Here's a striking chart from Eric Sundquist of the State Smart Transportation Initiative. For the past decade, state and federal governments have consistently overestimated future growth in U.S. road travel.
Vehicle miles traveled, forecasts vs. actual:
Sundquist and Tony Dutzik of the Frontier Group sifted through a half-dozen old U.S. Department of Transportation reports to Congress and found a trend: Since 1999, the agency has consistently predicted that U.S. vehicle-miles traveled would rise dramatically (based on state and local forecasts). In reality, the opposite happened. The growth in vehicle travel slowed and then appeared to "peak" around 2007.
There's still lots of debate about why Americans are driving less since 2007. Some of it may be due to temporary factors like the recession (unemployed people tend to drive less). Some may be due to long-term trends like the aging of America (retirees drive less). Or higher gasoline prices (a deterrent against road trips). Young people are also driving less, and there are a slew of theories as to why: It's getting more expensive to buy a car, the barriers to getting a license are increasing, Facebook makes it easier to hang out with friends without driving...
Few people were predicting that shift (though here's one exception) — so it's not surprising that official forecasts missed it. But what's notable is that the forecasts still haven't caught up. Even as the number of vehicle-miles traveled drooped after 2007, the agency kept predicting significant increases. (We'll see how these forecasts get updated in the agency's 2012 report, which is due out soon).
Now, the Department of Transportation doesn't generate these figures directly. It aggregates forecasts from state agencies, who get them from metropolitan planning organizations. But, for the last decade, those state and local forecasts have been overly optimistic — they essentially assume that it's still 1980 and driving is going to increase sharply for the foreseeable future.
Also, it's entirely possible that those government forecasts will prove more accurate over a longer time period: Maybe the recent "peak" in driving is only a blip, and vehicle-miles traveled will rebound sharply as the U.S. economy grows. But, for the time being, the divergence is significant.
Why does this matter? Predictions that prove too high can distort planning decisions. Policymakers, for instance, might think we need more road capacity than we actually do. Here's Sundquist: "1) [These forecasts] imply a level of 'needed' spending that is politically unachievable, 2) they can spur overbuilding on projects, draining resources from critical preservation multimodal investments, and conversely 3) they can discourage construction of lower-cost, lower-throughput streets that improve livability and property values."
Or, here's a more concrete example from the energy sector. Back in 2007, Congress passed a Renewable Fuel Standard requiring the nation to blend a greater share of ethanol and biofuels into gasoline each year. For reasons of chemistry and economics, things get tricky when the share of ethanol in gasoline goes above 10 percent. But, at the time, lawmakers didn't think this was a problem. They figured that driving — and, hence, gasoline use — would rise indefinitely. We wouldn't hit that 10 percent "blend wall" for a long time.
That prediction was wrong. Driving declined, gasoline use dropped, and the "blend wall" has become a major issue for the fuel standard. The EPA is now issuing waivers to loosen the rule, and the oil industry is pushing to repeal the rule entirely. Incorrect forecasts had consequences.
By the way, for contrast, here's the U.S. Energy Information Administration's most recent prediction for future road travel:
We'll have to see if that forecast does any better.
-- It turns out transportation planners in the United Kingdom have had the same problem in recent years. Their forecasts also missed the decline of driving — and have yet to catch up.