The Fall of the Creative Class? Not so fast.

A recent article in the inaugural issue of Thirty Two magazine purports to undercut my theory of the creative class and economic development. With an attention-getting headline, "The Fall of the Creative Class," writer Frank Bures relates how he and his wife moved to Madison, Wisconsin, for a "variety of not-very-well-thought-out reasons," among them the fact that Madison had "been deemed a 'Cre­ative Class' strong­hold by Richard Florida, the prophet of pros­per­ous cool."

Bures’ article is framed around anecdotes about his and his wife’s misadventures in Madison and ultimate relocation to Minneapolis. While it makes for a compelling narrative, are we really to believe he upended his entire life on the basis of a book he is so evidently skeptical of? His own moves to creative class hotbeds, Madison and Minneapolis, seem to undercut his claim that "the migra­tion of cre­ative work­ers to places that are tol­er­ant, open and diverse—was sim­ply not happening." 

The crux of his argument appears to be that my creative class theory simply apes the better known and longer-standing human capital approach. This, unfortunately, is where the piece really goes off the rails, relying heavily on questionable studies and cherry-picked negative comments from academics with their own personal axes to grind.  

Bures misunderstands the similarities and differences between the creative class and human capital theories of regional development. Both seek to gauge the underlying skills and capabilities of people and associate them with measures of economic growth and development. As I related in the original version of The Rise of the Creative Class, the human capital approach was what inspired me to develop an alternative measure of human skill. 

The two measures have meaningful, important and very useful differences. Human capital theory uses educational attainment (typically the percentage of adults with a college degree), a very broad measure that excludes such successful entrepreneurs as Bill Gates and Steve Jobs, who didn’t graduate from college. My creative class measure is based on the work people actually do, as measured by detailed Bureau of Labor Statistics data. This allows researchers and economic developers to zero in on the actual occupational categories – science and engineering, arts and culture, business and management, meds and eds – that make up the creative class and other occupational classes. My team and I have also done detailed empirical and statistical comparisons between the two.

The creative class is not just a proxy measure for college graduates. Roughly three-quarters of college grads in America work in creative class jobs, but four in ten members of the creative class—16.6 million workers—do not have college degrees.

Bures cites a 2009 study that purports to find no connection between my creative class theory and economic development, quoting one of the study’s authors as saying: “The mea­sure­ment of the creative class that Florida uses in his book does not cor­re­late with any known mea­sure of eco­nomic growth and devel­op­ment …the emperor has no clothes.” I asked a senior economist colleague of mine to rerun these numbers. The results were profoundly different. Where the study Bures relies on found a correlation of .001 between average wages and the creative class, my colleague found the correlation to be a whopping .84, significantly higher than that for human capital/ college grads (.65).

Other independent studies document the considerable role of the creative class in regional economic development. In a series of careful and detailed studies, some of which were published under the auspices of the Federal Reserve Bank of New York, the economists Jaison Abel of the New York Fed and Todd Gabe of the University of Maine found that the creative class is a distinct measure from educationally based human capital, and that the creative class adds considerable economic value on its own.

When Dutch economists Gerard Marlet and Clemens van Woerkens compared the human capital and creative class approaches and published their results in the journal Urban Studies, they concluded that my creative class measure “sets a new standard” for measuring skill and talent. "With our Dutch data set," they wrote, "we do find evidence that Florida’s Creative Class is a better predictor of city growth than traditional education standards. Therefore we conclude that Florida’s major contribution is his successful attempt to create a population category that is a better indicator for levels of human capital than average education levels or amounts of highly educated people.”

A newly released study titled “Education or Creativity: What Matters Most for Economic Performance?” in the journal Economic Geography (available online here) used advanced statistical models to compare the effects of the creative class and human capital across the 257 EU regions.

There is a large consensus among social researchers on the positive role that human capital plays in economic performances. The standard way to measure the human capital endowment is to consider the educational attainments of the resident population, usually the share of people with a university degree. Florida (2002) suggested a different measure of human capital—the “creative class”—based on the actual occupations of individuals in specific jobs like science, engineering, the arts, culture, and entertainment. … Our results indicate that highly educated people working in creative occupations are the most relevant component in explaining production efficiency. 

Bures invokes the old chicken and egg dilemma of what comes first, jobs or arts and cultural creativity, an overly simplistic formulation I wrote about here last month. As proof, he cites a study that purportedly uses "Granger causality tests" to tease out the relationship between arts funding and regional growth development. The study (which as far as I can tell has not been published) has little to do with my theory. Despite what Bures or others may try to claim, I’ve never said that arts spending is a driver of economic development. Rather, my argument is that open-minded, diverse and tolerant places where vibrant artistic, ethnically diverse and gay and lesbian communities have settled and thrived reflect underlying characteristics that are more likely to be accepting of new ideas, and better able to incubate new innovations and house and motivate entrepreneurial business ventures. Still, the findings that Bures summarizes are decidedly mixed. Arts funding seemingly does contribute to economic development in four of the 15 metros covered by the study, including his own chosen Minneapolis-St. Paul.

But when all is said and done, artistic and cultural creativity do add to regional economies. At its best, innovation — whether it occurs in leading cities or in highly successful companies like Apple — means mixing new designs (artistic and cultural creativity) with technological creativity and ultimately with economic creativity (entrepreneurship) to create revolutionary new products, build new firms, and create and transform whole industries. A detailed study I conducted with my colleagues Charlotta Mellander and Kevin Stolarick, published in the Journal of Economic Geography, found three occupational groups to be associated with increased regional wages: science and engineering;  management and business; and arts, design, media and entertainment. 

I actually laid all this out in a 3,000 word letter I sent Bures, after he contacted me to get my take on his article. In the course of his essay, Bures says my "answers didn’t really shed any more light than [my] books." Below, I've copied the text of the letter I sent him. You be the judge.

Full Text of My Response Letter to Thirty Two Magazine

Thanks for your interest in my research. I have taken some time to go through your questions. The bottom line is that the Creative Class approach has been subjected to vigorous debate, critique and vetting and careful (and some not-so-careful) empirical analysis. That analysis has used a wide range of statistical techniques.

While some have questioned the view (which is the nature of scientific progress ala Kuhn), the best, most careful and detailed research, including a wide range of independent studies, confirms the differential migration and clustering of the Creative Class and its effects on regional economic development measured as income or wages, when controlling for other factors. Much of my own work on this subject, conducted jointly with Kevin Stolarick and Charlotta Mellander among others, has been subject to detailed peer-review and published in the highly regarded Journal of Economic Geography among other academic outlets, which counts eminent urban economists like Edward Glaeser and economic geographers on its board and is a high-impact journal both in economics and geography. I’ve consulted with a wide range of other scholars and summarize my own and their key conclusions below. I hope you find this helpful to your article.

Can he show me an urban area that has experienced measurable economic growth as a result of an influx of creative individuals?

There are many. In fact, there is a wide consensus among economists, economic geographers and sociologists that the United States has witnessed a substantial migration and sorting of more highly-skilled, highly educated and “creative individuals” over the past several decades. This is detailed in sources as varied across the ideological spectrum as Christopher Berry and Ed Glaeser’s work on the divergence of human capital, and Glaeser’s book Triumph of the City, Bill Bishop’s The Big Sort, Charles Murray’s new book Coming Apart, and my own Atlantic essay on “The Means Migration.” There are many, many more. This migration has shaped the rise of some cities and metro areas while adding to the challenges of others. In terms of specific metros, places such as greater Washington DC, Greater Boston, Greater NY, Greater San Francisco and innumerable others have benefited from this migration. Also, New York City Mayor Michael Bloomberg has noted the importance not just of regional flows or migrations of people to regional growth but of international flows in his op ed in the Financial Times this past week, entitled “Cities Must Be Cool, Creative and in Control,” where he writes:

As individuals and capital become ever more mobile, cities are in competition for people, visitors and business. Until recently, “competitiveness” was outside a mayor’s domain because the factors defining it were decided at the national level. But today, with more than half the world’s population living in cities – generating about 80 per cent of global output – and businesses formulating growth strategies around urban markets, cities cannot afford to cede their futures to national governments.”


In this respect, part of what sets cities such as New York and London apart cannot be captured by rankings. Recent college graduates are flocking to Brooklyn not merely because of employment opportunities, but because it is where some of the most exciting things in the world are happening – in music, art, design, food, shops, technology and green industry. Economists may not say it this way but the truth of the matter is: being cool counts. When people can find inspiration in a community that also offers great parks, safe streets and extensive mass transit, they vote with their feet.”

How does he address the criticism that he is just describing the effects of growth rather than the causes?

There is a huge literature on this. Economists are in wide agreement that human capital broadly defined is the major cause or determinant of economic growth. This has been shown in a host of studies at the national level by Robert Barro and many at the local level. Robert Lucas, the Nobel Prize winning economist from the University of Chicago, notes that the clustering of human capital in cities is the primary determinant of economic growth in his seminal essay on “The Mechanics of Economic Development.” Human capital is a measure of skill broadly. So there is no debate on this: human capital causes economic growth.

The debate is over how to measure human capital. Most economists operationalize this as education level. The creative class measure is an alternative measure of skill that operationalizes skill by occupation, using the occupational codes from the Bureau of Labor Statistics. I discuss my reasoning for this in numerous places including The Rise of the Creative Class and Flight of the Creative Class, in my response to Ed Glaeser’s review of Rise, and my essay “Revenge of the Squelchers” in Next American City.  The basic reasoning is that occupation captures not what people study but what they actually do, and also that it is a more heterogeneous measure than simply tracking the share of adults with college degrees. Pragmatically, one can use it to identity occupational cluster strengths in a region, in a way comparable to Michael Porter’s use of industrial cluster analysis to identify industrial cluster strengths.

My Creative Class measures are highly correlated with the conventional education based one at about .9, so they tend to measure the same thing.

A detailed study by McGranahan and Wojan used more recently released O*NET data on actual occupational skill to test the definition of the creative class. When they did so, they made some minor revisions and refinements, but found that the original creative class measure does a very, very robust job of capturing occupations which are based on underlying creative skill.

That said, there are important differences that suggest the importance and utility of the Creative Class or occupational approach.  Across the entire United States, nearly three-fourths (72.2 percent, to be exact) of adults with college degrees are members of the Creative Class. But less than 60 percent (59.3 percent) of the members of the Creative Class have college degrees, according to a detailed analysis by Kevin Stolarick and Elizabeth Currid-Halkett. In other words, four in ten members of the Creative Class—16.6 million workers— do not have college degrees. As Stolarick and Currid-Halkett write: “Thus, while some correlation would be expected, our results indicate that human capital and the Creative Class do not necessarily capture the same people nor is a measure of each’s respective presence in a regional economy indicative of similar trends.”[1]

There is a substantial body of research that shows that the Creative Class measure operates in addition to and through other channels than the standard education-based human capital measure and outperforms it in many instances. A large-scale study by Stolarick, Mellander, and me shows that the Creative Class has a bigger effect on wages—a key element of regional productivity— whereas education tends to have a greater effect on income. Independent research by economist Todd Gabe and others backs this up, showing that the Creative Class continues to have a substantial effect on regional economic growth when controlling for the effects of education and other factors. More to the point, having a Creative Class job also brings economic benefits that extend beyond those of going to college. A college graduate working in the same occupation as a non-college graduate earns approximately 50 percent higher wages. But having a Creative Class job adds another 16 percent, about the same as another 1.5 years of additional education, according to Gabe’s research.[2] Even more important, just counting years of education ignores a lot of people who do very creative work, including world shaping entrepreneurs such as Steve Jobs and Bill Gates, or artists and others who did not complete college.

In some sense the argument about what causes what is a non-argument. Economic theory tells us that human capital is the engine of growth.  The whole question of chicken and egg in this sense is moot; growth is an ongoing iterative process.

One thing my research tries to get at that the conventional literature does not, is what causes the concentration of the creative class and of human capital broadly. Here, I argue that it is openness and low barriers to entry that matter. This idea is catching on. Scott Page has written an important book on the economics of it, The Difference. Ronald Ingelhart’s detailed World Values Surveys have identified openness toward various types of people including gays and lesbians as a key component of post-materialist cultures of advanced nations, and economists Marcus Noland and Howard Peck have identified (more backward) attitudes toward homosexuality to be associated with lagging development in the Middle East. My Black Box paper with Mellander and Stolarick shows the effect of these factors using structural equation models, while my housing paper with Mellander, “There Goes the Metro,” documents their effects on housing prices over and above income and conventional human capital.

Is there some statistical evidence that young creative professionals are migrating to creative cities?

Yes, detailed studies by Joe Cortright for CEOs for Cities show the migratory pattern of young creatives to certain regions and to urban centers of those regions.

The question is not one of age, but of skill and occupation broadly. I continuously point out in my studies that Creative Class theory is not about age but about skill. I do mention that previously economic development practice had focused on families, which is important, but that in today’s society single people make up roughly 50 percent of all households, as Eric Klinenberg’s Going Solo points out. I also note via Cortright’s studies that young people are the demographic group most likely to move, so attracting them can be beneficial.

More recent research zeros in not just on age, education or occupation/class but the actual skills that lie behind work. In other words, they provide a fundamental measure of human capital as skill. This work is based on the Occupational Information Network (or O*NET) database developed by the Bureau of Labor Statistics, which provides richly detailed information on the mix and level of skills required for more than 800 occupations. Various research teams including my own have used these data to identify the fundamental skills underlying Creative Class and other types of work, to chart the economic returns to these core skills, and ultimately to map their distribution across the US economic landscape.

There are three core types of skills, according to this line of research. The first is one we are all familiar with—basic physical skill of the sort associated with traditional work. Its attributes include good hand-to-eye coordination, strength, and dexterity. 

The second two types of skills are those associated with Creative Class work. This second basic skill type—cognitive skill—is reasonably well understood. It involves the ability to acquire knowledge, process information, and solve problems. This basic intellectual and analytical horsepower has been identified as the core skill underpinning the knowledge economy by writers from Peter Drucker and Daniel Bell to Robert Reich and Charles Murray.

However, there is a third type of skill set that is even more critical. The O*NET data defines its core attributes as the “capacities used to work with people to achieve goals.” You can call it “social intelligence.” Its salient characteristics are discernment, communications abilities, leadership, awareness, and the like. Highly developed social skills include the capacity to bring the right people together on a project, persuasion, social perceptiveness, the ability to help develop other people, and a developed sense of empathy. These are the leadership skills that are needed to innovate, mobilize resources, build effective organizations, and launch new firms.

Recent research by economists William Strange, Marigee Bacolod, and Bernardo Bluk; the economic geographer Allen Scott of UCLA and my own team has examined the geographic distribution of these skills. Jobs requiring physical skill cluster in smaller and medium-sized metro areas—industrial centers where land for factories is relatively inexpensive. Jobs featuring analytic skills are sparse in these places and heavily concentrated in the largest metro areas, indicating rising benefits from having larger numbers of well-educated, highly intelligent people working close together. And jobs requiring the highest level of social skill are the most concentrated in the very largest metro areas. In fact, these skills seem to grow ever-more essential as local economies grow larger and more complex. What this research has helped us understand is that it is not just the accumulation of knowledge or cognitive ability that drives the growth of cities, but the additional clustering of social-intelligence skill. This clustering of social-intelligence skill increases the quality of the combinations and recombinations that drive innovation and economic growth.

Has he run economic models, or Granger causality tests, using the standard measures of economic growth to test the Creative Capital theory? If not, why?

There have been many, many tests of the creative class theory, including some my own research team and I have run. I am not aware of any that use Granger causality tests, but they do use varied types of regression models and our own research uses well-specified structural equation models.  

In the original edition of Rise, I cited the research of Robert Cushing, a former University of Texas sociologist and statistician. Cushing undertook a systematic comparison of the effect of the three main theories of regional growth—Glaeser’s human capital theory, Putnam’s social capital theory and my own, which he referred to as the creative capital theory. He built statistical models to determine the effect of these factors on regional growth between 1990 and 2000. To do so, he included separate measures of education and human capital; occupation, wages and hours worked; poverty and income inequality; innovation and high-tech industry; and creativity and diversity for the period 1970–1990. His results were striking. He found no evidence that social capital leads to regional economic growth; in fact the effects were negative. In a related study he found that leading high-tech regions had higher income and higher levels of growth but scored below average on almost every measure of social capital. They had less trust, less reliance on faith-based institutions, fewer clubs, less volunteering, less interest in traditional politics and less civic leadership, but much higher levels of “protest politics” and “diversity of friendships.” In his own words, “conventional political involvement and social capital seem to relate negatively to technological development and higher economic growth.” Using creative occupations, bohemians, the Milken High-Tech Index and innovations as indicators of creative capital, he found that the creative capital theory produced formidable results, with the predictive power of the Bohemian and Innovation Indexes being particularly high. He concluded: “[The] creative capital model generates equally impressive results as the human capital model and perhaps better.”[3]

Several other independent statistical studies also confirm the creative class approach. A detailed study of regional development in the Netherlands found the Creative Class to considerably outperform the standard human capital measure in accounting for employment growth, leading its authors to conclude that Creative Class sets a “new standard” for measuring skill and talent especially when considering regional labor productivity. “With our Dutch dataset we do find evidence that Florida’s creative class is a better predictor of city growth than traditional education standards, “they write. “Therefore we conclude that Florida’s major contribution is his successful attempt to create a population category that is a better indicator for levels of human capital than average education levels or amounts of highly educated people. The point is, as Florida stated, not which or how much education people can boast of, but what they really do in working life.”

Economist Todd Gabe has found the Creative Class to have a strong positive effect on regional earnings as well as to lessen unemployment, especially since the crisis. Research by Wojan and McGranahan focusing on rural counties examines the relationship between measures of growth over time and Creative Class indicators at the beginning of the period. They uncover pretty strong effects related to the creative economy, controlling for a wide range of other factors. Potential endogenity should be of less concern in the research using individual-level data. Gabe’s research published in the Handbook of Creative Cities and elsewhere does so and identifies the strong effect of Creative Class occupations on reducing unemployment and enhancing earnings.

So the effects of the creative class on employment and earnings are clearly documented not just by my own research but by independent research in the field.

One last point: a growing number of scientists have come to favor simulation or adaptive agent models over more conventional regression techniques. Robert Axtell of George Mason University, a leading figure in adaptive agent modeling, has also built adaptive agent models which show the growth effects of the clustering of Creative Class individuals ala a basic Lucas-Jacobs clustering model. These models show that the migration and clustering of highly skilled creative class like individuals shapes the growth and development of cities over long time horizons.

I hope these responses help with your article.

[1] Kevin Stolarick and Elizabeth Currid-Halkett, Creativity and the Crisis: The Impact of Creative Workers on Regional Unemployment,Martin Prosperity Institute, September 2011.

[2] Todd Gabe, “The Value of Creativity,” in David Emanuel Andersson, Åke Emanuel Andersson, and Charlotta Mellander, eds., Handbook of Creative Cities (Cheltenham, UK: Edward Elgar, 2011), pp. 128–145.

[3] See Robert Cushing, “Creative Capital, Diversity and Urban Growth.” Unpublished manuscript, Austin, Texas, December 2001.


About the Author

Most Popular

  1. A photo of a police officer in El Paso, Texas.

    What New Research Says About Race and Police Shootings

    Two new studies have revived the long-running debate over how police respond to white criminal suspects versus African Americans.

  2. Equity

    The Problem With Research on Racial Bias and Police Shootings

    Despite new research on police brutality, we still have no idea whether violence toward African Americans is fueled by racial prejudice. That has consequences.

  3. A map of Minneapolis from the late 19th century.

    When Minneapolis Segregated

    In the early 1900s, racial housing covenants in the Minnesota city blocked home sales to minorities, establishing patterns of inequality that persist today.

  4. photo: A protester stands on a damaged bus stop near the Third Police Precinct on May 28 in Minneapolis during a protest over the death of George Floyd.

    In Minneapolis Protests, Bus Drivers Take a Side

    The city’s transit union issued a statement of support for members who balked at assisting police during demonstrations over the killing of George Floyd.

  5. A map of population density in Tokyo, circa 1926.

    How to Detect the Distortions of Maps

    All maps have biases. A new online exhibit explores the history of map distortions, from intentional propaganda to basic data literacy.