Sunday, May 16th, 2010
The Brookings Institution just released a gigantic report called “The State of Metropolitan America.” This is a hugely ambitious and important study. Brookings examined changes in the characteristics of the top 100 metro areas in the United States from 2000-2008 across a wide range of domains with all sorts of interesting findings. The report is so overwhelming that I don’t think anyone can claim to have fully digested it yet. So my own comments should be seen as preliminary. Congrats to Brookings for this undertaking.
Increasing Urban Diversity
Among Brookings key findings is something I’ve been stressing for some time, namely how different metro areas are. This diversity only seems to be increasing over times. As they put it:
In some ways, large metropolitan areas actually became more different from each other in the 2000s.
The 100 largest metro areas span a wide range of social, demographic, and economic experience. Across the nine subject areas of this report, enormous differences separate the metropolitan areas with the highest and lowest rankings in 2008.
We’ll come back to the policy implications of this.
New Urban Typologies
Among the big splash items in the report – and clearly one Brookings is stressing – is their new typology of American metros. Before I describe that, it’s worth backing up to see the way cities were often grouped together in the past, namely regionally. Brookings did a report back in 2006 called “The Vital Center” which talked about the critical importance of the Great Lakes region. Here’s a graphic showing that study’s geographic construct:
It’s a clearly regional construct. It’s also obvious it is that this report was written by someone from Michigan.
Here’s how Brookings looks at cities today. Pay particular attention the classifications of cities in the Great Lakes area.
Note the many variations of urban areas within the Great Lakes. Before discussing the Brookings typology in detail, it’s worth looking at what they said about this particular matter:
The notion of a unified ‘Rust Belt’ stretching across large portions of the Northeast and Midwest overlooks the important factors that distinguish populations in Rochester, Cleveland, Indianapolis, and Chicago from one another.
Amen. Brookings is silent on what this report means for their previous construct, but it is notable that they used the term “Rust Belt” not “Great Lakes”. This particular excerpt makes it sound like they have decided to go in a very different direction.
Brookings characterizes metros into seven types based on three characteristics: population growth, educational attainment, and diversity. Their list is:
- Next Frontier – Scoring high on all three categories, these are the ones Brookings say are the most demographically advantaged. It includes places like Seattle, Denver, and the Texas Triangle, and are mostly in the West.
- New Heartland – Similar to Next Frontier but less diverse, including Portland, Columbus, and Charlotte.
- Diverse Giant – Slow growing, but educated and diverse regions, mostly made up of America’s Tier One cities like New York and Chicago.
- Border Growth – Areas mostly along the Mexican border with strong growth from immigrants, but low educational attainment
- Mid-Sized Magnets – Similar to border growth, but apparently growing from domestic migration, since they are less diverse.
- Skilled Anchor – Cities like Cincinnati and Pittsburgh that are educated, but growing slowly and without much diversity.
- Industrial Core – Classic Rust Belt type cities ranging from Cleveland to Birmingham with low growth, low educational attainment, and low diversity.
Like all classification schemes, this one has its questionable cases, but I think it broadly captures some important differences between cities. It also makes the very important point that cities ought to be viewed by their attributes as much as by their location. Birmingham and Memphis may be in the South, but they are every bit as much Rust Belt cities as their Midwest brethren.
Now as it turns out, most of these classifications are geographically clustered. Only Diverse Giant and New Heartland are really a national phenomenon. Still, the Midwest area does have a diversity of places.
I was particularly struck by how Brookings Midwest classifications matched almost exactly my classification from a year ago into Global Cities, Successful Cities, Stable Cities, and Struggling Cities. Here’s the cross mapping:
|Chicago||Diverse Giant||Global City|
|Des Moines||New Heartland||Successful|
|Kansas City||New Heartland||Successful|
|Minneapolis-St. Paul||New Heartland||Successful|
|Grand Rapids||Industrial Core||Stable|
|St. Louis||Skilled Anchor||Stable|
As you can see, we only disagree on two cities. And if I must say so myself, I think my classification is more accurate for those two. Unlike Brookings, I used a single variable – population growth – with thresholds for bucketing. But I did cheat by putting Chicago in its own category arbitrarily. So Brookings has me there. Chicago’s population growth would have put it as “Stable” in my schema.
The only difference between Skilled Anchor and Diverse Giant is diversity. But I don’t think that’s what really separates a Chicago or New York from St. Louis or Milwaukee. The real difference is their economic structure. The Diverse Giant group is actually mostly so-called “global cities.” They’ve got a skilled core of very high value services that is going well surrounded by a large zone that is at best hanging in there. Other than perhaps in Miami, the diversity angle seems to be a legacy of their status as port of entry more than anything. If we looked at this economically, you’d probably boot Honolulu (an outlier in any US scheme), and add Boston. But of course, as I said, any classification done by metrics generates border cases.
Local Policy Implications
Let’s turn to the policy implications of this, first on the local side. Does this mean that the idea of traditional regional policy is dead? Jim Russell seems to think so. I’m not so sure.
I think there’s a difference between typologies of cities and how you integrate economically. Places in the New Heartland class, for example, might have much to learn from each other, might benchmark against each other, etc., but it doesn’t necessarily go beyond that. And while cities in a region may be very different, that doesn’t mean they can’t collaborate or economically integrate. Indeed, being different helps there through better enabling specialization.
I’m basically skeptical of the mega-regional concept, but can appreciate the virtues of both approaches. For example, Longworth’s pan-Midwest cooperation approach could be very useful in areas like creating a high speed rail network or promoting better collaboration between Big Ten universities. And it seems reasonable that regional relationships with Chicago would make a lot of sense. But Russell is right that so long as the world sees the Midwest region as a region, as the “Rust Belt”, it cripples these places. The world needs to understand that Columbus isn’t Toledo. Right now only Chicago, because of its longstanding huge size and history as having an independent standing in the world, can punch through the scarlet letter that is the Rust Belt label. Columbus can’t do that easily. In a sense, to start recovering, we have to blow up the idea – externally at least – that everyplace in the Midwest shares the same meta-narrative. We also have to recognize, as Russell stresses, that economic networks are now global in nature, and there are all sorts of diverse circuits in which cities can participate.
Federal Policy Implications
As a DC think tank, Brookings obviously stresses federal policy. In this case, they promote a bifurcated view: a federal policy that deals with common themes and issue, and state and local policy that deals with the different and granular issues.
National policy makers have the unique obligation to address aspects of the five new realities that affect all metropolitan areas, or are simply beyond metropolitan area’s own capacity to tackle. As this report demonstrates, however, different challenges assume varying levels of prominence in different types of metropolitan areas. Leaders at the state, regional, and local levels must now more than ever understand and respond purposefully to the demographic, social, and economic changes most affecting their places.
National policy will be necessary, but not sufficient, for addressing the wide range of challenges facing metropolitan areas. Indeed, the increasingly distinct profiles of major metro areas along the key dimensions outlined in this report demand that their own agendas – at the state, regional, and local levels – confront the issues most pressing to their own futures.
The Brookings five new realities are Growth and Outward Expansion, Population Diversification, Aging of the Population, Uneven Higher Educational Attainment, and Income Polarization.
I recently wrote a piece on a federal policy for cities in which I said that federal policy itself must address the increasingly unique needs of metros. I’m not sure that we can just keep the common elements at the national level punt the distinct items to the state and local level. Generally I’m a fan of devolution. However, in this case what we are really talking about is devolving de facto to the states. And as we’ve seen repeatedly, state governments are to varying degrees implicitly or explicitly hostile to cities. There are almost no effective or empowered regional entities in America and municipalities are often hamstrung by state laws that limit their ability to deal with their own problems. In a choice between federal vs. state government, I’m usually a state and local kind of guy, but when it comes to metro policy, I’d probably have to say the feds are more likely to get it right, alas.
Beyond that, the new geography Brookings highlights goes a long way towards explaining why we’re gridlocked on so much policy at the national level and why we’ve got to grasp the nettle of urban/sub-national diversity at the federal level. The problem isn’t just one of partisan gridlock, it is that these different metros and their states really do live in different worlds and as a result have very different policy points of view. Back in the Eisenhower era, pretty much everybody – even big city mayors – wanted freeways. That created policy consensus. Today, there are not only philosophical disagreements, there are also legitimate differences of character and need between these metro areas. I don’t think we’ve fully grasped the implications of this. Along with things like income polarization, what’s happening is that the American commonwealth is tearing apart. Our fortunes are no longer as linked as they once were. A rising tide won’t necessarily lift all boats, and what’s good for thee is not necessarily good for me.
This only promises to get worse over time. One of the incredible stats in this report is that less than 25% of children under 18 in Los Angeles are white. In a generation or so Los Angeles will be a Latin American city in character. This will add an ethnic dimension to the problem. One doesn’t have to scream “racism” to see this. Just look at ultra-politically correct Europe, where national-ethnic undertones are a big part of the debate over the bailout of Greece. The US is becoming like the EU in a way few people talk about, in that it is culturally separating. That’s one reason a federal bailout of California is unlikely outside of some executive fiat. And it is only going to get worse.
I won’t pretend I’ve got the answers, but bigtime thinking needs to go into how we find a framework to deal with this increasing diversity – and I’m not just talking about ethnic diversity. That’s one of the keys to breaking the gridlock on things like transport policy. Yes there are partisan differences, but it is more than just that.
Perhaps some of the answers are buried in our antebellum history, when there was much more state and regional than national allegiance. I don’t know. Perhaps the dominance of slavery as an issue in that era makes it less useful. Or maybe we can learn from the EU itself, though that model has been heavily criticized as undemocratic and it isn’t working too well at present. In any case, I do believe we need a federal policy around cities that addresses areas where there is divergence, and I believe we’ve got to find policy frameworks to make it actionable.
Lest I overstate the case on Brookings policy split, I should note that they did also write, “National policy responses must recognize the diverse starting points of metropolitan areas and, where necessary, ensure that interventions are tailored to those differing on-the-ground realities.” I would love to see them do a lot more research and thinking along those lines.
Highlights of the Report
I would like to highlight a few of the interesting facts that come out of the report. Again, this doesn’t even scratch the surface.
I wrote previously about how immigrants are revitalizing inner suburban areas, even in the Heartland. This report provides more evidence of that. Indianapolis was the #6 city in the entire US for both increase in Hispanic population and increase in Asian population. Columbus was #9 on both metrics. Now that’s on a low base to be sure, but there are tons of other places with low bases that didn’t see this growth. Cleveland, for example, ranked 95 out of 100 in its change in percentage of foreign born population and the percentage of foreign born residents in the city itself actually went down. This also demonstrates that some heartland cities aren’t just seeing the nearly ubiquitous Mexican immigration, but immigration from many different places.
It gets even better. Pittsburgh and Indianapolis both increased their college degree attainment by a robust 5.3 percentage points. That made them the #3 and #4 metros in the entire country for growth in the degreed. Just for the record, that’s called “brain gain.”
Another interesting metric was the change in under 18 population. This to me is a huge measure of demographic health. Those children are your city’s future. If you don’t have them, you’re having a going out of business sale. Alas, many Rust Belt cities experienced large declines in children.
That’s serious. Some have claimed that the Pittsburgh demographics are skewed by a high elderly population. Perhaps that’s true, but the decline in children shows that Pittsburgh is far from demographically healthy. Again, there is quite the regional contrast. Some regional cities had the opposite situation. Indy’s child count grew by 12% compared with a national average of only 2.5%. Interestingly, Chicago became the only Midwest city to become “majority minority” in its population under age 18. Here’s a broader national view of the percentage of households who are families with children:
Reversing the Great Migration, blacks continued to move to the South. Atlanta added 445,578 blacks, reinforcing its position as the capital of black America. The next nearest competitors were Dallas and Houston who both added in the 100,000’s. That shows the overwhelming locational preference for Atlanta, which passed Chicago to have the nation’s second largest black population after New York City.
Lastly, Minneapolis-St. Paul was #7 in America in the worsening of income inequality. Holy Scandinavian scandal, Batman!
There are tons more interesting facts where these came from. You’d be well served to read the report for yourself, and look at the fact sheets on your city available for download from the Brookings website.