Real income datapoint of the day

By Felix Salmon
May 11, 2010
Manhattan, Kansas, that is. Meanwhile, in much more educated and vibrant cities like Raleigh and Austin, real incomes fell substantially. What's going on here? Mike Mandel looks at the numbers:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Manhattan incomes rose by 35.5%, in real terms, between 2000 and 2008. Manhattan, Kansas, that is. Meanwhile, in much more educated and vibrant cities like Raleigh and Austin, real incomes fell substantially. What’s going on here? Mike Mandel looks at the numbers:

Brains and education did not seem to count too much in success in the last business cycle. Overall, the top ten cities, measured by growth in per capita income, had an average college graduate rate of 17.7% The bottom ten cities had a college graduate rate of 31.8%.

My feeling is that this is a historical anomaly, and largely a product of the beginning and end points that Mandel used: 2000 was the peak of the dot-com bubble, which artificially inflated tech salaries, while 2008 came at the end of a commodity boom which helped oil-rich states. The long-term trend is inescapable: the returns to education are large and growing, and if you’re not a college graduate and you don’t own your own company, it’s becoming increasingly difficult to maintain a middle-class lifestyle.

What’s more, Mandel’s outliers have to bee seen in the context of the bigger story about real wages, which he noted back in April: real wages in general have been falling, for the first time since the Great Depression. And with unemployment still at 10%, there’s not much hope that they’ll start rising again any time soon. If you want to see incomes go up in your city or region, your best hope is frankly just to get lucky, like Manhattan did. Because wages in the U.S. as a whole aren’t going anywhere.

(HT: Cowen)

Update: I had lunch today with Allison Schrager of the Economist, and she asked a good question: how did the percentage of college graduates in these cities change from 2000 to 2008? And what happened to the student population in Austin?

Update 2: Mark Beauchamp of EMSI makes some excellent points via email:

Mr. Mandel’s lead example of Metro Areas with the Biggest Real Per-Capita Income Gains was Houma-Bayou Cane- Thibodaux, LA — a metropolitan area that had 11% of the population of the San Jose MSA in 2002.

Between 2002 and 2009, its population grew 5.08% while its total employment grew 19%.  You have to have an income-per-capita increase in jobs like that.  An income-per-capita ratio favors regions that have explosive income growth (especially jobs that pay above the previous average), and population growth that lags behind the job growth. Conversely, the ratio will not favor areas that have a high amount of population growth with concurrent losses in total income.

The  biggest “loser” by this metric is the San Jose MSA, and during the same time period  its population grew by 86,269 people (5.16% growth) and lost 43,314 jobs (a 5% loss).  For purposes of scale, the San Jose MSA added the equivalent of half of the population of the Houma MSA and lost the equivalent of half Houma MSA’s workforce between 2002 and 2009.

Tying this in to education level is pretty silly — these are boom towns (oil and military), and likely with a high amount of young workers, early in their careers who didn’t bring spouses or children (extra, non-income-producing population, thus dampening the ratio).  This is like comparing the boom towns of the Western US with the established cities of the Eastern US during the Long Depression at the end of the 1800’s.

He also has some numbers for college graduates in Austin: they were 46% of the population in 2002, and 44% in 2009. So that doesn’t explain very much.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Your “historical anomaly” theory is a good one, but I’m not sure one should put full confidence in the idea that returns to higher education will be growing in the future; that returns to skill have increased over the past forty years might itself be an anomaly, and a regression to the mean would imply more of that. (My spidey sense tingles whenever anyone makes an observation that “such and such thing that one might reasonably expect, in absolute levels, to exceed other thing, has been declining relative to other thing” — e.g. “stocks of companies that were priced for bankruptcy in March 2009 have since outperformed stocks of companies that were not expected to go bankrupt”. Well, yeah. That’s not to say that returns to skill are negative, nor that financial weakness raises your stock price, only that the effects are less dramatic than they once were.)

Krugman wrote once — this was years ago — that as artificial intelligence improves, a lot of the “skills” are likely to be more amenable to replacement with capital than are things like “common sense” and social skills that are more relatively common among the less educated. This was a broad, several decade, not fully confident sort of prediction, and I would say based on vague impression that it hasn’t happened yet, but you could view this as the first stage of Krugman’s prediction coming true. At the moment, though, I’d bet on your explanation.

Posted by dWj | Report as abusive

Austin and Raleigh are very tech-heavy, so taking 2000 as your starting point will skew results. I was going to ask about San Jose, but it tops Mandel’s chart.

Look at the businesses that rode the housing/real-estate/credit bubble: construction, mortgage banking, realtors, MBS trading. How much specialized training does a loan originator at the shady lenders need? How much could they make, with the biggest paydays going to the least scrupulous?

This should not be a surprising statistic.

Posted by winstongator | Report as abusive

To me, the big downside to the knowledge economy is that a few geniuses can do everything, leaving lots of smart people hanging in the breeze. Google’s win created lots of losers including every search engine on the planet except Baidu, knowledge trades such as libraries and publishing and market research, the whole advertising sector and much of media including almost all newspapers and much television. All of these sectors now have massive gluts of workers.

Meanwhile, jobs in natural resources, building, and military remain very labor intensive and thus can absorb many people.

I would be in favor of several government-sponsored (but privately executed) Apollo-scale projects on different scientific problems of our time. If we are going to be printing and reflating, why not do something useful. This will also solve our education problem as kids will respond if money goes to the scientists and engineers. It is a national tragedy that Wall Street has drawn top physics graduates for so long while the ball has barely moved in physics in fifty years. The decades-long misadventure into string theory is not the central problem. New money is needed to make science and engineering as glamorous as it was tw decades ago.

Posted by DanHess | Report as abusive

I went to Manhattan, KS for college, and that instance of growth is at least partially driven by the expansion of nearby Fort Riley during that period.

Posted by gregbrown | Report as abusive

i am thinking we are seeing a decades long deflation of incomes. with a minor up tick in the mid 1990s. and a lot of the current deflation since 2000 is because jobs that required education were subject to being sent where they could hire much cheaper labor. and i like that idea of several Apollo like projects, the reason so many in physics and math went to wall street is really simple. not only do they pay well, but its where the jobs are. there are very few companies that do much in the way of research and development that need their skills and knowledge any more. most r&d in the private sector is only for projects that can show returns in a quarter or year at most. might have some impact on why we don’t have a lot more math and science grads

Posted by willid3 | Report as abusive