The Institute of Transportation Engineers has this marketing puff on its masthead: “A Community of Transportation Professionals, Your source for expertise, knowledge and ideas.” But in matters of forecasting traffic they often perform like a bunch of bumbling amateurs. Some of their worst malpractice comes in a Trip Generation Manual (see cover nearby) now in its 9th edition. A University of California Santa Cruz researcher Adam Millard Ball finds that the Manual overstates trip generation by land use an average of 55%!
City planners routinely use the ITE manual as the basis for estimating the traffic impact of new development. It gets referred to as the ‘bible’ of planners, and is now in its 9th edition.
Some ITE trip estimates by land use are just way off – churches it gets wrong with an excess of 129%, schools are 198% high, retailing 70%, apartments 108%. Trip generation by offices is exaggerated 27%, supermarkets 44%, all forms of housing by 56%.
The only land use underestimated in the ITE manual is medical – by 13%.
ITE is getting worse, Ball finds, the current 9th edition being off by much more than the 8th.
Researcher Ball’s findings are featured as the lead article in the latest issue of the University of California Transportation Center’s journal “Access” and a pdf copy is downloadable here:
The lousy forecasting is important because trip generation is the basis for most city planning – from zoning requirements through construction of city parking garages to the width of new roads. It then plays heavily into costs imposed on developers which get passed on in sale prices and rents of properties.
Sometimes it kills investment that otherwise would be viable.
Projects that should be financially self-supporting sometimes wangle City subsidies and transfer risks to taxpayers to make them happen. Would taxpayers have to underwrite the new hotel in downtown Frederick if the City made fewer demands for unnecessary facilities and excessive fees based on the shoddy ITE manual?
Fatalism about traffic another result
Not only that but exaggeration of future traffic of the kind encouraged by the ITE manual, may contribute to fatalism, the sense that we’re helpless to deal with traffic. If the ITE data suggests traffic is going to grow so much huge new spending will be required to cope, then we may be unwilling to make smaller, prudent investments in roads on grounds that they will make no difference.
Or we oppose new development, use NIMBY (not-in-my-backyard) political power to block development, and lose the local jobs, the housing and valuable facilities that development might otherwise have brought.
How does the ITE get trip generation so badly wrong?
Sources of ITE error – rubbish in, rubbish out
Ball attributes the gross inaccuracy of the ITE manual to the following incompetencies:
1. Data generated from an unrepresentative sample. ITE relies on reports of traffic generated by developments where there have been before and after traffic counts. Ball points out that this traffic data is only required for larger and more controversial developments, usually ones where the developer has been required to contribute considerably to road improvements. These are atypical of development generally.
2. Another source of sample bias lies in the ITE’s use of only voluntarily submitted reports. Smaller developers who have better things to do than sponsor traffic studies or help the ITE will be under-represented.
3. Further, Ball notes, the ITE itself biases the data by asking for reports only from “mature,” “economically healthy,” and “reasonably fully” occupied developments – cutting out a whole class of developments from the data. Developments that are unsuccessful, partially successful or slow to mature are all culled out.
4. A further fatal flaw is ITE’s micro/macro confusion. What is true at the micro level is often not true macro.
A development served by a cul de sac will, no doubt, generate net new trips in that cul de sac. To illustrate with a local example the Wegmans and Lowes and other retail and residential developments in the Market Square area of North Frederick are served exclusively by Worman’s Mill Road off MD26. New traffic on Wormans Mill traffic is somewhat predictable with the ITE methodology because entry and exit for vehicles is exclusively via it onto MD26 at the end of North Market St, MD355.
But from MD26 out and onto connecting US15 and North Market St offsetting effects come into play depressing traffic. Many of the Wormans Mill Rd trips to Wegmans, to use only the supermarket example, will be offset by reduced trips to other north Frederick area grocery stores – Giant less than a mile east on 26, Safeway a mile south on W7th St and Weis just to the west off Opossumtown Pike.
Therefore on Routes 15 and 26 new trips to the large new Wegmans may be largely offset by reductions in trips to competing grocery stores. With construction of dense 3- and 4- story townhouses and condos around narrow alleys in the northwest section there may actually be more walking and fewer vehicle trips – at least to the many new smaller service and retail establishments (we’re skeptical anyone with a car will ever forgo it for walking with a supermarket cart more than a couple of hundred yards.)
Ball writes that “ITE’s core premise, that development always generates new trips, is misleading in many circumstances.” The ITE trip generation manual overlooks offsetting trip reductions that often produce no net trip increase. Instead it crudely scales up the gross new trip rates in a procedure that Economics 101 – and for that matter commonsense – could detect as hugely misleading.
“Planning for phantom trips”
These faulty procedures mean that our city planners are “planning for phantom trips,” as Ball puts it.
“The implicit assumption in this (ITE) model is that new destination opportunities spur households to make new trips, rather than merely substituting new destinations for old ones. In practice, new development has much more complex effects on total trip making. Some trips will be completely new, as households take advantage of new employment or recreational opportunities. But most trips substitute for existing ones—they are diverted from existing locations as people change where they live, work, and shop in the light of new travel options.”
Ball writes that the whole language of the ITE engineers is misleading. Development itself, he points out, cannot generate a trip: “By calling this process ‘generation,’ we mislead people into thinking new trips are created as a direct result of development activities. But new land uses do not generate trips in a mechanistic way. Instead, people generate trips, in response to the characteristics of the built environment rather than the amount of development.“
Ball’s estimates of the extraordinary exaggeration of traffic by ITE compare the engineers’ trip generation projections with data from the US Census Bureau’s National Household Travel Survey (NHTS.)
He finds the magnitude of ITE error is growing: “From 1990 to 2009, the number of annual trips…in the US increased by 89 million, according to the NHTS. But according to the ITE-based method, annual trips should have increased by 189 million—an estimate more than double that of the NHTS. The contrast is even starker in more recent years: an increase of 2 million trips between 2001 and 2009 according to NHTS, but 90 million by the ITE-based method.”
The ITE’s nonsense trip generation manual doesn’t come cheap either: $325 for members and $500 for non-members.
My own view is that the ITE should not be doing this stuff. They aren’t professionally qualified. Engineers are simply out of their intellectual depth forecasting traffic. Their expertise is in designing, and building roads in the case of civil engineers, in building bridges that stay up, hopefully, in the case of structural engineers, and in designing traffic signals and managing traffic count data in the case of traffic engineers.
Traffic forecasting isn’t engineering, it’s economics. Having engineers forecast traffic makes as much sense as having economists design bridges. But I should declare my bias here: I trained in economics at university, and taught it for my first two years out. I have a lot of respect for engineers… when they stick to engineering.
- editor 20150222
$39 million in state/local support for Frederick’s downtown hotel project
Dec 18, 2015
Some $39 million of state and local government funds read more
“Painful preservation” a Frederick News-Post editorial, HPC Chair reacts
Dec 15, 2015
Following is an important editorial on the Historic read more
Letter of Interest in position on the City Planning Commission
Dec 14, 2015
LETTER OF INTEREST Ms Grimes: I’d like to offer myself read more