Moving

While I am keeping this blog and will continue to use it for more experimental projects and opinions, I have recently joined UVA's Weldon Cooper Center for Public Service, and will be largely migrating over to its Stat Chat blog.  You may know them as the source of Virginia's population projections and the creator of the Racial Dot Map.  See my first post on the movements of millennials and their spatial gender gap:

What are the young people up to these days?

Age and Place

I want to start what may become a series of several posts looking at the movement of people throughout their lifetimes.  I'll leave you with this graphic for now - which shows the relative concentration of each age group by census tract.  Red means the age group is overrepresented, blue means it is underrepresented, yellow means it is proportionally represented.

There is, as you would expect, a drastic shift around age 20 to college towns, followed by a shift to central urban areas, and then a long dispersion out to suburban counties and rural areas.  The cycle for children, unsurprisingly, mirrors the cycle for those age 30 to 50ish.

Ten Walmarts (and other stores) that are larger than their downtowns

 1. Downtown Waynesboro and Walmart

I once witnessed a conversation between Waynesboro, VA's Director of Planning Michael Barnes and a citizen.  The citizen asserted that at Walmart, there was always lots of parking right in front of the store, whereas you had to park and walk if you went downtown.  He responded something to the effect of: "I bet if you drew a line from the middle of the Walmart parking lot to the door, walked around the store, checked out, and walked back to the car, you'd have gone farther than if you parked in Constitution Park lot [a parking lot in downtown Waynesboro that is never full] and visited every store on the 300 block of Main Street."

He was absolutely correct.  The sense we have of space while in a vehicle is radically different from the sense of space we have as a pedestrian.  Walkable places have an entirely different scale - a density of connections and destinations that provide a full city experience within a very short physical distance - often a quarter to a half mile (or a five to ten minute walk).  Automobile-oriented places provide the same grid of connections and destinations at a dramatically expanded scale.  The time distance is still the same - everything is accessible in five to ten minutes.  But that means something different in terms of space - more like five miles than a half mile.

There's no particular political agenda behind this post except to illustrate the difference in scale that exists between places that were meant to be experienced on foot and places meant to be experienced by car.  The stores that are superimposed over the main street blocks are at the exact same scale.

2. Staunton (with Walmart and Lowes)

 

3. Front Royal (Walmart and Lowes)

 

4. Winchester's pedestrian mall (with both Winchester Walmarts)

 

5. Harrisonburg's Court Square (with the townie Walmart)

 

6. Warrenton (Walmart and Home Depot)

 

7. Danville

8. Leesburg (Target, Costco, and Dockers)

 

9. Norfolk's Granby Street (you can also see Scope and MacArthur Mall), with the Battlefield Walmart, Sam's Club, and Garden Ridge - I almost stuck Kohl's in too because its parking lot is so huge when put together with the others.

 

10. Culpeper - Davis and Main Streets at the Amtrak station.

Dollar Density

Here's a quick reprise of the 3D density map from a while back.  Rather than each person taking up the same amount of 3-dimensional space, each dollar of household income takes up the same amount of space.  The color of the census tract indicates the median household income, with red being lower and green being higher.  So red tracts that are tall have lots of dollars, but a low median income (aka a high density of people that those dollars are divided among).

Excellent comment from Jim Bacon at baconsrebellion.com:

"The map vividly portrays the overwhelming economic dominance of the Chesapeake Crescent in Virginia’s economy — and Northern Virginia’s dominance within the crescent."

Paving Paradise

How much space does a car take up exactly?  The answer, of course, is that it depends - on the design of the place, the type of driving going on, the density, the tendency of the population to build new lanes and parking lots everywhere, etc.  The answer is important because people so frequently think of sprawl in terms of houses and house size, which is really fairly unimportant until you get into really dense configurations.  It's the automobiles that are taking up the space.  This is important to understand.

Cars are larger than houses* - at least in terms of their urban footprint. [*at the densities most Virginia municipalities are dealing with]

I think a helpful metric every county should know about itself is its "Pavement Per Vehicle" and/or "Pavement Per Capita" (either works since there is about one car per adult in Virginia).  Most counties have the data to calculate this pretty easily and I'd love to do a state index, but unfortunately most counties do not provide that data free of charge.

Three do, however: Charlottesville City, Albemarle County, and Richmond City.  Charlottesville and Richmond are two of the most compact cities in the state and both have lots of college students who tend to own cars but not drive them.  Their numbers should be similar and at the low end of the spectrum.  Albemarle is a largely suburban and exurban county with a population density close to that of the state as a whole.  It is fairly affluent, has major state and interstate highways, and should have a number at the higher end (though I think there will be more counties closer to Albemarle than closer to Charlottesville).

So here's the question: how many square feet of pavement dedicated to automobiles (parking spaces, driveways, roads, etc. - not including sidewalks or pedestrian trails) does the city or county have per vehicle?

The data on paved surface comes in all three cases from the municipalities - and props to them for making their data freely available.  Charlottesville boasts over 1100 acres - nearly 2 square miles - of automobile pavement.  Richmond clocks in with nearly ten square miles of it and Albemarle has over 14 square miles altogether.  The data on number of vehicles and residents comes from the 2009 NHTS and the Census, respectively.

Remember that each house today has an average of two to three vehicles if you're thinking about the paved space taken up by each new home.  I used ODU's lovely new basketball court to help you get a better idea of the scale.

Why should we care at all about this number?  Well paving places over carries a number of costs.

I. Sprawl

This one is too complicated to hash out in a post like this, but space is a finite commodity and large open spaces in cities disrupt the economy of density that makes cities economic engines.  Creating more auto space has a snowball effect in encouraging people to spread out more, creating the need for more lanes and more parking.  If sprawl isn't something you care about, just skip this one for now.

II. Real cost

How much did it actually cost to lay that pavement?  The answer is tough to estimate, but some back of the napkin calculations based on VDOT's averages and some standard development numbers put it somewhere around $18,000 per vehicle for Charlottesville and $45,000 per vehicle for Albemarle County.  The lifetime of that pavement is also a thorny question - for most pavement it's 20 to 30 years.  Annualized (without interest), we are looking at about $500/year per vehicle for Charlottesville City and $2,000/year per vehicle for Albemarle County.  I think it's actually quite a bit more thanks to the number of highways with large axle loads that require dramatically more maintenance (my numbers are based on parking lots), but we'll go with the lower, very round figures - the pavement to drive each vehicle is costing somewhere between $500 and $2000 a year - a number you should think about being tacked onto your other yearly maintenance and fees.  I'd love to hear a better estimate of this from someone at VDOT or a county public works department.

Who pays for this construction?  The answer is, of course, you.  But how?

1. Directly - in the case of private driveways, which make up a very low percentage of these numbers.

2. Through the cost of goods and services.  Stores pay to build parking lots, those costs become part of their overhead, and are then passed onto you in the cost of the item.  This accounts for about half of the pavement.

3. Property Tax.  Local roads and other infrastructure are often maintained by the city or county, which uses property tax revenues to pay for it.

4. The Gas Tax.  Theoretically covers major roads and highways.  It represents the most direct connection between the users of the pavement and the cost of it - making it the closest thing to a user fee rather than a tax.  However...

5. Sales and Income Tax.  In Virginia, legislators were unwilling to raise the state's absurdly low gas tax, despite the fact that it could no longer pay for the upkeep and construction of roads.  So instead... they lowered it some more!  And then made up the difference with an increase in the sales tax and money from the general fund.  This means people who drive less and live in more compact places directly subsidize those who drive a lot.

III. Environmental Cost

The added pavement is one of several pollution sources that is killing the Chesapeake Bay.  While it's not the largest source of pollution (agriculture is), it's the only source of bay pollution that is increasing every year.  Rain falls on impermeable surfaces, picks up chemicals and sediments, and runs off into the bay, creating a giant dead zone in the middle where life is unable to survive, and blocking the sunlight that helps the bay's abundant sea grasses to grow.  If you aren't familiar by now with the trajectory of the bay and the cultures that used to live off of it, it might be a good time to go read up.

The collapse of the oyster population - an economic engine for communities along the bay, which once provided over half the world's oysters.

 

http://www.cbf.org/how-we-save-the-bay/issues/polluted-runoff

Up until now, this has been just another unquantifiable cost to feel bad about, but the EPA is buckling down on stormwater and making municipalities take real steps to curb it, which means these are now becoming real financial costs, funded in many places by a stormwater utility fee assessed based on the paved surface on each parcel.

Additionally, paving places over makes them ugly and often ruins them for future agricultural use.  It also has implications for wildlife and other natural systems.  And it creates more air quality problems, though all of these will have to be the subjects of later posts.


Future Growth

I've mentioned before that the Weldon Cooper Center has us adding around 2.5 million people in the next 25 years.  If that adds a proportional number of cars to places that are somewhere in between being Charlottesville and being Albemarle (say 4000 square feet per vehicle - I bet that's about Virginia Beach's number), we could be looking at paving another 280 square miles of the Commonwealth.  That's an area the size of Goochland County.  Sweet dreams.

Income Distribution

More people are becoming aware of the spatial segregation of wealth these days - from liberal commentators on concentrated poverty to the oft-maligned Charles Murray and his "super-zips."  But while we talk frequently about median household income, we don't seem to look very often at the distribution of wealth at a county or census tract level.

The ACS (I'm largely using the 2012 five-year estimates here) has this data - with households divided into ten income brackets - but it doesn't get visualized frequently because it's hard to show ten quantities on a map.  So I wanted to calculate a standard deviation for wealth in each county and find census tracts that had high and low incomes with low variance.

I used family income because it better represents actual wealth and poverty - household and individual incomes include too many college students, young professionals with low incomes but few expenses, etc.  The problem with doing a standard deviation here is that the difference between making $25,000 a year and making $75,000 a year is a lot bigger than the difference between making $150,000 a year and $200,000.  So instead of using actual dollar numbers, I used the census bureau's ten categories - weighted one through ten.  Those categories - and the statewide distribution of families among them - can be seen in this histogram:



So a standard deviation of "2" means the data tends to be distributed two income brackets in either direction.

On this map, I've pulled out the census tracts that have a standard deviation below 1.85 and 2.1, representing relatively little variation.  In green are the tracts with a median family income over $80,000.  In red are the tracts with a median family income under $50,000.  There is a clear trend - suburban areas around the major metro areas are where wealth is concentrated, while inner city areas, as well as rural southwest and southside Virginia are more likely to have more concentrated poverty.  Tracts in gray have either a more typical median income or more spread-out distribution.

I've also mapped the standard deviation by county.  Blue counties have a narrower distribution of incomes, while red counties have more variation.  As you might expect, cities lean towards a wider range of incomes, while counties tend to have more homogeneity.  But those cities and counties at the top and bottom have a wide range of median family incomes.

Here are the top and bottom counties for variance (citizens in a wide range of income brackets):

One last thing piqued my interest.  Which counties have an income distribution that most resembles the statewide income spread?  Refer to the first graph in this post for the statewide numbers.

I calculated the variance for each county and came up with the following map:

 

If your county is darker, it means the income distribution of the state is well-represented.  If lighter, it means there is a significant difference between the state and the county.  Henrico County and Charlottesville City take the top spots as being most representative.   High on the list are Manassas, Virginia Beach, Albemarle, and Chesterfield Counties.  Charlottesville would be much different if I was using individual income (or even household income) thanks to the high number of university students and young professionals.

Here's Henrico and Charlottesville's histograms alongside the Commonwealth's.  Virginia is in blue, Henrico is in red, and Charlottesville is in green.:

 Bringing up the rear are Falls Church and Emporia, with Grayson County, Arlington, Loudoun County, Patrick County, Martinsville, and others close behind.  Here's Falls Church's and Emporia's incomes next to the state's.  The state is in blue still.  I don't think I need to tell you which city is which...