Nonprofit, nonpartisan journalism. Supported by readers.


Where the jobs are, and aren’t

Welcome to Public Numbers, wherein a variety of journalists will write Current Posts about numbers and public life — what we can learn from these numbers, and how to know when numbers are being misused. Joel Kramer, editor and CEO, is writing today’s entry, and you can look for him to write more.  But others will join in when the numbers move them, and we hope our readers will contribute to the conversation, too.

Where the jobs are, and aren’t

By Joel Kramer

Mike Meyers’ excellent piece in Monday’s Star Tribune about Minnesota’s dismal job growth is a vivid reminder of why we need journalists telling us what the government’s numbers mean, rather than relying just on the government’s own communications.

When the state Department of Employment and Economic Development released its last employment numbersin mid-October, which showed the state actually losing 6,300 jobs in September, here’s what DEED commisioner Dan McElroy had to say in the department’s press release:

“Because monthly numbers fluctuate, job creation and employment trends are best viewed over a longer timeframe. For example, Minnesota employers have added 9,800 jobs in the last year and 100,683 jobs over the last four years.”

McElroy’s assertion was technically accurate but misleading, because he did not compare the results to employment growth nationwide.  So, with the help of fiscal guru Jeff Van Wychen, I prepared this chart last week for publication in MinnPost today.



Over the five-year period ending in September 2007, employment in the United States grew more than 4 ½ times faster than in Minnesota.   In the previous five years, ending September 2002,  the U.S. very slightly outpaced Minnesota.  And in the five years before that, ending September 1997, Minnesota significantly outperformed the nation.

How much do these numbers matter?  If Minnesota had increased employment over the past five years at the national rate, we would have 140,000 more jobs than we now have.

Meyers beat me to the punch Monday morning with a similar analysis.  What’s interesting is that he asked the department to reanalyze the numbers for him, and they did — producing a kind of truth the state agency had not revealed before.

Why Minnesota is doing relatively poorly is a complex question. But one observation by Federal Reserve economist Toby Madden is worth noting.  Madden points out in Meyers’ story that population growth is a factor in job growth, and in the 1990s the state’s population growth was only 6 percent below the nation’s while more recently it’s 30 to 50 percent lower.

Of course, jobs can create population growth just as much as population growth can create jobs.  But this brings up a larger point — employment growth is not necessarily the best measure of the health of an economy. New jobs that pay poorly and don’t offer benefits, for example, can actually lower the average prosperity of the people who were in the state before those jobs arrived.

A better measure of prosperity is per capita income, and the trend is measured by per capita income growth. Minnesota remains above average in per capita income, but our recent performance is just as dismal as it is on the employment front.

In 2006, Minnesota ranked 46th among the states in per capita income growth, and we were below average in 2005 as well. As a result, we’ve dropped out of the top 10 most prosperous states. Since this is a per capita number, our population growth relative to the nation is not an explanation for this deteriorating performance.  If you think you can shed light on what does explain it, let me know.

Do you have an example of how numbers can inform us about life in Minnesota or the nation? Or an example of how numbers are being used to misinform us? Email us at numbers [at] minnpost [dot] com; please be specific and include links to the data and its uses. Links to charts and maps are great, too.

Comments (3)

  1. Submitted by Kathy Hansen on 11/09/2007 - 12:09 pm.

    The beta version of a great new “data” tool might be of interest. You can find it at Another amazing new resource is the US Census Burearu’s “DataFerrett” at


  2. Submitted by Bill Gleason on 11/10/2007 - 09:29 am.

    Excellent example of how to use numbers to give a misleading impression.

    One of your advertisers (the U of M) has done something similar recently. While bragging about improvements at the U, the conveniently neglect to point out that we are near the bottom by these same criteria when compared to a self-selected peer group.

    It is time to drop the “ambitious aspirations” of becoming “one of the top three public research universities in the world.” We should be aiming for a place in the middle of the BigTen (our real competitors) which will be ambitious aspiration enough given the current sad state of affairs.

  3. Submitted by Gerald Abrahamson on 11/10/2007 - 02:45 pm.

    An analysis of the state figures will show if current figures are, in fact, in line with historic trends–or if there is a variance. The federal unemployment rate shows at 4.7% (as of this date), but a long-term analysis shows the real unemployment rate to be about 8.0%. This is using BLS “actual” figures–not some numbers from weird sources.

Leave a Reply