Technical solutions for policy problems; Policy solutions for technical problems

The transportation sector is rife with technology/policy mismatches. We often seek new to develop or deploy technologies to solve what are ultimately policy problems; and we often try to implement policies to solve what are ultimately technical problems. Attaching the correct domain to the problem is the first step in solving it.

As an example of the first, consider any transportation investment aimed at expanding capacity to address congestion. The policy failure here is the lack of peak-period pricing (and secondarily location-specific pricing) that would ration scarce capacity. Instead we move to a ‘predict and provide’ mode, and where we fail to provide, we ration by queueing. Whether or not we should build a project to expand accessibility from a place, we should not expand capacity to expand accessibility only at a given time-of-day until we have properly priced the network, so that travelers consider the externalities they impose and pay the full marginal costs of their trip.

A Political Economy of Access: Infrastructure, Networks, Cities, and Institutions by David M. Levinson and David A. King
A Political Economy of Access: Infrastructure, Networks, Cities, and Institutions by David M. Levinson and David A. King

The best illustrations include your typical freeway widening. A widening by definition just expands capacity (which is only relevant in the peak hours), and does nothing to expand accessibility during the rest of the day, since no new places are connected. The road may very well be congested, which is what you would expect when you give something valuable away at below marginal cost. The capacity expansions often don’t have the desired effect on congestion reduction due to induced demand in the short run, and induced development in the longer run.

Transit projects are often billed as “congestion reducing”. This is the one of the worst reasons to invest in transit. It is as if we said a new highway reduces crowding on the bus or the bike lanes. Transit investments should be justified by the benefits they provide to existing and new transit users, not on some marginal third-party actor. Yet we see this claim all the time. [I know, the claim is marketing to induce non-users to see a benefit for the enormous subsidy they are providing, but it is intellectually dishonest]. I won’t say there is never a highway congestion reduction by-product, as if the project does gain new riders, at least some of them may have otherwise been driving a car during peak times, but that is a third-order effect that often does not appear in the traffic statistics.

As an example of out-of-sync policy approaches to technical problems, consider emissions reduction. Automobile emissions are a problem for many reasons, not the least of which is public health. Emissions can be thought of as a logical chain of calculations like this:

Tailpipe Emissions = (Number of Motor Vehicle Trips) * (Average Trip Length (Miles)) * (Fuel Efficiency (Gallons / Mile)) * (Emission Rate (Pollutants/Gallon))

The land use and transportation planning side of things (policy) tries affect behavior to reduce number of vehicle trips and average trip length. The technology side of the equation tries to improve fuel efficiency and reduce pollution per unit of fuel burned. There are policies that can encourage the development of the technology (CAFE standards) and technologies which can reduce average trip length, like elevators and skyscrapers. But the opportunity to practically eliminate emissions can only come from driving one of these numbers to zero.

Electric cars at Imperial - 2

My bet would be that it is far easier to change the power plant on the fleet (through electrification), and the pollutants per gallon through more efficient combustion if you continue to burn liquid fuel, than to get number of motor vehicle trips or trip length close to zero. In the US, we might be able to take baby steps towards behavioral change, but we are starting from an existing built environment, an existing economic make-up, and existing transportation networks that took a century to build and will take a century to unwind. We have climbed Mt. Auto, and atop Mt. Auto, Mt. Not-Auto is very distant.

In short I claim the vehicle fleet, which turns over every 10 or so years, is more malleable than land use, which turns over every 50 years, or road networks, which effectively don’t turn over, or travel behavior in the absence of pricing.

This is not to say people should not be encouraged to walk (or bike if masochistic) and locate near jobs if feasible. Similarly, developers should not be needlessly discouraged or prohibited from high densities so long as they internalize costs or pays off the losers. But the reasons behind that are fundamentally private (that is what people prefer) rather than for the good of the environment. Preferences too are malleable, to a point, but reprogramming 300 million individual, customized brains takes time and effort, and the population only turns over generationally.

I walk (even in the snow and ice). There are many reasons people may want to not drive their car, or not drive as much. Personal cost and pleasure are among them. But to promote those policies as serious efforts to address pollution problems, when technical solutions are available is Sisyphean. Induced demand rears its beloved head. A large part of the reductions in VMT from one set of sources will be eaten by another, even if we have macroscopic trends working (very slowly) in a favorable direction, like peak travel.

There are potentially effective travel demand management (TDM) strategies. Pricing (as per above) is the most effective, if only it could be implemented. By all means, impose a carbon tax, but that will have very small effects on overall behavior unless the tax itself is very large.

Sadly, most policies, short of pricing, aimed at TDM have been dismally unsuccessful. Land use changes are slow at best, and there is little guarantee that new development will do much to actually reduce vehicle emissions, most people don’t work or live on site, and new transit oriented developments are not in dense environments to do enough to enable people to forego much auto travel.

Similarly, improving the efficiency of roads (transportation systems management or TSM) through strategies like improved signal coordination and ramp metering are modest in their gains.

This is the classic hammer/nail conundrum. Some people want to help save the world. This is a wonderful and potentially socially-productive endeavor. Yet, their only tool is a hammer. Their conclusion is that saving the world is a nail, when it is really a snowglobe. We would be better served by finding nails for the hammer users, and rags for the snowglobe tenderers.

To get meta about this, policies which provide proper incentives can help match the wannabe heroes with the real Penelope Pitstops . Technologies can help with the matchmaking market as well, as has proven so popular in the dating world.

8 thoughts on “Technical solutions for policy problems; Policy solutions for technical problems

  1. I appreciate the pragmatic approach, You certainly bring up good points on behavioral change as it relates to generational ideals, land-use turnover change time, etc all vs time to turnover the vehicle fleet in the likelihood of that being the best shot at reducing emissions to near zero (or as low as possible). My question is – great, vehicles can turn over every 10-20 years and fit in to the American ethos of car-culture and existing land-use patterns. But how long does the power infrastructure change over that feeds the new, efficient vehicles? We have existing large-scale utility production facilities – a hefty amount of capital invested by municipally owned or cooperated entities with lifespans of 50+ years. Further, we have similar facilities invested in fossil fuel exploration/extraction/refining. The private entities (and social consciousness) will not be swayed or moved easily. We have a collective amount of capital invested in the money that’s still in the ground (futures investing, etc). I’d be curious to see your take on how these factors weigh in on how likely this fleet turnover will be able to affect climate change vs the other options you mention (which, it’s obvious you’re not opposed to, only that they are longer-term solutions to reverting to a smarter infrastructure/development pattern).

    Further, reducing vehicle emissions alone still doesn’t take care of other environmental issues like land-use change (loss of vegetation, farm-land, water resources due to low-density development), inefficiencies of detached structures, financial viability of the very places we’ve built, and larger societal problems caused at least somewhat by our land-use (depression, isolation, obesity, etc) and the costs they bring. How do those factor in when considering a truly holistic approach to policy and technical solutions?

    Again, great post.


    1. Well, I am not an expert on energy systems, BUT, based on what I have seen, I am a technological optimist in this area.

      I have seen curves on solar power that suggest the long-awaited dream of the 70s will soon be alive in America. My dad, (admittedly leading edge, but not one to pay more than necessary) living in Arizona, already has solar rooftop installation powering electrical and heating water. See .

      We have already seen a bending of the curve on Carbon in the US, as emissions are down to ~1990 levels, due primarily to substitution of natural gas for coal.

      So I think the energy problem will take care of itself (from the perspective of users, like transportation), obviously energy people have to actually do this, and inevitably some government subsidies will try to push this along faster than the market.

      Other land use changes:

      Vegetation, ok, but the world’s population will peak at 9 billion or so sometime this century, and planet-wide the amount of space for houses in particular is relatively small, and will mostly come out of farmland (since farmland tends to be adjacent to metro areas, rather than wilderness. Locally at least suburbs are far better for the environment than cities (though globally may be a different issue), since the land around a house can accommodate runoff etc. without treatment.

      Farmland, this is environmentally the most harmful stuff we do I think, and is a function of number of people and food productivity. Since people probably don’t want SLURM, we will continue to see real food. The conversion of nature to farmland is a problem, and charging for environmental consequences of run-off etc. is probably the best we can do unless we monetize land in a natural state (which people have tried to do, but nothing is convincing) or give rights to non-human life forms. There is an inherent values conflict between land for people and land for nature (sans people).

      Detached structures: The inefficiency you refer to is an energy loss and land loss I am guessing, so see above. This is a preferences thing as far as I can tell. Having lived in apartments growing up, I prefer not having neighbors on the ceiling.

      Financial viability: This is strictly an economic/political issue. I am not as pessimistic as StrongTowns on the money side of things, we could just raise taxes a bit and defer the large screen TV if we really want more infrastructure, its a municipal preference. Clearly many places are over-building for what the need, and that should be avoided, but people “waste” money all the time individually (what is “need” anyway, whether they should do that communally is a political question.

      So I say, get the prices right, and let people do what they want in that environment. If people chose social isolation (etc.), it might not be the choice I would make, but it’s their right to do it.


      1. Just a few responses, I agree with many of your thoughts but have a slightly different take on some but a discussion on nuance isn’t worth it 🙂

        Peaking population (if it does indeed happen) at 9 billion is 2b more than today – meaning if housing does come from farmland (assuming the US and developing nations urbanize at lower densities than previous generations), the resulting farmland to replace the old and add capacity for the extra ~28% population we’ve added makes for a pretty big loss. Just a different take.

        I am an optimist in the technology side as well, but I think we’ll see large jumps in energy usage thanks to the efficiency gains we see (Jevons). If the efficiencies (energy unit cost) come from lower emitting forms, great. But have we fuly understood what it would mean if everyone in India, China, Brazil, etc consumed the amount of energy we do in the US, even if it were from solar/wind/water sources (with lower emission sources like natural gas making up the other large chunk)? I’m not so sure we have (and I’m not making a case for leaving these huge populations in a lower state of existence, just asking the question on what it means for total CO2 output).

        From a municipal/financial side, I agree that places just need to strike the balance between level of service and tax rates (which boils down to the individual making the trade off between tax dollars or the TV). Unfortunately, I haven’t seen many areas in decline really be able to grapple with that choice (see the recent example of taxpayer outrage: ) but maybe I’m being overly pessimistic as SmartGrowth/etc is starting to get a foothold.

        Your example of living with neighbors on the ceiling provides a great example of 2 solutions to a problem. The noise they make could be solved by a political decision to allow people to live apart from each other (and the fees/consequences that come with it) or a technical one in better building codes and sound barriers to fix the root cause of the problem. Just an idea. And I don’t live in an apartment so I’m not judging, just thinking aloud.

        I agree wholeheartedly on getting the prices right. I think it would be unwise to under-price carbon as a means for change. I really like the way this Grist article ( ) states the situation and why a very low discount rate makes the most sense, but that’s just me. It would certainly have the effect of speeding up competing technologies and industries to meet demand for energy, housing, etc without the need for subsidization. Just my take.


  2. Behavior change on a large scale can happen within the time scale of the “fleet changeover”, but it is a serious effort and probably needs to be accompanied by technological improvement. See recycling or water clean up efforts. The problem with the issue of climate change is that it likely requires both voluntary behavior change and significant pricing and technological improvements for us to avoid the worst impacts.

    You’re wrong on the stormwater piece, see this EPA study: Basically on the whole, denser development is better because less land is made impervious per unit. Typically mowed lawns are only a small improvement over concrete.

    The carbon curve for the US is bending (depending on your inventory methodology), but the atmosphere knows no political boundaries. Cheap natural gas means we’re exporting our coal to be burned elsewhere: Here’s the curve that matters:


    1. See

      The issue is more complicated. Per unit area more pervious surface is better. Higher population density concentrations, keeping totals constant, as EPA uses, creates more pervious surface globally, since less rooftop and street etc. are built per person, but not locally, since cities are mostly impervious, and polluted storm water is mostly unfiltered.


      1. Total runoff is the useful metric I think, not amount of pervious area. How does it not create more runoff locally? According to EPA’s results, 8 houses developed at 1 unit per acre generates 149,000 cubic feet per year of runoff. 8 houses developed at 8 units per acre generates 39,000 cubic feet of runoff (see page 15). Likewise, the watershed level analysis on page 16 is instructive. Even if you develop a whole watershed at 20% impervious cover, you generate more runoff for the same amount of units built at a higher density.

        If we have to choose a growth policy for future development from a stormwater perspective, it should be higher density.


  3. Because the ground has absorptive capacity to filter runoff. At 8 units per acre, the water all goes to storm sewer. At 1 unit per acre, the ground might be able to handle the excess water, and it is filtered before it reaches the river, e.g.

    Isn’t this the entire rationale for storm water detention basins?


    1. I think this is the purpose of the term “runoff” in the EPA report, it is stormwater that is no infilltrated or filtered. Low density development has far less absorptive capacity than commonly thought. From the report:

      1. The “pervious” surface left in low-density development often acts like impervious surface.
      2. Density and imperviousness are not equivalent.
      3. Low-density developments often mean more off-site impervious infrastructure. (pages 7 & 8)

      This is the purpose of retention basins, in both built-up and less built-up (more lawn) environments. However, the total runoff (untreated, unfilltrated) leaving the site and entering waterbodies is less for an equivalent number of units at a higher density.

      From the report:

      “These results indicate when runoff is measured by the acre, limiting density does minimize water quality impacts compared to the higher-density scenarios. However, when measured by the house, higher densities produce less stormwater runoff.” (page 14)

      If we know we’re going to grow a certain amount in a region (or at least guess we are going to grow some), we shouldn’t measure stormwater impacts per each acre developed, but as a whole per the total amount of development we will experience. To best protect watersheds, this report concludes higher densities are usually better.

      “Taken together, these findings indicate that low-density development may not always be the preferred strategy for reducing stormwater runoff. In addition, the findings indicate that higher densities may better protect water quality—especially at the lot and watershed levels. Higher-density developments consume less land to accommodate the same number of houses as lower density. Consuming less land means less impervious cover is created within the watershed. To better protect watershed function, communities must preserve large, continuous areas of open space and protect sensitive ecological areas, regardless of how densely they develop.” (page 29)


Comments are closed.