Archive for the ‘The Atlantic’ Category

Richard Florida
by Richard Florida
Wed Aug 26th 2009 at 9:00am UTC

Stressed-Out States

Wednesday, August 26th, 2009

Stress is a fundamental fact of life these days. But which parts of the country have the most stressed-out people?

The map below from the the Gallup-Healthways Well-Being survey shows the stress levels for each of the 50 states. The map reflects the fraction of survey respondents who said they experienced stress “during a lot of the day yesterday” between January and June 2009.

Stress Map.jpg
While people in Kentucky, West Virginia, and Colorado are more stressed out than their counterparts in Hawaii, North Dakota, and Iowa, what strikes me most is how many Americans across-the-board report substantial levels of stress – from a low of 31.4 percent in Hawaii to a high of 44.9 percent in Kentucky. In half of all states, four in 10 residents or more report experiencing stress “during a lot of the day.”
Richard Florida
by Richard Florida
Sat Aug 22nd 2009 at 12:45pm UTC

City Residents Pay More… Taxes

Saturday, August 22nd, 2009

A new study by University of Michigan economist and MPI associate David Albouy, published in the Journal of Political Economy, finds that workers in expensive cities – including those in the Rustbelt and even hard-hit Detroit – pay a disproportionate share of federal taxes. Overall, urbanites pay 27 percent more in federal income taxes than workers with similar skills in a small city or rural area. Here’s a summary of the study.

“Workers in cities are generally paid higher wages than similarly skilled workers in smaller towns, so they’re taxed at higher rates. That may sound fair, until one considers the higher cost of living in cities, which means those higher wages don’t provide any extra buying power. The federal income tax system doesn’t account for cost of living. So the effect is that workers in expensive cities like New York, Los Angeles and Chicago pay more in taxes even though their real income is essentially the same as workers in smaller, cheaper places.

“The extra burden wouldn’t be so excessive if more federal tax dollars were returned to urban areas in the form of higher federal spending. But according to Albouy’s research, that’s not the case. His data show that more federal dollars are actually spent in rural areas, despite the fact that cities send far more cash to Washington. The net effect of all this is a transfer of $269 million from workers in high-cost areas to workers in lower cost rural areas in 2008 alone.

“Over the long haul, Albouy says, the larger tax burden causes workers to flee large urban centers in the Northeast and settle in less expensive places in the South. So to some extent, it may have been the federal tax system that put the rust on the rust belt.

“Detroit is a perfect example of a city that gets the short end of the stick.

“With its high wage levels, Detroit was, until recently, contributing far more in federal revenues per capita than most other places for over one hundred years,” Albouy said. The recent federal bailout to Detroit automakers “is peanuts relative to the extra billions the city has poured into Washington over the 20th Century.”

“Albouy says that city folk shouldn’t expect relief from this system anytime soon.

“Highly taxed areas tend to be in large cities inside of populous states, which have low Congressional representation per capita, making the prospect of reform daunting,” he writes.

The full study is here (PDF).

Richard Florida
by Richard Florida
Fri Aug 21st 2009 at 12:00pm UTC

The Bailout Maps

Friday, August 21st, 2009

The bailout is big. But, where exactly is it going?

Thanks to the efforts of ProPublica, we can track bailout funds by state. The map below, based on their data, shows the geographic distribution of bailout spending.

The bailout is massively concentrated in just a few states. Total bailout funding, according to the ProPublica data, is $476.5 billion to date. One state, New York, has captured $175 billion of that, more than a third. Michigan is next with $80.7 billion or 17 percent of the total, followed by North Carolina with $56.3 billion, Virginia with $54.9 billion, and California with $34.4 billion. The top three states accounted for 66 percent of bailout spending; the top five 84 percent; and the top two more than 10 percent.

How does the geography of the bailout look when we control for the size of state economies – say, by population and economic output? With the number-crunching help of Ronnie Sanders and map-making assistance of Scott Pennington, both of the Martin Prosperity Institute, we decided to take a look.

The second map shows the geography of bailout funding per person. The average per state, excluding the District of Columbia and based on the ProPublica data, is $1,570.34 bailout dollars per person. Again, two states dominate the tally – New York, where the bailout adds up to $8,978.83 per person, and Michigan where it’s $8,067.28. There are just five additional states where bailout funds top $1,000 per person: Virginia ($7,044.47), North Carolina ($6.104.69), Minnesota ($1,379.21), Connecticut ($1,085.33), and Iowa ($1,065.76).

The third map shows the geography of the bailout as a percent of state economic output or gross state product. The bailout, again based on the ProPublica data, was three percent of total state output, with each state on average receiving 1.8 percent of its GDP. Michigan takes the top spot here, with bailout funds equivalent to a whopping 21.1 percent of its total economic output. New York is next at 15.3 percent; followed by North Carolina, 14.1 percent; and Virginia, 13.8 percent. No other state received bailout funding that was more than three percent of its output.

By any measure, the bailout has been massively concentrated geographically.

Richard Florida
by Richard Florida
Wed Aug 19th 2009 at 6:37pm UTC

Economics and Ideology

Wednesday, August 19th, 2009

Political scientist Andrew Gelman has some great graphs on the connection between economics and ideology. Comparing income levels, ideology, and party identification, he and collaborator Daniel Lee found the connection between income and party identification was strongest among conservative Republicans. But the relationship was “close to zero” for liberals. Liberal Dems were spread across all income groups, while conservative Dems had much lower income levels.

My reading is that class continues to play a considerable role in American politics: With the exception of liberal Dems who draw from across the spectrum of classes, the parties and their key factions increasingly represent class blocs. Gelman notes that the connection between economic status and party/ideology underpins America’s increasingly polarized policy debates. He’s right. In the current zero-sum economic climate, it’s only likely to get worse.

Richard Florida
by Richard Florida
Tue Aug 18th 2009 at 9:30am UTC

Unemployment and Happiness

Tuesday, August 18th, 2009

How has the economic crisis affected the happiness and well-being of Americans? Newly released data from the Gallup-Healthways Well-Being Index enables us to take a look.

At the national level, not so much: The mid-year 2009 score is 65.1, a moderate decline from 65.5 in 2008. (Catherine Rampell of Economix provides a nice summary of the survey methods, indicators, and key findings.)

But, rising unemployment appears to have a significant relationship to the happiness of states, according to our analysis of the Gallup-Healthways data.

Not surprisingly, the biggest declines in overall happiness occurred in work-related well-being. The Gallup-Heathways Well-Being Index is made up of six separate sub-indexes – life evaluation, emotional health, work environment, physical health, healthy behavior, and access to basic necessities. Five of these indexes fell between 2008 and 2009, with the biggest decline occurring in the work environment index: More than three-quarters of states saw their work environment score fall in 2009.

This is broadly in line with happiness research. It had been long thought that happiness essentially levels off after a moderate income level is crossed. But an influential study by Betsy Stevenson and Justin Wolfers found a strong association between happiness and economic conditions. A 2005 study found that a significant increase in Finland’s unemployment rate (from three to 17 percent) did not produce a significant drop in overall well-being.

The new Gallup-Healthways Index also covers the 50 states. Interestingly enough, the “happiest states” in 2009 – Hawaii, Utah, and Montana – were more or less the same as in 2008; the same is true of the “unhappiest states” – West Virginia, Kentucky, and Arkansas. Drilling down a little further, Utah topped the list in life evaluation, Hawaii in emotional health, Idaho in work environment, North Dakota in physical health, Vermont in healthy behavior, and Iowa in basic access.

Still, it’s clear that the economic crisis has been harder on some states that others. Older industrial states of Michigan, Indiana, and Ohio have seen their unemployment rates soar in the double digits, while the housing crisis has wreaked havoc on once fast-growing states like Florida and Arizona.

The availability of state-level data for before (2008) and after (2009) the crisis provides a useful lens for examining the effects of worsening economic conditions on state happiness.

So my collaborator, regional economist Charlotta Mellander, looked at the relationships between happiness and economic factors like output, income, and unemployment. Let me emphasize that what I am reporting here are correlations or associations. While these findings do not imply causation, they remain interesting nonetheless.

First off, the relationship between happiness and economic output has apparently become weaker. The relationship between the two which was correlated (.33) and statistically significant in 2008, is no longer so (.27 and not statistically significant in 2009).

Second, the relationship between income on happiness also seems to have weakened (falling from a correlation of .43 in 2008 to .30 in 2009 – both significant at the .01 level).

Third, unemployment appears to be the biggest short-run factor affecting state happiness. Two measures of unemployment – a higher state unemployment rate and a bigger increase in that rate between 2008 and 2009 – were associated with both lower levels of state well-being and a bigger drop in state well-being between 2008 and 2009.

The first chart graphs the relationship between 2009 state unemployment rate and state well-being. Hawaii and Utah, above the line; and West Virginia, Kentucky, and Arkansas below it, are clearly outliers. Still, the fitted line shows a reasonably close association between unemployment and happiness among states. The correlation coefficient of -.44 between the two (statistically significant at the .01 level) lends additional support to this.

The second chart graphs the relationship between state happiness and the change in the unemployment rate between 2008 and 2009. Hawaii and Utah, and West Virginia, Kentucky, and Arkansas are again outliers. But the fitted line shows a clear association between the two. And while the correlation coefficient between the two is weaker than above (-.34 and statistically significant at the 0.05 level) it nonetheless supports the association.

The connection between state happiness and unemployment also came through when we looked at the relationships between the change in state well-being between 2008 and 2009 and the two measures of unemployment – the 2009 unemployment rate and change in unemployment between 2008 and 2009. The correlations for each are statistically significant (-.30 for the 2009 unemployment rate and -.34 for change in the unemployment rate between 2008 and 2009, both significant at the .05 level).

Given all of this, it’s safe to say that unemployment plays a reasonably big role in the happiness – or should I say, unhappiness – of states.

Richard Florida
by Richard Florida
Sat Aug 15th 2009 at 9:30am UTC

This is Your State’s Personality on Drugs

Saturday, August 15th, 2009

Yesterday, we looked at the relationship between drug use and the concentrations of certain kinds of jobs in states. We saw that cocaine is more likely to be used in states where lawyers make up a larger share of the workforce, while marijuana use is associated with states with higher concentrations of artists, scientists, architects, and educators.

Today, we turn to the relationships between drug use and personality. Psychologists define personality according to the five types – extraversion, agreeableness, conscientiousness, neuroticism, and openness-to-experience. A study by psychologists Jason Rentfrow, Sam Gosling, and Jeff Potter has developed and analyzed data on the state-wide concentration of these five major personality types. Charlotta Mellander and I then worked with Rentfrow to examine the relationship between drug use and the state-wide concentration of personality types. The charts below graph the results.

Basically, drug use was positively and significantly associated with one personality type – openness-to-experience (.33**). It was negatively and significantly associated with three others – agreeableness (-.41**), conscientiousness (-.29*), and extraversion (.-52**).

Rentfrow explains our results this way.

I find it helpful to think about these regional differences as reflecting different psychosocial climates/scenes, and one question we can ask is what underlies these climates/scenes?

Openness, for example, is associated with curiosity and trying new things, so it would make sense that Open regions are places where more people have experimented/used drugs than places low in Openness.

Conscientiousness is associated with order, structure, caution, and obedience, so it would make sense that there would be less experimenting with drugs in places where there are large numbers of conscientious people.

Low levels of agreeableness are associated with aggression and antisocial behavior, so it’s conceivable that places with large numbers of disagreeable people will also be places with comparatively high drug use.

The one dimension that is inconsistent with what I would expect is Extraversion. Hans Eysenck proposed that extraversion is driven by arousal. Whereas introverts have higher levels of internal arousal, which motivates them to avoid social contact because it generates more arousal, extraverts are low in internal arousal and seek out stimulating activities to increase their level of arousal (it’s like introverts are anxious and avoid stimulation and extraverts are bored and seek it out). With that in mind, I was expecting to see that stimulants were more commonly used in places where Extraversion is high and that marijuana was used more in places with fewer extraverts. That’s not what we’ve found, though.

Correlation coefficient: .38**

Correlation coefficient: .38**

Correlation coefficient: .40**

Correlation coefficient: -.52**

Correlation coefficient: -.41**

Correlation coefficient: -.29*

Note: * indicates statistical significance at the .05 level; ** indicates significance at the .01 level.

Richard Florida
by Richard Florida
Fri Aug 14th 2009 at 9:30am UTC

This Is Your Occupation on Drugs

Friday, August 14th, 2009

Yesterday, we looked at the relationship between drug use and class. We found that drug use was significantly associated with the percentage of the creative class in a state, and negatively so with the percentage of people employed in the working class.

Today, I dig a bit deeper into the relationship between drug use and specific types of professional, knowledge-based, and creative jobs – management, business and finance, architecture and engineering, science, health-care, education, and arts and entertainment. The patterns here are quite interesting.

Occupations sort relatively neatly along the lines of marijuana versus cocaine use. The short of it is that marijuana use is more positively associated with science (.35), education (.38), artistic professions (.35), and engineering and architecture (.29), while cocaine use is positively associated with lawyers (.41) and, to a lesser extent, with business and finance occupations (.27), computer jobs (.25), and management fields (.26).

Drug use overall is significantly associated with the state-wide concentrations of three major types of occupations – science (.35), architecture and engineering (.34), and arts, design, and entertainment (.33). And, in all three cases, this correlation appears to be driven by marijuana use; none of them are significantly associated with cocaine. Management occupations are also positively associated with overall drug use, though the correlation (.26) is somewhat weaker.

Here’s what my colleague and collaborator, Cambridge University psychologist Jason Rentfrow, had to say about our results:

I think it’s interesting that cocaine is high for finance, law, and quant professions. Although we can’t infer whether it’s people in those jobs actually doing drugs, those professions are generally regarded as intense and lavish. So it’s interesting that an expensive stimulant like cocaine is used more often in places where comparatively large numbers of people work in intense and high-paying jobs… It’s also interesting that marijuana is popular in places with artists, designers, and architects because those are jobs that encourage divergent thinking and marijuana is a psychoactive drug that’s associated with creativity.

What I think is particularly interesting about the results is that most professions possess elements of income, education, and personality. Even in those cases where lawyers and architects make similar amounts of money, they’re very different lines of work and appeal to different types of people.

Correlation coefficient: .41**

Correlation coefficient: .35*

Correlation coefficient: .29*

Correlation coefficient: .32*

Correlation coefficient: .38**

Note: * indicates statistical significance at the .05 level; ** indicates significance at the .01 level.

Richard Florida
by Richard Florida
Thu Aug 13th 2009 at 9:30am UTC

Drug Use and Class

Thursday, August 13th, 2009

Yesterday, we looked at the relationship between drug use and economic patterns. We saw that drug use was associated with both higher levels of state economic output as well as higher levels of unemployment.

Today, I turn to the relationships between drug use and economic class. My colleague Charlotta Mellander charted the relationships between drug use and the percentage of a state’s economy that is made up of two classes: the creative class – that is, people who work in knowledge-based, artistic, and professional occupations; and the working class – those who work in production, transportation, and construction jobs.

While the associations between drug use overall are weak, the patterns for marijuana and cocaine are significant. Take the creative class: Both marijuana and cocaine use are positively and significantly related to states with higher concentrations of the creative class.

Correlation coefficient: 39**

Correlation coefficient: 36**

Now look at the results for the working class, where the pattern is reversed. Both marijuana and cocaine are negatively and significantly related to the concentration of working class jobs in state.

Correlation coefficient: -.35**

Correlation coefficient: -.36**

Note: * indicates statistical significance at the .05 level; ** indicates significance at the .01 level.

Richard Florida
by Richard Florida
Wed Aug 12th 2009 at 9:30am UTC

This is Your Economy on Drugs

Wednesday, August 12th, 2009

Yesterday, I looked at the relationship between drug use and politics. We saw that states that voted for Obama had higher levels of marijuana and cocaine use than those that voted for McCain. But perhaps economic factors lie behind those political trends. We know that Obama drew from less affluent minority voters and also from more well-educated, creative class voters. Perhaps the associations between drug use and voting patterns reflect deeper economic patterns.

The conventional wisdom is that economic hardship is a key factor in drug use. Anyone who watches crime shows like The Wire gets this picture really fast.

To get a first approximation of this, we examined the relationship between drug use and unemployment. Not surprisingly, the use of illegal drugs is correlated with state unemployment (.31). And the correlations are even a bit higher when we look at marijuana (.36) and cocaine use (.36).

Correlation coefficient: 0.31*

But things get more interesting when we look at the relationship between drug use and economic development. While there is no relationship between economic output and illegal drug use overall, there is a significant relationship between state economic output and marijuana, and an even stronger correlation between economic output and cocaine use, as the charts below show.

Correlation coefficient: 0.31*

Correlation coefficient:  0.61**

But there’s more to the story. Tomorrow, I turn to the relationship between drug use and the class structure of state economies.

Note: * indicates statistical significance at the .05 level; ** indicates significance at the .01 level.

Richard Florida
by Richard Florida
Tue Aug 11th 2009 at 1:00pm UTC

This is Your Candidate on Drugs

Tuesday, August 11th, 2009

Ryan Grim’s new book, This is Your Country on Drugs, has revived interest in drug use and drug policy. Around the time it hit the streets, this map of drug use by state (via Map Scroll) started circulating around the Internet.

As it turns out, the map is based on detailed data from the National Survey of Drug Use and Health on the use of various types of “illegal drugs” by state.

So, with this treasure trove of data in hand, and with the help of two colleagues, the Swedish regional economist, Charlotta Mellander, and Cambridge University personality psychologist, Jason Rentfrow, we decided to take a look at the relationship between drug use and various political, economic, and psychological characteristics of states.

There’s lots and lots of research that examines the effects of factors like income, poverty, and race on the propensity to use drugs. But our team has been focusing on the role of psycho-social as well as economic factors on state and regional outcomes. A pioneering study by Rentfrow, Sam Gosling, and Jeff Porter identified the effects of personality factors on state-level economic and social outcomes. So we wanted to extend this line of research to see if and how these various economic, demographic, and personality factors might be related to drug use. We are knee-deep in a more extensive research project, but our preliminary results looked so interesting we thought we would report them and encourage feedback.

Some of the results reinforce the conventional wisdom, but others are surprising – at least for us.

Let’s start with an indicator of politics that’s sure to spark some interest – whether a state voted for Obama or McCain in 2008.

When it comes to the use of illegal drugs overall, there’s no real correlation. But that changes when we look at marijuana and cocaine. Both are significantly and positively related to with Obama states. The converse is true of McCain states, where the correlations are negative. Let me reiterate that these are provisional results which point to general relationships – or should I say associations – which could have many causes.

Conservative commentators might take this as evidence of the anything-goes, libertine lifestyles of “latte liberals” and of the need to return to more traditional, “all-American,” working class values. But that misses the bigger point. There are real differences in the economic and social environments of Obama and McCain states, as John Judis and Ruy Ruy Teixeira’s Emerging Democratic Majority, and Andrew Gelman and his collaborator’s Red State, Blue State, Rich State, Poor State, along with other studies have shown – particularly in their levels of development, economic and occupational structure, and, I would add, in their psycho-social environments as well.

Tomorrow, we’ll start to dig a little deeper into economic correlates of drug use. And, later this week, I’ll look at the relationships between drug use and certain kinds of occupations, and also to the personality types of states.

Correlation coefficient: .42**

Correlation coefficient: -.44**

Correlation coefficient: .37**

Correlation coefficient: -.36**

Note:  * indicates statistical significance at the .05 level; ** indicates significance at the .01 level.