Demographic Trends and Traffic Projections

Allen E. Pisarski

James A. Bacon

Back in October, Northern Virginia commuting guru Alan E. Pisarski updated the Washington Metro board of directors on the latest trends in commuting behavior. He made a number of important points that should temper the agitation of those who believe that Virginia’s under-funded transportation system is doomed to be overwhelmed by increasing levels of traffic.

(The following commentary is based upon data appearing in Pisarski’s PowerPoint presentation, which was shared with me. I did not hear the  presentation. The analysis is mine.)

For starters, the national average travel time did not get worse over the past 10 years. Despite significant population growth, the average commute went from 25.5 minutes in 2000 to 25.3 minutes in 2010.That bears repeating: By this basic measure, commuting got no worse. Many people thought it would, but it didn’t.

Furthermore, many of the demographic forces that propelled an increase in the number of Vehicle Miles Traveled in past decades are largely (though not completely) spent:

  • Population growth is slowing
  • Household formation is slowing
  • Labor force growth is slowing
  • Migration from state to state is slowing
  • Growth in driver’s licenses is saturated
  • Growth in car ownership is saturated

I have examine all of these trends before except one — the slowing growth in the labor force. But this is fundamental. As Baby Boomers retires, succeeding generations are barely large enough to replace them. The workforce will continue to grow, but at an exceedingly slow rate. Why does that matter? Because workers account for peak traffic demand when they commute to work.

If we combine these demographic trends with the rising cost of automobile ownership, there is no reason to expect Vehicle Miles Traveled to increase nationally at anywhere near the pace that it has in previous decades. Bottom line: Throw out all national traffic projections based upon the extrapolation of past trends.

Admittedly, metro-level traffic projections are a different story. Population and economic growth in Northern Virginia, in particular, has outpaced the national averages, and it may well continue to do so. But that’s not the sure thing it seemed to be a few  years ago. NoVa’s economy grew in tandem with federal spending. At some point, federal spending will level off. It may even decline. Indeed, if you believe my Boomergeddon thesis, it will crash. While it’s possible that NoVa will reinvent itself, as, say Silicon Valley has, there is a degree of uncertainty and risk that did not exist before.

Pisarski presented one other data set that should make Virginia’s policy makers perk up and take notice. Virginians are more likely than commuters in any other state in the country to leave their county of residence to go to work: 51.3% compared to a national average of 27.4%. That’s really extraordinary — and it’s an indictment of our collective failure to build communities with a balance of housing and jobs.

Share this article


(comments below)


(comments below)


10 responses to “Demographic Trends and Traffic Projections”

  1. Quoting national averages for commute time as part of an analysis of NoVa traffic is like quoting national average temperature as part of an analysis for NoVa weather.

    Yes, I did read your disclaimer about national trends and local trends.

    The county statistic is badly flawed too. In Virginia, counties do not include cities. So, I have to leave the county of Henrico to go to my job in the City of Richmond. In other states, like Illinois, counties encompass cities. So, I can live in Schamburg, IL and commute 30 miles to downtown Chicago without ever leaving Cook County. Of the 42 independent cities in the United States, 39 are in Virginia (Baltimore, St Louis and Carson City, NV are the other three).

    Please refer Mr. Pisarski to Ed Risse’s web site for a review of core confusing words.

    If that doesn’t help, perhaps this list of cities within the County of Los Angeles might provide a clue:

    More people live in Los Angeles County than entire state of Virginia. I would bet that very few of them cross a county boundary to work. I assume that this fact would help prove that Los Angeles has no traffic problems based on your “county – transit” statistical analysis.

    The size of counties also counts. The average number of counties per state is 62. Virginia has 95. California is four times larger than Virginia. It has 4.5 times the population of Virginia. And it has 58 counties as opposed to Virginia’s 95.

    Given that Virginia is, as usual, a bizarre outlier – the only way to peoperly measure the number of municipalities is by county and county equivalents. Virginia has 95 counties and 39 independent cities for a total of 134 counties and county equivalents. That places Virginia third on the list after Texas and Georgia.

    An additional problem with Virginia’s governance philosophy of small counties and lots of independent cities is that most of the local jurisdictions are small in population and budget. Therefore, it would be something of a struggle for these jurisdictions to pay for the overhead of a roads department and other associated locality-operated government efforts.

    In summary, years of odd thinking by the Clown Show in Richmond has rendered Virginia with a governance structure that is both unique and uniquely backwards. Using statistics such as “county line commute” crossings is either mistaken or disingenuous.

  2. methinks Mr. Groveton has forgotten the concept of an MPO which includes the metro NOVA region and does studies and traffic modelling and projections for the region – as well as maintaining a fiscally-constrained construction plan – actually 2 of them – a short range and a long range plan.

    Groveton is sounding more and more like he has lots of blame AND ….LOTS of EXCUSES as to why poor old NoVa has no options to help itself so he blames the Clown Show.

    I’d say a Clown Show is a region that plans a METRO that is god awful expensive… tolls the hell out of commuters on one road… begs money from the Feds and the State ..and blames Rova and Richmond…. for the problems.

    tsk tsk

    I think NoVa should GROW UP…and take some responsibility … instead of whining and blaming others…

  3. Groveton, You caught me — I may have used the commute-outside-the-county figures carelessly, even though I know full well that Virginia’s system of local government is unique and any comparisons are very tricky.

    It is interesting to note, however, that Maryland ranks No. 2 in the country, according to Pisarski, with 47% of workers commuting to work in a different county. Maryland’s system of government is not like ours.

    The numbers may be meaningful, but they may not. Without further analysis, I will retract my statement that the numbers represent “an indictment of our collective failure to build communities with a balance of housing and jobs.” I simply don’t know enough to be able to back up that claim.

  4. LarryG:

    I am sorry that you were not taught statistics in whatever school system you attended. That’s unfortunate. As a graduate of Fairfax County Public Schools, I was taught statistics.

    For example, I was taught that a comparison of means between two populations is only useful if both populations fall along the same (or similar) probability curves.

    For example, if I write “1” on one side of a fair coin and “2” on the other what are the odds of me tossing a “2” over large number of tosses? 1 out of 2? 50%. Now, what are the odds of me rolling a two on a fair die? Also 50%? No? Why? Because a coin has two sides and a standard die has six? Very good.

    Now, back to Bacon …

    “Virginians are more likely than commuters in any other state in the country to leave their county of residence to go to work: 51.3% compared to a national average of 27.4%. That’s really extraordinary — and it’s an indictment of our collective failure to build communities with a balance of housing and jobs.”.

    The only way this comparison makes sense is if Virginia falls along the same probability curve as the other 49 states. If that were true, then a comparison between the probability of a commuter crossing a county line on the way to work in Virginia could be compared to the probability of a commuter crossing a county line on the way to work in other states.

    Now, let’s start easy here. If Virginia had only one county for the entire state – would it be fair to compare county crossings in Virginia to the other states? Obviously not.

    What if Virginia had a county for every zip code in the state? Would that make the comparison fair? Obviously not.

    So, what are the facts? Is Virginia sufficiently similar in county configuration to the other 49 states so that county crossings is a valid comparative statistic?

    I say no.

    As I pointed out, Virginia has a near national monopoly on independent cities. Those 39 independent cities would be part of a bigger county in virtually every other US state. So, the odds of crossing a county line to go to work in Virginia are wildly distorted by the fact that 39 of America’s 42 independent cities are in Virginia.

    Are you following this, LarryG?

    Now, what about the size of counties? Wouldn’t you think that the odds of crossing a county boundary on the way to work would go up as the size of counties goes down? If NoVa were one county – what would be different with regard to traffic jams, commuting times, human population settlements, etc? Nothing. But what would change with regard to the county crossing statistic? It would fall dramatically. When it comes to county size and county crossing statistics, size does matter.

    Again, Virginia is an outlier. Virginia is the 35th largest state by land area but it has the third most municipalities. Even if we disregard the independent cities and only count the counties, Virginia would be tied for 10th most counties.

    LarryG – did you ever play that game, “Step on a crack, break your Momma’s back?”. It was pretty easy to avoid hurting dear old Mom when you were on a standard sidewalk with big slabs of concrete. Much harder when you you were on a brick sidewalk. In fact, damn near impossible to avoid stepping on cracks when the pieces are so little. Kind of like being damn near impossible to get from anywhere to anywhere without crossing a county line when the state you are in is full of little counties.

    Now, back to the actual point …

    Do you think that comparing county-crossings between Virginia and the nation as a whole could lead a reasonable person to declare:

    ” it’s an indictment of our collective failure to build communities with a balance of housing and jobs.”?

  5. re: indictment

    well.. if NoVa were unique in the country in terms of an urban area with density and jobs and a substantial commuting workforce… I might buy it but virtually every urbanized area in the country – the metro area – regardless of county or even state boundaries… has twice daily rush hours of commuters coming in from exurbia and leaving in the afternoon for exurbia.

    if you had a map of the US with all the urban areas colored green for non-rush and and let them go red for rush hour – they’d all light up …. first in the East and then cascade to the west coast…. in the morning…

    and then in the evening… another cross-country cascade.

    Why Groveton thinks the clown show in Richmond is responsible for NoVa’s very common problem… is… well.. it’s not statstics.. it’s simple recognition of the realities.

    and I know of no urban area where building more roads has erased rush-hour rash….

    the problem with urban areas is …area-wide network congestion and building new roads also diverts it – moves it to the next bottleneck.

    if we are really serious about the idea of reducing congestion in NoVa – we need to model it as a network and see what changes would actually result in overall less network congestion and how much it would cost.

    right now… we’re in the dumb zone… build something… whether it actually results in improved performance… just build it.

    this is why I advocate for every new road – a mandatory investor-grade analysis of how much people are actually willing to pay to use it – even if it is decided it won’t be tolled.

    determining “need” based on how much people are willing to pay is a much better approach to understanding real need.

  6. actually more than investor grade studies …

    network wide analysis… so that if a new road is proposed to reduce congestion – we find it if it is really reducing congestion or just moving it somewhere else.

    Often what new roads do is just provide more places for cars to wait for the restricted bottlenecks …..

    that’s a fundamental difference with rail transit which basically restricts the rail network – to what the rail cars themselves can carry but they also benefit from being able to add cars to carry many more people instead of one new car for each new passenger.

    what HOT Lanes SHOULD DO is ALLOW some version of Bus Rapid Transit since they will have priority access to the less/least congested roads – at least en-route.

  7. Arlington county is working on their mix of jobs and housing. They plan to buy an office bilding in the high rent Ballston area for $25.5 million, taking it by eminent domain if necessary, and then convert it to a homeless shelter.

    I can practically guarantee Larry that HOT lanes will be managed to maximize revenue, and not traffic throughput.

    “but they also benefit from being able to add cars to carry many more people instead of one new car for each new passenger.”

    Wrong. Both VRE and metro are limited by the length of the statio platforms. And the Escalator capacity, because passengers jam up on the platform after exiting the trains. Even if more cars wee possible, more passengers take more time to unload, and that increases the lag teime necessary between trains. Besides that, most cars are travelling way less than full almost all the time. Virtually every auto that is moving has a load ratio of sat least 25%, which is probably better than the average load ratio for Metro.

    It is no more necessary to add one auto for each passenger than it is for Metro to add one car for each passenger. Give drivers the same kind of subsidy and retirement plan that Metro has, and you could fill up those empty auto seats in a hurry. Incentives matter.

  8. we find it if it is really reducing congestion or just moving it somewhere else.
    What else is new?
    Consider a small town with a factory. Even if the streets are never crowded, there will be congestion at the factory gate at shift change. If that factory is instead a bunch of independent workshops, congestion goes away.

    Dispersion is the only suggestion for congestion.

  9. Maybe “Distribution is the suggestion for congestion.”

  10. re: adding cars and station platforms, escalators, lag time, etc, et al

    all true – but still easy to expand than highways are since expanding METRO does not need/use near the physical footprint that roads do.

    re: congestion dispersion and shift changes at physical workplaces.

    staggered work start/finish times can dramatically change the congestion at the entrance…..

    congestion is caused in part by the requirement that everyone be at a particular place at a particular time.

    if it costs nothing to institute that requirement then no problem.

    if it costs lost time because of the resulting congestion – then who should own those costs?

Leave a Reply