Excerpt from “The Planners Guide to Advocacy Modeling” by Alan Horowitz
Chapter 12. Shielding and Weaponizing Your Travel Model
Earlier chapters of this book describe methods of building and calibrating a travel model to assure that it produces the desired policy outputs. However, a model will not be particularly useful in advancing a political agenda unless certain steps are taken to protect the model from criticism and to maintain exclusivity. Doing these things is difficult in a public environment. This chapter details all steps to be taken to assure your model is properly shielded and weaponized.
Building Model Credibility through Big Documents and Public Forums
Earlier chapters of this book show how model insensitivity to certain policies can work to the agency’s advantage. Admit to no deficiencies. Your model is always state-of-the-practice. An unwanted enhancement is unproven research. Cite other models around the country that do it the same way (there are always some). Tout the model as the best way to analyze all transportation plans. Publish very large reports with lots of nifty, preferably irrelevant, graphics and lots of mathematics to show how much thought has gone into the modeling effort. Be indignant when anybody suggests the model could be improved. Discredit any attempts by interest groups to challenge transportation plans by pointing out their lack of expertise or their lack of any evidence from their own travel model.
Managing Outside Expertise
The importance of mitigating bothersome experts cannot be overstressed. Many large urban areas have university faculty with travel forecasting expertise. Many of these people can be easily discredited if travel forecasting is not their principal teaching or research specialty. Experts at consulting firms can be neutralized by a mere hint of distain. Under no circumstance should a modeling expert be invited to join a technical committee or attend a public participation event. Intimidation and defamation have little long-term value and may help to build opposing coalitions. Honoring or otherwise ingratiating yourself with the expert may buy some time, but it may not provide long-term benefits. So as not to appear biased, standing policies that broadly marginalize all experts are better than ad hoc policies that target one particularly troublesome expert. State DOTs have modeling experts, but they can be neutralized by interagency agreements to stay away from each other’s territory. Participating in the development of a state-wide model might encourage DOT staff to look closely at your local model, so such participation should be minimized.
Controlling the Peer Review Process
Solicited, as well as unsolicited, criticism by experts is to be avoided, including any peer reviews. Peer reviews of travel models are most often conducted by the Travel Model Improvement Program of FHWA and FTA. TMIP chooses unbiased reviewers with considerable expertise, with the peer review managed and documented by a third-party contractor to TMIP. Peer reviews are open forums, often providing an opportunity for public discourse. Peer reviews, by nature, are difficult to control and may produce suggested improvements that run counter to agency objectives.
The most important principles of peer review are the three D’s: deny, delay, and deflect. There is no advantage to having a peer review for a model designed to advocate for a specific planning outcome, beyond public validation of what already exists. Therefore, the need for a peer review should be denied, if possible. However, political pressure might force your agency to give an impression it is making a sincere effort at a review. In those cases, it is necessary to control the substance and flow of information. The least option is to delay any external review by doing an internal audit of the model. To help make the internal audit seem credible, your model could, for instance, be compared with a few models at similar agencies. Proper selection of those other agencies/models is essential to the successful conclusion of the internal review. Should an internal review not satisfy critics, then there might be a need to bring in outside experts. Do not use TMIP’s services, which are problematically unbiased. Identify a few modelers from similar agencies as yours to participate. Do not bring in anybody from academia. The location of the review should be at a remote location, preferably outside the boundaries of the MPO and even more preferably outside your state. Such a location will discourage attendance from representatives of special interest groups, who might want to avail themselves of any “open meeting” requirements. These actions should be sufficient to deflect criticism of the absence of a TMIP-style peer review.
Protecting the Model from Prying
The travel model, including all input data, plus survey data should be treated as proprietary. Technical assistance for a fee may be provided to special interest groups and outside agencies in lieu of giving them access to the model. Preference should be given to a very expensive, modeling platform, so as to discourage special interest groups from buying their own modeling capability. To avoid leaks, all model development must be done in-house.
Preventing Access to Data
Never share data with anyone, not even the state DOT. Even though all such data were collected at public expense and subject to open-records laws, it is important to deny access to these data. However, you should be willing to provide data summaries, to anyone who wishes to reimburse your agency for its time and trouble, since summaries are unlikely to be helpful in countering findings from your travel model. You should be ready with excuses for not providing raw data, such as “the data have not been verified” or “there are privacy concerns”. You may be permitted to charge a processing fee for fulfilling open records requests. Since most special interest groups have tight budgets, it usually takes only a slightly exorbitant fee to discourage further inquiries. Threats of a lawsuit are unlikely to materialize, since legal budgets of special interest groups are also meager. Universities and outside public agencies avoid lawsuits, as a rule. Stonewalling is best accomplished with professionalism and politeness.
Conclusion
Shielding and weaponizing a travel model require a multifaceted and consistent strategy, applied over a long period of time. Even the smallest breach in this effort can cause your model to lose credibility, thereby undermining its ability to advance political objectives.
[Author’s note: This article is satire. I am not advocating doing any of this. However, everything in this article is based on actual events I have observed over my career. Alan Horowitz, Whitefish Bay, October 23, 2020]
ja ja ja, how sad that life has to be this way
I’m glad I read it all the way through — and the author’s note twice. 🙂
I sent the following e-mail message to a friend who has often observed misuse of travel models:
We cannot shame these bad actors into using their models ethically, so what’s to be done? Those of us in professional organizations have Canons of Ethics, but these are routinely ignored. Perhaps the only viable counter to such bad behavior is to enable those “Thousand Friends” with their own modeling capabilities. There are few people out there like David Hartgen who are willing to call out politically slanted forecasts.