Induced demands

20 posts / 0 new
Last post
hrashid
Induced demands

TMIPers, greetings!
In a congested urban roadway network, if new capacity is added, there will be new travel demand and soon the new facility will be as (or more) congested.
Paraphrasing what I read on "induced demands" in this forum, or elsewhere. It's been a while this topic has been discussed here, and I couldn't find any formal synthesis of the past discussions. If I remember correctly, few members suggested that such phenomenon can be analyzed within a travel demand model framework, of course a whole lot depends on how you define 'new' demand - a. same number of overall trips, but shifts in mode choice; b. new trips from latent demands, trips that otherwise wouldn't exist without new capacity; c. new trips from land developments due to added roadway capacity. This is not exhaustive, there can be other iterations/possibilities for new demands.

My objective is to understand current practices/experiences to address this issue - by regional transportation planners using a travel demand model. Appreciate any responses, I will provide an anonymized summary. Feel free to respond directly to me (my contact below). Thanks in advance!

Harun.

[cid:image002.jpg@01D3EDF5.3282CE00]

HARUN RASHID, AICP
Transportation Planner
Northern Virginia Transportation Authority
3040 Williams Drive, Suite 200, Fairfax, VA 22031.
Harun.Rashid@thenovaauthority.org
Office: 703-642-4659 | Cell: 571-355-3938
www.TheNoVaAuthority.org

elizabethsall

Harun,
The California Governor's Office of Planning and Research compiled some
resources on induced demand from
increases in highway capacity (see drop-down for "Induced VMT from Highway
Capacity").

Whether or not a travel model captures induced demand depends on the model
and how they are set up. Some do not capture any dimensions you mentioned.
Some capture all.

Even if the model can mechanistically capture induced travel, it should be
validated against external datapoints.

Cheers,
--Elizabeth Sall

On Mon, May 13, 2019 at 12:47 PM hicecesap <
Harun.Rashid@thenovaauthority.org> wrote:

> TMIPers, greetings!
> In a congested urban roadway network, if new capacity is added, there will
> be new travel demand and soon the new facility will be as (or more)
> congested.
> Paraphrasing what I read on "induced demands" in this forum, or elsewhere.
> It's been a while this topic has been discussed here, and I couldn't find
> any formal synthesis of the past discussions. If I remember correctly, few
> members suggested that such phenomenon can be analyzed within a travel
> demand model framework, of course a whole lot depends on how you define
> 'new' demand - a. same number of overall trips, but shifts in mode choice;
> b. new trips from latent demands, trips that otherwise wouldn't exist
> without new capacity; c. new trips from land developments due to added
> roadway capacity. This is not exhaustive, there can be other
> iterations/possibilities for new demands.
>
> My objective is to understand current practices/experiences to address
> this issue - by regional transportation planners using a travel demand
> model. Appreciate any responses, I will provide an anonymized summary. Feel
> free to respond directly to me (my contact below). Thanks in advance!
>
> Harun.
>
> [cid:image002.jpg@01D3EDF5.3282CE00]
>
> HARUN RASHID, AICP
> Transportation Planner
> Northern Virginia Transportation Authority
> 3040 Williams Drive, Suite 200, Fairfax, VA 22031.
> Harun.Rashid@thenovaauthority.org
> Office: 703-642-4659 | Cell: 571-355-3938
> www.TheNoVaAuthority.org
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>

Mark Bradley

I think of six main potential sources of "induced demand":

1. Route choice - people diverting from other routes
2. Departure time choice - people shifting (back) towards the peak period.
3. Destination choice - People traveling to different destinations than they would otherwise
4. Mode choice - People switching to car (or from HOV to SOV). In the longer term, this can be amplified if car ownership is affected.
5. Trip generation - People making additional trips or tours instead of staying home.
6. Land use shifts - In the longer term, households and/or jobs moving to take advantage of better accessibility (if it persists over time)

Most 4-step models can represent shifts in 1, 3 and 4. Activity-based models and some advanced trip-based models can also predict shifts in 2 and 5. Handling land use shifts in a consistent way requires an integrated land use model.

I would only call 5 and 6 "induced demand", while 1-4 are more like "diverted demand", but that's mostly semantics. If you're focusing just on the number of vehicles using a specific facility during the peak period, they are all sources of potential new demand. The relative influence of these factors depends on things such as the availability of alternative routes and good alternative modes, but I suspect that route switching and destination switching are the main sources of diverted demand in most US regions outside of the largest urban cores. The lower the availability of good alternative routes, the more important departure time shifting becomes.

I have heard debates in the past about whether "latent demand" really exists or not, but I don't think that makes much sense. As long as we live in something resembling our current economic and social system, if you provide a faster and/or cheaper and/or more reliable way to get to someplace that people want to go to at the time that they want to go there, then more people are going to want to use that way to get there. On an economist's supply-demand curve, there is no special kind of demand called "latent demand" - it's all demand that will manifest if the price (generalized cost) is right..

..........................................
Mark Bradley
Senior Director

RSG
www.rsginc.com

From: Harun.Rashid=thenovaauthority.org@mg.tmip.org On Behalf Of hicecesap
Sent: Monday, May 13, 2019 12:46 PM
To: TMIP
Subject: [TMIP] Induced demands

TMIPers, greetings!
In a congested urban roadway network, if new capacity is added, there will be new travel demand and soon the new facility will be as (or more) congested.
Paraphrasing what I read on "induced demands" in this forum, or elsewhere. It's been a while this topic has been discussed here, and I couldn't find any formal synthesis of the past discussions. If I remember correctly, few members suggested that such phenomenon can be analyzed within a travel demand model framework, of course a whole lot depends on how you define 'new' demand - a. same number of overall trips, but shifts in mode choice; b. new trips from latent demands, trips that otherwise wouldn't exist without new capacity; c. new trips from land developments due to added roadway capacity. This is not exhaustive, there can be other iterations/possibilities for new demands.

My objective is to understand current practices/experiences to address this issue - by regional transportation planners using a travel demand model. Appreciate any responses, I will provide an anonymized summary. Feel free to respond directly to me (my contact below). Thanks in advance!

Harun.

[cid:image002.jpg@01D3EDF5.3282CE00]

HARUN RASHID, AICP
Transportation Planner
Northern Virginia Transportation Authority
3040 Williams Drive, Suite 200, Fairfax, VA 22031.
Harun.Rashid@thenovaauthority.org
Office: 703-642-4659 | Cell: 571-355-3938
www.TheNoVaAuthority.org
--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

jabraham

Hi Harun.

I've been building spatial economic and land use models using the PECAS framework for a couple of decades, and before that using the MEPLAN framework. Fifteen or so years ago "induced demand" was a big buzzword and driver of land use modelling. Agencies wanted to be able to predict the impact of infrastructure on land development patterns, so wanted a land use model to be able to predict "induced demand".

I think it's an outdated concept now. Let me explain why.

Induced demand is a classic economic response — if you lower the price of something, more people will consume something, and there is a consumer surplus benefit. In the case of land use and transport economic systems, it's more complex, because the transportation system is primarily a supporting/complimentary service to a whole lot of other things, including new buildings built on land over time. So, induced demand cannot be analyzed properly just as a transport phenomena; it needs to be analyzed at a higher level.

Induced demand has economic value that results in a long term economic rearrangement. A transport demand model by itself does not represent economic interactions nor developer response (shifting patterns of building construction) over time. Modern activity based models (ABMs) can give us a better sense of short term induced demand (mode shifts, route shifts, etc.) but are usually worse at analyzing long term spatial economics because their run times are so long. ABMs are not typically not used in a tight integration with land use and spatial economic models because they are too cumbersome, and because many (but not all!) of the agencies who have recently built ABMs have spent all their modelling money and effort on ABMs and so haven't recently invested in spatial economic models.

So, I don't think it's a good idea to say "such phenomenon can be analyzed within a travel demand model framework". A portion can be analyzed within a travel demand model framework, but you need a land use model based on spatial economics to really analyze induced demand, one that includes a representation of business-to-business interactions and freight and service flows. With such a model, instead of focusing on induced demand itself, the agency typically wants to go one step further and focus on the economic value provided by the transportation system, rather than on proxy measures like "travel time" or "induced demand'.

The best practice is to have a high level spatial economic model and land use model linked with a transport model for understanding and forecasting big picture future patterns and economic impacts, and to recognize induced demand as a normal response to lower costs (e.g. lower travel time) within a slow moving adaptive spatial economic system.

Nowadays the agencies that have invested in using spatial economic and land use models to analyze policy are focused on economic performance, and induced demand is really just a side effect, an incidental part of the analysis, rather than the focus of the analysis. They don't even talk about it directly anymore, they've moved on. The main analysis is on economic spatial benefit measures through consumer surplus analysis. Meanwhile, the agencies that don't have land use models, or who have predictive land use models not based on economic interactions, realize they are not seeing the full picture of induced demand, so couldn't properly analyze it even if they tried.

I think that's why you found "It's been a while this topic has been discussed here". We've all either moved away from it (in the case of those who are only doing transport models or who invested in predictive land use models), or much beyond it to economic benefit measures (for those invested in spatial interaction based land use models).

So, I worry you are asking the wrong question. Or, an outdated question. But, I'd like to hear others' opinions on it, so please share the promised summary when it's ready.

--
John Abraham
HBA Specto Incorporated
jea@hbaspecto.com
403-232-1060

> On May 13, 2019, at 12:46 PM, hicecesap wrote:
>
> TMIPers, greetings!
> In a congested urban roadway network, if new capacity is added, there will be new travel demand and soon the new facility will be as (or more) congested.
> Paraphrasing what I read on "induced demands" in this forum, or elsewhere. It's been a while this topic has been discussed here, and I couldn't find any formal synthesis of the past discussions. If I remember correctly, few members suggested that such phenomenon can be analyzed within a travel demand model framework, of course a whole lot depends on how you define 'new' demand - a. same number of overall trips, but shifts in mode choice; b. new trips from latent demands, trips that otherwise wouldn't exist without new capacity; c. new trips from land developments due to added roadway capacity. This is not exhaustive, there can be other iterations/possibilities for new demands.
>
> My objective is to understand current practices/experiences to address this issue - by regional transportation planners using a travel demand model. Appreciate any responses, I will provide an anonymized summary. Feel free to respond directly to me (my contact below). Thanks in advance!
>
> Harun.
>
> [cid:image002.jpg@01D3EDF5.3282CE00]
>
> HARUN RASHID, AICP
> Transportation Planner
> Northern Virginia Transportation Authority
> 3040 Williams Drive, Suite 200, Fairfax, VA 22031.
> Harun.Rashid@thenovaauthority.org
> Office: 703-642-4659 | Cell: 571-355-3938
> www.TheNoVaAuthority.org --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

Alex Bettinardi

Not an answer to the question; but a desire for our industry to clarify these terms / definitions:

I pose that we are not using the term "induced" demand correctly:
See the definitions linked below or complete your own search in google or a physical dictionary.

Induce - https://www.google.com/search?source=hp&ei=hSDbXNvAB_6w0PEPooqvyAw&q=def...

Latent - https://www.google.com/search?q=define+latent&rlz=1C1GCEA_enUS806US806&o...

While there may be cases of truly induced demand (like something special like an IKEA shows up, and you flood the store to buy a bunch of furniture and meatballs you didn't realize you needed), in almost all cases what we are really talking about is latent demand. Demand that has been suppressed (hidden / concealed) by high congestion and is now returning with the new capacity. This is especially the case in the example given below where the transportation infrastructure is being expanded and immediately fills up. The new highway isn't attracting new trips, it is serving trips that were previously put off or served in a different way.

My office would appreciate it if the industry could correct it's use of these terms.

Alex Bettinardi, P.E.
503.986.4104
http://www.oregon.gov/ODOT/Planning/Pages/default.aspx

From: easall=gmail.com@mg.tmip.org On Behalf Of elizabethsall
Sent: Tuesday, May 14, 2019 1:04 PM
To: TMIP
Subject: Re: [TMIP] Induced demands

Harun,
The California Governor's Office of Planning and Research compiled some
resources on induced demand from
increases in highway capacity (see drop-down for "Induced VMT from Highway
Capacity").

Whether or not a travel model captures induced demand depends on the model
and how they are set up. Some do not capture any dimensions you mentioned.
Some capture all.

Even if the model can mechanistically capture induced travel, it should be
validated against external datapoints.

Cheers,
--Elizabeth Sall

On Mon, May 13, 2019 at 12:47 PM hicecesap <
Harun.Rashid@thenovaauthority.org> wrote:

> TMIPers, greetings!
> In a congested urban roadway network, if new capacity is added, there will
> be new travel demand and soon the new facility will be as (or more)
> congested.
> Paraphrasing what I read on "induced demands" in this forum, or elsewhere.
> It's been a while this topic has been discussed here, and I couldn't find
> any formal synthesis of the past discussions. If I remember correctly, few
> members suggested that such phenomenon can be analyzed within a travel
> demand model framework, of course a whole lot depends on how you define
> 'new' demand - a. same number of overall trips, but shifts in mode choice;
> b. new trips from latent demands, trips that otherwise wouldn't exist
> without new capacity; c. new trips from land developments due to added
> roadway capacity. This is not exhaustive, there can be other
> iterations/possibilities for new demands.
>
> My objective is to understand current practices/experiences to address
> this issue - by regional transportation planners using a travel demand
> model. Appreciate any responses, I will provide an anonymized summary. Feel
> free to respond directly to me (my contact below). Thanks in advance!
>
> Harun.
>
> [cid:image002.jpg@01D3EDF5.3282CE00]
>
> HARUN RASHID, AICP
> Transportation Planner
> Northern Virginia Transportation Authority
> 3040 Williams Drive, Suite 200, Fairfax, VA 22031.
> Harun.Rashid@thenovaauthority.org
> Office: 703-642-4659 | Cell: 571-355-3938
> www.TheNoVaAuthority.org
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>
--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

nmarshallvt

Induced travel is a real world phenomenon  based on economic theory that has been studied extensively.  When the price of any good drops, the demand increases. For road travel, most consumers consider the travel time the largest component of the "price". Therefore, if travel time drops, demand increases. We call this induced travel.

Susan Handy and Marlon Boarnet summarized the research literature in 2014. They conclude:

"Thus, the best estimate for the long-run effect of highway capacity on VMT is an elasticity close to 1.0, implying that in congested metropolitan areas, adding new capacity to the existing system of limited-access highways is unlikely to reduce congestion or associated GHG in the long-run."

https://www.arb.ca.gov/cc/sb375/policies/policies.htm

https://www.arb.ca.gov/cc/sb375/policies/hwycapacity/highway_capacity_bkgd.pdf

Induced travel studies look at changes in real world travel in before and after conditions. In contrast, the term "latent demand" wants us to instead focus on an impossible fantasy where there is no congestion. This has not occurred and cannot occur in the real world. Therefore, 

latent demand cannot be quantified or studied. When networks without capacity constraints have been modeled, it has been done to demonstrate that building our way out of congestion is impossible.

Joe Cortright recently made an instructive post using the analogy of "free" ice cream day. Along with a photo of a long line of people queued up waiting for their "free" ice cream, he writes:

"As you’re standing in line waiting for your “free” ice cream cone, give a little thought to the parallels between that line and your typical rush hour traffic jam.  In both cases, you’re waiting in line for the same reason–the price is too low, and demand is overwhelming supply.  This is the valuable lesson that Ben and Jerry are providing in the fundamentals of transportation economics."

http://cityobservatory.org/2019/04/

Going back to the original question about modeling and induced travel, Mark Bradley suggested that travel demand models account for induced route choice and induced destination choice. As I discussed in a recent thread on this forum, static assignment does not properly account for induced route choice because it does not constrain flow to capacity.  Then, the resulting travel time matrices underestimate the affects of congestion, so that the models also fail to properly account for induced destination choice. These problems can be corrected by replacing the static assignment with regional DTA with feedback.

Some modelers cite the failure of the static assignment models to account for induced travel as evidence that induced travel isn't real. If there is a discrepancy between the research literature and the model, the model should be fixed so that it matches the data.

Some modelers defend these model errors by saying the models are accounting for latent demand. This is a fantasy which leads to wasteful spending. I recently was talking to a senior modeling consultant who does a lot of work for a large state DOT. He complained that the DOT requires them to design for latent demand even when this demand could never occur due to upstream bottlenecks that will not be addressed. 

Norm Marshall, Smart Mobility Inc. 

nmarshall@smartmobility.com

 

Alex Bettinardi

Thank you Bernie,

You point out a good historic example, I think that helps explain how the industry got so used to using induced demand as a term to cover all shifting. I think you capture this correctly, that here is a pie chart for any given new roadway or lanes or new land development. There is a percent of that pie that is latent and a percent of the pie chart that is induced.

I am urging the industry to acknowledge that where we currently at a point in history (at least in the US) where we don’t expand to the degree where the majority of the pie is induce. In almost all projects we are now primarily latent due to the high congestion in our urban areas. Historic examples might indeed show that demand is induced, but that is now the minority as opposed to latent.

I get what you are saying that there is a mix. I propose that the high majority of common day examples is latent. I think what would be the best solution would be to properly talk about how much of the demand is latent and how much is induced. I think a next best solution might be to refer to this as “latent / induced demand” and make it clear that it’s some percentage of both. I think the worst case is that everything is labeled as induced. I think in present day labeling everything as latent is a better (more correct) solution than labeling everything as induced, since latent is the majority, but I get that is a judgment call based on what type of projects you deal with and see on a day-to-day basis and where you are on the globe.

So maybe we can all agree to lengthen the title a little bit and go with - “latent / induced demand”.

Alex Bettinardi, P.E.
503.986.4104
http://www.oregon.gov/ODOT/Planning/Pages/default.aspx

From: ברנרד אלפרן
Sent: Wednesday, May 15, 2019 5:06 AM
To: BETTINARDI Alexander O * Alex ; tmip@mg.tmip.org
Subject: FW: [TMIP] Induced demands

Hi Alex,

I completely understand (and agree with) your desire for the industry to clearly define the terms it uses, but (very) respectfully disagree with your suggestion.

One of the primary meanings of “induce” is “to cause something to happen” (see https://dictionary.cambridge.org/dictionary/english/induce). That “something” could be something new, not necessarily something that was “suppressed” by the lack of a cause (or inducement). A classic example in the transportation field is the Long Island Expressway (the current official name of I-495 in Nassau and Suffolk Counties, New York). This 71-mile east-west highway across Queens, Nassau, and Suffolk Counties was constructed in stages. The westernmost (closest to Manhattan) section opened to traffic in 1940 and the easternmost section was opened in 1972. During the course of those 32 years, the highway induced significantly faster-paced and lower density development in Eastern Queens, and in Nassau and Suffolk Counties than could or would have happened otherwise. Legend has it that travel times from Long Island to Manhattan were no shorter on the day of the road’s completion than on the day that its construction started. Traffic reporters in the 1970’s would regularly refer to the Long Island Expressway as “the big LIE” (some referred to it as “the world’s longest parking lot”).

Clearly, the bulk of the trips congesting the Expressway had not been suppressed since before 1940. They were new trips, “induced” by the faster and lower density development on Long Island that was, in part, caused by the highway. They weren’t “latent” trips, they were new trips that didn’t exist before.

Very often when a transportation facility construction or expansion results in higher amounts of travel in the corridor served, it is difficult to parse out how much of that increased travel demand is latent, and how much is induced (and how much is simply diverted from other parallel facilities). Hence the confusion. As long as we understand that the increase in demand has multiple components, that’s probably the best we can do.

It’s not for nothing that a very senior manager at USDOT was known to say “whoever labelled Economics ‘the dismal science’ never heard of travel demand forecasting”. It is, and always will be, at least as much art as science.

Yours truly,

Bernie Alpern

Bernard Alpern, Transportation Planner
Eyal Kraus Roadway and Traffic Engineers
Beit Ha-Dfus Street 22, 2nd floor
Jerusalem, Israel 9548326
Office: +972-2-532-8814 Direct: +972-2-501-1754 Mobile: +972-58-699-9391
alpern@ekroads.co.il
[cid:image001.png@01D2FFE1.A915AC20]

---------- Forwarded message ---------
From: Alex Bettinardi >
Date: Wed, May 15, 2019 at 1:41 PM
Subject: Re: [TMIP] Induced demands
To: TMIP >

Not an answer to the question; but a desire for our industry to clarify these terms / definitions:

I pose that we are not using the term "induced" demand correctly:
See the definitions linked below or complete your own search in google or a physical dictionary.

Induce - https://www.google.com/search?source=hp&ei=hSDbXNvAB_6w0PEPooqvyAw&q=def...

Latent - https://www.google.com/search?q=define+latent&rlz=1C1GCEA_enUS806US806&o...

While there may be cases of truly induced demand (like something special like an IKEA shows up, and you flood the store to buy a bunch of furniture and meatballs you didn't realize you needed), in almost all cases what we are really talking about is latent demand. Demand that has been suppressed (hidden / concealed) by high congestion and is now returning with the new capacity. This is especially the case in the example given below where the transportation infrastructure is being expanded and immediately fills up. The new highway isn't attracting new trips, it is serving trips that were previously put off or served in a different way.

My office would appreciate it if the industry could correct it's use of these terms.

Alex Bettinardi, P.E.
503.986.4104
http://www.oregon.gov/ODOT/Planning/Pages/default.aspx

From: easall=gmail.com@mg.tmip.org On Behalf Of elizabethsall
Sent: Tuesday, May 14, 2019 1:04 PM
To: TMIP
Subject: Re: [TMIP] Induced demands

Harun,
The California Governor's Office of Planning and Research compiled some
resources on induced demand from
increases in highway capacity (see drop-down for "Induced VMT from Highway
Capacity").

Whether or not a travel model captures induced demand depends on the model
and how they are set up. Some do not capture any dimensions you mentioned.
Some capture all.

Even if the model can mechanistically capture induced travel, it should be
validated against external datapoints.

Cheers,
--Elizabeth Sall

On Mon, May 13, 2019 at 12:47 PM hicecesap <
Harun.Rashid@thenovaauthority.org> wrote:

> TMIPers, greetings!
> In a congested urban roadway network, if new capacity is added, there will
> be new travel demand and soon the new facility will be as (or more)
> congested.
> Paraphrasing what I read on "induced demands" in this forum, or elsewhere.
> It's been a while this topic has been discussed here, and I couldn't find
> any formal synthesis of the past discussions. If I remember correctly, few
> members suggested that such phenomenon can be analyzed within a travel
> demand model framework, of course a whole lot depends on how you define
> 'new' demand - a. same number of overall trips, but shifts in mode choice;
> b. new trips from latent demands, trips that otherwise wouldn't exist
> without new capacity; c. new trips from land developments due to added
> roadway capacity. This is not exhaustive, there can be other
> iterations/possibilities for new demands.
>
> My objective is to understand current practices/experiences to address
> this issue - by regional transportation planners using a travel demand
> model. Appreciate any responses, I will provide an anonymized summary. Feel
> free to respond directly to me (my contact below). Thanks in advance!
>
> Harun.
>
> [cid:image002.jpg@01D3EDF5.3282CE00]
>
> HARUN RASHID, AICP
> Transportation Planner
> Northern Virginia Transportation Authority
> 3040 Williams Drive, Suite 200, Fairfax, VA 22031.
> Harun.Rashid@thenovaauthority.org
> Office: 703-642-4659 | Cell: 571-355-3938
> www.TheNoVaAuthority.org
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>
--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

quxoxalow

It seems to me that both terms, "induced" and "latent", are problematic in
this context. Both imply that but for some action on the part of a road
authority demand is somehow static. Demand is always a function of a
variety of factors including preferences, income and most immediately,
price. The very notion of a demand curve implies that as price (including
time costs of congestion delays) changes demand will also change.

If building, or widening, roads induces demand (this sounds like a
conjuror's trick) then it must also be true that failing to build, or widen
roads (at no direct cost to the users) must suppress demand. Both notions
seem a bit absurd. As a supplier of roads the authority cannot define the
demand conditions in the market, but rather responds, or does not respond,
to those conditions.

If road authorities, as public monopolists, behaved as private monopolists
then they would likely under-supply the market and charge monopolist
prices, resulting in larger than normal profits. Instead they mostly
under-supply the market and charge no prices, resulting in larger than
normal levels of congestion.

From a modeling perspective it seems that Mark Bradley and John Abraham
have covered the main approaches in practice designed to more
comprehensively handle the full range of travel demand responses.

Matthew

On Wed, May 15, 2019 at 8:00 AM Alex Bettinardi <
alexander.o.bettinardi@odot.state.or.us> wrote:

> Thank you Bernie,
>
> You point out a good historic example, I think that helps explain how the
> industry got so used to using induced demand as a term to cover all
> shifting. I think you capture this correctly, that here is a pie chart for
> any given new roadway or lanes or new land development. There is a percent
> of that pie that is latent and a percent of the pie chart that is induced.
>
> I am urging the industry to acknowledge that where we currently at a point
> in history (at least in the US) where we don’t expand to the degree where
> the majority of the pie is induce. In almost all projects we are now
> primarily latent due to the high congestion in our urban areas. Historic
> examples might indeed show that demand is induced, but that is now the
> minority as opposed to latent.
>
> I get what you are saying that there is a mix. I propose that the high
> majority of common day examples is latent. I think what would be the best
> solution would be to properly talk about how much of the demand is latent
> and how much is induced. I think a next best solution might be to refer to
> this as “latent / induced demand” and make it clear that it’s some
> percentage of both. I think the worst case is that everything is labeled as
> induced. I think in present day labeling everything as latent is a better
> (more correct) solution than labeling everything as induced, since latent
> is the majority, but I get that is a judgment call based on what type of
> projects you deal with and see on a day-to-day basis and where you are on
> the globe.
>
> So maybe we can all agree to lengthen the title a little bit and go with -
> “latent / induced demand”.
>
> Alex Bettinardi, P.E.
> 503.986.4104
> http://www.oregon.gov/ODOT/Planning/Pages/default.aspx
>
> From: ברנרד אלפרן
> Sent: Wednesday, May 15, 2019 5:06 AM
> To: BETTINARDI Alexander O * Alex ; tmip@mg.tmip.org
> Subject: FW: [TMIP] Induced demands
>
> Hi Alex,
>
> I completely understand (and agree with) your desire for the industry to
> clearly define the terms it uses, but (very) respectfully disagree with
> your suggestion.
>
> One of the primary meanings of “induce” is “to cause something to happen”
> (see https://dictionary.cambridge.org/dictionary/english/induce). That
> “something” could be something new, not necessarily something that was
> “suppressed” by the lack of a cause (or inducement). A classic example in
> the transportation field is the Long Island Expressway (the current
> official name of I-495 in Nassau and Suffolk Counties, New York). This
> 71-mile east-west highway across Queens, Nassau, and Suffolk Counties was
> constructed in stages. The westernmost (closest to Manhattan) section
> opened to traffic in 1940 and the easternmost section was opened in 1972.
> During the course of those 32 years, the highway induced significantly
> faster-paced and lower density development in Eastern Queens, and in Nassau
> and Suffolk Counties than could or would have happened otherwise. Legend
> has it that travel times from Long Island to Manhattan were no shorter on
> the day of the road’s completion than on the day that its construction
> started. Traffic reporters in the 1970’s would regularly refer to the Long
> Island Expressway as “the big LIE” (some referred to it as “the world’s
> longest parking lot”).
>
> Clearly, the bulk of the trips congesting the Expressway had not been
> suppressed since before 1940. They were new trips, “induced” by the faster
> and lower density development on Long Island that was, in part, caused by
> the highway. They weren’t “latent” trips, they were new trips that didn’t
> exist before.
>
> Very often when a transportation facility construction or expansion
> results in higher amounts of travel in the corridor served, it is difficult
> to parse out how much of that increased travel demand is latent, and how
> much is induced (and how much is simply diverted from other parallel
> facilities). Hence the confusion. As long as we understand that the
> increase in demand has multiple components, that’s probably the best we can
> do.
>
> It’s not for nothing that a very senior manager at USDOT was known to say
> “whoever labelled Economics ‘the dismal science’ never heard of travel
> demand forecasting”. It is, and always will be, at least as much art as
> science.
>
> Yours truly,
>
> Bernie Alpern
>
> Bernard Alpern, Transportation Planner
> Eyal Kraus Roadway and Traffic Engineers
> Beit Ha-Dfus Street 22, 2nd floor
> Jerusalem, Israel 9548326
> Office: +972-2-532-8814 Direct: +972-2-501-1754 Mobile: +972-58-699-9391
> alpern@ekroads.co.il
> [cid:image001.png@01D2FFE1.A915AC20]
>
> ---------- Forwarded message ---------
> From: Alex Bettinardi >
> Date: Wed, May 15, 2019 at 1:41 PM
> Subject: Re: [TMIP] Induced demands
> To: TMIP >
>
> Not an answer to the question; but a desire for our industry to clarify
> these terms / definitions:
>
> I pose that we are not using the term "induced" demand correctly:
> See the definitions linked below or complete your own search in google or
> a physical dictionary.
>
> Induce -
> https://www.google.com/search?source=hp&ei=hSDbXNvAB_6w0PEPooqvyAw&q=def
> ...
>
> Latent -
> https://www.google.com/search?q=define+latent&rlz=1C1GCEA_enUS806US806&o
> ...
>
> While there may be cases of truly induced demand (like something special
> like an IKEA shows up, and you flood the store to buy a bunch of furniture
> and meatballs you didn't realize you needed), in almost all cases what we
> are really talking about is latent demand. Demand that has been suppressed
> (hidden / concealed) by high congestion and is now returning with the new
> capacity. This is especially the case in the example given below where the
> transportation infrastructure is being expanded and immediately fills up.
> The new highway isn't attracting new trips, it is serving trips that were
> previously put off or served in a different way.
>
> My office would appreciate it if the industry could correct it's use of
> these terms.
>
> Alex Bettinardi, P.E.
> 503.986.4104
> http://www.oregon.gov/ODOT/Planning/Pages/default.aspx
>
> From: easall=gmail.com@mg.tmip.org On Behalf Of elizabethsall
> Sent: Tuesday, May 14, 2019 1:04 PM
> To: TMIP
> Subject: Re: [TMIP] Induced demands
>
> Harun,
> The California Governor's Office of Planning and Research compiled some
> resources on induced demand from
> increases in highway capacity (see drop-down for "Induced VMT from Highway
> Capacity").
>
> Whether or not a travel model captures induced demand depends on the model
> and how they are set up. Some do not capture any dimensions you mentioned.
> Some capture all.
>
> Even if the model can mechanistically capture induced travel, it should be
> validated against external datapoints.
>
> Cheers,
> --Elizabeth Sall
>
> On Mon, May 13, 2019 at 12:47 PM hicecesap <
> Harun.Rashid@thenovaauthority.org> wrote:
>
> > TMIPers, greetings!
> > In a congested urban roadway network, if new capacity is added, there
> will
> > be new travel demand and soon the new facility will be as (or more)
> > congested.
> > Paraphrasing what I read on "induced demands" in this forum, or
> elsewhere.
> > It's been a while this topic has been discussed here, and I couldn't find
> > any formal synthesis of the past discussions. If I remember correctly,
> few
> > members suggested that such phenomenon can be analyzed within a travel
> > demand model framework, of course a whole lot depends on how you define
> > 'new' demand - a. same number of overall trips, but shifts in mode
> choice;
> > b. new trips from latent demands, trips that otherwise wouldn't exist
> > without new capacity; c. new trips from land developments due to added
> > roadway capacity. This is not exhaustive, there can be other
> > iterations/possibilities for new demands.
> >
> > My objective is to understand current practices/experiences to address
> > this issue - by regional transportation planners using a travel demand
> > model. Appreciate any responses, I will provide an anonymized summary.
> Feel
> > free to respond directly to me (my contact below). Thanks in advance!
> >
> > Harun.
> >
> > [cid:image002.jpg@01D3EDF5.3282CE00]
> >
> > HARUN RASHID, AICP
> > Transportation Planner
> > Northern Virginia Transportation Authority
> > 3040 Williams Drive, Suite 200, Fairfax, VA 22031.
> > Harun.Rashid@thenovaauthority.org
> > Office: 703-642-4659 | Cell: 571-355-3938
> > www.TheNoVaAuthority.org
> > --
> > Full post: https://tmip.org/content/induced-demands
> > Manage my subscriptions: https://tmip.org/mailinglist
> > Stop emails for this post:
> https://tmip.org/mailinglist/unsubscribe/13101
> >
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>

balpern2

Very interesting, Alex.

I guess I would say that here in Israel it's a mixed bag. Population growth
is still strong (almost 2%/year nationwide, with considerably higher rates
in some areas). But land in the large metropolitan areas has become very
scarce. Therefore, one of the policy goals of the Government is to provide
high-quality rail connections to relatively remote areas, to *induce*
people to spread out more and not all try to crowd into the main cities in
the country's center, further burdening the infrastructure (transportation
and otherwise) there.

In the Center, however, congestion rules and public transit, while
relatively ubiquitous, is not very attractive. I'm sure that travel is
being suppressed, as it is in Southern California, the Greater Washington
area, and other very congested urban areas in the US where the bulk of
personal travel takes place in automobiles. Again, because of the scarcity
of land, the Government is investing more than ever before in upgrading
public transit, building light (and eventually heavy) urban rail systems.
There is still a long way to go in this regard....

Have a nice day out in beautiful Oregon!

Bernie

On Wed, May 15, 2019 at 6:00 PM Alex Bettinardi <
alexander.o.bettinardi@odot.state.or.us> wrote:

> Thank you Bernie,
>
> You point out a good historic example, I think that helps explain how the
> industry got so used to using induced demand as a term to cover all
> shifting. I think you capture this correctly, that here is a pie chart for
> any given new roadway or lanes or new land development. There is a percent
> of that pie that is latent and a percent of the pie chart that is induced.
>
> I am urging the industry to acknowledge that where we currently at a point
> in history (at least in the US) where we don’t expand to the degree where
> the majority of the pie is induce. In almost all projects we are now
> primarily latent due to the high congestion in our urban areas. Historic
> examples might indeed show that demand is induced, but that is now the
> minority as opposed to latent.
>
> I get what you are saying that there is a mix. I propose that the high
> majority of common day examples is latent. I think what would be the best
> solution would be to properly talk about how much of the demand is latent
> and how much is induced. I think a next best solution might be to refer to
> this as “latent / induced demand” and make it clear that it’s some
> percentage of both. I think the worst case is that everything is labeled as
> induced. I think in present day labeling everything as latent is a better
> (more correct) solution than labeling everything as induced, since latent
> is the majority, but I get that is a judgment call based on what type of
> projects you deal with and see on a day-to-day basis and where you are on
> the globe.
>
> So maybe we can all agree to lengthen the title a little bit and go with -
> “latent / induced demand”.
>
> Alex Bettinardi, P.E.
> 503.986.4104
> http://www.oregon.gov/ODOT/Planning/Pages/default.aspx
>
> From: ברנרד אלפרן
> Sent: Wednesday, May 15, 2019 5:06 AM
> To: BETTINARDI Alexander O * Alex ; tmip@mg.tmip.org
> Subject: FW: [TMIP] Induced demands
>
> Hi Alex,
>
> I completely understand (and agree with) your desire for the industry to
> clearly define the terms it uses, but (very) respectfully disagree with
> your suggestion.
>
> One of the primary meanings of “induce” is “to cause something to happen”
> (see https://dictionary.cambridge.org/dictionary/english/induce). That
> “something” could be something new, not necessarily something that was
> “suppressed” by the lack of a cause (or inducement). A classic example in
> the transportation field is the Long Island Expressway (the current
> official name of I-495 in Nassau and Suffolk Counties, New York). This
> 71-mile east-west highway across Queens, Nassau, and Suffolk Counties was
> constructed in stages. The westernmost (closest to Manhattan) section
> opened to traffic in 1940 and the easternmost section was opened in 1972.
> During the course of those 32 years, the highway induced significantly
> faster-paced and lower density development in Eastern Queens, and in Nassau
> and Suffolk Counties than could or would have happened otherwise. Legend
> has it that travel times from Long Island to Manhattan were no shorter on
> the day of the road’s completion than on the day that its construction
> started. Traffic reporters in the 1970’s would regularly refer to the Long
> Island Expressway as “the big LIE” (some referred to it as “the world’s
> longest parking lot”).
>
> Clearly, the bulk of the trips congesting the Expressway had not been
> suppressed since before 1940. They were new trips, “induced” by the faster
> and lower density development on Long Island that was, in part, caused by
> the highway. They weren’t “latent” trips, they were new trips that didn’t
> exist before.
>
> Very often when a transportation facility construction or expansion
> results in higher amounts of travel in the corridor served, it is difficult
> to parse out how much of that increased travel demand is latent, and how
> much is induced (and how much is simply diverted from other parallel
> facilities). Hence the confusion. As long as we understand that the
> increase in demand has multiple components, that’s probably the best we can
> do.
>
> It’s not for nothing that a very senior manager at USDOT was known to say
> “whoever labelled Economics ‘the dismal science’ never heard of travel
> demand forecasting”. It is, and always will be, at least as much art as
> science.
>
> Yours truly,
>
> Bernie Alpern
>
> Bernard Alpern, Transportation Planner
> Eyal Kraus Roadway and Traffic Engineers
> Beit Ha-Dfus Street 22, 2nd floor
> Jerusalem, Israel 9548326
> Office: +972-2-532-8814 Direct: +972-2-501-1754 Mobile: +972-58-699-9391
> alpern@ekroads.co.il
> [cid:image001.png@01D2FFE1.A915AC20]
>
> ---------- Forwarded message ---------
> From: Alex Bettinardi >
> Date: Wed, May 15, 2019 at 1:41 PM
> Subject: Re: [TMIP] Induced demands
> To: TMIP >
>
> Not an answer to the question; but a desire for our industry to clarify
> these terms / definitions:
>
> I pose that we are not using the term "induced" demand correctly:
> See the definitions linked below or complete your own search in google or
> a physical dictionary.
>
> Induce -
> https://www.google.com/search?source=hp&ei=hSDbXNvAB_6w0PEPooqvyAw&q=def
> ...
>
> Latent -
> https://www.google.com/search?q=define+latent&rlz=1C1GCEA_enUS806US806&o
> ...
>
> While there may be cases of truly induced demand (like something special
> like an IKEA shows up, and you flood the store to buy a bunch of furniture
> and meatballs you didn't realize you needed), in almost all cases what we
> are really talking about is latent demand. Demand that has been suppressed
> (hidden / concealed) by high congestion and is now returning with the new
> capacity. This is especially the case in the example given below where the
> transportation infrastructure is being expanded and immediately fills up.
> The new highway isn't attracting new trips, it is serving trips that were
> previously put off or served in a different way.
>
> My office would appreciate it if the industry could correct it's use of
> these terms.
>
> Alex Bettinardi, P.E.
> 503.986.4104
> http://www.oregon.gov/ODOT/Planning/Pages/default.aspx
>
> From: easall=gmail.com@mg.tmip.org On Behalf Of elizabethsall
> Sent: Tuesday, May 14, 2019 1:04 PM
> To: TMIP
> Subject: Re: [TMIP] Induced demands
>
> Harun,
> The California Governor's Office of Planning and Research compiled some
> resources on induced demand from
> increases in highway capacity (see drop-down for "Induced VMT from Highway
> Capacity").
>
> Whether or not a travel model captures induced demand depends on the model
> and how they are set up. Some do not capture any dimensions you mentioned.
> Some capture all.
>
> Even if the model can mechanistically capture induced travel, it should be
> validated against external datapoints.
>
> Cheers,
> --Elizabeth Sall
>
> On Mon, May 13, 2019 at 12:47 PM hicecesap <
> Harun.Rashid@thenovaauthority.org> wrote:
>
> > TMIPers, greetings!
> > In a congested urban roadway network, if new capacity is added, there
> will
> > be new travel demand and soon the new facility will be as (or more)
> > congested.
> > Paraphrasing what I read on "induced demands" in this forum, or
> elsewhere.
> > It's been a while this topic has been discussed here, and I couldn't find
> > any formal synthesis of the past discussions. If I remember correctly,
> few
> > members suggested that such phenomenon can be analyzed within a travel
> > demand model framework, of course a whole lot depends on how you define
> > 'new' demand - a. same number of overall trips, but shifts in mode
> choice;
> > b. new trips from latent demands, trips that otherwise wouldn't exist
> > without new capacity; c. new trips from land developments due to added
> > roadway capacity. This is not exhaustive, there can be other
> > iterations/possibilities for new demands.
> >
> > My objective is to understand current practices/experiences to address
> > this issue - by regional transportation planners using a travel demand
> > model. Appreciate any responses, I will provide an anonymized summary.
> Feel
> > free to respond directly to me (my contact below). Thanks in advance!
> >
> > Harun.
> >
> > [cid:image002.jpg@01D3EDF5.3282CE00]
> >
> > HARUN RASHID, AICP
> > Transportation Planner
> > Northern Virginia Transportation Authority
> > 3040 Williams Drive, Suite 200, Fairfax, VA 22031.
> > Harun.Rashid@thenovaauthority.org
> > Office: 703-642-4659 | Cell: 571-355-3938
> > www.TheNoVaAuthority.org
> > --
> > Full post: https://tmip.org/content/induced-demands
> > Manage my subscriptions: https://tmip.org/mailinglist
> > Stop emails for this post:
> https://tmip.org/mailinglist/unsubscribe/13101
> >
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>

nytydewuq

When the price of any good drops, the quantity demanded increases – demand does not increase, i.e. the demand curve does not shift.

From: nmarshall=smartmobility.com@mg.tmip.org On Behalf Of nmarshallvt
Sent: Wednesday, May 15, 2019 7:59 AM
To: TMIP
Subject: Re: [TMIP] Induced demands

Induced travel is a real world phenomenon based on economic theory that has been studied extensively. When the price of any good drops, the demand increases. For road travel, most consumers consider the travel time the largest component of the "price". Therefore, if travel time drops, demand increases. We call this induced travel.

Susan Handy and Marlon Boarnet summarized the research literature in 2014. They conclude:

"Thus, the best estimate for the long-run effect of highway capacity on VMT is an elasticity close to 1.0, implying that in congested metropolitan areas, adding new capacity to the existing system of limited-access highways is unlikely to reduce congestion or associated GHG in the long-run."

https://www.arb.ca.gov/cc/sb375/policies/policies.htm

https://www.arb.ca.gov/cc/sb375/policies/hwycapacity/highway_capacity_bk...

Induced travel studies look at changes in real world travel in before and after conditions. In contrast, the term "latent demand" wants us to instead focus on an impossible fantasy where there is no congestion. This has not occurred and cannot occur in the real world. Therefore,

latent demand cannot be quantified or studied. When networks without capacity constraints have been modeled, it has been done to demonstrate that building our way out of congestion is impossible.

Joe Cortright recently made an instructive post using the analogy of "free" ice cream day. Along with a photo of a long line of people queued up waiting for their "free" ice cream, he writes:

"As you’re standing in line waiting for your “free” ice cream cone, give a little thought to the parallels between that line and your typical rush hour traffic jam. In both cases, you’re waiting in line for the same reason–the price is too low, and demand is overwhelming supply. This is the valuable lesson that Ben and Jerry are providing in the fundamentals of transportation economics."

http://cityobservatory.org/2019/04/

Going back to the original question about modeling and induced travel, Mark Bradley suggested that travel demand models account for induced route choice and induced destination choice. As I discussed in a recent thread on this forum, static assignment does not properly account for induced route choice because it does not constrain flow to capacity. Then, the resulting travel time matrices underestimate the affects of congestion, so that the models also fail to properly account for induced destination choice. These problems can be corrected by replacing the static assignment with regional DTA with feedback.

Some modelers cite the failure of the static assignment models to account for induced travel as evidence that induced travel isn't real. If there is a discrepancy between the research literature and the model, the model should be fixed so that it matches the data.

Some modelers defend these model errors by saying the models are accounting for latent demand. This is a fantasy which leads to wasteful spending. I recently was talking to a senior modeling consultant who does a lot of work for a large state DOT. He complained that the DOT requires them to design for latent demand even when this demand could never occur due to upstream bottlenecks that will not be addressed.

Norm Marshall, Smart Mobility Inc.

nmarshall@smartmobility.com

--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

vipoqijyt

The terms induced and latent demand seem to be stirring up a semantic brawl
here.

The issue is really about understanding that when you make public
investments in infrastructure, whether it be in roads or rail, you change
the relative prices of travel at different times of day, modes, and routes
(Anthony Downs' triple convergence). In the longer-term you also change
relative prices via multi-modal accessibility changes that influence
outcomes such as location choices of households and firms, rents and prices
in the real estate market, and real estate development feasibility in
different locations. It is not really so helpful to quibble about whether
these impacts are induced or latent demand - but just to acknowledge that
there is considerable evidence that these feedbacks from shifts in relative
prices of travel have short-term and long-term impacts, and that ignoring
them can lead to overly optimistic assessments of the impacts on congestion
of road projects, and under-estimates of the ridership impacts of rail
projects, depending on many other factors as well. That is the case for
increasing the use of land use models including UrbanSim, tightly coupled
with travel models, to capture these short and long-term impacts.

Paul

On Wed, May 15, 2019 at 8:55 AM nytydewuq wrote:

> When the price of any good drops, the quantity demanded increases – demand
> does not increase, i.e. the demand curve does not shift.
>
> From: nmarshall=smartmobility.com@mg.tmip.org On Behalf Of nmarshallvt
> Sent: Wednesday, May 15, 2019 7:59 AM
> To: TMIP
> Subject: Re: [TMIP] Induced demands
>
> Induced travel is a real world phenomenon based on economic theory that
> has been studied extensively. When the price of any good drops, the demand
> increases. For road travel, most consumers consider the travel time the
> largest component of the "price". Therefore, if travel time drops, demand
> increases. We call this induced travel.
>
> Susan Handy and Marlon Boarnet summarized the research literature in 2014.
> They conclude:
>
> "Thus, the best estimate for the long-run effect of highway capacity on
> VMT is an elasticity close to 1.0, implying that in congested metropolitan
> areas, adding new capacity to the existing system of limited-access
> highways is unlikely to reduce congestion or associated GHG in the
> long-run."
>
> https://www.arb.ca.gov/cc/sb375/policies/policies.htm
>
> https://www.arb.ca.gov/cc/sb375/policies/hwycapacity/highway_capacity_bk...
>
>
> Induced travel studies look at changes in real world travel in before and
> after conditions. In contrast, the term "latent demand" wants us to instead
> focus on an impossible fantasy where there is no congestion. This has not
> occurred and cannot occur in the real world. Therefore,
>
> latent demand cannot be quantified or studied. When networks without
> capacity constraints have been modeled, it has been done to demonstrate
> that building our way out of congestion is impossible.
>
> Joe Cortright recently made an instructive post using the analogy of
> "free" ice cream day. Along with a photo of a long line of people queued up
> waiting for their "free" ice cream, he writes:
>
> "As you’re standing in line waiting for your “free” ice cream cone, give a
> little thought to the parallels between that line and your typical rush
> hour traffic jam. In both cases, you’re waiting in line for the same
> reason–the price is too low, and demand is overwhelming supply. This is the
> valuable lesson that Ben and Jerry are providing in the fundamentals of
> transportation economics."
>
> http://cityobservatory.org/2019/04/
>
> Going back to the original question about modeling and induced travel,
> Mark Bradley suggested that travel demand models account for induced route
> choice and induced destination choice. As I discussed in a recent thread on
> this forum, static assignment does not properly account for induced route
> choice because it does not constrain flow to capacity. Then, the resulting
> travel time matrices underestimate the affects of congestion, so that the
> models also fail to properly account for induced destination choice. These
> problems can be corrected by replacing the static assignment with regional
> DTA with feedback.
>
> Some modelers cite the failure of the static assignment models to account
> for induced travel as evidence that induced travel isn't real. If there is
> a discrepancy between the research literature and the model, the model
> should be fixed so that it matches the data.
>
> Some modelers defend these model errors by saying the models are
> accounting for latent demand. This is a fantasy which leads to wasteful
> spending. I recently was talking to a senior modeling consultant who does a
> lot of work for a large state DOT. He complained that the DOT requires them
> to design for latent demand even when this demand could never occur due to
> upstream bottlenecks that will not be addressed.
>
> Norm Marshall, Smart Mobility Inc.
>
> nmarshall@smartmobility.com
>
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>

covomopez

Interestingly the definition of induced demand in UK DfT guidance (and practice) is NOT basedon changes in VMT resulting but on changes to the Demand Matrix (i.e the numbers of trips from A to B).This would include destination choice,  mode choice and time period choice, as well as trip generation arising from policy induced changes in land use.  But is would exclude route choice.Mark Bradley put this all much better earlier on in this discussion.

I doubt the validity of unit demand elasticities in response to capacity improvements and even if there aretransferable elasticites of this kind. 
Both cross-sectional and longitudinal studes are needed to establishrealistic elasticities for each stage of the demand model, as well as robustnesstests of alternative specifications as an integral part of model calibration.
But I think this is all quite well established. Admittedly the terminology is not used consistentlyacross the profession and in different countries, so the context needs to be set out clearly ininternational discussion groups. 
The current dialogue makes an excellent contribution to this process. 
Mr G HymanGeoffrey Hyman Consultancy.(previously UK DfT advisor on modelling and appraisal)
-----Original Message-----
From: nytydewuq
To: TMIP
Sent: Wed, 15 May 2019 16:54
Subject: Re: [TMIP] Induced demands

When the price of any good drops, the quantity demanded increases – demand does not increase, i.e. the demand curve does not shift.From: nmarshall=smartmobility.com@mg.tmip.org On Behalf Of nmarshallvt
Sent: Wednesday, May 15, 2019 7:59 AM
To: TMIP
Subject: Re: [TMIP] Induced demandsInduced travel is a real world phenomenon based on economic theory that has been studied extensively. When the price of any good drops, the demand increases. For road travel, most consumers consider the travel time the largest component of the "price". Therefore, if travel time drops, demand increases. We call this induced travel.Susan Handy and Marlon Boarnet summarized the research literature in 2014. They conclude:"Thus, the best estimate for the long-run effect of highway capacity on VMT is an elasticity close to 1.0, implying that in congested metropolitan areas, adding new capacity to the existing system of limited-access highways is unlikely to reduce congestion or associated GHG in the long-run."https://www.arb.ca.gov/cc/sb375/policies/policies.htmhttps://www.arb.ca.... travel studies look at changes in real world travel in before and after conditions. In contrast, the term "latent demand" wants us to instead focus on an impossible fantasy where there is no congestion. This has not occurred and cannot occur in the real world. Therefore,latent demand cannot be quantified or studied. When networks without capacity constraints have been modeled, it has been done to demonstrate that building our way out of congestion is impossible.Joe Cortright recently made an instructive post using the analogy of "free" ice cream day. Along with a photo of a long line of people queued up waiting for their "free" ice cream, he writes:"As you’re standing in line waiting for your “free” ice cream cone, give a little thought to the parallels between that line and your typical rush hour traffic jam. In both cases, you’re waiting in line for the same reason–the price is too low, and demand is overwhelming supply. This is the valuable lesson that Ben and Jerry are providing in the fundamentals of transportation economics."http://cityobservatory.org/2019/04/Going back to the original question about modeling and induced travel, Mark Bradley suggested that travel demand models account for induced route choice and induced destination choice. As I discussed in a recent thread on this forum, static assignment does not properly account for induced route choice because it does not constrain flow to capacity. Then, the resulting travel time matrices underestimate the affects of congestion, so that the models also fail to properly account for induced destination choice. These problems can be corrected by replacing the static assignment with regional DTA with feedback.Some modelers cite the failure of the static assignment models to account for induced travel as evidence that induced travel isn't real. If there is a discrepancy between the research literature and the model, the model should be fixed so that it matches the data.Some modelers defend these model errors by saying the models are accounting for latent demand. This is a fantasy which leads to wasteful spending. I recently was talking to a senior modeling consultant who does a lot of work for a large state DOT. He complained that the DOT requires them to design for latent demand even when this demand could never occur due to upstream bottlenecks that will not be addressed.Norm Marshall, Smart Mobility Inc.nmarshall@smartmobility.com--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

Alex Bettinardi

I (personally) do not agree with your (Norm’s) argument or definitions.
But it is not up to me to decide, it’s up to the industry.

I would be curious to poll the modelers, planners, engineers and see where the industry is on this? Maybe TMIP could construct such a (quick) survey.

Alex Bettinardi, P.E.
503.986.4104
http://www.oregon.gov/ODOT/Planning/Pages/default.aspx

From: nmarshall=smartmobility.com@mg.tmip.org On Behalf Of nmarshallvt
Sent: Wednesday, May 15, 2019 7:59 AM
To: TMIP
Subject: Re: [TMIP] Induced demands

Induced travel is a real world phenomenon based on economic theory that has been studied extensively. When the price of any good drops, the demand increases. For road travel, most consumers consider the travel time the largest component of the "price". Therefore, if travel time drops, demand increases. We call this induced travel.

Susan Handy and Marlon Boarnet summarized the research literature in 2014. They conclude:

"Thus, the best estimate for the long-run effect of highway capacity on VMT is an elasticity close to 1.0, implying that in congested metropolitan areas, adding new capacity to the existing system of limited-access highways is unlikely to reduce congestion or associated GHG in the long-run."

https://www.arb.ca.gov/cc/sb375/policies/policies.htm

https://www.arb.ca.gov/cc/sb375/policies/hwycapacity/highway_capacity_bk...

Induced travel studies look at changes in real world travel in before and after conditions. In contrast, the term "latent demand" wants us to instead focus on an impossible fantasy where there is no congestion. This has not occurred and cannot occur in the real world. Therefore,

latent demand cannot be quantified or studied. When networks without capacity constraints have been modeled, it has been done to demonstrate that building our way out of congestion is impossible.

Joe Cortright recently made an instructive post using the analogy of "free" ice cream day. Along with a photo of a long line of people queued up waiting for their "free" ice cream, he writes:

"As you’re standing in line waiting for your “free” ice cream cone, give a little thought to the parallels between that line and your typical rush hour traffic jam. In both cases, you’re waiting in line for the same reason–the price is too low, and demand is overwhelming supply. This is the valuable lesson that Ben and Jerry are providing in the fundamentals of transportation economics."

http://cityobservatory.org/2019/04/

Going back to the original question about modeling and induced travel, Mark Bradley suggested that travel demand models account for induced route choice and induced destination choice. As I discussed in a recent thread on this forum, static assignment does not properly account for induced route choice because it does not constrain flow to capacity. Then, the resulting travel time matrices underestimate the affects of congestion, so that the models also fail to properly account for induced destination choice. These problems can be corrected by replacing the static assignment with regional DTA with feedback.

Some modelers cite the failure of the static assignment models to account for induced travel as evidence that induced travel isn't real. If there is a discrepancy between the research literature and the model, the model should be fixed so that it matches the data.

Some modelers defend these model errors by saying the models are accounting for latent demand. This is a fantasy which leads to wasteful spending. I recently was talking to a senior modeling consultant who does a lot of work for a large state DOT. He complained that the DOT requires them to design for latent demand even when this demand could never occur due to upstream bottlenecks that will not be addressed.

Norm Marshall, Smart Mobility Inc.

nmarshall@smartmobility.com

--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

Alex Bettinardi

Words matter – this is about clarity in messaging.
Which is why it stirs up such strong feelings on both sides.

Alex Bettinardi, P.E.
503.986.4104
http://www.oregon.gov/ODOT/Planning/Pages/default.aspx

From: waddell=berkeley.edu@mg.tmip.org On Behalf Of vipoqijyt
Sent: Wednesday, May 15, 2019 10:59 AM
To: TMIP
Subject: Re: [TMIP] Induced demands

The terms induced and latent demand seem to be stirring up a semantic brawl
here.

The issue is really about understanding that when you make public
investments in infrastructure, whether it be in roads or rail, you change
the relative prices of travel at different times of day, modes, and routes
(Anthony Downs' triple convergence). In the longer-term you also change
relative prices via multi-modal accessibility changes that influence
outcomes such as location choices of households and firms, rents and prices
in the real estate market, and real estate development feasibility in
different locations. It is not really so helpful to quibble about whether
these impacts are induced or latent demand - but just to acknowledge that
there is considerable evidence that these feedbacks from shifts in relative
prices of travel have short-term and long-term impacts, and that ignoring
them can lead to overly optimistic assessments of the impacts on congestion
of road projects, and under-estimates of the ridership impacts of rail
projects, depending on many other factors as well. That is the case for
increasing the use of land use models including UrbanSim, tightly coupled
with travel models, to capture these short and long-term impacts.

Paul

On Wed, May 15, 2019 at 8:55 AM nytydewuq wrote:

> When the price of any good drops, the quantity demanded increases – demand
> does not increase, i.e. the demand curve does not shift.
>
> From: nmarshall=smartmobility.com@mg.tmip.org On Behalf Of nmarshallvt
> Sent: Wednesday, May 15, 2019 7:59 AM
> To: TMIP
> Subject: Re: [TMIP] Induced demands
>
> Induced travel is a real world phenomenon based on economic theory that
> has been studied extensively. When the price of any good drops, the demand
> increases. For road travel, most consumers consider the travel time the
> largest component of the "price". Therefore, if travel time drops, demand
> increases. We call this induced travel.
>
> Susan Handy and Marlon Boarnet summarized the research literature in 2014.
> They conclude:
>
> "Thus, the best estimate for the long-run effect of highway capacity on
> VMT is an elasticity close to 1.0, implying that in congested metropolitan
> areas, adding new capacity to the existing system of limited-access
> highways is unlikely to reduce congestion or associated GHG in the
> long-run."
>
> https://www.arb.ca.gov/cc/sb375/policies/policies.htm
>
> https://www.arb.ca.gov/cc/sb375/policies/hwycapacity/highway_capacity_bk...
>
>
> Induced travel studies look at changes in real world travel in before and
> after conditions. In contrast, the term "latent demand" wants us to instead
> focus on an impossible fantasy where there is no congestion. This has not
> occurred and cannot occur in the real world. Therefore,
>
> latent demand cannot be quantified or studied. When networks without
> capacity constraints have been modeled, it has been done to demonstrate
> that building our way out of congestion is impossible.
>
> Joe Cortright recently made an instructive post using the analogy of
> "free" ice cream day. Along with a photo of a long line of people queued up
> waiting for their "free" ice cream, he writes:
>
> "As you’re standing in line waiting for your “free” ice cream cone, give a
> little thought to the parallels between that line and your typical rush
> hour traffic jam. In both cases, you’re waiting in line for the same
> reason–the price is too low, and demand is overwhelming supply. This is the
> valuable lesson that Ben and Jerry are providing in the fundamentals of
> transportation economics."
>
> http://cityobservatory.org/2019/04/
>
> Going back to the original question about modeling and induced travel,
> Mark Bradley suggested that travel demand models account for induced route
> choice and induced destination choice. As I discussed in a recent thread on
> this forum, static assignment does not properly account for induced route
> choice because it does not constrain flow to capacity. Then, the resulting
> travel time matrices underestimate the affects of congestion, so that the
> models also fail to properly account for induced destination choice. These
> problems can be corrected by replacing the static assignment with regional
> DTA with feedback.
>
> Some modelers cite the failure of the static assignment models to account
> for induced travel as evidence that induced travel isn't real. If there is
> a discrepancy between the research literature and the model, the model
> should be fixed so that it matches the data.
>
> Some modelers defend these model errors by saying the models are
> accounting for latent demand. This is a fantasy which leads to wasteful
> spending. I recently was talking to a senior modeling consultant who does a
> lot of work for a large state DOT. He complained that the DOT requires them
> to design for latent demand even when this demand could never occur due to
> upstream bottlenecks that will not be addressed.
>
> Norm Marshall, Smart Mobility Inc.
>
> nmarshall@smartmobility.com
>
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>
--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

Becky_Knudson

Congestion delay is a price mechanism in units of time. Users pay the cost willingly (albeit begrudgingly). When time-costs rise beyond their ability or willingness to pay, they make different choices and adapt to fit their time/money budgets and preferences. Transportation is only one among many time/budget expenditure decisions made by households and businesses.

I understand the passion and interest in discussing latent demand, induced demand, pent up demand, and so on. However, at times I am confounded by the “bad” reputation associated with this topic. The purpose of a transportation system is to support economic activity. Building capacity to accommodate economic activity is the responsibility of the public sector. Challenges arise when users are not paying the full cost of the system (quantity demanded is greater than quantity supplied) and negative externalities arise, such as emissions, noise, unreliability, property damage, personal injury and fatalities.

This is a very complex problem to solve with layers of trade-offs. It is impractical to focus on only one aspect of the system. For example, if safety were our only concern, we would close the roads to all users. If vehicle emissions were our only concern, we would require all vehicles use emission free power sources. If freight movement was our only concern, we would optimize system use to meet that goal. We can’t apply simple solutions to a complex problem caused by millions of independent economic agents with unique needs and preferences.

“Free” ice cream is not free, the units of cost merely shift from cash to time. However, people are free to choose to wait in line. When I was young and poor, I waited in line for free ice cream, now I don’t.

Becky Knudson || Senior Transportation Economist || Oregon DOT || TPAU || Phone: 503.986.4113 || rebecca.a.knudson@odot.state.or.us ||
From: eanderson=azmag.gov@mg.tmip.org On Behalf Of nytydewuq
Sent: Wednesday, May 15, 2019 8:54 AM
To: TMIP
Subject: Re: [TMIP] Induced demands

When the price of any good drops, the quantity demanded increases – demand does not increase, i.e. the demand curve does not shift.

From: nmarshall=smartmobility.com@mg.tmip.org On Behalf Of nmarshallvt
Sent: Wednesday, May 15, 2019 7:59 AM
To: TMIP
Subject: Re: [TMIP] Induced demands

Induced travel is a real world phenomenon based on economic theory that has been studied extensively. When the price of any good drops, the demand increases. For road travel, most consumers consider the travel time the largest component of the "price". Therefore, if travel time drops, demand increases. We call this induced travel.

Susan Handy and Marlon Boarnet summarized the research literature in 2014. They conclude:

"Thus, the best estimate for the long-run effect of highway capacity on VMT is an elasticity close to 1.0, implying that in congested metropolitan areas, adding new capacity to the existing system of limited-access highways is unlikely to reduce congestion or associated GHG in the long-run."

https://www.arb.ca.gov/cc/sb375/policies/policies.htm

https://www.arb.ca.gov/cc/sb375/policies/hwycapacity/highway_capacity_bk...

Induced travel studies look at changes in real world travel in before and after conditions. In contrast, the term "latent demand" wants us to instead focus on an impossible fantasy where there is no congestion. This has not occurred and cannot occur in the real world. Therefore,

latent demand cannot be quantified or studied. When networks without capacity constraints have been modeled, it has been done to demonstrate that building our way out of congestion is impossible.

Joe Cortright recently made an instructive post using the analogy of "free" ice cream day. Along with a photo of a long line of people queued up waiting for their "free" ice cream, he writes:

"As you’re standing in line waiting for your “free” ice cream cone, give a little thought to the parallels between that line and your typical rush hour traffic jam. In both cases, you’re waiting in line for the same reason–the price is too low, and demand is overwhelming supply. This is the valuable lesson that Ben and Jerry are providing in the fundamentals of transportation economics."

http://cityobservatory.org/2019/04/

Going back to the original question about modeling and induced travel, Mark Bradley suggested that travel demand models account for induced route choice and induced destination choice. As I discussed in a recent thread on this forum, static assignment does not properly account for induced route choice because it does not constrain flow to capacity. Then, the resulting travel time matrices underestimate the affects of congestion, so that the models also fail to properly account for induced destination choice. These problems can be corrected by replacing the static assignment with regional DTA with feedback.

Some modelers cite the failure of the static assignment models to account for induced travel as evidence that induced travel isn't real. If there is a discrepancy between the research literature and the model, the model should be fixed so that it matches the data.

Some modelers defend these model errors by saying the models are accounting for latent demand. This is a fantasy which leads to wasteful spending. I recently was talking to a senior modeling consultant who does a lot of work for a large state DOT. He complained that the DOT requires them to design for latent demand even when this demand could never occur due to upstream bottlenecks that will not be addressed.

Norm Marshall, Smart Mobility Inc.

nmarshall@smartmobility.com

--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

Alex Bettinardi

As the lighter of this fire storm, I feel I have the self-proclaimed right of deciding the winner.
Nicely done Matthew. That was the best response on the thread/topic.

While the response doesn’t determine the winner (is it latent or induced); I think it does the best job of redirecting and addressing the real factors at play. I was much in the same headspace as I started to reconsider all these arguments. I went back to other utilities in my mind. What if water became so cheap that you could dump it on your lawn for hours on end with no economic feedback; would that be latent or induced – does it matter. What does society determine is the correct amount of water use per household before the extra use of water becomes latent or demand?

Conclusion it doesn’t matter. So I vote that we, as an industry, talk about these issues more constructively as other’s on the thread have proposed. So here’s the deal – I’ll lay down my weapons and go down this path of peace and harmony where we all work towards solutions that benefit society. I will rest peacefully knowing that we (all of us) will no longer will throw around the terms latent and induced so freely and without full consideration of what those terms really mean.

Ronald Milam – in this new space of peace and reduced usage of latent and induced, I suggest that your report (which is probably amazing) needs a title revision, and probably a global find and replace.

Alex Bettinardi, P.E.
503.986.4104
http://www.oregon.gov/ODOT/Planning/Pages/default.aspx

From: mkitch67=gmail.com@mg.tmip.org On Behalf Of quxoxalow
Sent: Wednesday, May 15, 2019 8:51 AM
To: TMIP
Subject: Re: [TMIP] Induced demands

It seems to me that both terms, "induced" and "latent", are problematic in
this context. Both imply that but for some action on the part of a road
authority demand is somehow static. Demand is always a function of a
variety of factors including preferences, income and most immediately,
price. The very notion of a demand curve implies that as price (including
time costs of congestion delays) changes demand will also change.

If building, or widening, roads induces demand (this sounds like a
conjuror's trick) then it must also be true that failing to build, or widen
roads (at no direct cost to the users) must suppress demand. Both notions
seem a bit absurd. As a supplier of roads the authority cannot define the
demand conditions in the market, but rather responds, or does not respond,
to those conditions.

If road authorities, as public monopolists, behaved as private monopolists
then they would likely under-supply the market and charge monopolist
prices, resulting in larger than normal profits. Instead they mostly
under-supply the market and charge no prices, resulting in larger than
normal levels of congestion.

From a modeling perspective it seems that Mark Bradley and John Abraham
have covered the main approaches in practice designed to more
comprehensively handle the full range of travel demand responses.

Matthew

On Wed, May 15, 2019 at 8:00 AM Alex Bettinardi <
alexander.o.bettinardi@odot.state.or.us> wrote:

> Thank you Bernie,
>
> You point out a good historic example, I think that helps explain how the
> industry got so used to using induced demand as a term to cover all
> shifting. I think you capture this correctly, that here is a pie chart for
> any given new roadway or lanes or new land development. There is a percent
> of that pie that is latent and a percent of the pie chart that is induced.
>
> I am urging the industry to acknowledge that where we currently at a point
> in history (at least in the US) where we don’t expand to the degree where
> the majority of the pie is induce. In almost all projects we are now
> primarily latent due to the high congestion in our urban areas. Historic
> examples might indeed show that demand is induced, but that is now the
> minority as opposed to latent.
>
> I get what you are saying that there is a mix. I propose that the high
> majority of common day examples is latent. I think what would be the best
> solution would be to properly talk about how much of the demand is latent
> and how much is induced. I think a next best solution might be to refer to
> this as “latent / induced demand” and make it clear that it’s some
> percentage of both. I think the worst case is that everything is labeled as
> induced. I think in present day labeling everything as latent is a better
> (more correct) solution than labeling everything as induced, since latent
> is the majority, but I get that is a judgment call based on what type of
> projects you deal with and see on a day-to-day basis and where you are on
> the globe.
>
> So maybe we can all agree to lengthen the title a little bit and go with -
> “latent / induced demand”.
>
> Alex Bettinardi, P.E.
> 503.986.4104
> http://www.oregon.gov/ODOT/Planning/Pages/default.aspx
>
> From: ברנרד אלפרן
> Sent: Wednesday, May 15, 2019 5:06 AM
> To: BETTINARDI Alexander O * Alex ; tmip@mg.tmip.org
> Subject: FW: [TMIP] Induced demands
>
> Hi Alex,
>
> I completely understand (and agree with) your desire for the industry to
> clearly define the terms it uses, but (very) respectfully disagree with
> your suggestion.
>
> One of the primary meanings of “induce” is “to cause something to happen”
> (see https://dictionary.cambridge.org/dictionary/english/induce). That
> “something” could be something new, not necessarily something that was
> “suppressed” by the lack of a cause (or inducement). A classic example in
> the transportation field is the Long Island Expressway (the current
> official name of I-495 in Nassau and Suffolk Counties, New York). This
> 71-mile east-west highway across Queens, Nassau, and Suffolk Counties was
> constructed in stages. The westernmost (closest to Manhattan) section
> opened to traffic in 1940 and the easternmost section was opened in 1972.
> During the course of those 32 years, the highway induced significantly
> faster-paced and lower density development in Eastern Queens, and in Nassau
> and Suffolk Counties than could or would have happened otherwise. Legend
> has it that travel times from Long Island to Manhattan were no shorter on
> the day of the road’s completion than on the day that its construction
> started. Traffic reporters in the 1970’s would regularly refer to the Long
> Island Expressway as “the big LIE” (some referred to it as “the world’s
> longest parking lot”).
>
> Clearly, the bulk of the trips congesting the Expressway had not been
> suppressed since before 1940. They were new trips, “induced” by the faster
> and lower density development on Long Island that was, in part, caused by
> the highway. They weren’t “latent” trips, they were new trips that didn’t
> exist before.
>
> Very often when a transportation facility construction or expansion
> results in higher amounts of travel in the corridor served, it is difficult
> to parse out how much of that increased travel demand is latent, and how
> much is induced (and how much is simply diverted from other parallel
> facilities). Hence the confusion. As long as we understand that the
> increase in demand has multiple components, that’s probably the best we can
> do.
>
> It’s not for nothing that a very senior manager at USDOT was known to say
> “whoever labelled Economics ‘the dismal science’ never heard of travel
> demand forecasting”. It is, and always will be, at least as much art as
> science.
>
> Yours truly,
>
> Bernie Alpern
>
> Bernard Alpern, Transportation Planner
> Eyal Kraus Roadway and Traffic Engineers
> Beit Ha-Dfus Street 22, 2nd floor
> Jerusalem, Israel 9548326
> Office: +972-2-532-8814 Direct: +972-2-501-1754 Mobile: +972-58-699-9391
> alpern@ekroads.co.il
> [cid:image001.png@01D2FFE1.A915AC20]
>
> ---------- Forwarded message ---------
> From: Alex Bettinardi >
> Date: Wed, May 15, 2019 at 1:41 PM
> Subject: Re: [TMIP] Induced demands
> To: TMIP >
>
> Not an answer to the question; but a desire for our industry to clarify
> these terms / definitions:
>
> I pose that we are not using the term "induced" demand correctly:
> See the definitions linked below or complete your own search in google or
> a physical dictionary.
>
> Induce -
> https://www.google.com/search?source=hp&ei=hSDbXNvAB_6w0PEPooqvyAw&q=def
> ...
>
> Latent -
> https://www.google.com/search?q=define+latent&rlz=1C1GCEA_enUS806US806&o
> ...
>
> While there may be cases of truly induced demand (like something special
> like an IKEA shows up, and you flood the store to buy a bunch of furniture
> and meatballs you didn't realize you needed), in almost all cases what we
> are really talking about is latent demand. Demand that has been suppressed
> (hidden / concealed) by high congestion and is now returning with the new
> capacity. This is especially the case in the example given below where the
> transportation infrastructure is being expanded and immediately fills up.
> The new highway isn't attracting new trips, it is serving trips that were
> previously put off or served in a different way.
>
> My office would appreciate it if the industry could correct it's use of
> these terms.
>
> Alex Bettinardi, P.E.
> 503.986.4104
> http://www.oregon.gov/ODOT/Planning/Pages/default.aspx
>
> From: easall=gmail.com@mg.tmip.org On Behalf Of elizabethsall
> Sent: Tuesday, May 14, 2019 1:04 PM
> To: TMIP
> Subject: Re: [TMIP] Induced demands
>
> Harun,
> The California Governor's Office of Planning and Research compiled some
> resources on induced demand from
> increases in highway capacity (see drop-down for "Induced VMT from Highway
> Capacity").
>
> Whether or not a travel model captures induced demand depends on the model
> and how they are set up. Some do not capture any dimensions you mentioned.
> Some capture all.
>
> Even if the model can mechanistically capture induced travel, it should be
> validated against external datapoints.
>
> Cheers,
> --Elizabeth Sall
>
> On Mon, May 13, 2019 at 12:47 PM hicecesap <
> Harun.Rashid@thenovaauthority.org> wrote:
>
> > TMIPers, greetings!
> > In a congested urban roadway network, if new capacity is added, there
> will
> > be new travel demand and soon the new facility will be as (or more)
> > congested.
> > Paraphrasing what I read on "induced demands" in this forum, or
> elsewhere.
> > It's been a while this topic has been discussed here, and I couldn't find
> > any formal synthesis of the past discussions. If I remember correctly,
> few
> > members suggested that such phenomenon can be analyzed within a travel
> > demand model framework, of course a whole lot depends on how you define
> > 'new' demand - a. same number of overall trips, but shifts in mode
> choice;
> > b. new trips from latent demands, trips that otherwise wouldn't exist
> > without new capacity; c. new trips from land developments due to added
> > roadway capacity. This is not exhaustive, there can be other
> > iterations/possibilities for new demands.
> >
> > My objective is to understand current practices/experiences to address
> > this issue - by regional transportation planners using a travel demand
> > model. Appreciate any responses, I will provide an anonymized summary.
> Feel
> > free to respond directly to me (my contact below). Thanks in advance!
> >
> > Harun.
> >
> > [cid:image002.jpg@01D3EDF5.3282CE00]
> >
> > HARUN RASHID, AICP
> > Transportation Planner
> > Northern Virginia Transportation Authority
> > 3040 Williams Drive, Suite 200, Fairfax, VA 22031.
> > Harun.Rashid@thenovaauthority.org
> > Office: 703-642-4659 | Cell: 571-355-3938
> > www.TheNoVaAuthority.org
> > --
> > Full post: https://tmip.org/content/induced-demands
> > Manage my subscriptions: https://tmip.org/mailinglist
> > Stop emails for this post:
> https://tmip.org/mailinglist/unsubscribe/13101
> >
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>
--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101

KenCervenka

The following observations are not about “how to predict” changes over time, but rather how to analyze actual observed changes over time. But it is not clear to me there is “some other listserv” that could offer anything useful on this subject, so….

Perhaps this particular thread has run its course, and there is nothing more to be said that can lead to anything constructive, but I am (nevertheless) struck by the following:
-- Rebecca noted that “The purpose of a transportation system is to support economic activity.”
-- Norm offered up some real-world examples where major (and immediate) reductions of road capacity in congested areas did not lead to any major traffic problems, and asserts that “when urban freeways have been removed, there have been great benefits to the cities.”

First question, which I suppose is primarily for Norm: what is your definition of a quantifiable “benefit” to a city, and what (if any) relationship does this have to “increased economic activity” for that city?

The second set of questions are for anyone who dares to answer, or at least offer comment:
I did a quick-check of Inrix’s web site for the findings from their “2018 Global Traffic Scorecard,” and observe 115 urban areas in the world that had more than 100 hours lost to the average driver in 2018, due to traffic congestion (Bogota is the “winner” with 272 hours). Now I would surely think these 115 urban areas had some very different levels of population growth over the past 20 (or so) years, as well as very different levels of increases (maybe even a few decreases) in free or tolled road capacity over the same number of years. Plus, for that matter, some very different levels of “transit service improvements” over the same number of years. But has anyone tried to do some type of “comparative analysis” that attempts to identify what (if any) aggregate relationship exists in regards to increases (or decreases) in actual road capacity, traffic congestion, population growth, and a clear description of whatever constitutes “increased economic activity”? Or some other definition for “benefit”? Not what comes out of model-based scenario/sensitivity testing, but rather focused heavily on actual observed changes, for dozens of different urban areas? I think it may be close to impossible to draw useful observation-based conclusions because no one will ever really know how much (for example) regional population growth (or “economic activity”) in a specific urban area might have been different if “actual increases in road capacity” were either more or less than what was actually provided, but surely there is some academic researcher out there that has at least pursued this type of exploratory investigation in a meaningful “comparison of different urban areas” way?

About all I can conclude right now is that I am quite happy to not be a transportation planner in Bogota, or in any of the “top 20” congested urban areas in the world, who tells the local decision-makers to not “worry” about traffic congestion over the next 10 years, in the face of continued increases in urban population over those same 10 years, because everything will eventually work itself out, without new transportation investments! Or that it would be even better to reduce road capacity.

*****************************
From: rebecca.a.knudson
Sent: Friday, May 17, 2019 3:23 PM
Congestion delay is a price mechanism in units of time. Users pay the cost willingly (albeit begrudgingly). When time-costs rise beyond their ability or willingness to pay, they make different choices and adapt to fit their time/money budgets and preferences. Transportation is only one among many time/budget expenditure decisions made by households and businesses.

I understand the passion and interest in discussing latent demand, induced demand, pent up demand, and so on. However, at times I am confounded by the “bad” reputation associated with this topic. The purpose of a transportation system is to support economic activity. Building capacity to accommodate economic activity is the responsibility of the public sector. Challenges arise when users are not paying the full cost of the system (quantity demanded is greater than quantity supplied) and negative externalities arise, such as emissions, noise, unreliability, property damage, personal injury and fatalities.

This is a very complex problem to solve with layers of trade-offs. It is impractical to focus on only one aspect of the system. For example, if safety were our only concern, we would close the roads to all users. If vehicle emissions were our only concern, we would require all vehicles use emission free power sources. If freight movement was our only concern, we would optimize system use to meet that goal. We can’t apply simple solutions to a complex problem caused by millions of independent economic agents with unique needs and preferences.

“Free” ice cream is not free, the units of cost merely shift from cash to time. However, people are free to choose to wait in line. When I was young and poor, I waited in line for free ice cream, now I don’t.

From: Norm Marshall
Sent: Friday, May 17, 2019 3:22 PM
I appreciate that reduced highway capacity has been brought into this discussion. When urban freeways have been removed, there have been great benefits to the cities and the expected traffic problems have failed to materialize. This has surprised traffic engineers. Sam Schwartz wrote in his 2015 book Street Smart: The Rise of Cities and the Fall of Cars:

"On Saturday morning, December 15, 1973, the forty-year-old West Side Highway, which runs along the Hudson River and connects the Brooklyn Battery Tunnel with the Henry Hudson Parkway— both built by the unstoppable Robert Moses— collapsed under the weight of a truck carrying more than thirty tons of asphalt. ... The immediate problem— What to do with eighty thousand cars a day that would have to find an alternate route?— fell to me."

"I would love to take credit for coming up with a brilliant solution that saved the city, but the truth is both more mundane and a lot more interesting. The predicted traffic disaster never appeared. Somehow, those eighty thousand cars went somewhere, but to this day we have no idea where. Or how, two years later, twenty-five thousand more people were getting into Manhattan’s Central Business District."

Traffic similarly "disappeared" when the earthquake-damaged Central Freeway in San Francisco was removed in 1996.

Inspired by the success of both unplanned and intentional urban freeway removals in the U.S. and abroad, many initiatives are underway with goals of removing urban freeways. For example, NYSDOT recently selected a "Community Grid" alternative as the preferred choice to replace elevated I-81 through central Syracuse.

A major obstacle to these projects is that just as the static assignment models fail to properly account for induced travel, they also fail to properly account for reduced travel. Studies with static models generally just show that the existing traffic won't fit on a downsized facility, and fail to account fully either for traffic re-assignment or changes in destinations.

In 2016, I analyzed a range of alternatives in Little Rock ranging from freeway expansion to removal using Dynamic Traffic Assignment (DTA). The modeling showed significant induced travel with freeway expansion which would create new bottlenecks beyond the limits of construction. The modeling also showed that the downsizing alternative would not cause significant traffic problems.

The report is here:

http://www.arpanel.org/improve30crossing (especially the "Second Report")

I presented this work at the 2017 TRB Planning Applications Conference. The abstract is here:

https://www.trbappcon.org/oldsite/2017conf/PresentationDetailsca73.html?...

The link to the presentation no longer works.

Norm Marshall

nmarshall@smartmobility.com

--

nmarshallvt

Ken asks about benefits of freeway removal. I second Juan de Dios Ortuzar's recommendation of the wikipedia site https://en.wikipedia.org/wiki/Freeway_removal. It has links to completed freeway removal projects and also projects that are under discussion. 

There are few cases where a freeway carrying a large traffic volume has been removed in the U.S. - the West Side Highway in New York City, the Central and Emarcadero Freeways in San Franciso, and the Central Artery in Boston (the "Big Dig"). The immediate areas around all of these are thriving economically. The west side of Manhattan has very high property values and the $20 billion Hudson Yards project is under construction. The Embarcadero area is one of the great tourist attractions in the U.S., with 15 million visitors annually. A quick Google search finds that 1-bedroom condos on Octavia Boulevard (which replaced the Central Freeway) are priced at about $1 million. There has been a huge real estate boom in the part of Boston opened up by the Big Dig. In 2015, the Boston Globe wrote: "The Shawmut Peninsula is some of the most sought-after and desirable urban real estate in the country." Obviously not all of this economic success in these 3 properous cities is because of the freeway removal, but freeway removal also is not holding the areas back.

Each of these cases is unique, but the results likely are at least partially transferable to other cities with hot real estate markets. For example, advocates of downsizing I-30 in Austin have estimated that their conceptual plan would free up a large amount of very valuable land for development as well as increasing property values substantially in areas proximate to I-30.

The transformation could be even greater in cities that are not thriving, e.g. Syracuse where it is likely that the elevated freeway will be removed. I have done a fair amount of work there and have always been struck by what a poor decision it was to build an elevated highway between the CBD on the west and Syracuse University and medical district on the east. Cut off from their natural customers, the CBD became depressed. If the freeway comes down, I predict an urban renaissance will occur there over the subsequent 20 years.

Ken's mention of Inrix is giving me an excuse to plug another one of my papers. I didn't try to understand international differences, but I did a cross-sectional analysis of U.S. urban areas combining INRIX data with data take from TTI's Urban Mobility Study, and presented the paper at the 2016 TRB Annual Meeting.

https://trid.trb.org/view/1392295

Here's an excerpt from the abstract: In a regression analysis of 74 regions, it is found that more arterial capacity is strongly related to less congestion, but that more freeway capacity is not. The public policy implications are that it is critical that an adequate network of streets be constructed in growing areas rather than relying too much on a system of freeways. In already-congested areas, arterial capacity improvements likely would be more effective at reducing congestion than adding freeway capacity. Otherwise, the regression model suggests that congestion is more a sign of regional success than a problem than can be solved. Only two other independent variables were found to be highly significant in predicting congestion. Higher incomes increase congestion. Higher incomes attract population growth, which also increases congestion.

In 1992 Anthony Downs coined the term triple convergence in his book Stuck in Traffic: Coping with Peak-Hour Traffic Congestion. Triple convergence describes how peak period traffic congestion is inevitable because drivers will compensate for capacity increases by (a) shifting routes, (b) shifting travel time of travel, and (c) shifting travel mode. After capacity expansion, the new equilibrium will be just as congested as the old equilibrium. Downs writes that drivers will choose “limited-access roads that are faster than local streets if they are not congested”, but the attractiveness of such routes will cause them to become congested “to the point where they have no advantage over the alternate routes.” 

Responding to inevitable peak period urban freeway congestion through widening that will attract more local traffic to the freeway is very expensive and ultimately futile, and also has significant disbenefits to the adjacent urban areas including traffic, noise, air pollution, poor environment for pedestrians, and depressed property values.

Norm Marshall nmarshall@smartmobility.com

winufuwub

Hi Ken et al.,

I am copying my colleagues Juan Carlos Muñoz (Director of our Centre for
Sustainable Urban Transport, CEDEUS) and Ricardo Hurtubia, who has a joint
appointment with us and our School of Architecture (and worked closely in
the past with Francisco Martinez, mentioned by some in the thread) to give
their opinion on these two questions.

First, I am in total agreement with Norm's opinion (I said something
similar in the thread when commenting Clare's) and my answer to you is that
a city without urban freeways and with less space and priority given to the
private car, is a far more livable city. If its planners give priority,
instead, to public transport and "active modes" - such as walking and
cycling - congestion, environmental pollution, noise and accidents reduce
significantly (all negative externalities mainly caused by the cars) and
that is indeed a great benefit in terms of sustainability and quality of
life to all its citizens.

On the other hand, the "economic activity" requires that freight can move
efficiently; so this is a case, which cannot be denied, for better road
infrastructure in certain areas, but not - as it tends to be in many places
- dedicated or justified on the basis of car use.

Juan de Dios

J. de D. Ortúzar
Emeritus Professor
Department of Transport Engineering and Logistics
Pontificia Universidad Catolica de Chile
www.ing.puc.cl/jos

El mié., 22 may. 2019 a las 18:06, KenCervenka ()
escribió:

> The following observations are not about “how to predict” changes over
> time, but rather how to analyze actual observed changes over time. But it
> is not clear to me there is “some other listserv” that could offer anything
> useful on this subject, so….
>
> Perhaps this particular thread has run its course, and there is nothing
> more to be said that can lead to anything constructive, but I am
> (nevertheless) struck by the following:
> -- Rebecca noted that “The purpose of a transportation system is to
> support economic activity.”
> -- Norm offered up some real-world examples where major (and immediate)
> reductions of road capacity in congested areas did not lead to any major
> traffic problems, and asserts that “when urban freeways have been removed,
> there have been great benefits to the cities.”
>
> First question, which I suppose is primarily for Norm: what is your
> definition of a quantifiable “benefit” to a city, and what (if any)
> relationship does this have to “increased economic activity” for that city?
>
> The second set of questions are for anyone who dares to answer, or at
> least offer comment:
> I did a quick-check of Inrix’s web site for the findings from their “2018
> Global Traffic Scorecard,” and observe 115 urban areas in the world that
> had more than 100 hours lost to the average driver in 2018, due to traffic
> congestion (Bogota is the “winner” with 272 hours). Now I would surely
> think these 115 urban areas had some very different levels of population
> growth over the past 20 (or so) years, as well as very different levels of
> increases (maybe even a few decreases) in free or tolled road capacity over
> the same number of years. Plus, for that matter, some very different levels
> of “transit service improvements” over the same number of years. But has
> anyone tried to do some type of “comparative analysis” that attempts to
> identify what (if any) aggregate relationship exists in regards to
> increases (or decreases) in actual road capacity, traffic congestion,
> population growth, and a clear description of whatever constitutes
> “increased economic activity”? Or some other definition for “benefit”? Not
> what comes out of model-based scenario/sensitivity testing, but rather
> focused heavily on actual observed changes, for dozens of different urban
> areas? I think it may be close to impossible to draw useful
> observation-based conclusions because no one will ever really know how much
> (for example) regional population growth (or “economic activity”) in a
> specific urban area might have been different if “actual increases in road
> capacity” were either more or less than what was actually provided, but
> surely there is some academic researcher out there that has at least
> pursued this type of exploratory investigation in a meaningful “comparison
> of different urban areas” way?
>
> About all I can conclude right now is that I am quite happy to not be a
> transportation planner in Bogota, or in any of the “top 20” congested urban
> areas in the world, who tells the local decision-makers to not “worry”
> about traffic congestion over the next 10 years, in the face of continued
> increases in urban population over those same 10 years, because everything
> will eventually work itself out, without new transportation investments! Or
> that it would be even better to reduce road capacity.
>
> *****************************
> From: rebecca.a.knudson
> Sent: Friday, May 17, 2019 3:23 PM
> Congestion delay is a price mechanism in units of time. Users pay the cost
> willingly (albeit begrudgingly). When time-costs rise beyond their ability
> or willingness to pay, they make different choices and adapt to fit their
> time/money budgets and preferences. Transportation is only one among many
> time/budget expenditure decisions made by households and businesses.
>
> I understand the passion and interest in discussing latent demand, induced
> demand, pent up demand, and so on. However, at times I am confounded by the
> “bad” reputation associated with this topic. The purpose of a
> transportation system is to support economic activity. Building capacity to
> accommodate economic activity is the responsibility of the public sector.
> Challenges arise when users are not paying the full cost of the system
> (quantity demanded is greater than quantity supplied) and negative
> externalities arise, such as emissions, noise, unreliability, property
> damage, personal injury and fatalities.
>
> This is a very complex problem to solve with layers of trade-offs. It is
> impractical to focus on only one aspect of the system. For example, if
> safety were our only concern, we would close the roads to all users. If
> vehicle emissions were our only concern, we would require all vehicles use
> emission free power sources. If freight movement was our only concern, we
> would optimize system use to meet that goal. We can’t apply simple
> solutions to a complex problem caused by millions of independent economic
> agents with unique needs and preferences.
>
> “Free” ice cream is not free, the units of cost merely shift from cash to
> time. However, people are free to choose to wait in line. When I was young
> and poor, I waited in line for free ice cream, now I don’t.
>
> From: Norm Marshall
> Sent: Friday, May 17, 2019 3:22 PM
> I appreciate that reduced highway capacity has been brought into this
> discussion. When urban freeways have been removed, there have been great
> benefits to the cities and the expected traffic problems have failed to
> materialize. This has surprised traffic engineers. Sam Schwartz wrote in
> his 2015 book Street Smart: The Rise of Cities and the Fall of Cars:
>
> "On Saturday morning, December 15, 1973, the forty-year-old West Side
> Highway, which runs along the Hudson River and connects the Brooklyn
> Battery Tunnel with the Henry Hudson Parkway— both built by the unstoppable
> Robert Moses— collapsed under the weight of a truck carrying more than
> thirty tons of asphalt. ... The immediate problem— What to do with eighty
> thousand cars a day that would have to find an alternate route?— fell to
> me."
>
> "I would love to take credit for coming up with a brilliant solution that
> saved the city, but the truth is both more mundane and a lot more
> interesting. The predicted traffic disaster never appeared. Somehow, those
> eighty thousand cars went somewhere, but to this day we have no idea where.
> Or how, two years later, twenty-five thousand more people were getting into
> Manhattan’s Central Business District."
>
> Traffic similarly "disappeared" when the earthquake-damaged Central
> Freeway in San Francisco was removed in 1996.
>
> Inspired by the success of both unplanned and intentional urban freeway
> removals in the U.S. and abroad, many initiatives are underway with goals
> of removing urban freeways. For example, NYSDOT recently selected a
> "Community Grid" alternative as the preferred choice to replace elevated
> I-81 through central Syracuse.
>
> A major obstacle to these projects is that just as the static assignment
> models fail to properly account for induced travel, they also fail to
> properly account for reduced travel. Studies with static models generally
> just show that the existing traffic won't fit on a downsized facility, and
> fail to account fully either for traffic re-assignment or changes in
> destinations.
>
> In 2016, I analyzed a range of alternatives in Little Rock ranging from
> freeway expansion to removal using Dynamic Traffic Assignment (DTA). The
> modeling showed significant induced travel with freeway expansion which
> would create new bottlenecks beyond the limits of construction. The
> modeling also showed that the downsizing alternative would not cause
> significant traffic problems.
>
> The report is here:
>
> http://www.arpanel.org/improve30crossing (especially the "Second Report")
>
> I presented this work at the 2017 TRB Planning Applications Conference.
> The abstract is here:
>
> https://www.trbappcon.org/oldsite/2017conf/PresentationDetailsca73.html
> ?...
>
> The link to the presentation no longer works.
>
> Norm Marshall
>
> nmarshall@smartmobility.com
>
> --
> --
> Full post: https://tmip.org/content/induced-demands
> Manage my subscriptions: https://tmip.org/mailinglist
> Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
>

Claire Bozic

I’m sure I’m not saying anything new, but I feel inspired. Yes, there is induced demand or latent demand and our travel models reflect this. Transportation projects can encourage development, change destinations, change modes, change time of travel and change routes. The best projects will have the largest impacts. Induced or latent demand is a signal of transportation project success. Why do we celebrate induced demand in the case of transit projects, but lament induced demand in the case of road projects?

The view that economic activity is just trucks carrying freight is myopic. Access to better quality opportunities, and business access to larger markets and employee pools are critical components of economic activity. The quest to provide access to more destinations and reduce the time between locations was born when commerce began and predates automobiles by far. We built roads to support these activities and traversed them by foot, ox-carts, horses and carts, bicycles, buses, and personal vehicles – each technology designed to reduce travel time and expand opportunities. Roads allowed the growth of commerce and empires. In more recent times when autos were uncommon, people walked or took public transportation in the largest cities in the world. In this idyllic environment, people were not content to carry out the daily business of life only in a walkable area or as far as they could go in a reasonable bus ride, and many voted against these constraints by purchasing vehicles. Why desire to turn the clock back?

Travelers wish to access better destinations not more VMT. They want the best destination within the budgeted travel minutes. For both driving and taking public transportation, travelers don’t know how many miles were traveled and so we don’t ask that question on travel surveys. When capacity is changed, trips change destinations, and use other routes and modes. The daily business of life must carry on in the face of changing choice sets. The fact that a large increase in congestion wasn’t measured in examples where a single road was removed doesn’t disaffirm the fact that when it happens travelers may be pushed to worse destinations, routes or modes than they would have otherwise chosen. If your only measure is vehicle miles traveled, this is a good outcome. Earnest urban planners, drivers, transit riders, community residents, and pedestrians could honestly disagree about whether the outcomes were good or bad. The system optimal solution and the user optimal solution are not the same.

Examples where individual roads were removed without catastrophic mobility impacts leave me unconvinced that removing the expressway system from an urban area has only positive impacts. Given the density of our road systems and the multitude of alternative routes and destinations, the impacts of a single deletion can be distributed over a large area and are therefore individually small. What happens here in Chicago, the freight hub of the United States, when the expressway system is removed and inter-city expressways end abruptly at the urban boundary, nowhere near the rail freight yards? What happens to people in disinvested areas who travel longer distances to get to jobs? What about warehouse distribution to stores where we shop? But what about the reduced pollution, crashes, noise and other “benefits?” There are many ways to address these problems on their own. Removing the expressway system doesn’t seem a practical response.

From: Kenneth.Cervenka=dot.gov@mg.tmip.org On Behalf Of KenCervenka
Sent: Wednesday, May 22, 2019 5:05 PM
To: TMIP
Subject: Re: [TMIP] Induced demands

The following observations are not about “how to predict” changes over time, but rather how to analyze actual observed changes over time. But it is not clear to me there is “some other listserv” that could offer anything useful on this subject, so….

Perhaps this particular thread has run its course, and there is nothing more to be said that can lead to anything constructive, but I am (nevertheless) struck by the following:
-- Rebecca noted that “The purpose of a transportation system is to support economic activity.”
-- Norm offered up some real-world examples where major (and immediate) reductions of road capacity in congested areas did not lead to any major traffic problems, and asserts that “when urban freeways have been removed, there have been great benefits to the cities.”

First question, which I suppose is primarily for Norm: what is your definition of a quantifiable “benefit” to a city, and what (if any) relationship does this have to “increased economic activity” for that city?

The second set of questions are for anyone who dares to answer, or at least offer comment:
I did a quick-check of Inrix’s web site for the findings from their “2018 Global Traffic Scorecard,” and observe 115 urban areas in the world that had more than 100 hours lost to the average driver in 2018, due to traffic congestion (Bogota is the “winner” with 272 hours). Now I would surely think these 115 urban areas had some very different levels of population growth over the past 20 (or so) years, as well as very different levels of increases (maybe even a few decreases) in free or tolled road capacity over the same number of years. Plus, for that matter, some very different levels of “transit service improvements” over the same number of years. But has anyone tried to do some type of “comparative analysis” that attempts to identify what (if any) aggregate relationship exists in regards to increases (or decreases) in actual road capacity, traffic congestion, population growth, and a clear description of whatever constitutes “increased economic activity”? Or some other definition for “benefit”? Not what comes out of model-based scenario/sensitivity testing, but rather focused heavily on actual observed changes, for dozens of different urban areas? I think it may be close to impossible to draw useful observation-based conclusions because no one will ever really know how much (for example) regional population growth (or “economic activity”) in a specific urban area might have been different if “actual increases in road capacity” were either more or less than what was actually provided, but surely there is some academic researcher out there that has at least pursued this type of exploratory investigation in a meaningful “comparison of different urban areas” way?

About all I can conclude right now is that I am quite happy to not be a transportation planner in Bogota, or in any of the “top 20” congested urban areas in the world, who tells the local decision-makers to not “worry” about traffic congestion over the next 10 years, in the face of continued increases in urban population over those same 10 years, because everything will eventually work itself out, without new transportation investments! Or that it would be even better to reduce road capacity.

*****************************
From: rebecca.a.knudson
Sent: Friday, May 17, 2019 3:23 PM
Congestion delay is a price mechanism in units of time. Users pay the cost willingly (albeit begrudgingly). When time-costs rise beyond their ability or willingness to pay, they make different choices and adapt to fit their time/money budgets and preferences. Transportation is only one among many time/budget expenditure decisions made by households and businesses.

I understand the passion and interest in discussing latent demand, induced demand, pent up demand, and so on. However, at times I am confounded by the “bad” reputation associated with this topic. The purpose of a transportation system is to support economic activity. Building capacity to accommodate economic activity is the responsibility of the public sector. Challenges arise when users are not paying the full cost of the system (quantity demanded is greater than quantity supplied) and negative externalities arise, such as emissions, noise, unreliability, property damage, personal injury and fatalities.

This is a very complex problem to solve with layers of trade-offs. It is impractical to focus on only one aspect of the system. For example, if safety were our only concern, we would close the roads to all users. If vehicle emissions were our only concern, we would require all vehicles use emission free power sources. If freight movement was our only concern, we would optimize system use to meet that goal. We can’t apply simple solutions to a complex problem caused by millions of independent economic agents with unique needs and preferences.

“Free” ice cream is not free, the units of cost merely shift from cash to time. However, people are free to choose to wait in line. When I was young and poor, I waited in line for free ice cream, now I don’t.

From: Norm Marshall
Sent: Friday, May 17, 2019 3:22 PM
I appreciate that reduced highway capacity has been brought into this discussion. When urban freeways have been removed, there have been great benefits to the cities and the expected traffic problems have failed to materialize. This has surprised traffic engineers. Sam Schwartz wrote in his 2015 book Street Smart: The Rise of Cities and the Fall of Cars:

"On Saturday morning, December 15, 1973, the forty-year-old West Side Highway, which runs along the Hudson River and connects the Brooklyn Battery Tunnel with the Henry Hudson Parkway— both built by the unstoppable Robert Moses— collapsed under the weight of a truck carrying more than thirty tons of asphalt. ... The immediate problem— What to do with eighty thousand cars a day that would have to find an alternate route?— fell to me."

"I would love to take credit for coming up with a brilliant solution that saved the city, but the truth is both more mundane and a lot more interesting. The predicted traffic disaster never appeared. Somehow, those eighty thousand cars went somewhere, but to this day we have no idea where. Or how, two years later, twenty-five thousand more people were getting into Manhattan’s Central Business District."

Traffic similarly "disappeared" when the earthquake-damaged Central Freeway in San Francisco was removed in 1996.

Inspired by the success of both unplanned and intentional urban freeway removals in the U.S. and abroad, many initiatives are underway with goals of removing urban freeways. For example, NYSDOT recently selected a "Community Grid" alternative as the preferred choice to replace elevated I-81 through central Syracuse.

A major obstacle to these projects is that just as the static assignment models fail to properly account for induced travel, they also fail to properly account for reduced travel. Studies with static models generally just show that the existing traffic won't fit on a downsized facility, and fail to account fully either for traffic re-assignment or changes in destinations.

In 2016, I analyzed a range of alternatives in Little Rock ranging from freeway expansion to removal using Dynamic Traffic Assignment (DTA). The modeling showed significant induced travel with freeway expansion which would create new bottlenecks beyond the limits of construction. The modeling also showed that the downsizing alternative would not cause significant traffic problems.

The report is here:

http://www.arpanel.org/improve30crossing (especially the "Second Report")

I presented this work at the 2017 TRB Planning Applications Conference. The abstract is here:

https://www.trbappcon.org/oldsite/2017conf/PresentationDetailsca73.html?...

The link to the presentation no longer works.

Norm Marshall

nmarshall@smartmobility.com

--
--
Full post: https://tmip.org/content/induced-demands
Manage my subscriptions: https://tmip.org/mailinglist
Stop emails for this post: https://tmip.org/mailinglist/unsubscribe/13101
CAUTION: This email originated from outside of the organization. Do not click links or open attachments unless you recognize the sender and know the content is safe.