Working Smart for the Money

A new study (PDF) from the Bureau of Labor Statistics provides important insight on states where workers toil the longest hours and make the most money. The study by Dante DeAntonio uses data from the Current Employment Statistics -- a monthly survey of more than 400,000 U.S. business establishments -- to provide estimates for employment, hours, and earnings for all 50 U.S. states. Catherine Rampell summarized some key findings of the study earlier this week over at Economix.

Take a look at the map of the hardest-working states in terms of hours worked. Nevada tops the list with an average of 37 hours per week. Wyoming, Louisiana, Texas, Kentucky, and Alabama all average more than 36 hours per week. At the opposite end of the spectrum are Montana, the Dakotas, Hawaii, and New Hampshire which average less than 33 hours per week.

Now look at the map of the highest-earning states in terms of average hourly earnings. The top earner is D.C. followed by Connecticut, Massachusetts, New York, New Jersey, Washington, Alaska, California, and Maryland. The lowest-earning states are South Dakota, Mississippi, Arkansas, Oklahoma, and West Virginia.

What's most striking is there's virtually no overlap at all between the two maps.

So, I decided to take a closer look at some key economic and demographic factors that might be behind the variation in state earnings and working hours. With the help of Charlotta Mellander, we ran a series of scatter-graphs and performed a simple correlation analysis. I remind readers of the usual qualifiers: our analysis is based on correlations which point only to associations between variables and do not identify causality. Still, some patterns are striking.

First and foremost, we find a total lack of  correlation between hours worked and earnings across the 50 states -- there was no statistical significance at all for the correlation between these two variables.

Second, we find that when it comes to state earning power, working smarter trumps working harder across the board.

The connection between human capital and economic development has been theorized and documented empirically by economists like Gary Becker, Robert Barro, Robert Lucas, Edward Glaeser, and many others. Human capital (that is, the percentage of a state's workforce with a bachelor's degree and above) is closely associated with earnings -- the correlation being .65. We ran a scatter-plot for these two variables. Connecticut, Massachusetts, New York, Washington, and New Jersey all stand way above the line, while Mississippi, Montana, and the Dakotas fall far below it.

Earnings are also closely associated with the kinds of work that states specialize in. The correlation between creative class jobs and hourly earnings is .79. The line on the scatter-graph runs quite steeply upward. Clearly, the transition to smarter, more knowledge-based state economies has a big effect on earnings levels.

On the other hand, state earnings are negatively correlated with blue-collar, working-class jobs (-.67). The slope of the line is steeply negative. Interestingly enough, industrial states like Michigan, Illinois, Ohio, and Indiana all stand above the line. Below the line are states like Mississippi, South Dakota, Oklahoma, Idaho, and West Virginia.

Some believe that immigrants pose an economic burden on states -- that they drive wages down and consume high levels of public services. Our analysis suggests this view is wrong. We find that state earnings are positively associated with the percentage of immigrants (.64). There appears to be a geographic component to this. Northeastern states -- Connecticut, Massachusetts, New York, and New Jersey -- all stand above the line, while California, Texas, Nevada, Arizona, and Florida stand below it. This may also reflect the industrial and employment structures of these state economies.

Earnings were also higher in states with higher concentrations of gay people (.55). Some might say that higher earning states attract more immigrants or gays. But my own view is that openness affects wages by enabling states to better compete for more highly educated and highly skilled workers across demographic, ethnic, and racial categories.

Smarter states also work less. Working hours are negatively associated with state human capital levels (a correlation of -.59) and also with creative class work (a correlation of -.33). Working hours are, however, positively associated with blue-collar, working-class jobs (.4).

How do work effort and earnings affect the happiness and well-being of states? Interestingly enough, it appears that hours worked play a bigger role. The correlation between state well-being and earnings is less than .3 (.29), while the association between it and hours worked is more than .4 and negative (-.45).

Our analysis reinforces a simple fact that working smarter, and not working harder, is what brings higher earnings to states. In contrast to the view, held by some, that immigrants drag down wages, states with higher levels of immigrants have higher earnings. So do states with higher levels of gays and lesbians. It's hard to say exactly what's causing what here -- if immigrants or gay people are attracted to richer states that offer more opportunity and higher wages, or if they are part of an economic system that generates higher earnings. My own view is that more open and tolerant states are better able to compete for a wider range of talented and skilled workers across the board. And, smarter states not only generate higher earnings, they afford a greater level of happiness and well-being to their residents.

It's time to get over the notion that simply working harder brings wealth and economic development. The structure and composition of jobs matter greatly. At a time when job creation is at the top of the agenda, this is something policy-makers need to factor into their thinking about exactly what kinds of jobs we wish to create.