By Ian Savage

Can we expect further railroad regulation and increased regulatory enforcement to have a significant effect on ameliorating risk? Or would legal actions to ensure that the full costs of external harms are passed along to the shippers of hazardous materials be more effective? Savage provides an economic perspective on regulations governing the movement of hazardous materials by rail.

Virginia Issues & Answers, Spring 2015

Download a PDF of this article

Print This Post Print This Post

Winter 2015 Issue

The fiery derailment and subsequent oil spill into the James River at Lynchburg, Virginia on April 30, 2014 has prompted a public debate concerning the carrying of hazardous materials by rail. The debate extends beyond Virginia as railroads have be- come, in the absence of additional pipeline capacity, the predominant method for long-distance transportation of oil extracted from the rapidly expanding oilfields in the Bakken formation of North Dakota and neighboring states and Canadian provinces. Bakken crude oil appears to be particularly volatile.

The debate is not new, although the type of cargo is. In 2005, a collision at Graniteville, South Carolina, led to nine deaths af- ter a tank car filled with chlorine ruptured (“Collision of Norfolk Southern Freight Train 192 with standing Norfolk Southern Lo- cal Train P22”). The subsequent federal Rail Safety Improvement Act of 2008 required installation of “positive train control” on

mainlines where materials that are poisonous or toxic by inhala- tion are carried (“Rail Safety Improvement Act”). Positive train control is a system in which the locations of trains are monitored and brakes can be applied to avoid collisions. Earlier public policy interventions were in response to the increased carrying of chemicals in the 1960s and 1970s and munitions and explosives in the early 1900s. These interventions included promulgation of rules on the design of railroad tank cars.

All of the occurrences described thus far have in common something that economists call “externalities,” which are spillover or unintended effects on communities adjacent to railroads. It is worth stating that most railroad collisions and derailments are not of this type. Equipment and track are routinely damaged or destroyed but there is little or no harm to the surrounding com- munities (“The Economics of Railroad Safety”). Certainly such incidents are a cause of concern to railroad management, railroad employees, and those who ship their goods by rail, but they should not elicit a public policy response. The parties involved have sufficient incentives to take appropriate action.

However, nearly all of the nightmare scenarios involve situa- tions where nearby communities may have to be evacuated, local residents may be killed or injured, or environmental damage occurs. These are not just hypothetical concerns, as demonstrated by numerous incidents most notably the derailment of a runaway train carrying Bakken crude at Lac-Mégantic, Quebec, in 2013 which resulted in the death of 47 local residents (Runaway and Main-Track Derailment”).

While not minimizing the human suffering, the mere pos- sibility of externalities need not require a regulatory response. Nobel Prize-winning economist Ronald Coase argued in 1960 that the party causing the externality has the proper incentives to control the risk if the parties harmed can easily obtain full compensation through legal processes (“The Problem of Social Cost”). A failure of the market processes only occurs if some harms cannot be fully recovered, or if the legal transactions costs are too high.

There is good reason to believe that the full costs are not transferred to the railroads. Some costs of emergency services and clean-up expenses may not be ordinarily recoverable. The public sector may be left on the hook for some portion of the remedia- tion of environmental damage. Local residents who are temporar- ily evacuated may receive payment for out-of-pocket expenses, but are not fully compensated for their inconvenience. Neighbor- ing farms and businesses will probably be fully compensated for damaged property and lost inventory, but not for forgone profits.

Even if all bystanders were fully compensated, there can still be problems. Railroads carry many products from the benign to the very hazardous. For the market to work the railroads would need to pass along in the form of prices the expected costs of liability settlements associated with that particular commod-
ity. Consider what would happen if the railroad just charged a standard mark-up on all types of hazardous materials to cover liability settlements. In these circumstances the freight rate for ultra-hazardous materials would be too low, so too much would be transported. The shippers of these goods would not face the correct incentives to consider whether there should be changes to the manufacturing location or the distribution network. The reverse would be true for less-hazardous goods, as the freight rate

would be too high and too little would be transported. Perhaps the problem with Bakken crude is that the price the railroads charge oil companies is too low.

Legal remedies are also flawed if the magnitude of the dam- ages forces the defendants into bankruptcy with the result that claims are not fully settled. Following the Lac-Mégantic tragedy, the Montreal, Maine, and Atlantic Railway declared bankruptcy as the claims exceeded its assets and its insurance coverage (“MM&A files for creditor protection after Lac-Mégantic rail di- saster”). Of course, liability also only occurs after the fact, which is why some legal scholars argue that before-the-fact regulation should be deployed in combination with after-the-fact liability.

Comprehensive safety regulation of the railroads dates from 1970. The push for regulation was the result of a couple of decades of poor financial conditions that had led to disinvest- ment from the railroad industry and an increase in derailments and collisions. The economic deregulation of the industry in 1980 brought a dramatic turnaround in both profitability and safety. Unlike the airline and trucking industries, where there was a debate as to whether economic deregulation had led to compromises in safety, the railroad industry has flourished after deregulation, reinvested in its infrastructure, and improved its safety record considerably.

Safety regulation should not be seen as a panacea. Press reports indicate that track defects in the Lynchburg area were known by both the railroad (“CSX found defect in track”) and government inspectors (“Inspectors find scores of defects on CSX rails across the state”) prior to the derailment. Moreover, most of the tank cars involved were built recently and to a higher stan- dard than the older cars that were involved in the Lac-Mégantic derailment (“Re-inventing the DOT 111”). So almost 50 years

Ian Savage is a member of the faculty of both the Depart- ment of Economics and the Transportation Center at Northwestern University. His research is concentrated in two areas: urban public transportation and transportation safety. He has conducted research into the safety perfor- mance and the effectiveness of safety regulations in most modes of transportation with a particular emphasis on the trucking and railroad industries.

Winter 2015 35

of federal regulation had little impact on both the causation and the consequences of the Lynchburg derailment.

Can we expect further regulation and increased enforcement of regulations to have a significant effect on ameliorating risk? Or would legal actions to ensure that the full costs of external harms are passed along to the shippers of hazardous materials be more effective? Faced with higher transportation costs, the shippers may demand more modern tank cars, choose modified routings, change their supply chain, or consider whether to reformulate their products.

The arguments thus far may have given the reader a skewed perspective on the most pressing rail safety issues. There is no doubt that the derailment at Lynchburg could have been much worse. However, a look at the tabulation of rail-related fatalities and injuries in Virginia in 2012 and 2013 highlights the fact that priority should be given to tackling other types of risks.

Table 1: Railroad Fatalities and Injuries in Virginia, 2012 and 2013

crossings?”). While the risk to motor vehicle occupants at cross- ings has dropped significantly, the number of pedestrians killed has remained stubbornly unchanged. In the U.S. in recent years, an average of 775 pedestrians have died each year on railroads

as users of grade crossings, trespassers on the tracks, or suicide victims. Pedestrians comprise 80 percent of total railroad deaths, with motor vehicle occupants at grade crossings representing another 16 percent (“Analysis of fatal train-pedestrian collisions in metropolitan Chicago 2004-2012”). So while fiery derail- ments may claim public attention, perhaps the biggest payoff in lifesaving would come from continuing aggressive measures to improve or eliminate grade crossings and to developing trespass and suicide prevention strategies.

Works cited

Coase, Ronald H. “The problem of social cost.” Journal of Law and Economics 3 (1960): 1-44.

“CSX found defect in track a day before Lynchburg derailment.” Richmond Times-Dispatch, 4 Jun. 2014. Web. 2 Oct. 2014.

“Inspectors find scores of defects on CSX rails across the state.” Tidewater Daily Press, 27 Sep. 2014. Web. 2 Oct. 2014.

“MM&A files for creditor protection after Lac-Mégantic rail disaster.” The Globe and Mail, 7 Aug. 2013. Web. 2 Oct. 2014.

Mok, Shannon and Savage, Ian. “Why has safety improved at rail-highway grade crossings?” Risk Analysis 25 (2005): 867-81.

National Transportation Safety Board. “Collision of Norfolk Southern Freight Train 192 with standing Norfolk Southern Local Train P22 with Subsequent Hazard- ous Materials Release at Graniteville, South Carolina, January 6, 2005.” Railroad Accident Report NTSB/RAR-05/04. 29 Nov. 2005. Web. 2 Oct. 2014.

“Re-inventing the DOT 111.” Railway Age, 7 Feb. 2014. Web 2 Oct. 2014.

Savage, Ian. The Economics of Railroad Safety. Boston: Kluwer Academic Publish- ers, 1998. Print.

Savage, Ian. “Analysis of fatal train-pedestrian collisions in metropolitan Chicago 2004-2012.” Proceedings of the 2014 Global Level Crossing Symposium, Urbana- Champaign: University of Illinois, 2014. Print.

Transportation Safety Board of Canada. “Runaway and Main-Track Derailment, Montreal, Maine & Atlantic Railway, Freight Train MMA-002, 06 July 2013.” Railway Investigation Report R13D0054. 19 Aug. 2014. Web. 2 Oct. 2014.

U.S. Congress. “Rail Safety Improvement Act.” Public Law 110–432, 16 Oct. 2008. Web. 2 Oct. 2014.



Employees and Contractors



Passengers on Trains



Others Legally on Railroad Property



Motor Vehicle Occupants at Grade Crossings



Pedestrians at Grade Crossings












Source: Federal Railroad Administration

Collisions with motor vehicles at grade crossings continue to be a major concern, although the risks have been reduced substantially in the past 40 years. The reduction is due to a combination of crossing closures, installation of gates and/or flashing lights where previously there were only warning signs, and a public education campaign under the “Operation Life- saver” banner (“Why has safety improved at rail-highway grade

Author Biography:

Paul Manna is associate professor of government and public policy…