the advantages of using occupation, as opposed to educational level, as a basic building block of a new Democratic electoral coalition. “For the first time, we are asking occupation to try to get at this—and so, I think there really is potential for Democrats to gain here,” he told the New York Times.
When was the last time you heard a major Democratic politician talk about the day-to-day struggles of retail workers, clerical workers, personal care workers, nurses’ aides, orderlies, or bartenders in the same way they talk about the struggles of auto workers or steel workers? Maybe it’s time they should.
CityLab editorial fellow Claire Tran contributed research and editorial assistance to this article.