Thursday, June 3, 2010

Peak Oil Stress Map

The map above is a first rough cut at where the stress of peak oil (or any oil shock) is likely to be greatest.  It comes from taking county level data from the Census Quick Facts and extracting two variables: the average travel time to work (from the 2000 census), and the median household income (from 2008 data).  The idea is that if average travel time is long, that probably indicates that people in that county need a lot of oil to run their cars.  On the other hand, if income is low, they are probably going to have more trouble paying for that oil.  So I divided the travel time by median income, and then rescaled that index by its own average and standard deviation to produce a map of where the problems are likely to be greatest.

This is a fairly rough methodology.  In particular:
  • It's average, county-level, data, so this may miss a lot of neighborhood level details
  • It ignores other calls on people's income (like housing costs)
  • It assumes that work travel time is a decent indicator of overall household oil need (in particular, it misses north-eastern dependence on heating oil).
Still, the broad geographical results are fairly interesting, I think. It seems likely to be a robust result that the south-east of the country is the worst affected region, and since this is the main geographical base of the modern Republican party, that probably has interesting political implications.  Drill, baby, drill.

Looking at the individual variables, here's median household income per county (on a scale of white being $10k/year, with fully saturated purple being $90k/year):

And here's average travel time to work for the working population over 16, on a scale of white being 10 minutes, and fully saturated cyan being 40 minutes:

You can see that the problem with the south is that incomes are not that high, but travel times are generally long.  So when you take the ratio, that's where the stress shows up worst:

Finally, here's a 3-D visualization looking over the south-eastern region, with height of county representing population density:

All visualizations were done in Google Earth with KML files generated by some custom Perl hackery.


Unknown said...

Travel times are a function of travel distance and travel speed, the latter being heavily influenced by congestion. Commutes are these days a minor share of total travel.

Another way at looking at peak oil stress might be per capital vehicle miles traveled (VMT). Metro-level data on estimated VMT are available from the Federal Highway Administration, and are compiled by the Texas Transportation Institute in their periodic Urban Mobility Report. Not sure if County-level data are available, but the metros are County aggregations, so it might be possible to put something together. said...

Nicely done as always. a very creative mash-up.

As I look across the colors, and as I glance also at the income level map, my mind wanders to the affordability of Electric Vehicles in a time of overall, economic descent. Decline of net energy and the effect it has on overall activity presents quite the challenge to EV makers, who face their own power-density issue with batteries.

Thanks for the maps.


Lars-Eric Bjerke said...


I am impressed by the way you have presented statistics lately.

Just one comment on the consequences of peak oil. According to Global Energy Systems it takes the equivalent of 7.4 calories of energy to put 1 calorie of food on the table in the US. The world’s food requirement contains energy equivalent to 12 millions barrels of oil per day (mbd). If the US has 5 % of the population, the energy to put food on the table in the US is equivalent to 4.5 mbd. I guess most of that is oil. My conclusion is that peak oil will lead to high food prices. To cut down on food could be more difficult that to cut down on driving.

jewishfarmer said...

I think something that might change your map is the Northeast's heavy dependence on heating oil. Heat is not optional in the Northeast, and while only 8% of all households in the US heating oil, there are regions of the Northeast where the percentages are near 80%. During 2008, the situation was probably direr her that it was in places with high commuting costs - commuters can carpool - even alternative heating fuels involve high investments that are a real struggle for low income people to invest in. I think the weight of oil heat in the northeast may well be an important way the shocks play out.


KLR said...

When is that 8% figure from? U.S. Distillate Fuel Oil and Kerosene Sales for residential use declined in volume an average 5.06% YOY 2003-2008. The trend down seems to have started in 2005, after decades of oscillating around 6 billion gals/year.

Stuart Staniford said...


Agreed on all points. In general, I am constrained by what I can figure out in a few hours for a blog post. For example, the National Household Travel Survey has massive amounts of data that I'm sure would support a better map, but it requires downloading half a gig of text files and then figuring out an arcane format. At some point, I'll get on top of it, but it will take finding a day or so to figure it out. At the moment I've mastered making maps of county level census data and I'm figuring out what I can do with that.

A couple of other points: congestion breaks both ways, since it also causes a drop in fuel consumption. Also, I think the urban mobility report is strictly limited to cities so couldn't really be used for a national map.

Stuart Staniford said...


My current working hypothesis is that plug-in hybrids and electric vehicles will be primarily the adaptation of the wealthy and the upper-middle class (just as the Prius is, I think, primarily a creature of middle-class liberal guilt). I think that the lower half of the class spectrum will be more apt to adapt by moving from SUVs and trucks down to small cars, and from there to scooters and small motorbikes (basically reversing the trajectory by which developing countries motorize).

Stuart Staniford said...


I agree (and actually mentioned the heating oil caveat in the piece at the outset).

Stuart Staniford said...


You way overstate the sensitivity of US food production on oil. Oil costs are a relatively minor part of the cost structure of US agriculture (eg see this graph for corn.

That said, I agree that high oil prices are likely to cause high food prices, but I think the primary transmission channel is via increased biofuel production.

Daniel Daugherty said...

While you admit it's a rough look, I'm surprised there's no mention of the reliance on A/C in the Southeast and Southwest -- the so-called "New South." Cities like Charlotte, Atlanta, Dallas, Phoenix and Jacksonville have seen major growth over the last 20 years thanks to the widespread availability of A/C and relatively cheap square footage. People moved to these places from all over the country. When they can no longer afford the long commutes and heating/cooling expenses on their McMansions, the New South is going to revert back into what it was -- a sparsely populated cultural and economic backwater. If there's any good news, it's that Republicans will be stuck with them.

Lars-Eric Bjerke said...


I might be wrong, but all the references I have looked at shows that the food chain requires a lot of oil.

David Pimentel of Cornell University estimates that the United States devotes about 17 percent of its fossil fuel consumption to the production and consumption of food: 6 percent for crop and livestock production, 6 percent for processing and packaging, and 5 percent for distribution and cooking. This figure has increased year by year.

According to the UN's Food and Agricultural Organization (FAO), in the U.S., up to 20 percent of the country's fossil fuel consumption goes into the food chain, , and points out that fossil fuel use by the food systems in the developed world "often rivals that of automobiles".

You may argue that the energy portion of the cost of food is small, however the same goes for petrol, which still is a small portion of the total transportation costs.

Stuart Staniford said...


Right, but there are three fossil fuels, and oil is only one of them.

Stuart Staniford said...


A/C has nothing to do with oil - it's powered by electricity, which comes mainly from coal/NG/nuclear/hydro and hardly at all from oil.

That said, I agree that the south-east of the country is vulnerable to climate change and seems likely to get pretty intolerable in the summer as the century goes on. The changes will seem slow, however, to people living through them.

Datamunger said...

Many of those folks in the southeast, though, are paying less for gas as per this map from Gas Buddy.

Frymaster Speck said...

A FWIW, on the whole heating oil in the northeast, here's my piece entitled "Your Granny in Taunton, Frozen Like a Popsicle".

Love to see the map when you can factor this in.


Sixbears said...

Nice overview. Don't think those SUV's will be going away, at least the ones already on the road. Instead, they'll travel shorter distances with more people in them. Return of the car pool.

Factoring in things like heating oil makes the Northeast look worse. Including AC expense makes the south look absolutely terrible. Few people in the south know how to live without it.

Of course, the way the economy is going, there'll be darn few jobs to commute to in the first place

tim said...

A key factor which has to be taken into consideration is access to public transit. The Northeast Metro areas like New York, Philadelphia,
DC, Boston, Baltimore all have a major advantage by having more developed public transit systems available. Brookings just completed a major 2 year study on transit access in 100 US Metro areas and potential commuting time to work.
The mind boggling thing about this study is that they found 70% of working age were only 3/4th mile from a transit stop! The problem is of course the time it takes to commute on public transit due to infrequent service, bad connections, buses stuck in traffic making constant passenger stops, and the lack of a local/express model. But the potential is there!
Again Southeast is the worst....